[PATCH qemu v4 00/14] Add tail agnostic behavior for rvv instructions

All of lore.kernel.org
 help / color / mirror / Atom feed

* [PATCH qemu v4 00/14] Add tail agnostic behavior for rvv instructions
@ 2022-03-24 19:03 ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

According to v-spec, tail agnostic behavior can be either kept as
undisturbed or set elements' bits to all 1s. To distinguish the
difference of tail policies, QEMU should be able to simulate the tail
agnostic behavior as "set tail elements' bits to all 1s". An option
'rvv_ta_all_1s' is added to enable the behavior, it is default as
disabled.

There are multiple possibility for agnostic elements according to
v-spec. The main intent of this patch-set tries to add option that
can distinguish between tail policies. Setting agnostic elements to
all 1s makes things simple and allow QEMU to express this.

We may explore other possibility of agnostic behavior by adding
other options in the future. Please understand that this patch-set
is limited.

v2 updates:
- Addressed comments from Weiwei Li
- Added commit tail agnostic on load / store instructions (which
   I forgot to include into the patch-set)

v3 updates:
- Missed the very 1st commit, adding it back

v4 updates:
- Renamed vlmax to total_elems
- Deal with tail element when vl_eq_vlmax == true

eopXD (14):
  target/riscv: rvv: Prune redundant ESZ, DSZ parameter passed
  target/riscv: rvv: Rename ambiguous esz
  target/riscv: rvv: Early exit when vstart >= vl
  target/riscv: rvv: Add tail agnostic for vv instructions
  target/riscv: rvv: Add tail agnostic for vector load / store
    instructions
  target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions
  target/riscv: rvv: Add tail agnostic for vector integer shift
    instructions
  target/riscv: rvv: Add tail agnostic for vector integer comparison
    instructions
  target/riscv: rvv: Add tail agnostic for vector integer merge and move
    instructions
  target/riscv: rvv: Add tail agnostic for vector fix-point arithmetic
    instructions
  target/riscv: rvv: Add tail agnostic for vector floating-point
    instructions
  target/riscv: rvv: Add tail agnostic for vector reduction instructions
  target/riscv: rvv: Add tail agnostic for vector mask instructions
  target/riscv: rvv: Add tail agnostic for vector permutation
    instructions

 target/riscv/cpu.c                      |    1 +
 target/riscv/cpu.h                      |    2 +
 target/riscv/cpu_helper.c               |    2 +
 target/riscv/insn_trans/trans_rvv.c.inc |  154 +++
 target/riscv/internals.h                |    5 +-
 target/riscv/translate.c                |    2 +
 target/riscv/vector_helper.c            | 1597 ++++++++++++++---------
 7 files changed, 1140 insertions(+), 623 deletions(-)

-- 
2.34.1


^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 00/14] Add tail agnostic behavior for rvv instructions
@ 2022-03-24 19:03 ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

According to v-spec, tail agnostic behavior can be either kept as
undisturbed or set elements' bits to all 1s. To distinguish the
difference of tail policies, QEMU should be able to simulate the tail
agnostic behavior as "set tail elements' bits to all 1s". An option
'rvv_ta_all_1s' is added to enable the behavior, it is default as
disabled.

There are multiple possibility for agnostic elements according to
v-spec. The main intent of this patch-set tries to add option that
can distinguish between tail policies. Setting agnostic elements to
all 1s makes things simple and allow QEMU to express this.

We may explore other possibility of agnostic behavior by adding
other options in the future. Please understand that this patch-set
is limited.

v2 updates:
- Addressed comments from Weiwei Li
- Added commit tail agnostic on load / store instructions (which
   I forgot to include into the patch-set)

v3 updates:
- Missed the very 1st commit, adding it back

v4 updates:
- Renamed vlmax to total_elems
- Deal with tail element when vl_eq_vlmax == true

eopXD (14):
  target/riscv: rvv: Prune redundant ESZ, DSZ parameter passed
  target/riscv: rvv: Rename ambiguous esz
  target/riscv: rvv: Early exit when vstart >= vl
  target/riscv: rvv: Add tail agnostic for vv instructions
  target/riscv: rvv: Add tail agnostic for vector load / store
    instructions
  target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions
  target/riscv: rvv: Add tail agnostic for vector integer shift
    instructions
  target/riscv: rvv: Add tail agnostic for vector integer comparison
    instructions
  target/riscv: rvv: Add tail agnostic for vector integer merge and move
    instructions
  target/riscv: rvv: Add tail agnostic for vector fix-point arithmetic
    instructions
  target/riscv: rvv: Add tail agnostic for vector floating-point
    instructions
  target/riscv: rvv: Add tail agnostic for vector reduction instructions
  target/riscv: rvv: Add tail agnostic for vector mask instructions
  target/riscv: rvv: Add tail agnostic for vector permutation
    instructions

 target/riscv/cpu.c                      |    1 +
 target/riscv/cpu.h                      |    2 +
 target/riscv/cpu_helper.c               |    2 +
 target/riscv/insn_trans/trans_rvv.c.inc |  154 +++
 target/riscv/internals.h                |    5 +-
 target/riscv/translate.c                |    2 +
 target/riscv/vector_helper.c            | 1597 ++++++++++++++---------
 7 files changed, 1140 insertions(+), 623 deletions(-)

-- 
2.34.1


^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 04/14] target/riscv: rvv: Add tail agnostic for vv instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-01  9:07 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

This is the first commit regarding the tail agnostic behavior.
Added option 'rvv_ta_all_1s' to enable the behavior, the option
is default to false.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/cpu.c                      |   1 +
 target/riscv/cpu.h                      |   2 +
 target/riscv/cpu_helper.c               |   2 +
 target/riscv/insn_trans/trans_rvv.c.inc |  10 +
 target/riscv/internals.h                |   5 +-
 target/riscv/translate.c                |   2 +
 target/riscv/vector_helper.c            | 315 ++++++++++++++----------
 7 files changed, 206 insertions(+), 131 deletions(-)

diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index ddda4906ff..cd4cf4b41e 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -810,6 +810,7 @@ static Property riscv_cpu_properties[] = {
     DEFINE_PROP_BOOL("x-aia", RISCVCPU, cfg.aia, false),
 
     DEFINE_PROP_UINT64("resetvec", RISCVCPU, cfg.resetvec, DEFAULT_RSTVEC),
+    DEFINE_PROP_BOOL("rvv_ta_all_1s", RISCVCPU, cfg.rvv_ta_all_1s, false),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index c069fe85fa..8c4a79b5a0 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -369,6 +369,7 @@ struct RISCVCPUConfig {
     bool ext_zhinxmin;
     bool ext_zve32f;
     bool ext_zve64f;
+    bool rvv_ta_all_1s;
 
     /* Vendor-specific custom extensions */
     bool ext_XVentanaCondOps;
@@ -516,6 +517,7 @@ FIELD(TB_FLAGS, XL, 20, 2)
 /* If PointerMasking should be applied */
 FIELD(TB_FLAGS, PM_MASK_ENABLED, 22, 1)
 FIELD(TB_FLAGS, PM_BASE_ENABLED, 23, 1)
+FIELD(TB_FLAGS, VTA, 24, 1)
 
 #ifdef TARGET_RISCV32
 #define riscv_cpu_mxl(env)  ((void)(env), MXL_RV32)
diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
index 1c60fb2e80..2941c88c31 100644
--- a/target/riscv/cpu_helper.c
+++ b/target/riscv/cpu_helper.c
@@ -65,6 +65,8 @@ void cpu_get_tb_cpu_state(CPURISCVState *env, target_ulong *pc,
         flags = FIELD_DP32(flags, TB_FLAGS, LMUL,
                     FIELD_EX64(env->vtype, VTYPE, VLMUL));
         flags = FIELD_DP32(flags, TB_FLAGS, VL_EQ_VLMAX, vl_eq_vlmax);
+        flags = FIELD_DP32(flags, TB_FLAGS, VTA,
+                    FIELD_EX64(env->vtype, VTYPE, VTA));
     } else {
         flags = FIELD_DP32(flags, TB_FLAGS, VILL, 1);
     }
diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 57953923d5..166ee9df2b 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1223,6 +1223,15 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
     tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         gvec_fn(s->sew, vreg_ofs(s, a->rd),
                 vreg_ofs(s, a->rs2), vreg_ofs(s, a->rs1),
                 MAXSZ(s), MAXSZ(s));
@@ -1231,6 +1240,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
                            cpu_env, s->cfg_ptr->vlen / 8,
diff --git a/target/riscv/internals.h b/target/riscv/internals.h
index dbb322bfa7..512c6c30cf 100644
--- a/target/riscv/internals.h
+++ b/target/riscv/internals.h
@@ -24,8 +24,9 @@
 /* share data between vector helpers and decode code */
 FIELD(VDATA, VM, 0, 1)
 FIELD(VDATA, LMUL, 1, 3)
-FIELD(VDATA, NF, 4, 4)
-FIELD(VDATA, WD, 4, 1)
+FIELD(VDATA, VTA, 4, 1)
+FIELD(VDATA, NF, 5, 4)
+FIELD(VDATA, WD, 5, 1)
 
 /* float point classify helpers */
 target_ulong fclass_h(uint64_t frs1);
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index fac998a6b5..7775dade26 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -94,6 +94,7 @@ typedef struct DisasContext {
      */
     int8_t lmul;
     uint8_t sew;
+    uint8_t vta;
     target_ulong vstart;
     bool vl_eq_vlmax;
     uint8_t ntemp;
@@ -1083,6 +1084,7 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
     ctx->vill = FIELD_EX32(tb_flags, TB_FLAGS, VILL);
     ctx->sew = FIELD_EX32(tb_flags, TB_FLAGS, SEW);
     ctx->lmul = sextract32(FIELD_EX32(tb_flags, TB_FLAGS, LMUL), 0, 3);
+    ctx->vta = FIELD_EX32(tb_flags, TB_FLAGS, VTA) && cpu->cfg.rvv_ta_all_1s;
     ctx->vstart = env->vstart;
     ctx->vl_eq_vlmax = FIELD_EX32(tb_flags, TB_FLAGS, VL_EQ_VLMAX);
     ctx->misa_mxl_max = env->misa_mxl_max;
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index d0452a7756..b9976f9d77 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -122,6 +122,11 @@ static inline int32_t vext_lmul(uint32_t desc)
     return sextract32(FIELD_EX32(simd_data(desc), VDATA, LMUL), 0, 3);
 }
 
+static inline uint32_t vext_vta(uint32_t desc)
+{
+    return FIELD_EX32(simd_data(desc), VDATA, VTA);
+}
+
 /*
  * Get the maximum number of elements can be operated.
  *
@@ -140,6 +145,20 @@ static inline uint32_t vext_max_elems(uint32_t desc, uint32_t log2_esz)
     return scale < 0 ? vlenb >> -scale : vlenb << scale;
 }
 
+/*
+ * Get number of total elements, including prestart, body and tail elements.
+ * Note that when LMUL < 1, the tail includes the elements past VLMAX that
+ * are held in the same vector register.
+ */
+static inline uint32_t vext_get_total_elems(RISCVCPU *cpu, target_ulong vtype)
+{
+    uint8_t sew = FIELD_EX64(vtype, VTYPE, VSEW);
+    int8_t lmul = sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3) < 0 ?
+                  0 : sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3);
+    return cpu->cfg.vlen >> (sew + 3 - lmul);
+}
+
+
 static inline target_ulong adjust_addr(CPURISCVState *env, target_ulong addr)
 {
     return (addr & env->cur_pmmask) | env->cur_pmbase;
@@ -172,6 +191,32 @@ static void probe_pages(CPURISCVState *env, target_ulong addr,
     }
 }
 
+static void vext_set_elems_1s(void *base, uint32_t is_agnostic, uint32_t cnt,
+                              uint32_t tot)
+{
+    if (is_agnostic == 0) {
+        /* policy undisturbed */
+        return;
+    }
+    if (tot - cnt == 0) {
+        return ;
+    }
+    memset(base, -1, tot - cnt);
+}
+
+/* Set agnostic elements to 1s */
+#define GEN_SET_ELEMS_1S(SET_ELEMS_1S_FN, ETYPE, H)                            \
+static void SET_ELEMS_1S_FN(void *vd, uint32_t is_agnostic, uint32_t idx,      \
+                            uint32_t cnt, uint32_t tot)                        \
+{                                                                              \
+    ETYPE *cur = ((ETYPE *)vd + H(idx));                                       \
+    vext_set_elems_1s(cur, is_agnostic, cnt, tot);                             \
+}
+GEN_SET_ELEMS_1S(vext_set_elems_1s_b, int8_t, H1)
+GEN_SET_ELEMS_1S(vext_set_elems_1s_h, int16_t, H2)
+GEN_SET_ELEMS_1S(vext_set_elems_1s_w, int32_t, H4)
+GEN_SET_ELEMS_1S(vext_set_elems_1s_d, int64_t, H8)
+
 static inline void vext_set_elem_mask(void *v0, int index,
                                       uint8_t value)
 {
@@ -197,6 +242,14 @@ static inline int vext_elem_mask(void *v0, int index)
 typedef void vext_ldst_elem_fn(CPURISCVState *env, target_ulong addr,
                                uint32_t idx, void *vd, uintptr_t retaddr);
 
+/* set bytes to all 1s for agnostic elements */
+typedef void vext_set_elems_1s_fn(void *vd, uint32_t vta, uint32_t idx,
+                                  uint32_t cnt, uint32_t tot);
+static vext_set_elems_1s_fn *vext_set_elems_1s_fns[4] = {
+    vext_set_elems_1s_b, vext_set_elems_1s_h,
+    vext_set_elems_1s_w, vext_set_elems_1s_d
+};
+
 #define GEN_VEXT_LD_ELEM(NAME, ETYPE, H, LDSUF)            \
 static void NAME(CPURISCVState *env, abi_ptr addr,         \
                  uint32_t idx, void *vd, uintptr_t retaddr)\
@@ -710,10 +763,12 @@ RVVCALL(OPIVV2, vsub_vv_d, OP_SSS_D, H8, H8, H8, DO_SUB)
 
 static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       opivv2_fn *fn)
+                       opivv2_fn *fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
 
     for (i = env->vstart; i < vl; i++) {
@@ -723,26 +778,28 @@ static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
         fn(vd, vs1, vs2, i);
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate the helpers for OPIVV */
-#define GEN_VEXT_VV(NAME)                                 \
+#define GEN_VEXT_VV(NAME, ESZ)                            \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     do_vext_vv(vd, v0, vs1, vs2, env, desc,               \
-               do_##NAME);                                \
+               do_##NAME, ESZ);                           \
 }
 
-GEN_VEXT_VV(vadd_vv_b)
-GEN_VEXT_VV(vadd_vv_h)
-GEN_VEXT_VV(vadd_vv_w)
-GEN_VEXT_VV(vadd_vv_d)
-GEN_VEXT_VV(vsub_vv_b)
-GEN_VEXT_VV(vsub_vv_h)
-GEN_VEXT_VV(vsub_vv_w)
-GEN_VEXT_VV(vsub_vv_d)
+GEN_VEXT_VV(vadd_vv_b, 1)
+GEN_VEXT_VV(vadd_vv_h, 2)
+GEN_VEXT_VV(vadd_vv_w, 4)
+GEN_VEXT_VV(vadd_vv_d, 8)
+GEN_VEXT_VV(vsub_vv_b, 1)
+GEN_VEXT_VV(vsub_vv_h, 2)
+GEN_VEXT_VV(vsub_vv_w, 4)
+GEN_VEXT_VV(vsub_vv_d, 8)
 
 typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
 
@@ -887,30 +944,30 @@ RVVCALL(OPIVV2, vwadd_wv_w, WOP_WSSS_W, H8, H4, H4, DO_ADD)
 RVVCALL(OPIVV2, vwsub_wv_b, WOP_WSSS_B, H2, H1, H1, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_h, WOP_WSSS_H, H4, H2, H2, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_w, WOP_WSSS_W, H8, H4, H4, DO_SUB)
-GEN_VEXT_VV(vwaddu_vv_b)
-GEN_VEXT_VV(vwaddu_vv_h)
-GEN_VEXT_VV(vwaddu_vv_w)
-GEN_VEXT_VV(vwsubu_vv_b)
-GEN_VEXT_VV(vwsubu_vv_h)
-GEN_VEXT_VV(vwsubu_vv_w)
-GEN_VEXT_VV(vwadd_vv_b)
-GEN_VEXT_VV(vwadd_vv_h)
-GEN_VEXT_VV(vwadd_vv_w)
-GEN_VEXT_VV(vwsub_vv_b)
-GEN_VEXT_VV(vwsub_vv_h)
-GEN_VEXT_VV(vwsub_vv_w)
-GEN_VEXT_VV(vwaddu_wv_b)
-GEN_VEXT_VV(vwaddu_wv_h)
-GEN_VEXT_VV(vwaddu_wv_w)
-GEN_VEXT_VV(vwsubu_wv_b)
-GEN_VEXT_VV(vwsubu_wv_h)
-GEN_VEXT_VV(vwsubu_wv_w)
-GEN_VEXT_VV(vwadd_wv_b)
-GEN_VEXT_VV(vwadd_wv_h)
-GEN_VEXT_VV(vwadd_wv_w)
-GEN_VEXT_VV(vwsub_wv_b)
-GEN_VEXT_VV(vwsub_wv_h)
-GEN_VEXT_VV(vwsub_wv_w)
+GEN_VEXT_VV(vwaddu_vv_b, 2)
+GEN_VEXT_VV(vwaddu_vv_h, 4)
+GEN_VEXT_VV(vwaddu_vv_w, 8)
+GEN_VEXT_VV(vwsubu_vv_b, 2)
+GEN_VEXT_VV(vwsubu_vv_h, 4)
+GEN_VEXT_VV(vwsubu_vv_w, 8)
+GEN_VEXT_VV(vwadd_vv_b, 2)
+GEN_VEXT_VV(vwadd_vv_h, 4)
+GEN_VEXT_VV(vwadd_vv_w, 8)
+GEN_VEXT_VV(vwsub_vv_b, 2)
+GEN_VEXT_VV(vwsub_vv_h, 4)
+GEN_VEXT_VV(vwsub_vv_w, 8)
+GEN_VEXT_VV(vwaddu_wv_b, 2)
+GEN_VEXT_VV(vwaddu_wv_h, 4)
+GEN_VEXT_VV(vwaddu_wv_w, 8)
+GEN_VEXT_VV(vwsubu_wv_b, 2)
+GEN_VEXT_VV(vwsubu_wv_h, 4)
+GEN_VEXT_VV(vwsubu_wv_w, 8)
+GEN_VEXT_VV(vwadd_wv_b, 2)
+GEN_VEXT_VV(vwadd_wv_h, 4)
+GEN_VEXT_VV(vwadd_wv_w, 8)
+GEN_VEXT_VV(vwsub_wv_b, 2)
+GEN_VEXT_VV(vwsub_wv_h, 4)
+GEN_VEXT_VV(vwsub_wv_w, 8)
 
 RVVCALL(OPIVX2, vwaddu_vx_b, WOP_UUU_B, H2, H1, DO_ADD)
 RVVCALL(OPIVX2, vwaddu_vx_h, WOP_UUU_H, H4, H2, DO_ADD)
@@ -1089,18 +1146,18 @@ RVVCALL(OPIVV2, vxor_vv_b, OP_SSS_B, H1, H1, H1, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_h, OP_SSS_H, H2, H2, H2, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_w, OP_SSS_W, H4, H4, H4, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_d, OP_SSS_D, H8, H8, H8, DO_XOR)
-GEN_VEXT_VV(vand_vv_b)
-GEN_VEXT_VV(vand_vv_h)
-GEN_VEXT_VV(vand_vv_w)
-GEN_VEXT_VV(vand_vv_d)
-GEN_VEXT_VV(vor_vv_b)
-GEN_VEXT_VV(vor_vv_h)
-GEN_VEXT_VV(vor_vv_w)
-GEN_VEXT_VV(vor_vv_d)
-GEN_VEXT_VV(vxor_vv_b)
-GEN_VEXT_VV(vxor_vv_h)
-GEN_VEXT_VV(vxor_vv_w)
-GEN_VEXT_VV(vxor_vv_d)
+GEN_VEXT_VV(vand_vv_b, 1)
+GEN_VEXT_VV(vand_vv_h, 2)
+GEN_VEXT_VV(vand_vv_w, 4)
+GEN_VEXT_VV(vand_vv_d, 8)
+GEN_VEXT_VV(vor_vv_b, 1)
+GEN_VEXT_VV(vor_vv_h, 2)
+GEN_VEXT_VV(vor_vv_w, 4)
+GEN_VEXT_VV(vor_vv_d, 8)
+GEN_VEXT_VV(vxor_vv_b, 1)
+GEN_VEXT_VV(vxor_vv_h, 2)
+GEN_VEXT_VV(vxor_vv_w, 4)
+GEN_VEXT_VV(vxor_vv_d, 8)
 
 RVVCALL(OPIVX2, vand_vx_b, OP_SSS_B, H1, H1, DO_AND)
 RVVCALL(OPIVX2, vand_vx_h, OP_SSS_H, H2, H2, DO_AND)
@@ -1346,22 +1403,22 @@ RVVCALL(OPIVV2, vmax_vv_b, OP_SSS_B, H1, H1, H1, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_h, OP_SSS_H, H2, H2, H2, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_w, OP_SSS_W, H4, H4, H4, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_d, OP_SSS_D, H8, H8, H8, DO_MAX)
-GEN_VEXT_VV(vminu_vv_b)
-GEN_VEXT_VV(vminu_vv_h)
-GEN_VEXT_VV(vminu_vv_w)
-GEN_VEXT_VV(vminu_vv_d)
-GEN_VEXT_VV(vmin_vv_b)
-GEN_VEXT_VV(vmin_vv_h)
-GEN_VEXT_VV(vmin_vv_w)
-GEN_VEXT_VV(vmin_vv_d)
-GEN_VEXT_VV(vmaxu_vv_b)
-GEN_VEXT_VV(vmaxu_vv_h)
-GEN_VEXT_VV(vmaxu_vv_w)
-GEN_VEXT_VV(vmaxu_vv_d)
-GEN_VEXT_VV(vmax_vv_b)
-GEN_VEXT_VV(vmax_vv_h)
-GEN_VEXT_VV(vmax_vv_w)
-GEN_VEXT_VV(vmax_vv_d)
+GEN_VEXT_VV(vminu_vv_b, 1)
+GEN_VEXT_VV(vminu_vv_h, 2)
+GEN_VEXT_VV(vminu_vv_w, 4)
+GEN_VEXT_VV(vminu_vv_d, 8)
+GEN_VEXT_VV(vmin_vv_b, 1)
+GEN_VEXT_VV(vmin_vv_h, 2)
+GEN_VEXT_VV(vmin_vv_w, 4)
+GEN_VEXT_VV(vmin_vv_d, 8)
+GEN_VEXT_VV(vmaxu_vv_b, 1)
+GEN_VEXT_VV(vmaxu_vv_h, 2)
+GEN_VEXT_VV(vmaxu_vv_w, 4)
+GEN_VEXT_VV(vmaxu_vv_d, 8)
+GEN_VEXT_VV(vmax_vv_b, 1)
+GEN_VEXT_VV(vmax_vv_h, 2)
+GEN_VEXT_VV(vmax_vv_w, 4)
+GEN_VEXT_VV(vmax_vv_d, 8)
 
 RVVCALL(OPIVX2, vminu_vx_b, OP_UUU_B, H1, H1, DO_MIN)
 RVVCALL(OPIVX2, vminu_vx_h, OP_UUU_H, H2, H2, DO_MIN)
@@ -1402,10 +1459,10 @@ RVVCALL(OPIVV2, vmul_vv_b, OP_SSS_B, H1, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_h, OP_SSS_H, H2, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_w, OP_SSS_W, H4, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_d, OP_SSS_D, H8, H8, H8, DO_MUL)
-GEN_VEXT_VV(vmul_vv_b)
-GEN_VEXT_VV(vmul_vv_h)
-GEN_VEXT_VV(vmul_vv_w)
-GEN_VEXT_VV(vmul_vv_d)
+GEN_VEXT_VV(vmul_vv_b, 1)
+GEN_VEXT_VV(vmul_vv_h, 2)
+GEN_VEXT_VV(vmul_vv_w, 4)
+GEN_VEXT_VV(vmul_vv_d, 8)
 
 static int8_t do_mulh_b(int8_t s2, int8_t s1)
 {
@@ -1509,18 +1566,18 @@ RVVCALL(OPIVV2, vmulhsu_vv_b, OP_SUS_B, H1, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVV2, vmulhsu_vv_h, OP_SUS_H, H2, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVV2, vmulhsu_vv_w, OP_SUS_W, H4, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVV2, vmulhsu_vv_d, OP_SUS_D, H8, H8, H8, do_mulhsu_d)
-GEN_VEXT_VV(vmulh_vv_b)
-GEN_VEXT_VV(vmulh_vv_h)
-GEN_VEXT_VV(vmulh_vv_w)
-GEN_VEXT_VV(vmulh_vv_d)
-GEN_VEXT_VV(vmulhu_vv_b)
-GEN_VEXT_VV(vmulhu_vv_h)
-GEN_VEXT_VV(vmulhu_vv_w)
-GEN_VEXT_VV(vmulhu_vv_d)
-GEN_VEXT_VV(vmulhsu_vv_b)
-GEN_VEXT_VV(vmulhsu_vv_h)
-GEN_VEXT_VV(vmulhsu_vv_w)
-GEN_VEXT_VV(vmulhsu_vv_d)
+GEN_VEXT_VV(vmulh_vv_b, 1)
+GEN_VEXT_VV(vmulh_vv_h, 2)
+GEN_VEXT_VV(vmulh_vv_w, 4)
+GEN_VEXT_VV(vmulh_vv_d, 8)
+GEN_VEXT_VV(vmulhu_vv_b, 1)
+GEN_VEXT_VV(vmulhu_vv_h, 2)
+GEN_VEXT_VV(vmulhu_vv_w, 4)
+GEN_VEXT_VV(vmulhu_vv_d, 8)
+GEN_VEXT_VV(vmulhsu_vv_b, 1)
+GEN_VEXT_VV(vmulhsu_vv_h, 2)
+GEN_VEXT_VV(vmulhsu_vv_w, 4)
+GEN_VEXT_VV(vmulhsu_vv_d, 8)
 
 RVVCALL(OPIVX2, vmul_vx_b, OP_SSS_B, H1, H1, DO_MUL)
 RVVCALL(OPIVX2, vmul_vx_h, OP_SSS_H, H2, H2, DO_MUL)
@@ -1579,22 +1636,22 @@ RVVCALL(OPIVV2, vrem_vv_b, OP_SSS_B, H1, H1, H1, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_h, OP_SSS_H, H2, H2, H2, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_w, OP_SSS_W, H4, H4, H4, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_d, OP_SSS_D, H8, H8, H8, DO_REM)
-GEN_VEXT_VV(vdivu_vv_b)
-GEN_VEXT_VV(vdivu_vv_h)
-GEN_VEXT_VV(vdivu_vv_w)
-GEN_VEXT_VV(vdivu_vv_d)
-GEN_VEXT_VV(vdiv_vv_b)
-GEN_VEXT_VV(vdiv_vv_h)
-GEN_VEXT_VV(vdiv_vv_w)
-GEN_VEXT_VV(vdiv_vv_d)
-GEN_VEXT_VV(vremu_vv_b)
-GEN_VEXT_VV(vremu_vv_h)
-GEN_VEXT_VV(vremu_vv_w)
-GEN_VEXT_VV(vremu_vv_d)
-GEN_VEXT_VV(vrem_vv_b)
-GEN_VEXT_VV(vrem_vv_h)
-GEN_VEXT_VV(vrem_vv_w)
-GEN_VEXT_VV(vrem_vv_d)
+GEN_VEXT_VV(vdivu_vv_b, 1)
+GEN_VEXT_VV(vdivu_vv_h, 2)
+GEN_VEXT_VV(vdivu_vv_w, 4)
+GEN_VEXT_VV(vdivu_vv_d, 8)
+GEN_VEXT_VV(vdiv_vv_b, 1)
+GEN_VEXT_VV(vdiv_vv_h, 2)
+GEN_VEXT_VV(vdiv_vv_w, 4)
+GEN_VEXT_VV(vdiv_vv_d, 8)
+GEN_VEXT_VV(vremu_vv_b, 1)
+GEN_VEXT_VV(vremu_vv_h, 2)
+GEN_VEXT_VV(vremu_vv_w, 4)
+GEN_VEXT_VV(vremu_vv_d, 8)
+GEN_VEXT_VV(vrem_vv_b, 1)
+GEN_VEXT_VV(vrem_vv_h, 2)
+GEN_VEXT_VV(vrem_vv_w, 4)
+GEN_VEXT_VV(vrem_vv_d, 8)
 
 RVVCALL(OPIVX2, vdivu_vx_b, OP_UUU_B, H1, H1, DO_DIVU)
 RVVCALL(OPIVX2, vdivu_vx_h, OP_UUU_H, H2, H2, DO_DIVU)
@@ -1639,15 +1696,15 @@ RVVCALL(OPIVV2, vwmulu_vv_w, WOP_UUU_W, H8, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_b, WOP_SUS_B, H2, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_h, WOP_SUS_H, H4, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_w, WOP_SUS_W, H8, H4, H4, DO_MUL)
-GEN_VEXT_VV(vwmul_vv_b)
-GEN_VEXT_VV(vwmul_vv_h)
-GEN_VEXT_VV(vwmul_vv_w)
-GEN_VEXT_VV(vwmulu_vv_b)
-GEN_VEXT_VV(vwmulu_vv_h)
-GEN_VEXT_VV(vwmulu_vv_w)
-GEN_VEXT_VV(vwmulsu_vv_b)
-GEN_VEXT_VV(vwmulsu_vv_h)
-GEN_VEXT_VV(vwmulsu_vv_w)
+GEN_VEXT_VV(vwmul_vv_b, 2)
+GEN_VEXT_VV(vwmul_vv_h, 4)
+GEN_VEXT_VV(vwmul_vv_w, 8)
+GEN_VEXT_VV(vwmulu_vv_b, 2)
+GEN_VEXT_VV(vwmulu_vv_h, 4)
+GEN_VEXT_VV(vwmulu_vv_w, 8)
+GEN_VEXT_VV(vwmulsu_vv_b, 2)
+GEN_VEXT_VV(vwmulsu_vv_h, 4)
+GEN_VEXT_VV(vwmulsu_vv_w, 8)
 
 RVVCALL(OPIVX2, vwmul_vx_b, WOP_SSS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmul_vx_h, WOP_SSS_H, H4, H2, DO_MUL)
@@ -1698,22 +1755,22 @@ RVVCALL(OPIVV3, vnmsub_vv_b, OP_SSS_B, H1, H1, H1, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_h, OP_SSS_H, H2, H2, H2, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_w, OP_SSS_W, H4, H4, H4, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_d, OP_SSS_D, H8, H8, H8, DO_NMSUB)
-GEN_VEXT_VV(vmacc_vv_b)
-GEN_VEXT_VV(vmacc_vv_h)
-GEN_VEXT_VV(vmacc_vv_w)
-GEN_VEXT_VV(vmacc_vv_d)
-GEN_VEXT_VV(vnmsac_vv_b)
-GEN_VEXT_VV(vnmsac_vv_h)
-GEN_VEXT_VV(vnmsac_vv_w)
-GEN_VEXT_VV(vnmsac_vv_d)
-GEN_VEXT_VV(vmadd_vv_b)
-GEN_VEXT_VV(vmadd_vv_h)
-GEN_VEXT_VV(vmadd_vv_w)
-GEN_VEXT_VV(vmadd_vv_d)
-GEN_VEXT_VV(vnmsub_vv_b)
-GEN_VEXT_VV(vnmsub_vv_h)
-GEN_VEXT_VV(vnmsub_vv_w)
-GEN_VEXT_VV(vnmsub_vv_d)
+GEN_VEXT_VV(vmacc_vv_b, 1)
+GEN_VEXT_VV(vmacc_vv_h, 2)
+GEN_VEXT_VV(vmacc_vv_w, 4)
+GEN_VEXT_VV(vmacc_vv_d, 8)
+GEN_VEXT_VV(vnmsac_vv_b, 1)
+GEN_VEXT_VV(vnmsac_vv_h, 2)
+GEN_VEXT_VV(vnmsac_vv_w, 4)
+GEN_VEXT_VV(vnmsac_vv_d, 8)
+GEN_VEXT_VV(vmadd_vv_b, 1)
+GEN_VEXT_VV(vmadd_vv_h, 2)
+GEN_VEXT_VV(vmadd_vv_w, 4)
+GEN_VEXT_VV(vmadd_vv_d, 8)
+GEN_VEXT_VV(vnmsub_vv_b, 1)
+GEN_VEXT_VV(vnmsub_vv_h, 2)
+GEN_VEXT_VV(vnmsub_vv_w, 4)
+GEN_VEXT_VV(vnmsub_vv_d, 8)
 
 #define OPIVX3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)             \
 static void do_##NAME(void *vd, target_long s1, void *vs2, int i)   \
@@ -1766,15 +1823,15 @@ RVVCALL(OPIVV3, vwmacc_vv_w, WOP_SSS_W, H8, H4, H4, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_b, WOP_SSU_B, H2, H1, H1, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_h, WOP_SSU_H, H4, H2, H2, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_w, WOP_SSU_W, H8, H4, H4, DO_MACC)
-GEN_VEXT_VV(vwmaccu_vv_b)
-GEN_VEXT_VV(vwmaccu_vv_h)
-GEN_VEXT_VV(vwmaccu_vv_w)
-GEN_VEXT_VV(vwmacc_vv_b)
-GEN_VEXT_VV(vwmacc_vv_h)
-GEN_VEXT_VV(vwmacc_vv_w)
-GEN_VEXT_VV(vwmaccsu_vv_b)
-GEN_VEXT_VV(vwmaccsu_vv_h)
-GEN_VEXT_VV(vwmaccsu_vv_w)
+GEN_VEXT_VV(vwmaccu_vv_b, 2)
+GEN_VEXT_VV(vwmaccu_vv_h, 4)
+GEN_VEXT_VV(vwmaccu_vv_w, 8)
+GEN_VEXT_VV(vwmacc_vv_b, 2)
+GEN_VEXT_VV(vwmacc_vv_h, 4)
+GEN_VEXT_VV(vwmacc_vv_w, 8)
+GEN_VEXT_VV(vwmaccsu_vv_b, 2)
+GEN_VEXT_VV(vwmaccsu_vv_h, 4)
+GEN_VEXT_VV(vwmaccsu_vv_w, 8)
 
 RVVCALL(OPIVX3, vwmaccu_vx_b, WOP_UUU_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccu_vx_h, WOP_UUU_H, H4, H2, DO_MACC)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 04/14] target/riscv: rvv: Add tail agnostic for vv instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

This is the first commit regarding the tail agnostic behavior.
Added option 'rvv_ta_all_1s' to enable the behavior, the option
is default to false.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/cpu.c                      |   1 +
 target/riscv/cpu.h                      |   2 +
 target/riscv/cpu_helper.c               |   2 +
 target/riscv/insn_trans/trans_rvv.c.inc |  10 +
 target/riscv/internals.h                |   5 +-
 target/riscv/translate.c                |   2 +
 target/riscv/vector_helper.c            | 315 ++++++++++++++----------
 7 files changed, 206 insertions(+), 131 deletions(-)

diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index ddda4906ff..cd4cf4b41e 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -810,6 +810,7 @@ static Property riscv_cpu_properties[] = {
     DEFINE_PROP_BOOL("x-aia", RISCVCPU, cfg.aia, false),
 
     DEFINE_PROP_UINT64("resetvec", RISCVCPU, cfg.resetvec, DEFAULT_RSTVEC),
+    DEFINE_PROP_BOOL("rvv_ta_all_1s", RISCVCPU, cfg.rvv_ta_all_1s, false),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index c069fe85fa..8c4a79b5a0 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -369,6 +369,7 @@ struct RISCVCPUConfig {
     bool ext_zhinxmin;
     bool ext_zve32f;
     bool ext_zve64f;
+    bool rvv_ta_all_1s;
 
     /* Vendor-specific custom extensions */
     bool ext_XVentanaCondOps;
@@ -516,6 +517,7 @@ FIELD(TB_FLAGS, XL, 20, 2)
 /* If PointerMasking should be applied */
 FIELD(TB_FLAGS, PM_MASK_ENABLED, 22, 1)
 FIELD(TB_FLAGS, PM_BASE_ENABLED, 23, 1)
+FIELD(TB_FLAGS, VTA, 24, 1)
 
 #ifdef TARGET_RISCV32
 #define riscv_cpu_mxl(env)  ((void)(env), MXL_RV32)
diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
index 1c60fb2e80..2941c88c31 100644
--- a/target/riscv/cpu_helper.c
+++ b/target/riscv/cpu_helper.c
@@ -65,6 +65,8 @@ void cpu_get_tb_cpu_state(CPURISCVState *env, target_ulong *pc,
         flags = FIELD_DP32(flags, TB_FLAGS, LMUL,
                     FIELD_EX64(env->vtype, VTYPE, VLMUL));
         flags = FIELD_DP32(flags, TB_FLAGS, VL_EQ_VLMAX, vl_eq_vlmax);
+        flags = FIELD_DP32(flags, TB_FLAGS, VTA,
+                    FIELD_EX64(env->vtype, VTYPE, VTA));
     } else {
         flags = FIELD_DP32(flags, TB_FLAGS, VILL, 1);
     }
diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 57953923d5..166ee9df2b 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1223,6 +1223,15 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
     tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         gvec_fn(s->sew, vreg_ofs(s, a->rd),
                 vreg_ofs(s, a->rs2), vreg_ofs(s, a->rs1),
                 MAXSZ(s), MAXSZ(s));
@@ -1231,6 +1240,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
                            cpu_env, s->cfg_ptr->vlen / 8,
diff --git a/target/riscv/internals.h b/target/riscv/internals.h
index dbb322bfa7..512c6c30cf 100644
--- a/target/riscv/internals.h
+++ b/target/riscv/internals.h
@@ -24,8 +24,9 @@
 /* share data between vector helpers and decode code */
 FIELD(VDATA, VM, 0, 1)
 FIELD(VDATA, LMUL, 1, 3)
-FIELD(VDATA, NF, 4, 4)
-FIELD(VDATA, WD, 4, 1)
+FIELD(VDATA, VTA, 4, 1)
+FIELD(VDATA, NF, 5, 4)
+FIELD(VDATA, WD, 5, 1)
 
 /* float point classify helpers */
 target_ulong fclass_h(uint64_t frs1);
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index fac998a6b5..7775dade26 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -94,6 +94,7 @@ typedef struct DisasContext {
      */
     int8_t lmul;
     uint8_t sew;
+    uint8_t vta;
     target_ulong vstart;
     bool vl_eq_vlmax;
     uint8_t ntemp;
@@ -1083,6 +1084,7 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
     ctx->vill = FIELD_EX32(tb_flags, TB_FLAGS, VILL);
     ctx->sew = FIELD_EX32(tb_flags, TB_FLAGS, SEW);
     ctx->lmul = sextract32(FIELD_EX32(tb_flags, TB_FLAGS, LMUL), 0, 3);
+    ctx->vta = FIELD_EX32(tb_flags, TB_FLAGS, VTA) && cpu->cfg.rvv_ta_all_1s;
     ctx->vstart = env->vstart;
     ctx->vl_eq_vlmax = FIELD_EX32(tb_flags, TB_FLAGS, VL_EQ_VLMAX);
     ctx->misa_mxl_max = env->misa_mxl_max;
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index d0452a7756..b9976f9d77 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -122,6 +122,11 @@ static inline int32_t vext_lmul(uint32_t desc)
     return sextract32(FIELD_EX32(simd_data(desc), VDATA, LMUL), 0, 3);
 }
 
+static inline uint32_t vext_vta(uint32_t desc)
+{
+    return FIELD_EX32(simd_data(desc), VDATA, VTA);
+}
+
 /*
  * Get the maximum number of elements can be operated.
  *
@@ -140,6 +145,20 @@ static inline uint32_t vext_max_elems(uint32_t desc, uint32_t log2_esz)
     return scale < 0 ? vlenb >> -scale : vlenb << scale;
 }
 
+/*
+ * Get number of total elements, including prestart, body and tail elements.
+ * Note that when LMUL < 1, the tail includes the elements past VLMAX that
+ * are held in the same vector register.
+ */
+static inline uint32_t vext_get_total_elems(RISCVCPU *cpu, target_ulong vtype)
+{
+    uint8_t sew = FIELD_EX64(vtype, VTYPE, VSEW);
+    int8_t lmul = sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3) < 0 ?
+                  0 : sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3);
+    return cpu->cfg.vlen >> (sew + 3 - lmul);
+}
+
+
 static inline target_ulong adjust_addr(CPURISCVState *env, target_ulong addr)
 {
     return (addr & env->cur_pmmask) | env->cur_pmbase;
@@ -172,6 +191,32 @@ static void probe_pages(CPURISCVState *env, target_ulong addr,
     }
 }
 
+static void vext_set_elems_1s(void *base, uint32_t is_agnostic, uint32_t cnt,
+                              uint32_t tot)
+{
+    if (is_agnostic == 0) {
+        /* policy undisturbed */
+        return;
+    }
+    if (tot - cnt == 0) {
+        return ;
+    }
+    memset(base, -1, tot - cnt);
+}
+
+/* Set agnostic elements to 1s */
+#define GEN_SET_ELEMS_1S(SET_ELEMS_1S_FN, ETYPE, H)                            \
+static void SET_ELEMS_1S_FN(void *vd, uint32_t is_agnostic, uint32_t idx,      \
+                            uint32_t cnt, uint32_t tot)                        \
+{                                                                              \
+    ETYPE *cur = ((ETYPE *)vd + H(idx));                                       \
+    vext_set_elems_1s(cur, is_agnostic, cnt, tot);                             \
+}
+GEN_SET_ELEMS_1S(vext_set_elems_1s_b, int8_t, H1)
+GEN_SET_ELEMS_1S(vext_set_elems_1s_h, int16_t, H2)
+GEN_SET_ELEMS_1S(vext_set_elems_1s_w, int32_t, H4)
+GEN_SET_ELEMS_1S(vext_set_elems_1s_d, int64_t, H8)
+
 static inline void vext_set_elem_mask(void *v0, int index,
                                       uint8_t value)
 {
@@ -197,6 +242,14 @@ static inline int vext_elem_mask(void *v0, int index)
 typedef void vext_ldst_elem_fn(CPURISCVState *env, target_ulong addr,
                                uint32_t idx, void *vd, uintptr_t retaddr);
 
+/* set bytes to all 1s for agnostic elements */
+typedef void vext_set_elems_1s_fn(void *vd, uint32_t vta, uint32_t idx,
+                                  uint32_t cnt, uint32_t tot);
+static vext_set_elems_1s_fn *vext_set_elems_1s_fns[4] = {
+    vext_set_elems_1s_b, vext_set_elems_1s_h,
+    vext_set_elems_1s_w, vext_set_elems_1s_d
+};
+
 #define GEN_VEXT_LD_ELEM(NAME, ETYPE, H, LDSUF)            \
 static void NAME(CPURISCVState *env, abi_ptr addr,         \
                  uint32_t idx, void *vd, uintptr_t retaddr)\
@@ -710,10 +763,12 @@ RVVCALL(OPIVV2, vsub_vv_d, OP_SSS_D, H8, H8, H8, DO_SUB)
 
 static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       opivv2_fn *fn)
+                       opivv2_fn *fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
 
     for (i = env->vstart; i < vl; i++) {
@@ -723,26 +778,28 @@ static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
         fn(vd, vs1, vs2, i);
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate the helpers for OPIVV */
-#define GEN_VEXT_VV(NAME)                                 \
+#define GEN_VEXT_VV(NAME, ESZ)                            \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     do_vext_vv(vd, v0, vs1, vs2, env, desc,               \
-               do_##NAME);                                \
+               do_##NAME, ESZ);                           \
 }
 
-GEN_VEXT_VV(vadd_vv_b)
-GEN_VEXT_VV(vadd_vv_h)
-GEN_VEXT_VV(vadd_vv_w)
-GEN_VEXT_VV(vadd_vv_d)
-GEN_VEXT_VV(vsub_vv_b)
-GEN_VEXT_VV(vsub_vv_h)
-GEN_VEXT_VV(vsub_vv_w)
-GEN_VEXT_VV(vsub_vv_d)
+GEN_VEXT_VV(vadd_vv_b, 1)
+GEN_VEXT_VV(vadd_vv_h, 2)
+GEN_VEXT_VV(vadd_vv_w, 4)
+GEN_VEXT_VV(vadd_vv_d, 8)
+GEN_VEXT_VV(vsub_vv_b, 1)
+GEN_VEXT_VV(vsub_vv_h, 2)
+GEN_VEXT_VV(vsub_vv_w, 4)
+GEN_VEXT_VV(vsub_vv_d, 8)
 
 typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
 
@@ -887,30 +944,30 @@ RVVCALL(OPIVV2, vwadd_wv_w, WOP_WSSS_W, H8, H4, H4, DO_ADD)
 RVVCALL(OPIVV2, vwsub_wv_b, WOP_WSSS_B, H2, H1, H1, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_h, WOP_WSSS_H, H4, H2, H2, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_w, WOP_WSSS_W, H8, H4, H4, DO_SUB)
-GEN_VEXT_VV(vwaddu_vv_b)
-GEN_VEXT_VV(vwaddu_vv_h)
-GEN_VEXT_VV(vwaddu_vv_w)
-GEN_VEXT_VV(vwsubu_vv_b)
-GEN_VEXT_VV(vwsubu_vv_h)
-GEN_VEXT_VV(vwsubu_vv_w)
-GEN_VEXT_VV(vwadd_vv_b)
-GEN_VEXT_VV(vwadd_vv_h)
-GEN_VEXT_VV(vwadd_vv_w)
-GEN_VEXT_VV(vwsub_vv_b)
-GEN_VEXT_VV(vwsub_vv_h)
-GEN_VEXT_VV(vwsub_vv_w)
-GEN_VEXT_VV(vwaddu_wv_b)
-GEN_VEXT_VV(vwaddu_wv_h)
-GEN_VEXT_VV(vwaddu_wv_w)
-GEN_VEXT_VV(vwsubu_wv_b)
-GEN_VEXT_VV(vwsubu_wv_h)
-GEN_VEXT_VV(vwsubu_wv_w)
-GEN_VEXT_VV(vwadd_wv_b)
-GEN_VEXT_VV(vwadd_wv_h)
-GEN_VEXT_VV(vwadd_wv_w)
-GEN_VEXT_VV(vwsub_wv_b)
-GEN_VEXT_VV(vwsub_wv_h)
-GEN_VEXT_VV(vwsub_wv_w)
+GEN_VEXT_VV(vwaddu_vv_b, 2)
+GEN_VEXT_VV(vwaddu_vv_h, 4)
+GEN_VEXT_VV(vwaddu_vv_w, 8)
+GEN_VEXT_VV(vwsubu_vv_b, 2)
+GEN_VEXT_VV(vwsubu_vv_h, 4)
+GEN_VEXT_VV(vwsubu_vv_w, 8)
+GEN_VEXT_VV(vwadd_vv_b, 2)
+GEN_VEXT_VV(vwadd_vv_h, 4)
+GEN_VEXT_VV(vwadd_vv_w, 8)
+GEN_VEXT_VV(vwsub_vv_b, 2)
+GEN_VEXT_VV(vwsub_vv_h, 4)
+GEN_VEXT_VV(vwsub_vv_w, 8)
+GEN_VEXT_VV(vwaddu_wv_b, 2)
+GEN_VEXT_VV(vwaddu_wv_h, 4)
+GEN_VEXT_VV(vwaddu_wv_w, 8)
+GEN_VEXT_VV(vwsubu_wv_b, 2)
+GEN_VEXT_VV(vwsubu_wv_h, 4)
+GEN_VEXT_VV(vwsubu_wv_w, 8)
+GEN_VEXT_VV(vwadd_wv_b, 2)
+GEN_VEXT_VV(vwadd_wv_h, 4)
+GEN_VEXT_VV(vwadd_wv_w, 8)
+GEN_VEXT_VV(vwsub_wv_b, 2)
+GEN_VEXT_VV(vwsub_wv_h, 4)
+GEN_VEXT_VV(vwsub_wv_w, 8)
 
 RVVCALL(OPIVX2, vwaddu_vx_b, WOP_UUU_B, H2, H1, DO_ADD)
 RVVCALL(OPIVX2, vwaddu_vx_h, WOP_UUU_H, H4, H2, DO_ADD)
@@ -1089,18 +1146,18 @@ RVVCALL(OPIVV2, vxor_vv_b, OP_SSS_B, H1, H1, H1, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_h, OP_SSS_H, H2, H2, H2, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_w, OP_SSS_W, H4, H4, H4, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_d, OP_SSS_D, H8, H8, H8, DO_XOR)
-GEN_VEXT_VV(vand_vv_b)
-GEN_VEXT_VV(vand_vv_h)
-GEN_VEXT_VV(vand_vv_w)
-GEN_VEXT_VV(vand_vv_d)
-GEN_VEXT_VV(vor_vv_b)
-GEN_VEXT_VV(vor_vv_h)
-GEN_VEXT_VV(vor_vv_w)
-GEN_VEXT_VV(vor_vv_d)
-GEN_VEXT_VV(vxor_vv_b)
-GEN_VEXT_VV(vxor_vv_h)
-GEN_VEXT_VV(vxor_vv_w)
-GEN_VEXT_VV(vxor_vv_d)
+GEN_VEXT_VV(vand_vv_b, 1)
+GEN_VEXT_VV(vand_vv_h, 2)
+GEN_VEXT_VV(vand_vv_w, 4)
+GEN_VEXT_VV(vand_vv_d, 8)
+GEN_VEXT_VV(vor_vv_b, 1)
+GEN_VEXT_VV(vor_vv_h, 2)
+GEN_VEXT_VV(vor_vv_w, 4)
+GEN_VEXT_VV(vor_vv_d, 8)
+GEN_VEXT_VV(vxor_vv_b, 1)
+GEN_VEXT_VV(vxor_vv_h, 2)
+GEN_VEXT_VV(vxor_vv_w, 4)
+GEN_VEXT_VV(vxor_vv_d, 8)
 
 RVVCALL(OPIVX2, vand_vx_b, OP_SSS_B, H1, H1, DO_AND)
 RVVCALL(OPIVX2, vand_vx_h, OP_SSS_H, H2, H2, DO_AND)
@@ -1346,22 +1403,22 @@ RVVCALL(OPIVV2, vmax_vv_b, OP_SSS_B, H1, H1, H1, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_h, OP_SSS_H, H2, H2, H2, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_w, OP_SSS_W, H4, H4, H4, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_d, OP_SSS_D, H8, H8, H8, DO_MAX)
-GEN_VEXT_VV(vminu_vv_b)
-GEN_VEXT_VV(vminu_vv_h)
-GEN_VEXT_VV(vminu_vv_w)
-GEN_VEXT_VV(vminu_vv_d)
-GEN_VEXT_VV(vmin_vv_b)
-GEN_VEXT_VV(vmin_vv_h)
-GEN_VEXT_VV(vmin_vv_w)
-GEN_VEXT_VV(vmin_vv_d)
-GEN_VEXT_VV(vmaxu_vv_b)
-GEN_VEXT_VV(vmaxu_vv_h)
-GEN_VEXT_VV(vmaxu_vv_w)
-GEN_VEXT_VV(vmaxu_vv_d)
-GEN_VEXT_VV(vmax_vv_b)
-GEN_VEXT_VV(vmax_vv_h)
-GEN_VEXT_VV(vmax_vv_w)
-GEN_VEXT_VV(vmax_vv_d)
+GEN_VEXT_VV(vminu_vv_b, 1)
+GEN_VEXT_VV(vminu_vv_h, 2)
+GEN_VEXT_VV(vminu_vv_w, 4)
+GEN_VEXT_VV(vminu_vv_d, 8)
+GEN_VEXT_VV(vmin_vv_b, 1)
+GEN_VEXT_VV(vmin_vv_h, 2)
+GEN_VEXT_VV(vmin_vv_w, 4)
+GEN_VEXT_VV(vmin_vv_d, 8)
+GEN_VEXT_VV(vmaxu_vv_b, 1)
+GEN_VEXT_VV(vmaxu_vv_h, 2)
+GEN_VEXT_VV(vmaxu_vv_w, 4)
+GEN_VEXT_VV(vmaxu_vv_d, 8)
+GEN_VEXT_VV(vmax_vv_b, 1)
+GEN_VEXT_VV(vmax_vv_h, 2)
+GEN_VEXT_VV(vmax_vv_w, 4)
+GEN_VEXT_VV(vmax_vv_d, 8)
 
 RVVCALL(OPIVX2, vminu_vx_b, OP_UUU_B, H1, H1, DO_MIN)
 RVVCALL(OPIVX2, vminu_vx_h, OP_UUU_H, H2, H2, DO_MIN)
@@ -1402,10 +1459,10 @@ RVVCALL(OPIVV2, vmul_vv_b, OP_SSS_B, H1, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_h, OP_SSS_H, H2, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_w, OP_SSS_W, H4, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_d, OP_SSS_D, H8, H8, H8, DO_MUL)
-GEN_VEXT_VV(vmul_vv_b)
-GEN_VEXT_VV(vmul_vv_h)
-GEN_VEXT_VV(vmul_vv_w)
-GEN_VEXT_VV(vmul_vv_d)
+GEN_VEXT_VV(vmul_vv_b, 1)
+GEN_VEXT_VV(vmul_vv_h, 2)
+GEN_VEXT_VV(vmul_vv_w, 4)
+GEN_VEXT_VV(vmul_vv_d, 8)
 
 static int8_t do_mulh_b(int8_t s2, int8_t s1)
 {
@@ -1509,18 +1566,18 @@ RVVCALL(OPIVV2, vmulhsu_vv_b, OP_SUS_B, H1, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVV2, vmulhsu_vv_h, OP_SUS_H, H2, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVV2, vmulhsu_vv_w, OP_SUS_W, H4, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVV2, vmulhsu_vv_d, OP_SUS_D, H8, H8, H8, do_mulhsu_d)
-GEN_VEXT_VV(vmulh_vv_b)
-GEN_VEXT_VV(vmulh_vv_h)
-GEN_VEXT_VV(vmulh_vv_w)
-GEN_VEXT_VV(vmulh_vv_d)
-GEN_VEXT_VV(vmulhu_vv_b)
-GEN_VEXT_VV(vmulhu_vv_h)
-GEN_VEXT_VV(vmulhu_vv_w)
-GEN_VEXT_VV(vmulhu_vv_d)
-GEN_VEXT_VV(vmulhsu_vv_b)
-GEN_VEXT_VV(vmulhsu_vv_h)
-GEN_VEXT_VV(vmulhsu_vv_w)
-GEN_VEXT_VV(vmulhsu_vv_d)
+GEN_VEXT_VV(vmulh_vv_b, 1)
+GEN_VEXT_VV(vmulh_vv_h, 2)
+GEN_VEXT_VV(vmulh_vv_w, 4)
+GEN_VEXT_VV(vmulh_vv_d, 8)
+GEN_VEXT_VV(vmulhu_vv_b, 1)
+GEN_VEXT_VV(vmulhu_vv_h, 2)
+GEN_VEXT_VV(vmulhu_vv_w, 4)
+GEN_VEXT_VV(vmulhu_vv_d, 8)
+GEN_VEXT_VV(vmulhsu_vv_b, 1)
+GEN_VEXT_VV(vmulhsu_vv_h, 2)
+GEN_VEXT_VV(vmulhsu_vv_w, 4)
+GEN_VEXT_VV(vmulhsu_vv_d, 8)
 
 RVVCALL(OPIVX2, vmul_vx_b, OP_SSS_B, H1, H1, DO_MUL)
 RVVCALL(OPIVX2, vmul_vx_h, OP_SSS_H, H2, H2, DO_MUL)
@@ -1579,22 +1636,22 @@ RVVCALL(OPIVV2, vrem_vv_b, OP_SSS_B, H1, H1, H1, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_h, OP_SSS_H, H2, H2, H2, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_w, OP_SSS_W, H4, H4, H4, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_d, OP_SSS_D, H8, H8, H8, DO_REM)
-GEN_VEXT_VV(vdivu_vv_b)
-GEN_VEXT_VV(vdivu_vv_h)
-GEN_VEXT_VV(vdivu_vv_w)
-GEN_VEXT_VV(vdivu_vv_d)
-GEN_VEXT_VV(vdiv_vv_b)
-GEN_VEXT_VV(vdiv_vv_h)
-GEN_VEXT_VV(vdiv_vv_w)
-GEN_VEXT_VV(vdiv_vv_d)
-GEN_VEXT_VV(vremu_vv_b)
-GEN_VEXT_VV(vremu_vv_h)
-GEN_VEXT_VV(vremu_vv_w)
-GEN_VEXT_VV(vremu_vv_d)
-GEN_VEXT_VV(vrem_vv_b)
-GEN_VEXT_VV(vrem_vv_h)
-GEN_VEXT_VV(vrem_vv_w)
-GEN_VEXT_VV(vrem_vv_d)
+GEN_VEXT_VV(vdivu_vv_b, 1)
+GEN_VEXT_VV(vdivu_vv_h, 2)
+GEN_VEXT_VV(vdivu_vv_w, 4)
+GEN_VEXT_VV(vdivu_vv_d, 8)
+GEN_VEXT_VV(vdiv_vv_b, 1)
+GEN_VEXT_VV(vdiv_vv_h, 2)
+GEN_VEXT_VV(vdiv_vv_w, 4)
+GEN_VEXT_VV(vdiv_vv_d, 8)
+GEN_VEXT_VV(vremu_vv_b, 1)
+GEN_VEXT_VV(vremu_vv_h, 2)
+GEN_VEXT_VV(vremu_vv_w, 4)
+GEN_VEXT_VV(vremu_vv_d, 8)
+GEN_VEXT_VV(vrem_vv_b, 1)
+GEN_VEXT_VV(vrem_vv_h, 2)
+GEN_VEXT_VV(vrem_vv_w, 4)
+GEN_VEXT_VV(vrem_vv_d, 8)
 
 RVVCALL(OPIVX2, vdivu_vx_b, OP_UUU_B, H1, H1, DO_DIVU)
 RVVCALL(OPIVX2, vdivu_vx_h, OP_UUU_H, H2, H2, DO_DIVU)
@@ -1639,15 +1696,15 @@ RVVCALL(OPIVV2, vwmulu_vv_w, WOP_UUU_W, H8, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_b, WOP_SUS_B, H2, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_h, WOP_SUS_H, H4, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_w, WOP_SUS_W, H8, H4, H4, DO_MUL)
-GEN_VEXT_VV(vwmul_vv_b)
-GEN_VEXT_VV(vwmul_vv_h)
-GEN_VEXT_VV(vwmul_vv_w)
-GEN_VEXT_VV(vwmulu_vv_b)
-GEN_VEXT_VV(vwmulu_vv_h)
-GEN_VEXT_VV(vwmulu_vv_w)
-GEN_VEXT_VV(vwmulsu_vv_b)
-GEN_VEXT_VV(vwmulsu_vv_h)
-GEN_VEXT_VV(vwmulsu_vv_w)
+GEN_VEXT_VV(vwmul_vv_b, 2)
+GEN_VEXT_VV(vwmul_vv_h, 4)
+GEN_VEXT_VV(vwmul_vv_w, 8)
+GEN_VEXT_VV(vwmulu_vv_b, 2)
+GEN_VEXT_VV(vwmulu_vv_h, 4)
+GEN_VEXT_VV(vwmulu_vv_w, 8)
+GEN_VEXT_VV(vwmulsu_vv_b, 2)
+GEN_VEXT_VV(vwmulsu_vv_h, 4)
+GEN_VEXT_VV(vwmulsu_vv_w, 8)
 
 RVVCALL(OPIVX2, vwmul_vx_b, WOP_SSS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmul_vx_h, WOP_SSS_H, H4, H2, DO_MUL)
@@ -1698,22 +1755,22 @@ RVVCALL(OPIVV3, vnmsub_vv_b, OP_SSS_B, H1, H1, H1, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_h, OP_SSS_H, H2, H2, H2, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_w, OP_SSS_W, H4, H4, H4, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_d, OP_SSS_D, H8, H8, H8, DO_NMSUB)
-GEN_VEXT_VV(vmacc_vv_b)
-GEN_VEXT_VV(vmacc_vv_h)
-GEN_VEXT_VV(vmacc_vv_w)
-GEN_VEXT_VV(vmacc_vv_d)
-GEN_VEXT_VV(vnmsac_vv_b)
-GEN_VEXT_VV(vnmsac_vv_h)
-GEN_VEXT_VV(vnmsac_vv_w)
-GEN_VEXT_VV(vnmsac_vv_d)
-GEN_VEXT_VV(vmadd_vv_b)
-GEN_VEXT_VV(vmadd_vv_h)
-GEN_VEXT_VV(vmadd_vv_w)
-GEN_VEXT_VV(vmadd_vv_d)
-GEN_VEXT_VV(vnmsub_vv_b)
-GEN_VEXT_VV(vnmsub_vv_h)
-GEN_VEXT_VV(vnmsub_vv_w)
-GEN_VEXT_VV(vnmsub_vv_d)
+GEN_VEXT_VV(vmacc_vv_b, 1)
+GEN_VEXT_VV(vmacc_vv_h, 2)
+GEN_VEXT_VV(vmacc_vv_w, 4)
+GEN_VEXT_VV(vmacc_vv_d, 8)
+GEN_VEXT_VV(vnmsac_vv_b, 1)
+GEN_VEXT_VV(vnmsac_vv_h, 2)
+GEN_VEXT_VV(vnmsac_vv_w, 4)
+GEN_VEXT_VV(vnmsac_vv_d, 8)
+GEN_VEXT_VV(vmadd_vv_b, 1)
+GEN_VEXT_VV(vmadd_vv_h, 2)
+GEN_VEXT_VV(vmadd_vv_w, 4)
+GEN_VEXT_VV(vmadd_vv_d, 8)
+GEN_VEXT_VV(vnmsub_vv_b, 1)
+GEN_VEXT_VV(vnmsub_vv_h, 2)
+GEN_VEXT_VV(vnmsub_vv_w, 4)
+GEN_VEXT_VV(vnmsub_vv_d, 8)
 
 #define OPIVX3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)             \
 static void do_##NAME(void *vd, target_long s1, void *vs2, int i)   \
@@ -1766,15 +1823,15 @@ RVVCALL(OPIVV3, vwmacc_vv_w, WOP_SSS_W, H8, H4, H4, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_b, WOP_SSU_B, H2, H1, H1, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_h, WOP_SSU_H, H4, H2, H2, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_w, WOP_SSU_W, H8, H4, H4, DO_MACC)
-GEN_VEXT_VV(vwmaccu_vv_b)
-GEN_VEXT_VV(vwmaccu_vv_h)
-GEN_VEXT_VV(vwmaccu_vv_w)
-GEN_VEXT_VV(vwmacc_vv_b)
-GEN_VEXT_VV(vwmacc_vv_h)
-GEN_VEXT_VV(vwmacc_vv_w)
-GEN_VEXT_VV(vwmaccsu_vv_b)
-GEN_VEXT_VV(vwmaccsu_vv_h)
-GEN_VEXT_VV(vwmaccsu_vv_w)
+GEN_VEXT_VV(vwmaccu_vv_b, 2)
+GEN_VEXT_VV(vwmaccu_vv_h, 4)
+GEN_VEXT_VV(vwmaccu_vv_w, 8)
+GEN_VEXT_VV(vwmacc_vv_b, 2)
+GEN_VEXT_VV(vwmacc_vv_h, 4)
+GEN_VEXT_VV(vwmacc_vv_w, 8)
+GEN_VEXT_VV(vwmaccsu_vv_b, 2)
+GEN_VEXT_VV(vwmaccsu_vv_h, 4)
+GEN_VEXT_VV(vwmaccsu_vv_w, 8)
 
 RVVCALL(OPIVX3, vwmaccu_vx_b, WOP_UUU_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccu_vx_h, WOP_UUU_H, H4, H2, DO_MACC)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 04/14] target/riscv: rvv: Add tail agnostic for vv instructions
  2022-03-24 19:03   ` ~eopxd
@ 2022-03-25  9:10     ` Weiwei Li
  -1 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25  9:10 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, Bin Meng, Alistair Francis, eop Chen,
	Palmer Dabbelt


在 2022/3/1 下午5:07, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> This is the first commit regarding the tail agnostic behavior.
> Added option 'rvv_ta_all_1s' to enable the behavior, the option
> is default to false.
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/cpu.c                      |   1 +
>   target/riscv/cpu.h                      |   2 +
>   target/riscv/cpu_helper.c               |   2 +
>   target/riscv/insn_trans/trans_rvv.c.inc |  10 +
>   target/riscv/internals.h                |   5 +-
>   target/riscv/translate.c                |   2 +
>   target/riscv/vector_helper.c            | 315 ++++++++++++++----------
>   7 files changed, 206 insertions(+), 131 deletions(-)
>
> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> index ddda4906ff..cd4cf4b41e 100644
> --- a/target/riscv/cpu.c
> +++ b/target/riscv/cpu.c
> @@ -810,6 +810,7 @@ static Property riscv_cpu_properties[] = {
>       DEFINE_PROP_BOOL("x-aia", RISCVCPU, cfg.aia, false),
>   
>       DEFINE_PROP_UINT64("resetvec", RISCVCPU, cfg.resetvec, DEFAULT_RSTVEC),
> +    DEFINE_PROP_BOOL("rvv_ta_all_1s", RISCVCPU, cfg.rvv_ta_all_1s, false),
>       DEFINE_PROP_END_OF_LIST(),
>   };
>   
> diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
> index c069fe85fa..8c4a79b5a0 100644
> --- a/target/riscv/cpu.h
> +++ b/target/riscv/cpu.h
> @@ -369,6 +369,7 @@ struct RISCVCPUConfig {
>       bool ext_zhinxmin;
>       bool ext_zve32f;
>       bool ext_zve64f;
> +    bool rvv_ta_all_1s;
>   
>       /* Vendor-specific custom extensions */
>       bool ext_XVentanaCondOps;
> @@ -516,6 +517,7 @@ FIELD(TB_FLAGS, XL, 20, 2)
>   /* If PointerMasking should be applied */
>   FIELD(TB_FLAGS, PM_MASK_ENABLED, 22, 1)
>   FIELD(TB_FLAGS, PM_BASE_ENABLED, 23, 1)
> +FIELD(TB_FLAGS, VTA, 24, 1)
>   
>   #ifdef TARGET_RISCV32
>   #define riscv_cpu_mxl(env)  ((void)(env), MXL_RV32)
> diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
> index 1c60fb2e80..2941c88c31 100644
> --- a/target/riscv/cpu_helper.c
> +++ b/target/riscv/cpu_helper.c
> @@ -65,6 +65,8 @@ void cpu_get_tb_cpu_state(CPURISCVState *env, target_ulong *pc,
>           flags = FIELD_DP32(flags, TB_FLAGS, LMUL,
>                       FIELD_EX64(env->vtype, VTYPE, VLMUL));
>           flags = FIELD_DP32(flags, TB_FLAGS, VL_EQ_VLMAX, vl_eq_vlmax);
> +        flags = FIELD_DP32(flags, TB_FLAGS, VTA,
> +                    FIELD_EX64(env->vtype, VTYPE, VTA));
>       } else {
>           flags = FIELD_DP32(flags, TB_FLAGS, VILL, 1);
>       }
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index 57953923d5..166ee9df2b 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -1223,6 +1223,15 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
>       tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */
> +            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                             vreg_ofs(s, a->rd), -1,
> +                             vlenb, vlenb);
> +        }
>           gvec_fn(s->sew, vreg_ofs(s, a->rd),
>                   vreg_ofs(s, a->rs2), vreg_ofs(s, a->rs1),
>                   MAXSZ(s), MAXSZ(s));
> @@ -1231,6 +1240,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
>   
>           data = FIELD_DP32(data, VDATA, VM, a->vm);
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                              vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
>                              cpu_env, s->cfg_ptr->vlen / 8,
> diff --git a/target/riscv/internals.h b/target/riscv/internals.h
> index dbb322bfa7..512c6c30cf 100644
> --- a/target/riscv/internals.h
> +++ b/target/riscv/internals.h
> @@ -24,8 +24,9 @@
>   /* share data between vector helpers and decode code */
>   FIELD(VDATA, VM, 0, 1)
>   FIELD(VDATA, LMUL, 1, 3)
> -FIELD(VDATA, NF, 4, 4)
> -FIELD(VDATA, WD, 4, 1)
> +FIELD(VDATA, VTA, 4, 1)
> +FIELD(VDATA, NF, 5, 4)
> +FIELD(VDATA, WD, 5, 1)
>   
>   /* float point classify helpers */
>   target_ulong fclass_h(uint64_t frs1);
> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
> index fac998a6b5..7775dade26 100644
> --- a/target/riscv/translate.c
> +++ b/target/riscv/translate.c
> @@ -94,6 +94,7 @@ typedef struct DisasContext {
>        */
>       int8_t lmul;
>       uint8_t sew;
> +    uint8_t vta;
>       target_ulong vstart;
>       bool vl_eq_vlmax;
>       uint8_t ntemp;
> @@ -1083,6 +1084,7 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
>       ctx->vill = FIELD_EX32(tb_flags, TB_FLAGS, VILL);
>       ctx->sew = FIELD_EX32(tb_flags, TB_FLAGS, SEW);
>       ctx->lmul = sextract32(FIELD_EX32(tb_flags, TB_FLAGS, LMUL), 0, 3);
> +    ctx->vta = FIELD_EX32(tb_flags, TB_FLAGS, VTA) && cpu->cfg.rvv_ta_all_1s;
>       ctx->vstart = env->vstart;
>       ctx->vl_eq_vlmax = FIELD_EX32(tb_flags, TB_FLAGS, VL_EQ_VLMAX);
>       ctx->misa_mxl_max = env->misa_mxl_max;
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index d0452a7756..b9976f9d77 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -122,6 +122,11 @@ static inline int32_t vext_lmul(uint32_t desc)
>       return sextract32(FIELD_EX32(simd_data(desc), VDATA, LMUL), 0, 3);
>   }
>   
> +static inline uint32_t vext_vta(uint32_t desc)
> +{
> +    return FIELD_EX32(simd_data(desc), VDATA, VTA);
> +}
> +
>   /*
>    * Get the maximum number of elements can be operated.
>    *
> @@ -140,6 +145,20 @@ static inline uint32_t vext_max_elems(uint32_t desc, uint32_t log2_esz)
>       return scale < 0 ? vlenb >> -scale : vlenb << scale;
>   }
>   
> +/*
> + * Get number of total elements, including prestart, body and tail elements.
> + * Note that when LMUL < 1, the tail includes the elements past VLMAX that
> + * are held in the same vector register.
> + */
> +static inline uint32_t vext_get_total_elems(RISCVCPU *cpu, target_ulong vtype)
> +{
> +    uint8_t sew = FIELD_EX64(vtype, VTYPE, VSEW);
> +    int8_t lmul = sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3) < 0 ?
> +                  0 : sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3);
> +    return cpu->cfg.vlen >> (sew + 3 - lmul);
> +}
> +

I have sent another patchset to  try to optimize a few point for rvv.  
One of the point I haven't done is that,

for some rvv instructions, lmul field in 'desc'( or the total desc) is 
assigned in trans_* but not used in helper function.

This field seems can be used here. I think It may be better to use desc 
and esz(just like vext_max_elems) here.

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 04/14] target/riscv: rvv: Add tail agnostic for vv instructions
@ 2022-03-25  9:10     ` Weiwei Li
  0 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25  9:10 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt


在 2022/3/1 下午5:07, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> This is the first commit regarding the tail agnostic behavior.
> Added option 'rvv_ta_all_1s' to enable the behavior, the option
> is default to false.
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/cpu.c                      |   1 +
>   target/riscv/cpu.h                      |   2 +
>   target/riscv/cpu_helper.c               |   2 +
>   target/riscv/insn_trans/trans_rvv.c.inc |  10 +
>   target/riscv/internals.h                |   5 +-
>   target/riscv/translate.c                |   2 +
>   target/riscv/vector_helper.c            | 315 ++++++++++++++----------
>   7 files changed, 206 insertions(+), 131 deletions(-)
>
> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> index ddda4906ff..cd4cf4b41e 100644
> --- a/target/riscv/cpu.c
> +++ b/target/riscv/cpu.c
> @@ -810,6 +810,7 @@ static Property riscv_cpu_properties[] = {
>       DEFINE_PROP_BOOL("x-aia", RISCVCPU, cfg.aia, false),
>   
>       DEFINE_PROP_UINT64("resetvec", RISCVCPU, cfg.resetvec, DEFAULT_RSTVEC),
> +    DEFINE_PROP_BOOL("rvv_ta_all_1s", RISCVCPU, cfg.rvv_ta_all_1s, false),
>       DEFINE_PROP_END_OF_LIST(),
>   };
>   
> diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
> index c069fe85fa..8c4a79b5a0 100644
> --- a/target/riscv/cpu.h
> +++ b/target/riscv/cpu.h
> @@ -369,6 +369,7 @@ struct RISCVCPUConfig {
>       bool ext_zhinxmin;
>       bool ext_zve32f;
>       bool ext_zve64f;
> +    bool rvv_ta_all_1s;
>   
>       /* Vendor-specific custom extensions */
>       bool ext_XVentanaCondOps;
> @@ -516,6 +517,7 @@ FIELD(TB_FLAGS, XL, 20, 2)
>   /* If PointerMasking should be applied */
>   FIELD(TB_FLAGS, PM_MASK_ENABLED, 22, 1)
>   FIELD(TB_FLAGS, PM_BASE_ENABLED, 23, 1)
> +FIELD(TB_FLAGS, VTA, 24, 1)
>   
>   #ifdef TARGET_RISCV32
>   #define riscv_cpu_mxl(env)  ((void)(env), MXL_RV32)
> diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
> index 1c60fb2e80..2941c88c31 100644
> --- a/target/riscv/cpu_helper.c
> +++ b/target/riscv/cpu_helper.c
> @@ -65,6 +65,8 @@ void cpu_get_tb_cpu_state(CPURISCVState *env, target_ulong *pc,
>           flags = FIELD_DP32(flags, TB_FLAGS, LMUL,
>                       FIELD_EX64(env->vtype, VTYPE, VLMUL));
>           flags = FIELD_DP32(flags, TB_FLAGS, VL_EQ_VLMAX, vl_eq_vlmax);
> +        flags = FIELD_DP32(flags, TB_FLAGS, VTA,
> +                    FIELD_EX64(env->vtype, VTYPE, VTA));
>       } else {
>           flags = FIELD_DP32(flags, TB_FLAGS, VILL, 1);
>       }
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index 57953923d5..166ee9df2b 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -1223,6 +1223,15 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
>       tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */
> +            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                             vreg_ofs(s, a->rd), -1,
> +                             vlenb, vlenb);
> +        }
>           gvec_fn(s->sew, vreg_ofs(s, a->rd),
>                   vreg_ofs(s, a->rs2), vreg_ofs(s, a->rs1),
>                   MAXSZ(s), MAXSZ(s));
> @@ -1231,6 +1240,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
>   
>           data = FIELD_DP32(data, VDATA, VM, a->vm);
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                              vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
>                              cpu_env, s->cfg_ptr->vlen / 8,
> diff --git a/target/riscv/internals.h b/target/riscv/internals.h
> index dbb322bfa7..512c6c30cf 100644
> --- a/target/riscv/internals.h
> +++ b/target/riscv/internals.h
> @@ -24,8 +24,9 @@
>   /* share data between vector helpers and decode code */
>   FIELD(VDATA, VM, 0, 1)
>   FIELD(VDATA, LMUL, 1, 3)
> -FIELD(VDATA, NF, 4, 4)
> -FIELD(VDATA, WD, 4, 1)
> +FIELD(VDATA, VTA, 4, 1)
> +FIELD(VDATA, NF, 5, 4)
> +FIELD(VDATA, WD, 5, 1)
>   
>   /* float point classify helpers */
>   target_ulong fclass_h(uint64_t frs1);
> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
> index fac998a6b5..7775dade26 100644
> --- a/target/riscv/translate.c
> +++ b/target/riscv/translate.c
> @@ -94,6 +94,7 @@ typedef struct DisasContext {
>        */
>       int8_t lmul;
>       uint8_t sew;
> +    uint8_t vta;
>       target_ulong vstart;
>       bool vl_eq_vlmax;
>       uint8_t ntemp;
> @@ -1083,6 +1084,7 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
>       ctx->vill = FIELD_EX32(tb_flags, TB_FLAGS, VILL);
>       ctx->sew = FIELD_EX32(tb_flags, TB_FLAGS, SEW);
>       ctx->lmul = sextract32(FIELD_EX32(tb_flags, TB_FLAGS, LMUL), 0, 3);
> +    ctx->vta = FIELD_EX32(tb_flags, TB_FLAGS, VTA) && cpu->cfg.rvv_ta_all_1s;
>       ctx->vstart = env->vstart;
>       ctx->vl_eq_vlmax = FIELD_EX32(tb_flags, TB_FLAGS, VL_EQ_VLMAX);
>       ctx->misa_mxl_max = env->misa_mxl_max;
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index d0452a7756..b9976f9d77 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -122,6 +122,11 @@ static inline int32_t vext_lmul(uint32_t desc)
>       return sextract32(FIELD_EX32(simd_data(desc), VDATA, LMUL), 0, 3);
>   }
>   
> +static inline uint32_t vext_vta(uint32_t desc)
> +{
> +    return FIELD_EX32(simd_data(desc), VDATA, VTA);
> +}
> +
>   /*
>    * Get the maximum number of elements can be operated.
>    *
> @@ -140,6 +145,20 @@ static inline uint32_t vext_max_elems(uint32_t desc, uint32_t log2_esz)
>       return scale < 0 ? vlenb >> -scale : vlenb << scale;
>   }
>   
> +/*
> + * Get number of total elements, including prestart, body and tail elements.
> + * Note that when LMUL < 1, the tail includes the elements past VLMAX that
> + * are held in the same vector register.
> + */
> +static inline uint32_t vext_get_total_elems(RISCVCPU *cpu, target_ulong vtype)
> +{
> +    uint8_t sew = FIELD_EX64(vtype, VTYPE, VSEW);
> +    int8_t lmul = sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3) < 0 ?
> +                  0 : sextract32(FIELD_EX64(vtype, VTYPE, VLMUL), 0, 3);
> +    return cpu->cfg.vlen >> (sew + 3 - lmul);
> +}
> +

I have sent another patchset to  try to optimize a few point for rvv.  
One of the point I haven't done is that,

for some rvv instructions, lmul field in 'desc'( or the total desc) is 
assigned in trans_* but not used in helper function.

This field seems can be used here. I think It may be better to use desc 
and esz(just like vext_max_elems) here.

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 05/14] target/riscv: rvv: Add tail agnostic for vector load / store instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07  7:10 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  9 +++++++
 target/riscv/vector_helper.c            | 32 +++++++++++++++++++++++++
 2 files changed, 41 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 166ee9df2b..fee616ef38 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -711,6 +711,7 @@ static bool ld_us_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, false);
 }
 
@@ -748,6 +749,7 @@ static bool st_us_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, true);
 }
 
@@ -774,6 +776,7 @@ static bool ld_us_mask_op(DisasContext *s, arg_vlm_v *a, uint8_t eew)
     /* EMUL = 1, NFIELDS = 1 */
     data = FIELD_DP32(data, VDATA, LMUL, 0);
     data = FIELD_DP32(data, VDATA, NF, 1);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, false);
 }
 
@@ -791,6 +794,7 @@ static bool st_us_mask_op(DisasContext *s, arg_vsm_v *a, uint8_t eew)
     /* EMUL = 1, NFIELDS = 1 */
     data = FIELD_DP32(data, VDATA, LMUL, 0);
     data = FIELD_DP32(data, VDATA, NF, 1);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, true);
 }
 
@@ -862,6 +866,7 @@ static bool ld_stride_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_stride_trans(a->rd, a->rs1, a->rs2, data, fn, s, false);
 }
 
@@ -891,6 +896,7 @@ static bool st_stride_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     fn = fns[eew];
     if (fn == NULL) {
         return false;
@@ -991,6 +997,7 @@ static bool ld_index_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_index_trans(a->rd, a->rs1, a->rs2, data, fn, s, false);
 }
 
@@ -1043,6 +1050,7 @@ static bool st_index_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_index_trans(a->rd, a->rs1, a->rs2, data, fn, s, true);
 }
 
@@ -1108,6 +1116,7 @@ static bool ldff_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldff_trans(a->rd, a->rs1, data, fn, s);
 }
 
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index b9976f9d77..c9e5a839a3 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -289,6 +289,9 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
 
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
         if (!vm && !vext_elem_mask(v0, i)) {
@@ -303,6 +306,11 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 #define GEN_VEXT_LD_STRIDE(NAME, ETYPE, LOAD_FN)                        \
@@ -348,6 +356,9 @@ vext_ldst_us(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < evl; i++, env->vstart++) {
@@ -359,6 +370,11 @@ vext_ldst_us(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 /*
@@ -458,6 +474,9 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
@@ -473,6 +492,11 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 #define GEN_VEXT_LD_INDEX(NAME, ETYPE, INDEX_FN, LOAD_FN)                  \
@@ -540,6 +564,9 @@ vext_ldff(void *vd, void *v0, target_ulong base,
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
     target_ulong addr, offset, remain;
 
     /* probe every access*/
@@ -595,6 +622,11 @@ ProbeSuccess:
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 #define GEN_VEXT_LDFF(NAME, ETYPE, LOAD_FN)               \
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 05/14] target/riscv: rvv: Add tail agnostic for vector load / store instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  9 +++++++
 target/riscv/vector_helper.c            | 32 +++++++++++++++++++++++++
 2 files changed, 41 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 166ee9df2b..fee616ef38 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -711,6 +711,7 @@ static bool ld_us_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, false);
 }
 
@@ -748,6 +749,7 @@ static bool st_us_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, true);
 }
 
@@ -774,6 +776,7 @@ static bool ld_us_mask_op(DisasContext *s, arg_vlm_v *a, uint8_t eew)
     /* EMUL = 1, NFIELDS = 1 */
     data = FIELD_DP32(data, VDATA, LMUL, 0);
     data = FIELD_DP32(data, VDATA, NF, 1);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, false);
 }
 
@@ -791,6 +794,7 @@ static bool st_us_mask_op(DisasContext *s, arg_vsm_v *a, uint8_t eew)
     /* EMUL = 1, NFIELDS = 1 */
     data = FIELD_DP32(data, VDATA, LMUL, 0);
     data = FIELD_DP32(data, VDATA, NF, 1);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_us_trans(a->rd, a->rs1, data, fn, s, true);
 }
 
@@ -862,6 +866,7 @@ static bool ld_stride_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_stride_trans(a->rd, a->rs1, a->rs2, data, fn, s, false);
 }
 
@@ -891,6 +896,7 @@ static bool st_stride_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     fn = fns[eew];
     if (fn == NULL) {
         return false;
@@ -991,6 +997,7 @@ static bool ld_index_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_index_trans(a->rd, a->rs1, a->rs2, data, fn, s, false);
 }
 
@@ -1043,6 +1050,7 @@ static bool st_index_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldst_index_trans(a->rd, a->rs1, a->rs2, data, fn, s, true);
 }
 
@@ -1108,6 +1116,7 @@ static bool ldff_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
     data = FIELD_DP32(data, VDATA, VM, a->vm);
     data = FIELD_DP32(data, VDATA, LMUL, emul);
     data = FIELD_DP32(data, VDATA, NF, a->nf);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     return ldff_trans(a->rd, a->rs1, data, fn, s);
 }
 
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index b9976f9d77..c9e5a839a3 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -289,6 +289,9 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
 
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
         if (!vm && !vext_elem_mask(v0, i)) {
@@ -303,6 +306,11 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 #define GEN_VEXT_LD_STRIDE(NAME, ETYPE, LOAD_FN)                        \
@@ -348,6 +356,9 @@ vext_ldst_us(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < evl; i++, env->vstart++) {
@@ -359,6 +370,11 @@ vext_ldst_us(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 /*
@@ -458,6 +474,9 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
@@ -473,6 +492,11 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 #define GEN_VEXT_LD_INDEX(NAME, ETYPE, INDEX_FN, LOAD_FN)                  \
@@ -540,6 +564,9 @@ vext_ldff(void *vd, void *v0, target_ulong base,
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
     uint32_t max_elems = vext_max_elems(desc, log2_esz);
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
+    uint32_t esz = 1 << log2_esz;
     target_ulong addr, offset, remain;
 
     /* probe every access*/
@@ -595,6 +622,11 @@ ProbeSuccess:
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    for (k = 0; k < nf; ++k) {
+        vext_set_elems_1s_fns[log2_esz](vd, vta, env->vl + k * total_elems,
+                                        env->vl * esz, total_elems * esz);
+    }
 }
 
 #define GEN_VEXT_LDFF(NAME, ETYPE, LOAD_FN)               \
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 06/14] target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07  7:32 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  23 ++
 target/riscv/vector_helper.c            | 312 +++++++++++++-----------
 2 files changed, 197 insertions(+), 138 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index fee616ef38..e8ef5b1e0e 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1296,6 +1296,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
 
     data = FIELD_DP32(data, VDATA, VM, vm);
     data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8,
                                       s->cfg_ptr->vlen / 8, data));
 
@@ -1332,6 +1333,15 @@ do_opivx_gvec(DisasContext *s, arg_rmrr *a, GVecGen2sFn *gvec_fn,
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         TCGv_i64 src1 = tcg_temp_new_i64();
 
         tcg_gen_ext_tl_i64(src1, get_gpr(s, a->rs1, EXT_SIGN));
@@ -1461,6 +1471,7 @@ static bool opivi_trans(uint32_t vd, uint32_t imm, uint32_t vs2, uint32_t vm,
 
     data = FIELD_DP32(data, VDATA, VM, vm);
     data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8,
                                       s->cfg_ptr->vlen / 8, data));
 
@@ -1490,6 +1501,15 @@ do_opivi_gvec(DisasContext *s, arg_rmrr *a, GVecGen2iFn *gvec_fn,
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         gvec_fn(s->sew, vreg_ofs(s, a->rd), vreg_ofs(s, a->rs2),
                 extract_imm(s, a->rs1, imm_mode), MAXSZ(s), MAXSZ(s));
         mark_vs_dirty(s);
@@ -1543,6 +1563,7 @@ static bool do_opivv_widen(DisasContext *s, arg_rmrr *a,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1),
                            vreg_ofs(s, a->rs2),
@@ -1624,6 +1645,7 @@ static bool do_opiwv_widen(DisasContext *s, arg_rmrr *a,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1),
                            vreg_ofs(s, a->rs2),
@@ -1702,6 +1724,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index c9e5a839a3..4be99e3be4 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -861,10 +861,12 @@ RVVCALL(OPIVX2, vrsub_vx_d, OP_SSS_D, H8, H8, DO_RSUB)
 
 static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       opivx2_fn fn)
+                       opivx2_fn fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
 
     for (i = env->vstart; i < vl; i++) {
@@ -874,30 +876,32 @@ static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
         fn(vd, s1, vs2, i);
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate the helpers for OPIVX */
-#define GEN_VEXT_VX(NAME)                                 \
+#define GEN_VEXT_VX(NAME, ESZ)                            \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     do_vext_vx(vd, v0, s1, vs2, env, desc,                \
-               do_##NAME);                                \
-}
-
-GEN_VEXT_VX(vadd_vx_b)
-GEN_VEXT_VX(vadd_vx_h)
-GEN_VEXT_VX(vadd_vx_w)
-GEN_VEXT_VX(vadd_vx_d)
-GEN_VEXT_VX(vsub_vx_b)
-GEN_VEXT_VX(vsub_vx_h)
-GEN_VEXT_VX(vsub_vx_w)
-GEN_VEXT_VX(vsub_vx_d)
-GEN_VEXT_VX(vrsub_vx_b)
-GEN_VEXT_VX(vrsub_vx_h)
-GEN_VEXT_VX(vrsub_vx_w)
-GEN_VEXT_VX(vrsub_vx_d)
+               do_##NAME, ESZ);                           \
+}
+
+GEN_VEXT_VX(vadd_vx_b, 1)
+GEN_VEXT_VX(vadd_vx_h, 2)
+GEN_VEXT_VX(vadd_vx_w, 4)
+GEN_VEXT_VX(vadd_vx_d, 8)
+GEN_VEXT_VX(vsub_vx_b, 1)
+GEN_VEXT_VX(vsub_vx_h, 2)
+GEN_VEXT_VX(vsub_vx_w, 4)
+GEN_VEXT_VX(vsub_vx_d, 8)
+GEN_VEXT_VX(vrsub_vx_b, 1)
+GEN_VEXT_VX(vrsub_vx_h, 2)
+GEN_VEXT_VX(vrsub_vx_w, 4)
+GEN_VEXT_VX(vrsub_vx_d, 8)
 
 void HELPER(vec_rsubs8)(void *d, void *a, uint64_t b, uint32_t desc)
 {
@@ -1025,30 +1029,30 @@ RVVCALL(OPIVX2, vwadd_wx_w, WOP_WSSS_W, H8, H4, DO_ADD)
 RVVCALL(OPIVX2, vwsub_wx_b, WOP_WSSS_B, H2, H1, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_h, WOP_WSSS_H, H4, H2, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_w, WOP_WSSS_W, H8, H4, DO_SUB)
-GEN_VEXT_VX(vwaddu_vx_b)
-GEN_VEXT_VX(vwaddu_vx_h)
-GEN_VEXT_VX(vwaddu_vx_w)
-GEN_VEXT_VX(vwsubu_vx_b)
-GEN_VEXT_VX(vwsubu_vx_h)
-GEN_VEXT_VX(vwsubu_vx_w)
-GEN_VEXT_VX(vwadd_vx_b)
-GEN_VEXT_VX(vwadd_vx_h)
-GEN_VEXT_VX(vwadd_vx_w)
-GEN_VEXT_VX(vwsub_vx_b)
-GEN_VEXT_VX(vwsub_vx_h)
-GEN_VEXT_VX(vwsub_vx_w)
-GEN_VEXT_VX(vwaddu_wx_b)
-GEN_VEXT_VX(vwaddu_wx_h)
-GEN_VEXT_VX(vwaddu_wx_w)
-GEN_VEXT_VX(vwsubu_wx_b)
-GEN_VEXT_VX(vwsubu_wx_h)
-GEN_VEXT_VX(vwsubu_wx_w)
-GEN_VEXT_VX(vwadd_wx_b)
-GEN_VEXT_VX(vwadd_wx_h)
-GEN_VEXT_VX(vwadd_wx_w)
-GEN_VEXT_VX(vwsub_wx_b)
-GEN_VEXT_VX(vwsub_wx_h)
-GEN_VEXT_VX(vwsub_wx_w)
+GEN_VEXT_VX(vwaddu_vx_b, 2)
+GEN_VEXT_VX(vwaddu_vx_h, 4)
+GEN_VEXT_VX(vwaddu_vx_w, 8)
+GEN_VEXT_VX(vwsubu_vx_b, 2)
+GEN_VEXT_VX(vwsubu_vx_h, 4)
+GEN_VEXT_VX(vwsubu_vx_w, 8)
+GEN_VEXT_VX(vwadd_vx_b, 2)
+GEN_VEXT_VX(vwadd_vx_h, 4)
+GEN_VEXT_VX(vwadd_vx_w, 8)
+GEN_VEXT_VX(vwsub_vx_b, 2)
+GEN_VEXT_VX(vwsub_vx_h, 4)
+GEN_VEXT_VX(vwsub_vx_w, 8)
+GEN_VEXT_VX(vwaddu_wx_b, 2)
+GEN_VEXT_VX(vwaddu_wx_h, 4)
+GEN_VEXT_VX(vwaddu_wx_w, 8)
+GEN_VEXT_VX(vwsubu_wx_b, 2)
+GEN_VEXT_VX(vwsubu_wx_h, 4)
+GEN_VEXT_VX(vwsubu_wx_w, 8)
+GEN_VEXT_VX(vwadd_wx_b, 2)
+GEN_VEXT_VX(vwadd_wx_h, 4)
+GEN_VEXT_VX(vwadd_wx_w, 8)
+GEN_VEXT_VX(vwsub_wx_b, 2)
+GEN_VEXT_VX(vwsub_wx_h, 4)
+GEN_VEXT_VX(vwsub_wx_w, 8)
 
 /* Vector Integer Add-with-Carry / Subtract-with-Borrow Instructions */
 #define DO_VADC(N, M, C) (N + M + C)
@@ -1059,6 +1063,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
     uint32_t vl = env->vl;                                    \
+    uint32_t esz = sizeof(ETYPE);                             \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -1069,6 +1077,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
         *((ETYPE *)vd + H(i)) = DO_OP(s2, s1, carry);         \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,   \
+                                     total_elems * esz);      \
 }
 
 GEN_VEXT_VADC_VVM(vadc_vvm_b, uint8_t,  H1, DO_VADC)
@@ -1086,6 +1097,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,        \
                   CPURISCVState *env, uint32_t desc)                     \
 {                                                                        \
     uint32_t vl = env->vl;                                               \
+    uint32_t esz = sizeof(ETYPE);                                        \
+    uint32_t total_elems =                                               \
+        vext_get_total_elems(env_archcpu(env), env->vtype);              \
+    uint32_t vta = vext_vta(desc);                                       \
     uint32_t i;                                                          \
                                                                          \
     for (i = env->vstart; i < vl; i++) {                                 \
@@ -1095,6 +1110,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,        \
         *((ETYPE *)vd + H(i)) = DO_OP(s2, (ETYPE)(target_long)s1, carry);\
     }                                                                    \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                                        \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,              \
+                                     total_elems * esz);                 \
 }
 
 GEN_VEXT_VADC_VXM(vadc_vxm_b, uint8_t,  H1, DO_VADC)
@@ -1117,6 +1135,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
 {                                                             \
     uint32_t vl = env->vl;                                    \
     uint32_t vm = vext_vm(desc);                              \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -1126,6 +1147,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
         vext_set_elem_mask(vd, i, DO_OP(s2, s1, carry));      \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    if (vta) {                                                \
+        for (; i < total_elems; i++) {                        \
+            vext_set_elem_mask(vd, i, 1);                     \
+        }                                                     \
+    }                                                         \
 }
 
 GEN_VEXT_VMADC_VVM(vmadc_vvm_b, uint8_t,  H1, DO_MADC)
@@ -1144,6 +1171,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,          \
 {                                                               \
     uint32_t vl = env->vl;                                      \
     uint32_t vm = vext_vm(desc);                                \
+    uint32_t total_elems =                                      \
+        vext_get_total_elems(env_archcpu(env), env->vtype);     \
+    uint32_t vta = vext_vta(desc);                              \
     uint32_t i;                                                 \
                                                                 \
     for (i = env->vstart; i < vl; i++) {                        \
@@ -1153,6 +1183,12 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,          \
                 DO_OP(s2, (ETYPE)(target_long)s1, carry));      \
     }                                                           \
     env->vstart = 0;                                            \
+    /* set tail elements to 1s */                               \
+    if (vta) {                                                  \
+        for (; i < total_elems; i++) {                          \
+            vext_set_elem_mask(vd, i, 1);                       \
+        }                                                       \
+    }                                                           \
 }
 
 GEN_VEXT_VMADC_VXM(vmadc_vxm_b, uint8_t,  H1, DO_MADC)
@@ -1203,18 +1239,18 @@ RVVCALL(OPIVX2, vxor_vx_b, OP_SSS_B, H1, H1, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_h, OP_SSS_H, H2, H2, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_w, OP_SSS_W, H4, H4, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_d, OP_SSS_D, H8, H8, DO_XOR)
-GEN_VEXT_VX(vand_vx_b)
-GEN_VEXT_VX(vand_vx_h)
-GEN_VEXT_VX(vand_vx_w)
-GEN_VEXT_VX(vand_vx_d)
-GEN_VEXT_VX(vor_vx_b)
-GEN_VEXT_VX(vor_vx_h)
-GEN_VEXT_VX(vor_vx_w)
-GEN_VEXT_VX(vor_vx_d)
-GEN_VEXT_VX(vxor_vx_b)
-GEN_VEXT_VX(vxor_vx_h)
-GEN_VEXT_VX(vxor_vx_w)
-GEN_VEXT_VX(vxor_vx_d)
+GEN_VEXT_VX(vand_vx_b, 1)
+GEN_VEXT_VX(vand_vx_h, 2)
+GEN_VEXT_VX(vand_vx_w, 4)
+GEN_VEXT_VX(vand_vx_d, 8)
+GEN_VEXT_VX(vor_vx_b, 1)
+GEN_VEXT_VX(vor_vx_h, 2)
+GEN_VEXT_VX(vor_vx_w, 4)
+GEN_VEXT_VX(vor_vx_d, 8)
+GEN_VEXT_VX(vxor_vx_b, 1)
+GEN_VEXT_VX(vxor_vx_h, 2)
+GEN_VEXT_VX(vxor_vx_w, 4)
+GEN_VEXT_VX(vxor_vx_d, 8)
 
 /* Vector Single-Width Bit Shift Instructions */
 #define DO_SLL(N, M)  (N << (M))
@@ -1468,22 +1504,22 @@ RVVCALL(OPIVX2, vmax_vx_b, OP_SSS_B, H1, H1, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_h, OP_SSS_H, H2, H2, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_w, OP_SSS_W, H4, H4, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_d, OP_SSS_D, H8, H8, DO_MAX)
-GEN_VEXT_VX(vminu_vx_b)
-GEN_VEXT_VX(vminu_vx_h)
-GEN_VEXT_VX(vminu_vx_w)
-GEN_VEXT_VX(vminu_vx_d)
-GEN_VEXT_VX(vmin_vx_b)
-GEN_VEXT_VX(vmin_vx_h)
-GEN_VEXT_VX(vmin_vx_w)
-GEN_VEXT_VX(vmin_vx_d)
-GEN_VEXT_VX(vmaxu_vx_b)
-GEN_VEXT_VX(vmaxu_vx_h)
-GEN_VEXT_VX(vmaxu_vx_w)
-GEN_VEXT_VX(vmaxu_vx_d)
-GEN_VEXT_VX(vmax_vx_b)
-GEN_VEXT_VX(vmax_vx_h)
-GEN_VEXT_VX(vmax_vx_w)
-GEN_VEXT_VX(vmax_vx_d)
+GEN_VEXT_VX(vminu_vx_b, 1)
+GEN_VEXT_VX(vminu_vx_h, 2)
+GEN_VEXT_VX(vminu_vx_w, 4)
+GEN_VEXT_VX(vminu_vx_d, 8)
+GEN_VEXT_VX(vmin_vx_b, 1)
+GEN_VEXT_VX(vmin_vx_h, 2)
+GEN_VEXT_VX(vmin_vx_w, 4)
+GEN_VEXT_VX(vmin_vx_d, 8)
+GEN_VEXT_VX(vmaxu_vx_b, 1)
+GEN_VEXT_VX(vmaxu_vx_h, 2)
+GEN_VEXT_VX(vmaxu_vx_w, 4)
+GEN_VEXT_VX(vmaxu_vx_d, 8)
+GEN_VEXT_VX(vmax_vx_b, 1)
+GEN_VEXT_VX(vmax_vx_h, 2)
+GEN_VEXT_VX(vmax_vx_w, 4)
+GEN_VEXT_VX(vmax_vx_d, 8)
 
 /* Vector Single-Width Integer Multiply Instructions */
 #define DO_MUL(N, M) (N * M)
@@ -1627,22 +1663,22 @@ RVVCALL(OPIVX2, vmulhsu_vx_b, OP_SUS_B, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVX2, vmulhsu_vx_h, OP_SUS_H, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVX2, vmulhsu_vx_w, OP_SUS_W, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVX2, vmulhsu_vx_d, OP_SUS_D, H8, H8, do_mulhsu_d)
-GEN_VEXT_VX(vmul_vx_b)
-GEN_VEXT_VX(vmul_vx_h)
-GEN_VEXT_VX(vmul_vx_w)
-GEN_VEXT_VX(vmul_vx_d)
-GEN_VEXT_VX(vmulh_vx_b)
-GEN_VEXT_VX(vmulh_vx_h)
-GEN_VEXT_VX(vmulh_vx_w)
-GEN_VEXT_VX(vmulh_vx_d)
-GEN_VEXT_VX(vmulhu_vx_b)
-GEN_VEXT_VX(vmulhu_vx_h)
-GEN_VEXT_VX(vmulhu_vx_w)
-GEN_VEXT_VX(vmulhu_vx_d)
-GEN_VEXT_VX(vmulhsu_vx_b)
-GEN_VEXT_VX(vmulhsu_vx_h)
-GEN_VEXT_VX(vmulhsu_vx_w)
-GEN_VEXT_VX(vmulhsu_vx_d)
+GEN_VEXT_VX(vmul_vx_b, 1)
+GEN_VEXT_VX(vmul_vx_h, 2)
+GEN_VEXT_VX(vmul_vx_w, 4)
+GEN_VEXT_VX(vmul_vx_d, 8)
+GEN_VEXT_VX(vmulh_vx_b, 1)
+GEN_VEXT_VX(vmulh_vx_h, 2)
+GEN_VEXT_VX(vmulh_vx_w, 4)
+GEN_VEXT_VX(vmulh_vx_d, 8)
+GEN_VEXT_VX(vmulhu_vx_b, 1)
+GEN_VEXT_VX(vmulhu_vx_h, 2)
+GEN_VEXT_VX(vmulhu_vx_w, 4)
+GEN_VEXT_VX(vmulhu_vx_d, 8)
+GEN_VEXT_VX(vmulhsu_vx_b, 1)
+GEN_VEXT_VX(vmulhsu_vx_h, 2)
+GEN_VEXT_VX(vmulhsu_vx_w, 4)
+GEN_VEXT_VX(vmulhsu_vx_d, 8)
 
 /* Vector Integer Divide Instructions */
 #define DO_DIVU(N, M) (unlikely(M == 0) ? (__typeof(N))(-1) : N / M)
@@ -1701,22 +1737,22 @@ RVVCALL(OPIVX2, vrem_vx_b, OP_SSS_B, H1, H1, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_h, OP_SSS_H, H2, H2, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_w, OP_SSS_W, H4, H4, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_d, OP_SSS_D, H8, H8, DO_REM)
-GEN_VEXT_VX(vdivu_vx_b)
-GEN_VEXT_VX(vdivu_vx_h)
-GEN_VEXT_VX(vdivu_vx_w)
-GEN_VEXT_VX(vdivu_vx_d)
-GEN_VEXT_VX(vdiv_vx_b)
-GEN_VEXT_VX(vdiv_vx_h)
-GEN_VEXT_VX(vdiv_vx_w)
-GEN_VEXT_VX(vdiv_vx_d)
-GEN_VEXT_VX(vremu_vx_b)
-GEN_VEXT_VX(vremu_vx_h)
-GEN_VEXT_VX(vremu_vx_w)
-GEN_VEXT_VX(vremu_vx_d)
-GEN_VEXT_VX(vrem_vx_b)
-GEN_VEXT_VX(vrem_vx_h)
-GEN_VEXT_VX(vrem_vx_w)
-GEN_VEXT_VX(vrem_vx_d)
+GEN_VEXT_VX(vdivu_vx_b, 1)
+GEN_VEXT_VX(vdivu_vx_h, 2)
+GEN_VEXT_VX(vdivu_vx_w, 4)
+GEN_VEXT_VX(vdivu_vx_d, 8)
+GEN_VEXT_VX(vdiv_vx_b, 1)
+GEN_VEXT_VX(vdiv_vx_h, 2)
+GEN_VEXT_VX(vdiv_vx_w, 4)
+GEN_VEXT_VX(vdiv_vx_d, 8)
+GEN_VEXT_VX(vremu_vx_b, 1)
+GEN_VEXT_VX(vremu_vx_h, 2)
+GEN_VEXT_VX(vremu_vx_w, 4)
+GEN_VEXT_VX(vremu_vx_d, 8)
+GEN_VEXT_VX(vrem_vx_b, 1)
+GEN_VEXT_VX(vrem_vx_h, 2)
+GEN_VEXT_VX(vrem_vx_w, 4)
+GEN_VEXT_VX(vrem_vx_d, 8)
 
 /* Vector Widening Integer Multiply Instructions */
 RVVCALL(OPIVV2, vwmul_vv_b, WOP_SSS_B, H2, H1, H1, DO_MUL)
@@ -1747,15 +1783,15 @@ RVVCALL(OPIVX2, vwmulu_vx_w, WOP_UUU_W, H8, H4, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_b, WOP_SUS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_h, WOP_SUS_H, H4, H2, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_w, WOP_SUS_W, H8, H4, DO_MUL)
-GEN_VEXT_VX(vwmul_vx_b)
-GEN_VEXT_VX(vwmul_vx_h)
-GEN_VEXT_VX(vwmul_vx_w)
-GEN_VEXT_VX(vwmulu_vx_b)
-GEN_VEXT_VX(vwmulu_vx_h)
-GEN_VEXT_VX(vwmulu_vx_w)
-GEN_VEXT_VX(vwmulsu_vx_b)
-GEN_VEXT_VX(vwmulsu_vx_h)
-GEN_VEXT_VX(vwmulsu_vx_w)
+GEN_VEXT_VX(vwmul_vx_b, 2)
+GEN_VEXT_VX(vwmul_vx_h, 4)
+GEN_VEXT_VX(vwmul_vx_w, 8)
+GEN_VEXT_VX(vwmulu_vx_b, 2)
+GEN_VEXT_VX(vwmulu_vx_h, 4)
+GEN_VEXT_VX(vwmulu_vx_w, 8)
+GEN_VEXT_VX(vwmulsu_vx_b, 2)
+GEN_VEXT_VX(vwmulsu_vx_h, 4)
+GEN_VEXT_VX(vwmulsu_vx_w, 8)
 
 /* Vector Single-Width Integer Multiply-Add Instructions */
 #define OPIVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)   \
@@ -1828,22 +1864,22 @@ RVVCALL(OPIVX3, vnmsub_vx_b, OP_SSS_B, H1, H1, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_h, OP_SSS_H, H2, H2, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_w, OP_SSS_W, H4, H4, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_d, OP_SSS_D, H8, H8, DO_NMSUB)
-GEN_VEXT_VX(vmacc_vx_b)
-GEN_VEXT_VX(vmacc_vx_h)
-GEN_VEXT_VX(vmacc_vx_w)
-GEN_VEXT_VX(vmacc_vx_d)
-GEN_VEXT_VX(vnmsac_vx_b)
-GEN_VEXT_VX(vnmsac_vx_h)
-GEN_VEXT_VX(vnmsac_vx_w)
-GEN_VEXT_VX(vnmsac_vx_d)
-GEN_VEXT_VX(vmadd_vx_b)
-GEN_VEXT_VX(vmadd_vx_h)
-GEN_VEXT_VX(vmadd_vx_w)
-GEN_VEXT_VX(vmadd_vx_d)
-GEN_VEXT_VX(vnmsub_vx_b)
-GEN_VEXT_VX(vnmsub_vx_h)
-GEN_VEXT_VX(vnmsub_vx_w)
-GEN_VEXT_VX(vnmsub_vx_d)
+GEN_VEXT_VX(vmacc_vx_b, 1)
+GEN_VEXT_VX(vmacc_vx_h, 2)
+GEN_VEXT_VX(vmacc_vx_w, 4)
+GEN_VEXT_VX(vmacc_vx_d, 8)
+GEN_VEXT_VX(vnmsac_vx_b, 1)
+GEN_VEXT_VX(vnmsac_vx_h, 2)
+GEN_VEXT_VX(vnmsac_vx_w, 4)
+GEN_VEXT_VX(vnmsac_vx_d, 8)
+GEN_VEXT_VX(vmadd_vx_b, 1)
+GEN_VEXT_VX(vmadd_vx_h, 2)
+GEN_VEXT_VX(vmadd_vx_w, 4)
+GEN_VEXT_VX(vmadd_vx_d, 8)
+GEN_VEXT_VX(vnmsub_vx_b, 1)
+GEN_VEXT_VX(vnmsub_vx_h, 2)
+GEN_VEXT_VX(vnmsub_vx_w, 4)
+GEN_VEXT_VX(vnmsub_vx_d, 8)
 
 /* Vector Widening Integer Multiply-Add Instructions */
 RVVCALL(OPIVV3, vwmaccu_vv_b, WOP_UUU_B, H2, H1, H1, DO_MACC)
@@ -1877,18 +1913,18 @@ RVVCALL(OPIVX3, vwmaccsu_vx_w, WOP_SSU_W, H8, H4, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_b, WOP_SUS_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_h, WOP_SUS_H, H4, H2, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_w, WOP_SUS_W, H8, H4, DO_MACC)
-GEN_VEXT_VX(vwmaccu_vx_b)
-GEN_VEXT_VX(vwmaccu_vx_h)
-GEN_VEXT_VX(vwmaccu_vx_w)
-GEN_VEXT_VX(vwmacc_vx_b)
-GEN_VEXT_VX(vwmacc_vx_h)
-GEN_VEXT_VX(vwmacc_vx_w)
-GEN_VEXT_VX(vwmaccsu_vx_b)
-GEN_VEXT_VX(vwmaccsu_vx_h)
-GEN_VEXT_VX(vwmaccsu_vx_w)
-GEN_VEXT_VX(vwmaccus_vx_b)
-GEN_VEXT_VX(vwmaccus_vx_h)
-GEN_VEXT_VX(vwmaccus_vx_w)
+GEN_VEXT_VX(vwmaccu_vx_b, 2)
+GEN_VEXT_VX(vwmaccu_vx_h, 4)
+GEN_VEXT_VX(vwmaccu_vx_w, 8)
+GEN_VEXT_VX(vwmacc_vx_b, 2)
+GEN_VEXT_VX(vwmacc_vx_h, 4)
+GEN_VEXT_VX(vwmacc_vx_w, 8)
+GEN_VEXT_VX(vwmaccsu_vx_b, 2)
+GEN_VEXT_VX(vwmaccsu_vx_h, 4)
+GEN_VEXT_VX(vwmaccsu_vx_w, 8)
+GEN_VEXT_VX(vwmaccus_vx_b, 2)
+GEN_VEXT_VX(vwmaccus_vx_h, 4)
+GEN_VEXT_VX(vwmaccus_vx_w, 8)
 
 /* Vector Integer Merge and Move Instructions */
 #define GEN_VEXT_VMV_VV(NAME, ETYPE, H)                              \
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 06/14] target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  23 ++
 target/riscv/vector_helper.c            | 312 +++++++++++++-----------
 2 files changed, 197 insertions(+), 138 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index fee616ef38..e8ef5b1e0e 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1296,6 +1296,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
 
     data = FIELD_DP32(data, VDATA, VM, vm);
     data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8,
                                       s->cfg_ptr->vlen / 8, data));
 
@@ -1332,6 +1333,15 @@ do_opivx_gvec(DisasContext *s, arg_rmrr *a, GVecGen2sFn *gvec_fn,
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         TCGv_i64 src1 = tcg_temp_new_i64();
 
         tcg_gen_ext_tl_i64(src1, get_gpr(s, a->rs1, EXT_SIGN));
@@ -1461,6 +1471,7 @@ static bool opivi_trans(uint32_t vd, uint32_t imm, uint32_t vs2, uint32_t vm,
 
     data = FIELD_DP32(data, VDATA, VM, vm);
     data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
     desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8,
                                       s->cfg_ptr->vlen / 8, data));
 
@@ -1490,6 +1501,15 @@ do_opivi_gvec(DisasContext *s, arg_rmrr *a, GVecGen2iFn *gvec_fn,
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         gvec_fn(s->sew, vreg_ofs(s, a->rd), vreg_ofs(s, a->rs2),
                 extract_imm(s, a->rs1, imm_mode), MAXSZ(s), MAXSZ(s));
         mark_vs_dirty(s);
@@ -1543,6 +1563,7 @@ static bool do_opivv_widen(DisasContext *s, arg_rmrr *a,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1),
                            vreg_ofs(s, a->rs2),
@@ -1624,6 +1645,7 @@ static bool do_opiwv_widen(DisasContext *s, arg_rmrr *a,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1),
                            vreg_ofs(s, a->rs2),
@@ -1702,6 +1724,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index c9e5a839a3..4be99e3be4 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -861,10 +861,12 @@ RVVCALL(OPIVX2, vrsub_vx_d, OP_SSS_D, H8, H8, DO_RSUB)
 
 static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       opivx2_fn fn)
+                       opivx2_fn fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
 
     for (i = env->vstart; i < vl; i++) {
@@ -874,30 +876,32 @@ static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
         fn(vd, s1, vs2, i);
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate the helpers for OPIVX */
-#define GEN_VEXT_VX(NAME)                                 \
+#define GEN_VEXT_VX(NAME, ESZ)                            \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     do_vext_vx(vd, v0, s1, vs2, env, desc,                \
-               do_##NAME);                                \
-}
-
-GEN_VEXT_VX(vadd_vx_b)
-GEN_VEXT_VX(vadd_vx_h)
-GEN_VEXT_VX(vadd_vx_w)
-GEN_VEXT_VX(vadd_vx_d)
-GEN_VEXT_VX(vsub_vx_b)
-GEN_VEXT_VX(vsub_vx_h)
-GEN_VEXT_VX(vsub_vx_w)
-GEN_VEXT_VX(vsub_vx_d)
-GEN_VEXT_VX(vrsub_vx_b)
-GEN_VEXT_VX(vrsub_vx_h)
-GEN_VEXT_VX(vrsub_vx_w)
-GEN_VEXT_VX(vrsub_vx_d)
+               do_##NAME, ESZ);                           \
+}
+
+GEN_VEXT_VX(vadd_vx_b, 1)
+GEN_VEXT_VX(vadd_vx_h, 2)
+GEN_VEXT_VX(vadd_vx_w, 4)
+GEN_VEXT_VX(vadd_vx_d, 8)
+GEN_VEXT_VX(vsub_vx_b, 1)
+GEN_VEXT_VX(vsub_vx_h, 2)
+GEN_VEXT_VX(vsub_vx_w, 4)
+GEN_VEXT_VX(vsub_vx_d, 8)
+GEN_VEXT_VX(vrsub_vx_b, 1)
+GEN_VEXT_VX(vrsub_vx_h, 2)
+GEN_VEXT_VX(vrsub_vx_w, 4)
+GEN_VEXT_VX(vrsub_vx_d, 8)
 
 void HELPER(vec_rsubs8)(void *d, void *a, uint64_t b, uint32_t desc)
 {
@@ -1025,30 +1029,30 @@ RVVCALL(OPIVX2, vwadd_wx_w, WOP_WSSS_W, H8, H4, DO_ADD)
 RVVCALL(OPIVX2, vwsub_wx_b, WOP_WSSS_B, H2, H1, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_h, WOP_WSSS_H, H4, H2, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_w, WOP_WSSS_W, H8, H4, DO_SUB)
-GEN_VEXT_VX(vwaddu_vx_b)
-GEN_VEXT_VX(vwaddu_vx_h)
-GEN_VEXT_VX(vwaddu_vx_w)
-GEN_VEXT_VX(vwsubu_vx_b)
-GEN_VEXT_VX(vwsubu_vx_h)
-GEN_VEXT_VX(vwsubu_vx_w)
-GEN_VEXT_VX(vwadd_vx_b)
-GEN_VEXT_VX(vwadd_vx_h)
-GEN_VEXT_VX(vwadd_vx_w)
-GEN_VEXT_VX(vwsub_vx_b)
-GEN_VEXT_VX(vwsub_vx_h)
-GEN_VEXT_VX(vwsub_vx_w)
-GEN_VEXT_VX(vwaddu_wx_b)
-GEN_VEXT_VX(vwaddu_wx_h)
-GEN_VEXT_VX(vwaddu_wx_w)
-GEN_VEXT_VX(vwsubu_wx_b)
-GEN_VEXT_VX(vwsubu_wx_h)
-GEN_VEXT_VX(vwsubu_wx_w)
-GEN_VEXT_VX(vwadd_wx_b)
-GEN_VEXT_VX(vwadd_wx_h)
-GEN_VEXT_VX(vwadd_wx_w)
-GEN_VEXT_VX(vwsub_wx_b)
-GEN_VEXT_VX(vwsub_wx_h)
-GEN_VEXT_VX(vwsub_wx_w)
+GEN_VEXT_VX(vwaddu_vx_b, 2)
+GEN_VEXT_VX(vwaddu_vx_h, 4)
+GEN_VEXT_VX(vwaddu_vx_w, 8)
+GEN_VEXT_VX(vwsubu_vx_b, 2)
+GEN_VEXT_VX(vwsubu_vx_h, 4)
+GEN_VEXT_VX(vwsubu_vx_w, 8)
+GEN_VEXT_VX(vwadd_vx_b, 2)
+GEN_VEXT_VX(vwadd_vx_h, 4)
+GEN_VEXT_VX(vwadd_vx_w, 8)
+GEN_VEXT_VX(vwsub_vx_b, 2)
+GEN_VEXT_VX(vwsub_vx_h, 4)
+GEN_VEXT_VX(vwsub_vx_w, 8)
+GEN_VEXT_VX(vwaddu_wx_b, 2)
+GEN_VEXT_VX(vwaddu_wx_h, 4)
+GEN_VEXT_VX(vwaddu_wx_w, 8)
+GEN_VEXT_VX(vwsubu_wx_b, 2)
+GEN_VEXT_VX(vwsubu_wx_h, 4)
+GEN_VEXT_VX(vwsubu_wx_w, 8)
+GEN_VEXT_VX(vwadd_wx_b, 2)
+GEN_VEXT_VX(vwadd_wx_h, 4)
+GEN_VEXT_VX(vwadd_wx_w, 8)
+GEN_VEXT_VX(vwsub_wx_b, 2)
+GEN_VEXT_VX(vwsub_wx_h, 4)
+GEN_VEXT_VX(vwsub_wx_w, 8)
 
 /* Vector Integer Add-with-Carry / Subtract-with-Borrow Instructions */
 #define DO_VADC(N, M, C) (N + M + C)
@@ -1059,6 +1063,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
     uint32_t vl = env->vl;                                    \
+    uint32_t esz = sizeof(ETYPE);                             \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -1069,6 +1077,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
         *((ETYPE *)vd + H(i)) = DO_OP(s2, s1, carry);         \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,   \
+                                     total_elems * esz);      \
 }
 
 GEN_VEXT_VADC_VVM(vadc_vvm_b, uint8_t,  H1, DO_VADC)
@@ -1086,6 +1097,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,        \
                   CPURISCVState *env, uint32_t desc)                     \
 {                                                                        \
     uint32_t vl = env->vl;                                               \
+    uint32_t esz = sizeof(ETYPE);                                        \
+    uint32_t total_elems =                                               \
+        vext_get_total_elems(env_archcpu(env), env->vtype);              \
+    uint32_t vta = vext_vta(desc);                                       \
     uint32_t i;                                                          \
                                                                          \
     for (i = env->vstart; i < vl; i++) {                                 \
@@ -1095,6 +1110,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,        \
         *((ETYPE *)vd + H(i)) = DO_OP(s2, (ETYPE)(target_long)s1, carry);\
     }                                                                    \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                                        \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,              \
+                                     total_elems * esz);                 \
 }
 
 GEN_VEXT_VADC_VXM(vadc_vxm_b, uint8_t,  H1, DO_VADC)
@@ -1117,6 +1135,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
 {                                                             \
     uint32_t vl = env->vl;                                    \
     uint32_t vm = vext_vm(desc);                              \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -1126,6 +1147,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
         vext_set_elem_mask(vd, i, DO_OP(s2, s1, carry));      \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    if (vta) {                                                \
+        for (; i < total_elems; i++) {                        \
+            vext_set_elem_mask(vd, i, 1);                     \
+        }                                                     \
+    }                                                         \
 }
 
 GEN_VEXT_VMADC_VVM(vmadc_vvm_b, uint8_t,  H1, DO_MADC)
@@ -1144,6 +1171,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,          \
 {                                                               \
     uint32_t vl = env->vl;                                      \
     uint32_t vm = vext_vm(desc);                                \
+    uint32_t total_elems =                                      \
+        vext_get_total_elems(env_archcpu(env), env->vtype);     \
+    uint32_t vta = vext_vta(desc);                              \
     uint32_t i;                                                 \
                                                                 \
     for (i = env->vstart; i < vl; i++) {                        \
@@ -1153,6 +1183,12 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,          \
                 DO_OP(s2, (ETYPE)(target_long)s1, carry));      \
     }                                                           \
     env->vstart = 0;                                            \
+    /* set tail elements to 1s */                               \
+    if (vta) {                                                  \
+        for (; i < total_elems; i++) {                          \
+            vext_set_elem_mask(vd, i, 1);                       \
+        }                                                       \
+    }                                                           \
 }
 
 GEN_VEXT_VMADC_VXM(vmadc_vxm_b, uint8_t,  H1, DO_MADC)
@@ -1203,18 +1239,18 @@ RVVCALL(OPIVX2, vxor_vx_b, OP_SSS_B, H1, H1, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_h, OP_SSS_H, H2, H2, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_w, OP_SSS_W, H4, H4, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_d, OP_SSS_D, H8, H8, DO_XOR)
-GEN_VEXT_VX(vand_vx_b)
-GEN_VEXT_VX(vand_vx_h)
-GEN_VEXT_VX(vand_vx_w)
-GEN_VEXT_VX(vand_vx_d)
-GEN_VEXT_VX(vor_vx_b)
-GEN_VEXT_VX(vor_vx_h)
-GEN_VEXT_VX(vor_vx_w)
-GEN_VEXT_VX(vor_vx_d)
-GEN_VEXT_VX(vxor_vx_b)
-GEN_VEXT_VX(vxor_vx_h)
-GEN_VEXT_VX(vxor_vx_w)
-GEN_VEXT_VX(vxor_vx_d)
+GEN_VEXT_VX(vand_vx_b, 1)
+GEN_VEXT_VX(vand_vx_h, 2)
+GEN_VEXT_VX(vand_vx_w, 4)
+GEN_VEXT_VX(vand_vx_d, 8)
+GEN_VEXT_VX(vor_vx_b, 1)
+GEN_VEXT_VX(vor_vx_h, 2)
+GEN_VEXT_VX(vor_vx_w, 4)
+GEN_VEXT_VX(vor_vx_d, 8)
+GEN_VEXT_VX(vxor_vx_b, 1)
+GEN_VEXT_VX(vxor_vx_h, 2)
+GEN_VEXT_VX(vxor_vx_w, 4)
+GEN_VEXT_VX(vxor_vx_d, 8)
 
 /* Vector Single-Width Bit Shift Instructions */
 #define DO_SLL(N, M)  (N << (M))
@@ -1468,22 +1504,22 @@ RVVCALL(OPIVX2, vmax_vx_b, OP_SSS_B, H1, H1, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_h, OP_SSS_H, H2, H2, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_w, OP_SSS_W, H4, H4, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_d, OP_SSS_D, H8, H8, DO_MAX)
-GEN_VEXT_VX(vminu_vx_b)
-GEN_VEXT_VX(vminu_vx_h)
-GEN_VEXT_VX(vminu_vx_w)
-GEN_VEXT_VX(vminu_vx_d)
-GEN_VEXT_VX(vmin_vx_b)
-GEN_VEXT_VX(vmin_vx_h)
-GEN_VEXT_VX(vmin_vx_w)
-GEN_VEXT_VX(vmin_vx_d)
-GEN_VEXT_VX(vmaxu_vx_b)
-GEN_VEXT_VX(vmaxu_vx_h)
-GEN_VEXT_VX(vmaxu_vx_w)
-GEN_VEXT_VX(vmaxu_vx_d)
-GEN_VEXT_VX(vmax_vx_b)
-GEN_VEXT_VX(vmax_vx_h)
-GEN_VEXT_VX(vmax_vx_w)
-GEN_VEXT_VX(vmax_vx_d)
+GEN_VEXT_VX(vminu_vx_b, 1)
+GEN_VEXT_VX(vminu_vx_h, 2)
+GEN_VEXT_VX(vminu_vx_w, 4)
+GEN_VEXT_VX(vminu_vx_d, 8)
+GEN_VEXT_VX(vmin_vx_b, 1)
+GEN_VEXT_VX(vmin_vx_h, 2)
+GEN_VEXT_VX(vmin_vx_w, 4)
+GEN_VEXT_VX(vmin_vx_d, 8)
+GEN_VEXT_VX(vmaxu_vx_b, 1)
+GEN_VEXT_VX(vmaxu_vx_h, 2)
+GEN_VEXT_VX(vmaxu_vx_w, 4)
+GEN_VEXT_VX(vmaxu_vx_d, 8)
+GEN_VEXT_VX(vmax_vx_b, 1)
+GEN_VEXT_VX(vmax_vx_h, 2)
+GEN_VEXT_VX(vmax_vx_w, 4)
+GEN_VEXT_VX(vmax_vx_d, 8)
 
 /* Vector Single-Width Integer Multiply Instructions */
 #define DO_MUL(N, M) (N * M)
@@ -1627,22 +1663,22 @@ RVVCALL(OPIVX2, vmulhsu_vx_b, OP_SUS_B, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVX2, vmulhsu_vx_h, OP_SUS_H, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVX2, vmulhsu_vx_w, OP_SUS_W, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVX2, vmulhsu_vx_d, OP_SUS_D, H8, H8, do_mulhsu_d)
-GEN_VEXT_VX(vmul_vx_b)
-GEN_VEXT_VX(vmul_vx_h)
-GEN_VEXT_VX(vmul_vx_w)
-GEN_VEXT_VX(vmul_vx_d)
-GEN_VEXT_VX(vmulh_vx_b)
-GEN_VEXT_VX(vmulh_vx_h)
-GEN_VEXT_VX(vmulh_vx_w)
-GEN_VEXT_VX(vmulh_vx_d)
-GEN_VEXT_VX(vmulhu_vx_b)
-GEN_VEXT_VX(vmulhu_vx_h)
-GEN_VEXT_VX(vmulhu_vx_w)
-GEN_VEXT_VX(vmulhu_vx_d)
-GEN_VEXT_VX(vmulhsu_vx_b)
-GEN_VEXT_VX(vmulhsu_vx_h)
-GEN_VEXT_VX(vmulhsu_vx_w)
-GEN_VEXT_VX(vmulhsu_vx_d)
+GEN_VEXT_VX(vmul_vx_b, 1)
+GEN_VEXT_VX(vmul_vx_h, 2)
+GEN_VEXT_VX(vmul_vx_w, 4)
+GEN_VEXT_VX(vmul_vx_d, 8)
+GEN_VEXT_VX(vmulh_vx_b, 1)
+GEN_VEXT_VX(vmulh_vx_h, 2)
+GEN_VEXT_VX(vmulh_vx_w, 4)
+GEN_VEXT_VX(vmulh_vx_d, 8)
+GEN_VEXT_VX(vmulhu_vx_b, 1)
+GEN_VEXT_VX(vmulhu_vx_h, 2)
+GEN_VEXT_VX(vmulhu_vx_w, 4)
+GEN_VEXT_VX(vmulhu_vx_d, 8)
+GEN_VEXT_VX(vmulhsu_vx_b, 1)
+GEN_VEXT_VX(vmulhsu_vx_h, 2)
+GEN_VEXT_VX(vmulhsu_vx_w, 4)
+GEN_VEXT_VX(vmulhsu_vx_d, 8)
 
 /* Vector Integer Divide Instructions */
 #define DO_DIVU(N, M) (unlikely(M == 0) ? (__typeof(N))(-1) : N / M)
@@ -1701,22 +1737,22 @@ RVVCALL(OPIVX2, vrem_vx_b, OP_SSS_B, H1, H1, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_h, OP_SSS_H, H2, H2, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_w, OP_SSS_W, H4, H4, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_d, OP_SSS_D, H8, H8, DO_REM)
-GEN_VEXT_VX(vdivu_vx_b)
-GEN_VEXT_VX(vdivu_vx_h)
-GEN_VEXT_VX(vdivu_vx_w)
-GEN_VEXT_VX(vdivu_vx_d)
-GEN_VEXT_VX(vdiv_vx_b)
-GEN_VEXT_VX(vdiv_vx_h)
-GEN_VEXT_VX(vdiv_vx_w)
-GEN_VEXT_VX(vdiv_vx_d)
-GEN_VEXT_VX(vremu_vx_b)
-GEN_VEXT_VX(vremu_vx_h)
-GEN_VEXT_VX(vremu_vx_w)
-GEN_VEXT_VX(vremu_vx_d)
-GEN_VEXT_VX(vrem_vx_b)
-GEN_VEXT_VX(vrem_vx_h)
-GEN_VEXT_VX(vrem_vx_w)
-GEN_VEXT_VX(vrem_vx_d)
+GEN_VEXT_VX(vdivu_vx_b, 1)
+GEN_VEXT_VX(vdivu_vx_h, 2)
+GEN_VEXT_VX(vdivu_vx_w, 4)
+GEN_VEXT_VX(vdivu_vx_d, 8)
+GEN_VEXT_VX(vdiv_vx_b, 1)
+GEN_VEXT_VX(vdiv_vx_h, 2)
+GEN_VEXT_VX(vdiv_vx_w, 4)
+GEN_VEXT_VX(vdiv_vx_d, 8)
+GEN_VEXT_VX(vremu_vx_b, 1)
+GEN_VEXT_VX(vremu_vx_h, 2)
+GEN_VEXT_VX(vremu_vx_w, 4)
+GEN_VEXT_VX(vremu_vx_d, 8)
+GEN_VEXT_VX(vrem_vx_b, 1)
+GEN_VEXT_VX(vrem_vx_h, 2)
+GEN_VEXT_VX(vrem_vx_w, 4)
+GEN_VEXT_VX(vrem_vx_d, 8)
 
 /* Vector Widening Integer Multiply Instructions */
 RVVCALL(OPIVV2, vwmul_vv_b, WOP_SSS_B, H2, H1, H1, DO_MUL)
@@ -1747,15 +1783,15 @@ RVVCALL(OPIVX2, vwmulu_vx_w, WOP_UUU_W, H8, H4, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_b, WOP_SUS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_h, WOP_SUS_H, H4, H2, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_w, WOP_SUS_W, H8, H4, DO_MUL)
-GEN_VEXT_VX(vwmul_vx_b)
-GEN_VEXT_VX(vwmul_vx_h)
-GEN_VEXT_VX(vwmul_vx_w)
-GEN_VEXT_VX(vwmulu_vx_b)
-GEN_VEXT_VX(vwmulu_vx_h)
-GEN_VEXT_VX(vwmulu_vx_w)
-GEN_VEXT_VX(vwmulsu_vx_b)
-GEN_VEXT_VX(vwmulsu_vx_h)
-GEN_VEXT_VX(vwmulsu_vx_w)
+GEN_VEXT_VX(vwmul_vx_b, 2)
+GEN_VEXT_VX(vwmul_vx_h, 4)
+GEN_VEXT_VX(vwmul_vx_w, 8)
+GEN_VEXT_VX(vwmulu_vx_b, 2)
+GEN_VEXT_VX(vwmulu_vx_h, 4)
+GEN_VEXT_VX(vwmulu_vx_w, 8)
+GEN_VEXT_VX(vwmulsu_vx_b, 2)
+GEN_VEXT_VX(vwmulsu_vx_h, 4)
+GEN_VEXT_VX(vwmulsu_vx_w, 8)
 
 /* Vector Single-Width Integer Multiply-Add Instructions */
 #define OPIVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)   \
@@ -1828,22 +1864,22 @@ RVVCALL(OPIVX3, vnmsub_vx_b, OP_SSS_B, H1, H1, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_h, OP_SSS_H, H2, H2, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_w, OP_SSS_W, H4, H4, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_d, OP_SSS_D, H8, H8, DO_NMSUB)
-GEN_VEXT_VX(vmacc_vx_b)
-GEN_VEXT_VX(vmacc_vx_h)
-GEN_VEXT_VX(vmacc_vx_w)
-GEN_VEXT_VX(vmacc_vx_d)
-GEN_VEXT_VX(vnmsac_vx_b)
-GEN_VEXT_VX(vnmsac_vx_h)
-GEN_VEXT_VX(vnmsac_vx_w)
-GEN_VEXT_VX(vnmsac_vx_d)
-GEN_VEXT_VX(vmadd_vx_b)
-GEN_VEXT_VX(vmadd_vx_h)
-GEN_VEXT_VX(vmadd_vx_w)
-GEN_VEXT_VX(vmadd_vx_d)
-GEN_VEXT_VX(vnmsub_vx_b)
-GEN_VEXT_VX(vnmsub_vx_h)
-GEN_VEXT_VX(vnmsub_vx_w)
-GEN_VEXT_VX(vnmsub_vx_d)
+GEN_VEXT_VX(vmacc_vx_b, 1)
+GEN_VEXT_VX(vmacc_vx_h, 2)
+GEN_VEXT_VX(vmacc_vx_w, 4)
+GEN_VEXT_VX(vmacc_vx_d, 8)
+GEN_VEXT_VX(vnmsac_vx_b, 1)
+GEN_VEXT_VX(vnmsac_vx_h, 2)
+GEN_VEXT_VX(vnmsac_vx_w, 4)
+GEN_VEXT_VX(vnmsac_vx_d, 8)
+GEN_VEXT_VX(vmadd_vx_b, 1)
+GEN_VEXT_VX(vmadd_vx_h, 2)
+GEN_VEXT_VX(vmadd_vx_w, 4)
+GEN_VEXT_VX(vmadd_vx_d, 8)
+GEN_VEXT_VX(vnmsub_vx_b, 1)
+GEN_VEXT_VX(vnmsub_vx_h, 2)
+GEN_VEXT_VX(vnmsub_vx_w, 4)
+GEN_VEXT_VX(vnmsub_vx_d, 8)
 
 /* Vector Widening Integer Multiply-Add Instructions */
 RVVCALL(OPIVV3, vwmaccu_vv_b, WOP_UUU_B, H2, H1, H1, DO_MACC)
@@ -1877,18 +1913,18 @@ RVVCALL(OPIVX3, vwmaccsu_vx_w, WOP_SSU_W, H8, H4, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_b, WOP_SUS_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_h, WOP_SUS_H, H4, H2, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_w, WOP_SUS_W, H8, H4, DO_MACC)
-GEN_VEXT_VX(vwmaccu_vx_b)
-GEN_VEXT_VX(vwmaccu_vx_h)
-GEN_VEXT_VX(vwmaccu_vx_w)
-GEN_VEXT_VX(vwmacc_vx_b)
-GEN_VEXT_VX(vwmacc_vx_h)
-GEN_VEXT_VX(vwmacc_vx_w)
-GEN_VEXT_VX(vwmaccsu_vx_b)
-GEN_VEXT_VX(vwmaccsu_vx_h)
-GEN_VEXT_VX(vwmaccsu_vx_w)
-GEN_VEXT_VX(vwmaccus_vx_b)
-GEN_VEXT_VX(vwmaccus_vx_h)
-GEN_VEXT_VX(vwmaccus_vx_w)
+GEN_VEXT_VX(vwmaccu_vx_b, 2)
+GEN_VEXT_VX(vwmaccu_vx_h, 4)
+GEN_VEXT_VX(vwmaccu_vx_w, 8)
+GEN_VEXT_VX(vwmacc_vx_b, 2)
+GEN_VEXT_VX(vwmacc_vx_h, 4)
+GEN_VEXT_VX(vwmacc_vx_w, 8)
+GEN_VEXT_VX(vwmaccsu_vx_b, 2)
+GEN_VEXT_VX(vwmaccsu_vx_h, 4)
+GEN_VEXT_VX(vwmaccsu_vx_w, 8)
+GEN_VEXT_VX(vwmaccus_vx_b, 2)
+GEN_VEXT_VX(vwmaccus_vx_h, 4)
+GEN_VEXT_VX(vwmaccus_vx_w, 8)
 
 /* Vector Integer Merge and Move Instructions */
 #define GEN_VEXT_VMV_VV(NAME, ETYPE, H)                              \
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 06/14] target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions
  2022-03-24 19:03   ` ~eopxd
@ 2022-03-25 10:32     ` Weiwei Li
  -1 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:32 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, Bin Meng, Alistair Francis, eop Chen,
	Palmer Dabbelt


在 2022/3/7 下午3:32, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc |  23 ++
>   target/riscv/vector_helper.c            | 312 +++++++++++++-----------
>   2 files changed, 197 insertions(+), 138 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index fee616ef38..e8ef5b1e0e 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -1296,6 +1296,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
>   
>       data = FIELD_DP32(data, VDATA, VM, vm);
>       data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +    data = FIELD_DP32(data, VDATA, VTA, s->vta);
>       desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8,
>                                         s->cfg_ptr->vlen / 8, data));
>   
> @@ -1332,6 +1333,15 @@ do_opivx_gvec(DisasContext *s, arg_rmrr *a, GVecGen2sFn *gvec_fn,
>       }
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */

It seems better to as follows:

+            /*
+             * tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
Regards,
Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 06/14] target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions
@ 2022-03-25 10:32     ` Weiwei Li
  0 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:32 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt


在 2022/3/7 下午3:32, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc |  23 ++
>   target/riscv/vector_helper.c            | 312 +++++++++++++-----------
>   2 files changed, 197 insertions(+), 138 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index fee616ef38..e8ef5b1e0e 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -1296,6 +1296,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
>   
>       data = FIELD_DP32(data, VDATA, VM, vm);
>       data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +    data = FIELD_DP32(data, VDATA, VTA, s->vta);
>       desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8,
>                                         s->cfg_ptr->vlen / 8, data));
>   
> @@ -1332,6 +1333,15 @@ do_opivx_gvec(DisasContext *s, arg_rmrr *a, GVecGen2sFn *gvec_fn,
>       }
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */

It seems better to as follows:

+            /*
+             * tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
Regards,
Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 07/14] target/riscv: rvv: Add tail agnostic for vector integer shift instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07  9:38 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 10 ++++++++++
 target/riscv/vector_helper.c            | 14 ++++++++++++++
 2 files changed, 24 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index e8ef5b1e0e..ccbc55a2ab 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1849,6 +1849,15 @@ do_opivx_gvec_shift(DisasContext *s, arg_rmrr *a, GVecGen2sFn32 *gvec_fn,
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         TCGv_i32 src1 = tcg_temp_new_i32();
 
         tcg_gen_trunc_tl_i32(src1, get_gpr(s, a->rs1, EXT_NONE));
@@ -1907,6 +1916,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 4be99e3be4..9dedd1e29d 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -1263,6 +1263,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,                          \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(TS1);                                           \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint32_t i;                                                           \
                                                                           \
     for (i = env->vstart; i < vl; i++) {                                  \
@@ -1274,6 +1278,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,                          \
         *((TS1 *)vd + HS1(i)) = OP(s2, s1 & MASK);                        \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 GEN_VEXT_SHIFT_VV(vsll_vv_b, uint8_t,  uint8_t, H1, H1, DO_SLL, 0x7)
@@ -1298,6 +1305,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,      \
 {                                                           \
     uint32_t vm = vext_vm(desc);                            \
     uint32_t vl = env->vl;                                  \
+    uint32_t esz = sizeof(TD);                              \
+    uint32_t total_elems =                                  \
+        vext_get_total_elems(env_archcpu(env), env->vtype); \
+    uint32_t vta = vext_vta(desc);                          \
     uint32_t i;                                             \
                                                             \
     for (i = env->vstart; i < vl; i++) {                    \
@@ -1308,6 +1319,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,      \
         *((TD *)vd + HD(i)) = OP(s2, s1 & MASK);            \
     }                                                       \
     env->vstart = 0;                                        \
+    /* set tail elements to 1s */                           \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, \
+                                     total_elems * esz);    \
 }
 
 GEN_VEXT_SHIFT_VX(vsll_vx_b, uint8_t, int8_t, H1, H1, DO_SLL, 0x7)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 07/14] target/riscv: rvv: Add tail agnostic for vector integer shift instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 10 ++++++++++
 target/riscv/vector_helper.c            | 14 ++++++++++++++
 2 files changed, 24 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index e8ef5b1e0e..ccbc55a2ab 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1849,6 +1849,15 @@ do_opivx_gvec_shift(DisasContext *s, arg_rmrr *a, GVecGen2sFn32 *gvec_fn,
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         TCGv_i32 src1 = tcg_temp_new_i32();
 
         tcg_gen_trunc_tl_i32(src1, get_gpr(s, a->rs1, EXT_NONE));
@@ -1907,6 +1916,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 4be99e3be4..9dedd1e29d 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -1263,6 +1263,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,                          \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(TS1);                                           \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint32_t i;                                                           \
                                                                           \
     for (i = env->vstart; i < vl; i++) {                                  \
@@ -1274,6 +1278,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,                          \
         *((TS1 *)vd + HS1(i)) = OP(s2, s1 & MASK);                        \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 GEN_VEXT_SHIFT_VV(vsll_vv_b, uint8_t,  uint8_t, H1, H1, DO_SLL, 0x7)
@@ -1298,6 +1305,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,      \
 {                                                           \
     uint32_t vm = vext_vm(desc);                            \
     uint32_t vl = env->vl;                                  \
+    uint32_t esz = sizeof(TD);                              \
+    uint32_t total_elems =                                  \
+        vext_get_total_elems(env_archcpu(env), env->vtype); \
+    uint32_t vta = vext_vta(desc);                          \
     uint32_t i;                                             \
                                                             \
     for (i = env->vstart; i < vl; i++) {                    \
@@ -1308,6 +1319,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,      \
         *((TD *)vd + HD(i)) = OP(s2, s1 & MASK);            \
     }                                                       \
     env->vstart = 0;                                        \
+    /* set tail elements to 1s */                           \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, \
+                                     total_elems * esz);    \
 }
 
 GEN_VEXT_SHIFT_VX(vsll_vx_b, uint8_t, int8_t, H1, H1, DO_SLL, 0x7)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 08/14] target/riscv: rvv: Add tail agnostic for vector integer comparison instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07  9:43 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 9dedd1e29d..a972946bc2 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -1366,6 +1366,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
 {                                                             \
     uint32_t vm = vext_vm(desc);                              \
     uint32_t vl = env->vl;                                    \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -1377,6 +1380,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
         vext_set_elem_mask(vd, i, DO_OP(s2, s1));             \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    if (vta) {                                                \
+        for (; i < total_elems; i++) {                        \
+            vext_set_elem_mask(vd, i, 1);                     \
+        }                                                     \
+    }                                                         \
 }
 
 GEN_VEXT_CMP_VV(vmseq_vv_b, uint8_t,  H1, DO_MSEQ)
@@ -1415,6 +1424,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,   \
 {                                                                   \
     uint32_t vm = vext_vm(desc);                                    \
     uint32_t vl = env->vl;                                          \
+    uint32_t total_elems =                                          \
+        vext_get_total_elems(env_archcpu(env), env->vtype);         \
+    uint32_t vta = vext_vta(desc);                                  \
     uint32_t i;                                                     \
                                                                     \
     for (i = env->vstart; i < vl; i++) {                            \
@@ -1426,6 +1438,12 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,   \
                 DO_OP(s2, (ETYPE)(target_long)s1));                 \
     }                                                               \
     env->vstart = 0;                                                \
+    /* set tail elements to 1s */                                   \
+    if (vta) {                                                      \
+        for (; i < total_elems; i++) {                              \
+            vext_set_elem_mask(vd, i, 1);                           \
+        }                                                           \
+    }                                                               \
 }
 
 GEN_VEXT_CMP_VX(vmseq_vx_b, uint8_t,  H1, DO_MSEQ)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 08/14] target/riscv: rvv: Add tail agnostic for vector integer comparison instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 9dedd1e29d..a972946bc2 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -1366,6 +1366,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
 {                                                             \
     uint32_t vm = vext_vm(desc);                              \
     uint32_t vl = env->vl;                                    \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -1377,6 +1380,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
         vext_set_elem_mask(vd, i, DO_OP(s2, s1));             \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    if (vta) {                                                \
+        for (; i < total_elems; i++) {                        \
+            vext_set_elem_mask(vd, i, 1);                     \
+        }                                                     \
+    }                                                         \
 }
 
 GEN_VEXT_CMP_VV(vmseq_vv_b, uint8_t,  H1, DO_MSEQ)
@@ -1415,6 +1424,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,   \
 {                                                                   \
     uint32_t vm = vext_vm(desc);                                    \
     uint32_t vl = env->vl;                                          \
+    uint32_t total_elems =                                          \
+        vext_get_total_elems(env_archcpu(env), env->vtype);         \
+    uint32_t vta = vext_vta(desc);                                  \
     uint32_t i;                                                     \
                                                                     \
     for (i = env->vstart; i < vl; i++) {                            \
@@ -1426,6 +1438,12 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,   \
                 DO_OP(s2, (ETYPE)(target_long)s1));                 \
     }                                                               \
     env->vstart = 0;                                                \
+    /* set tail elements to 1s */                                   \
+    if (vta) {                                                      \
+        for (; i < total_elems; i++) {                              \
+            vext_set_elem_mask(vd, i, 1);                           \
+        }                                                           \
+    }                                                               \
 }
 
 GEN_VEXT_CMP_VX(vmseq_vx_b, uint8_t,  H1, DO_MSEQ)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 09/14] target/riscv: rvv: Add tail agnostic for vector integer merge and move instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07  9:53 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 40 +++++++++++++++++++++++++
 target/riscv/vector_helper.c            | 28 +++++++++++++++++
 2 files changed, 68 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index ccbc55a2ab..24784b6f5f 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -2113,11 +2113,21 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
         /* vmv.v.v has rs2 = 0 and vm = 1 */
         vext_check_sss(s, a->rd, a->rs1, 0, 1)) {
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             tcg_gen_gvec_mov(s->sew, vreg_ofs(s, a->rd),
                              vreg_ofs(s, a->rs1),
                              MAXSZ(s), MAXSZ(s));
         } else {
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_gvec_2_ptr * const fns[4] = {
                 gen_helper_vmv_v_v_b, gen_helper_vmv_v_v_h,
                 gen_helper_vmv_v_v_w, gen_helper_vmv_v_v_d,
@@ -2153,6 +2163,15 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
         s1 = get_gpr(s, a->rs1, EXT_SIGN);
 
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             tcg_gen_gvec_dup_tl(s->sew, vreg_ofs(s, a->rd),
                                 MAXSZ(s), MAXSZ(s), s1);
         } else {
@@ -2160,6 +2179,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
             TCGv_i64 s1_i64 = tcg_temp_new_i64();
             TCGv_ptr dest = tcg_temp_new_ptr();
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_vmv_vx * const fns[4] = {
                 gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
                 gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
@@ -2190,6 +2210,15 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
         vext_check_ss(s, a->rd, 0, 1)) {
         int64_t simm = sextract64(a->rs1, 0, 5);
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             tcg_gen_gvec_dup_imm(s->sew, vreg_ofs(s, a->rd),
                                  MAXSZ(s), MAXSZ(s), simm);
             mark_vs_dirty(s);
@@ -2198,6 +2227,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
             TCGv_i64 s1;
             TCGv_ptr dest;
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_vmv_vx * const fns[4] = {
                 gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
                 gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
@@ -2770,6 +2800,15 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
         TCGv_i64 t1;
 
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             t1 = tcg_temp_new_i64();
             /* NaN-box f[rs1] */
             do_nanbox(s, t1, cpu_fpr[a->rs1]);
@@ -2781,6 +2820,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
             TCGv_ptr dest;
             TCGv_i32 desc;
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_vmv_vx * const fns[3] = {
                 gen_helper_vmv_v_x_h,
                 gen_helper_vmv_v_x_w,
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index a972946bc2..89927a424f 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -1964,6 +1964,10 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
                   uint32_t desc)                                     \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
@@ -1971,6 +1975,9 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
         *((ETYPE *)vd + H(i)) = s1;                                  \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMV_VV(vmv_v_v_b, int8_t,  H1)
@@ -1983,12 +1990,19 @@ void HELPER(NAME)(void *vd, uint64_t s1, CPURISCVState *env,         \
                   uint32_t desc)                                     \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
         *((ETYPE *)vd + H(i)) = (ETYPE)s1;                           \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMV_VX(vmv_v_x_b, int8_t,  H1)
@@ -2001,6 +2015,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
                   CPURISCVState *env, uint32_t desc)                 \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
@@ -2008,6 +2026,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
         *((ETYPE *)vd + H(i)) = *(vt + H(i));                        \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMERGE_VV(vmerge_vvm_b, int8_t,  H1)
@@ -2020,6 +2041,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
                   void *vs2, CPURISCVState *env, uint32_t desc)      \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
@@ -2029,6 +2054,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
         *((ETYPE *)vd + H(i)) = d;                                   \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMERGE_VX(vmerge_vxm_b, int8_t,  H1)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 09/14] target/riscv: rvv: Add tail agnostic for vector integer merge and move instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 40 +++++++++++++++++++++++++
 target/riscv/vector_helper.c            | 28 +++++++++++++++++
 2 files changed, 68 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index ccbc55a2ab..24784b6f5f 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -2113,11 +2113,21 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
         /* vmv.v.v has rs2 = 0 and vm = 1 */
         vext_check_sss(s, a->rd, a->rs1, 0, 1)) {
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             tcg_gen_gvec_mov(s->sew, vreg_ofs(s, a->rd),
                              vreg_ofs(s, a->rs1),
                              MAXSZ(s), MAXSZ(s));
         } else {
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_gvec_2_ptr * const fns[4] = {
                 gen_helper_vmv_v_v_b, gen_helper_vmv_v_v_h,
                 gen_helper_vmv_v_v_w, gen_helper_vmv_v_v_d,
@@ -2153,6 +2163,15 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
         s1 = get_gpr(s, a->rs1, EXT_SIGN);
 
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             tcg_gen_gvec_dup_tl(s->sew, vreg_ofs(s, a->rd),
                                 MAXSZ(s), MAXSZ(s), s1);
         } else {
@@ -2160,6 +2179,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
             TCGv_i64 s1_i64 = tcg_temp_new_i64();
             TCGv_ptr dest = tcg_temp_new_ptr();
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_vmv_vx * const fns[4] = {
                 gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
                 gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
@@ -2190,6 +2210,15 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
         vext_check_ss(s, a->rd, 0, 1)) {
         int64_t simm = sextract64(a->rs1, 0, 5);
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             tcg_gen_gvec_dup_imm(s->sew, vreg_ofs(s, a->rd),
                                  MAXSZ(s), MAXSZ(s), simm);
             mark_vs_dirty(s);
@@ -2198,6 +2227,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
             TCGv_i64 s1;
             TCGv_ptr dest;
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_vmv_vx * const fns[4] = {
                 gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
                 gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
@@ -2770,6 +2800,15 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
         TCGv_i64 t1;
 
         if (s->vl_eq_vlmax) {
+            if (s->vta && s->lmul < 0) {
+                /* tail elements may pass vlmax when lmul < 0
+                * set tail elements to 1s
+                */
+                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                                vreg_ofs(s, a->rd), -1,
+                                vlenb, vlenb);
+            }
             t1 = tcg_temp_new_i64();
             /* NaN-box f[rs1] */
             do_nanbox(s, t1, cpu_fpr[a->rs1]);
@@ -2781,6 +2820,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
             TCGv_ptr dest;
             TCGv_i32 desc;
             uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
+            data = FIELD_DP32(data, VDATA, VTA, s->vta);
             static gen_helper_vmv_vx * const fns[3] = {
                 gen_helper_vmv_v_x_h,
                 gen_helper_vmv_v_x_w,
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index a972946bc2..89927a424f 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -1964,6 +1964,10 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
                   uint32_t desc)                                     \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
@@ -1971,6 +1975,9 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
         *((ETYPE *)vd + H(i)) = s1;                                  \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMV_VV(vmv_v_v_b, int8_t,  H1)
@@ -1983,12 +1990,19 @@ void HELPER(NAME)(void *vd, uint64_t s1, CPURISCVState *env,         \
                   uint32_t desc)                                     \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
         *((ETYPE *)vd + H(i)) = (ETYPE)s1;                           \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMV_VX(vmv_v_x_b, int8_t,  H1)
@@ -2001,6 +2015,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
                   CPURISCVState *env, uint32_t desc)                 \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
@@ -2008,6 +2026,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
         *((ETYPE *)vd + H(i)) = *(vt + H(i));                        \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMERGE_VV(vmerge_vvm_b, int8_t,  H1)
@@ -2020,6 +2041,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
                   void *vs2, CPURISCVState *env, uint32_t desc)      \
 {                                                                    \
     uint32_t vl = env->vl;                                           \
+    uint32_t esz = sizeof(ETYPE);                                    \
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
+                                         env->vtype);                \
+    uint32_t vta = vext_vta(desc);                                   \
     uint32_t i;                                                      \
                                                                      \
     for (i = env->vstart; i < vl; i++) {                             \
@@ -2029,6 +2054,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
         *((ETYPE *)vd + H(i)) = d;                                   \
     }                                                                \
     env->vstart = 0;                                                 \
+    /* set tail elements to 1s */                                    \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
+                                     total_elems * esz);             \
 }
 
 GEN_VEXT_VMERGE_VX(vmerge_vxm_b, int8_t,  H1)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 09/14] target/riscv: rvv: Add tail agnostic for vector integer merge and move instructions
  2022-03-24 19:03   ` ~eopxd
@ 2022-03-25 10:38     ` Weiwei Li
  -1 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:38 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, Bin Meng, Alistair Francis, eop Chen,
	Palmer Dabbelt


在 2022/3/7 下午5:53, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc | 40 +++++++++++++++++++++++++
>   target/riscv/vector_helper.c            | 28 +++++++++++++++++
>   2 files changed, 68 insertions(+)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index ccbc55a2ab..24784b6f5f 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -2113,11 +2113,21 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
>           /* vmv.v.v has rs2 = 0 and vm = 1 */
>           vext_check_sss(s, a->rd, a->rs1, 0, 1)) {
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
similar to patch 6.
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
Not aligned here.
> +            }
>               tcg_gen_gvec_mov(s->sew, vreg_ofs(s, a->rd),
>                                vreg_ofs(s, a->rs1),
>                                MAXSZ(s), MAXSZ(s));
>           } else {
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_gvec_2_ptr * const fns[4] = {
>                   gen_helper_vmv_v_v_b, gen_helper_vmv_v_v_h,
>                   gen_helper_vmv_v_v_w, gen_helper_vmv_v_v_d,
> @@ -2153,6 +2163,15 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
>           s1 = get_gpr(s, a->rs1, EXT_SIGN);
>   
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
Not aligned too.
> +            }
>               tcg_gen_gvec_dup_tl(s->sew, vreg_ofs(s, a->rd),
>                                   MAXSZ(s), MAXSZ(s), s1);
>           } else {
> @@ -2160,6 +2179,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
>               TCGv_i64 s1_i64 = tcg_temp_new_i64();
>               TCGv_ptr dest = tcg_temp_new_ptr();
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_vmv_vx * const fns[4] = {
>                   gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
>                   gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
> @@ -2190,6 +2210,15 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
>           vext_check_ss(s, a->rd, 0, 1)) {
>           int64_t simm = sextract64(a->rs1, 0, 5);
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
here too.
> +            }
>               tcg_gen_gvec_dup_imm(s->sew, vreg_ofs(s, a->rd),
>                                    MAXSZ(s), MAXSZ(s), simm);
>               mark_vs_dirty(s);
> @@ -2198,6 +2227,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
>               TCGv_i64 s1;
>               TCGv_ptr dest;
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_vmv_vx * const fns[4] = {
>                   gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
>                   gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
> @@ -2770,6 +2800,15 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
>           TCGv_i64 t1;
>   
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
here too.
> +            }
>               t1 = tcg_temp_new_i64();
>               /* NaN-box f[rs1] */
>               do_nanbox(s, t1, cpu_fpr[a->rs1]);
> @@ -2781,6 +2820,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
>               TCGv_ptr dest;
>               TCGv_i32 desc;
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_vmv_vx * const fns[3] = {
>                   gen_helper_vmv_v_x_h,
>                   gen_helper_vmv_v_x_w,
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index a972946bc2..89927a424f 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -1964,6 +1964,10 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
>                     uint32_t desc)                                     \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
> @@ -1971,6 +1975,9 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
>           *((ETYPE *)vd + H(i)) = s1;                                  \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMV_VV(vmv_v_v_b, int8_t,  H1)
> @@ -1983,12 +1990,19 @@ void HELPER(NAME)(void *vd, uint64_t s1, CPURISCVState *env,         \
>                     uint32_t desc)                                     \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
>           *((ETYPE *)vd + H(i)) = (ETYPE)s1;                           \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMV_VX(vmv_v_x_b, int8_t,  H1)
> @@ -2001,6 +2015,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
>                     CPURISCVState *env, uint32_t desc)                 \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
> @@ -2008,6 +2026,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
>           *((ETYPE *)vd + H(i)) = *(vt + H(i));                        \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMERGE_VV(vmerge_vvm_b, int8_t,  H1)
> @@ -2020,6 +2041,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
>                     void *vs2, CPURISCVState *env, uint32_t desc)      \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
better aligned here too.
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
> @@ -2029,6 +2054,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
>           *((ETYPE *)vd + H(i)) = d;                                   \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMERGE_VX(vmerge_vxm_b, int8_t,  H1)

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 09/14] target/riscv: rvv: Add tail agnostic for vector integer merge and move instructions
@ 2022-03-25 10:38     ` Weiwei Li
  0 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:38 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt


在 2022/3/7 下午5:53, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc | 40 +++++++++++++++++++++++++
>   target/riscv/vector_helper.c            | 28 +++++++++++++++++
>   2 files changed, 68 insertions(+)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index ccbc55a2ab..24784b6f5f 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -2113,11 +2113,21 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
>           /* vmv.v.v has rs2 = 0 and vm = 1 */
>           vext_check_sss(s, a->rd, a->rs1, 0, 1)) {
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
similar to patch 6.
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
Not aligned here.
> +            }
>               tcg_gen_gvec_mov(s->sew, vreg_ofs(s, a->rd),
>                                vreg_ofs(s, a->rs1),
>                                MAXSZ(s), MAXSZ(s));
>           } else {
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_gvec_2_ptr * const fns[4] = {
>                   gen_helper_vmv_v_v_b, gen_helper_vmv_v_v_h,
>                   gen_helper_vmv_v_v_w, gen_helper_vmv_v_v_d,
> @@ -2153,6 +2163,15 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
>           s1 = get_gpr(s, a->rs1, EXT_SIGN);
>   
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
Not aligned too.
> +            }
>               tcg_gen_gvec_dup_tl(s->sew, vreg_ofs(s, a->rd),
>                                   MAXSZ(s), MAXSZ(s), s1);
>           } else {
> @@ -2160,6 +2179,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
>               TCGv_i64 s1_i64 = tcg_temp_new_i64();
>               TCGv_ptr dest = tcg_temp_new_ptr();
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_vmv_vx * const fns[4] = {
>                   gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
>                   gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
> @@ -2190,6 +2210,15 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
>           vext_check_ss(s, a->rd, 0, 1)) {
>           int64_t simm = sextract64(a->rs1, 0, 5);
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
here too.
> +            }
>               tcg_gen_gvec_dup_imm(s->sew, vreg_ofs(s, a->rd),
>                                    MAXSZ(s), MAXSZ(s), simm);
>               mark_vs_dirty(s);
> @@ -2198,6 +2227,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
>               TCGv_i64 s1;
>               TCGv_ptr dest;
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_vmv_vx * const fns[4] = {
>                   gen_helper_vmv_v_x_b, gen_helper_vmv_v_x_h,
>                   gen_helper_vmv_v_x_w, gen_helper_vmv_v_x_d,
> @@ -2770,6 +2800,15 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
>           TCGv_i64 t1;
>   
>           if (s->vl_eq_vlmax) {
> +            if (s->vta && s->lmul < 0) {
> +                /* tail elements may pass vlmax when lmul < 0
> +                * set tail elements to 1s
> +                */
> +                uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +                tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                                vreg_ofs(s, a->rd), -1,
> +                                vlenb, vlenb);
here too.
> +            }
>               t1 = tcg_temp_new_i64();
>               /* NaN-box f[rs1] */
>               do_nanbox(s, t1, cpu_fpr[a->rs1]);
> @@ -2781,6 +2820,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
>               TCGv_ptr dest;
>               TCGv_i32 desc;
>               uint32_t data = FIELD_DP32(0, VDATA, LMUL, s->lmul);
> +            data = FIELD_DP32(data, VDATA, VTA, s->vta);
>               static gen_helper_vmv_vx * const fns[3] = {
>                   gen_helper_vmv_v_x_h,
>                   gen_helper_vmv_v_x_w,
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index a972946bc2..89927a424f 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -1964,6 +1964,10 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
>                     uint32_t desc)                                     \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
> @@ -1971,6 +1975,9 @@ void HELPER(NAME)(void *vd, void *vs1, CPURISCVState *env,           \
>           *((ETYPE *)vd + H(i)) = s1;                                  \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMV_VV(vmv_v_v_b, int8_t,  H1)
> @@ -1983,12 +1990,19 @@ void HELPER(NAME)(void *vd, uint64_t s1, CPURISCVState *env,         \
>                     uint32_t desc)                                     \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
>           *((ETYPE *)vd + H(i)) = (ETYPE)s1;                           \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMV_VX(vmv_v_x_b, int8_t,  H1)
> @@ -2001,6 +2015,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
>                     CPURISCVState *env, uint32_t desc)                 \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
> @@ -2008,6 +2026,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,          \
>           *((ETYPE *)vd + H(i)) = *(vt + H(i));                        \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMERGE_VV(vmerge_vvm_b, int8_t,  H1)
> @@ -2020,6 +2041,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
>                     void *vs2, CPURISCVState *env, uint32_t desc)      \
>   {                                                                    \
>       uint32_t vl = env->vl;                                           \
> +    uint32_t esz = sizeof(ETYPE);                                    \
> +    uint32_t total_elems = vext_get_total_elems(env_archcpu(env),    \
> +                                         env->vtype);                \
better aligned here too.
> +    uint32_t vta = vext_vta(desc);                                   \
>       uint32_t i;                                                      \
>                                                                        \
>       for (i = env->vstart; i < vl; i++) {                             \
> @@ -2029,6 +2054,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1,               \
>           *((ETYPE *)vd + H(i)) = d;                                   \
>       }                                                                \
>       env->vstart = 0;                                                 \
> +    /* set tail elements to 1s */                                    \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,          \
> +                                     total_elems * esz);             \
>   }
>   
>   GEN_VEXT_VMERGE_VX(vmerge_vxm_b, int8_t,  H1)

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 10/14] target/riscv: rvv: Add tail agnostic for vector fix-point arithmetic instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07 10:04 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 220 ++++++++++++++++++-----------------
 1 file changed, 114 insertions(+), 106 deletions(-)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 89927a424f..067095e19a 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -2106,10 +2106,12 @@ static inline void
 vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
              CPURISCVState *env,
              uint32_t desc,
-             opivv2_rm_fn *fn)
+             opivv2_rm_fn *fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
 
     switch (env->vxrm) {
     case 0: /* rnu */
@@ -2129,15 +2131,17 @@ vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
                      env, vl, vm, 3, fn);
         break;
     }
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate helpers for fixed point instructions with OPIVV format */
-#define GEN_VEXT_VV_RM(NAME)                                    \
+#define GEN_VEXT_VV_RM(NAME, ESZ)                               \
 void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,     \
                   CPURISCVState *env, uint32_t desc)            \
 {                                                               \
     vext_vv_rm_2(vd, v0, vs1, vs2, env, desc,                   \
-                 do_##NAME);                                    \
+                 do_##NAME, ESZ);                               \
 }
 
 static inline uint8_t saddu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
@@ -2187,10 +2191,10 @@ RVVCALL(OPIVV2_RM, vsaddu_vv_b, OP_UUU_B, H1, H1, H1, saddu8)
 RVVCALL(OPIVV2_RM, vsaddu_vv_h, OP_UUU_H, H2, H2, H2, saddu16)
 RVVCALL(OPIVV2_RM, vsaddu_vv_w, OP_UUU_W, H4, H4, H4, saddu32)
 RVVCALL(OPIVV2_RM, vsaddu_vv_d, OP_UUU_D, H8, H8, H8, saddu64)
-GEN_VEXT_VV_RM(vsaddu_vv_b)
-GEN_VEXT_VV_RM(vsaddu_vv_h)
-GEN_VEXT_VV_RM(vsaddu_vv_w)
-GEN_VEXT_VV_RM(vsaddu_vv_d)
+GEN_VEXT_VV_RM(vsaddu_vv_b, 1)
+GEN_VEXT_VV_RM(vsaddu_vv_h, 2)
+GEN_VEXT_VV_RM(vsaddu_vv_w, 4)
+GEN_VEXT_VV_RM(vsaddu_vv_d, 8)
 
 typedef void opivx2_rm_fn(void *vd, target_long s1, void *vs2, int i,
                           CPURISCVState *env, int vxrm);
@@ -2223,10 +2227,12 @@ static inline void
 vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
              CPURISCVState *env,
              uint32_t desc,
-             opivx2_rm_fn *fn)
+             opivx2_rm_fn *fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
 
     switch (env->vxrm) {
     case 0: /* rnu */
@@ -2246,25 +2252,27 @@ vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
                      env, vl, vm, 3, fn);
         break;
     }
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate helpers for fixed point instructions with OPIVX format */
-#define GEN_VEXT_VX_RM(NAME)                              \
+#define GEN_VEXT_VX_RM(NAME, ESZ)                         \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
         void *vs2, CPURISCVState *env, uint32_t desc)     \
 {                                                         \
     vext_vx_rm_2(vd, v0, s1, vs2, env, desc,              \
-                 do_##NAME);                              \
+                 do_##NAME, ESZ);                         \
 }
 
 RVVCALL(OPIVX2_RM, vsaddu_vx_b, OP_UUU_B, H1, H1, saddu8)
 RVVCALL(OPIVX2_RM, vsaddu_vx_h, OP_UUU_H, H2, H2, saddu16)
 RVVCALL(OPIVX2_RM, vsaddu_vx_w, OP_UUU_W, H4, H4, saddu32)
 RVVCALL(OPIVX2_RM, vsaddu_vx_d, OP_UUU_D, H8, H8, saddu64)
-GEN_VEXT_VX_RM(vsaddu_vx_b)
-GEN_VEXT_VX_RM(vsaddu_vx_h)
-GEN_VEXT_VX_RM(vsaddu_vx_w)
-GEN_VEXT_VX_RM(vsaddu_vx_d)
+GEN_VEXT_VX_RM(vsaddu_vx_b, 1)
+GEN_VEXT_VX_RM(vsaddu_vx_h, 2)
+GEN_VEXT_VX_RM(vsaddu_vx_w, 4)
+GEN_VEXT_VX_RM(vsaddu_vx_d, 8)
 
 static inline int8_t sadd8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2310,19 +2318,19 @@ RVVCALL(OPIVV2_RM, vsadd_vv_b, OP_SSS_B, H1, H1, H1, sadd8)
 RVVCALL(OPIVV2_RM, vsadd_vv_h, OP_SSS_H, H2, H2, H2, sadd16)
 RVVCALL(OPIVV2_RM, vsadd_vv_w, OP_SSS_W, H4, H4, H4, sadd32)
 RVVCALL(OPIVV2_RM, vsadd_vv_d, OP_SSS_D, H8, H8, H8, sadd64)
-GEN_VEXT_VV_RM(vsadd_vv_b)
-GEN_VEXT_VV_RM(vsadd_vv_h)
-GEN_VEXT_VV_RM(vsadd_vv_w)
-GEN_VEXT_VV_RM(vsadd_vv_d)
+GEN_VEXT_VV_RM(vsadd_vv_b, 1)
+GEN_VEXT_VV_RM(vsadd_vv_h, 2)
+GEN_VEXT_VV_RM(vsadd_vv_w, 4)
+GEN_VEXT_VV_RM(vsadd_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vsadd_vx_b, OP_SSS_B, H1, H1, sadd8)
 RVVCALL(OPIVX2_RM, vsadd_vx_h, OP_SSS_H, H2, H2, sadd16)
 RVVCALL(OPIVX2_RM, vsadd_vx_w, OP_SSS_W, H4, H4, sadd32)
 RVVCALL(OPIVX2_RM, vsadd_vx_d, OP_SSS_D, H8, H8, sadd64)
-GEN_VEXT_VX_RM(vsadd_vx_b)
-GEN_VEXT_VX_RM(vsadd_vx_h)
-GEN_VEXT_VX_RM(vsadd_vx_w)
-GEN_VEXT_VX_RM(vsadd_vx_d)
+GEN_VEXT_VX_RM(vsadd_vx_b, 1)
+GEN_VEXT_VX_RM(vsadd_vx_h, 2)
+GEN_VEXT_VX_RM(vsadd_vx_w, 4)
+GEN_VEXT_VX_RM(vsadd_vx_d, 8)
 
 static inline uint8_t ssubu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
 {
@@ -2371,19 +2379,19 @@ RVVCALL(OPIVV2_RM, vssubu_vv_b, OP_UUU_B, H1, H1, H1, ssubu8)
 RVVCALL(OPIVV2_RM, vssubu_vv_h, OP_UUU_H, H2, H2, H2, ssubu16)
 RVVCALL(OPIVV2_RM, vssubu_vv_w, OP_UUU_W, H4, H4, H4, ssubu32)
 RVVCALL(OPIVV2_RM, vssubu_vv_d, OP_UUU_D, H8, H8, H8, ssubu64)
-GEN_VEXT_VV_RM(vssubu_vv_b)
-GEN_VEXT_VV_RM(vssubu_vv_h)
-GEN_VEXT_VV_RM(vssubu_vv_w)
-GEN_VEXT_VV_RM(vssubu_vv_d)
+GEN_VEXT_VV_RM(vssubu_vv_b, 1)
+GEN_VEXT_VV_RM(vssubu_vv_h, 2)
+GEN_VEXT_VV_RM(vssubu_vv_w, 4)
+GEN_VEXT_VV_RM(vssubu_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssubu_vx_b, OP_UUU_B, H1, H1, ssubu8)
 RVVCALL(OPIVX2_RM, vssubu_vx_h, OP_UUU_H, H2, H2, ssubu16)
 RVVCALL(OPIVX2_RM, vssubu_vx_w, OP_UUU_W, H4, H4, ssubu32)
 RVVCALL(OPIVX2_RM, vssubu_vx_d, OP_UUU_D, H8, H8, ssubu64)
-GEN_VEXT_VX_RM(vssubu_vx_b)
-GEN_VEXT_VX_RM(vssubu_vx_h)
-GEN_VEXT_VX_RM(vssubu_vx_w)
-GEN_VEXT_VX_RM(vssubu_vx_d)
+GEN_VEXT_VX_RM(vssubu_vx_b, 1)
+GEN_VEXT_VX_RM(vssubu_vx_h, 2)
+GEN_VEXT_VX_RM(vssubu_vx_w, 4)
+GEN_VEXT_VX_RM(vssubu_vx_d, 8)
 
 static inline int8_t ssub8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2429,19 +2437,19 @@ RVVCALL(OPIVV2_RM, vssub_vv_b, OP_SSS_B, H1, H1, H1, ssub8)
 RVVCALL(OPIVV2_RM, vssub_vv_h, OP_SSS_H, H2, H2, H2, ssub16)
 RVVCALL(OPIVV2_RM, vssub_vv_w, OP_SSS_W, H4, H4, H4, ssub32)
 RVVCALL(OPIVV2_RM, vssub_vv_d, OP_SSS_D, H8, H8, H8, ssub64)
-GEN_VEXT_VV_RM(vssub_vv_b)
-GEN_VEXT_VV_RM(vssub_vv_h)
-GEN_VEXT_VV_RM(vssub_vv_w)
-GEN_VEXT_VV_RM(vssub_vv_d)
+GEN_VEXT_VV_RM(vssub_vv_b, 1)
+GEN_VEXT_VV_RM(vssub_vv_h, 2)
+GEN_VEXT_VV_RM(vssub_vv_w, 4)
+GEN_VEXT_VV_RM(vssub_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssub_vx_b, OP_SSS_B, H1, H1, ssub8)
 RVVCALL(OPIVX2_RM, vssub_vx_h, OP_SSS_H, H2, H2, ssub16)
 RVVCALL(OPIVX2_RM, vssub_vx_w, OP_SSS_W, H4, H4, ssub32)
 RVVCALL(OPIVX2_RM, vssub_vx_d, OP_SSS_D, H8, H8, ssub64)
-GEN_VEXT_VX_RM(vssub_vx_b)
-GEN_VEXT_VX_RM(vssub_vx_h)
-GEN_VEXT_VX_RM(vssub_vx_w)
-GEN_VEXT_VX_RM(vssub_vx_d)
+GEN_VEXT_VX_RM(vssub_vx_b, 1)
+GEN_VEXT_VX_RM(vssub_vx_h, 2)
+GEN_VEXT_VX_RM(vssub_vx_w, 4)
+GEN_VEXT_VX_RM(vssub_vx_d, 8)
 
 /* Vector Single-Width Averaging Add and Subtract */
 static inline uint8_t get_round(int vxrm, uint64_t v, uint8_t shift)
@@ -2493,19 +2501,19 @@ RVVCALL(OPIVV2_RM, vaadd_vv_b, OP_SSS_B, H1, H1, H1, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_h, OP_SSS_H, H2, H2, H2, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_w, OP_SSS_W, H4, H4, H4, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_d, OP_SSS_D, H8, H8, H8, aadd64)
-GEN_VEXT_VV_RM(vaadd_vv_b)
-GEN_VEXT_VV_RM(vaadd_vv_h)
-GEN_VEXT_VV_RM(vaadd_vv_w)
-GEN_VEXT_VV_RM(vaadd_vv_d)
+GEN_VEXT_VV_RM(vaadd_vv_b, 1)
+GEN_VEXT_VV_RM(vaadd_vv_h, 2)
+GEN_VEXT_VV_RM(vaadd_vv_w, 4)
+GEN_VEXT_VV_RM(vaadd_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vaadd_vx_b, OP_SSS_B, H1, H1, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_h, OP_SSS_H, H2, H2, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_w, OP_SSS_W, H4, H4, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_d, OP_SSS_D, H8, H8, aadd64)
-GEN_VEXT_VX_RM(vaadd_vx_b)
-GEN_VEXT_VX_RM(vaadd_vx_h)
-GEN_VEXT_VX_RM(vaadd_vx_w)
-GEN_VEXT_VX_RM(vaadd_vx_d)
+GEN_VEXT_VX_RM(vaadd_vx_b, 1)
+GEN_VEXT_VX_RM(vaadd_vx_h, 2)
+GEN_VEXT_VX_RM(vaadd_vx_w, 4)
+GEN_VEXT_VX_RM(vaadd_vx_d, 8)
 
 static inline uint32_t aaddu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2530,19 +2538,19 @@ RVVCALL(OPIVV2_RM, vaaddu_vv_b, OP_UUU_B, H1, H1, H1, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_h, OP_UUU_H, H2, H2, H2, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_w, OP_UUU_W, H4, H4, H4, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_d, OP_UUU_D, H8, H8, H8, aaddu64)
-GEN_VEXT_VV_RM(vaaddu_vv_b)
-GEN_VEXT_VV_RM(vaaddu_vv_h)
-GEN_VEXT_VV_RM(vaaddu_vv_w)
-GEN_VEXT_VV_RM(vaaddu_vv_d)
+GEN_VEXT_VV_RM(vaaddu_vv_b, 1)
+GEN_VEXT_VV_RM(vaaddu_vv_h, 2)
+GEN_VEXT_VV_RM(vaaddu_vv_w, 4)
+GEN_VEXT_VV_RM(vaaddu_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vaaddu_vx_b, OP_UUU_B, H1, H1, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_h, OP_UUU_H, H2, H2, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_w, OP_UUU_W, H4, H4, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_d, OP_UUU_D, H8, H8, aaddu64)
-GEN_VEXT_VX_RM(vaaddu_vx_b)
-GEN_VEXT_VX_RM(vaaddu_vx_h)
-GEN_VEXT_VX_RM(vaaddu_vx_w)
-GEN_VEXT_VX_RM(vaaddu_vx_d)
+GEN_VEXT_VX_RM(vaaddu_vx_b, 1)
+GEN_VEXT_VX_RM(vaaddu_vx_h, 2)
+GEN_VEXT_VX_RM(vaaddu_vx_w, 4)
+GEN_VEXT_VX_RM(vaaddu_vx_d, 8)
 
 static inline int32_t asub32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
 {
@@ -2566,19 +2574,19 @@ RVVCALL(OPIVV2_RM, vasub_vv_b, OP_SSS_B, H1, H1, H1, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_h, OP_SSS_H, H2, H2, H2, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_w, OP_SSS_W, H4, H4, H4, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_d, OP_SSS_D, H8, H8, H8, asub64)
-GEN_VEXT_VV_RM(vasub_vv_b)
-GEN_VEXT_VV_RM(vasub_vv_h)
-GEN_VEXT_VV_RM(vasub_vv_w)
-GEN_VEXT_VV_RM(vasub_vv_d)
+GEN_VEXT_VV_RM(vasub_vv_b, 1)
+GEN_VEXT_VV_RM(vasub_vv_h, 2)
+GEN_VEXT_VV_RM(vasub_vv_w, 4)
+GEN_VEXT_VV_RM(vasub_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vasub_vx_b, OP_SSS_B, H1, H1, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_h, OP_SSS_H, H2, H2, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_w, OP_SSS_W, H4, H4, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_d, OP_SSS_D, H8, H8, asub64)
-GEN_VEXT_VX_RM(vasub_vx_b)
-GEN_VEXT_VX_RM(vasub_vx_h)
-GEN_VEXT_VX_RM(vasub_vx_w)
-GEN_VEXT_VX_RM(vasub_vx_d)
+GEN_VEXT_VX_RM(vasub_vx_b, 1)
+GEN_VEXT_VX_RM(vasub_vx_h, 2)
+GEN_VEXT_VX_RM(vasub_vx_w, 4)
+GEN_VEXT_VX_RM(vasub_vx_d, 8)
 
 static inline uint32_t asubu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2603,19 +2611,19 @@ RVVCALL(OPIVV2_RM, vasubu_vv_b, OP_UUU_B, H1, H1, H1, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_h, OP_UUU_H, H2, H2, H2, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_w, OP_UUU_W, H4, H4, H4, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_d, OP_UUU_D, H8, H8, H8, asubu64)
-GEN_VEXT_VV_RM(vasubu_vv_b)
-GEN_VEXT_VV_RM(vasubu_vv_h)
-GEN_VEXT_VV_RM(vasubu_vv_w)
-GEN_VEXT_VV_RM(vasubu_vv_d)
+GEN_VEXT_VV_RM(vasubu_vv_b, 1)
+GEN_VEXT_VV_RM(vasubu_vv_h, 2)
+GEN_VEXT_VV_RM(vasubu_vv_w, 4)
+GEN_VEXT_VV_RM(vasubu_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vasubu_vx_b, OP_UUU_B, H1, H1, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_h, OP_UUU_H, H2, H2, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_w, OP_UUU_W, H4, H4, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_d, OP_UUU_D, H8, H8, asubu64)
-GEN_VEXT_VX_RM(vasubu_vx_b)
-GEN_VEXT_VX_RM(vasubu_vx_h)
-GEN_VEXT_VX_RM(vasubu_vx_w)
-GEN_VEXT_VX_RM(vasubu_vx_d)
+GEN_VEXT_VX_RM(vasubu_vx_b, 1)
+GEN_VEXT_VX_RM(vasubu_vx_h, 2)
+GEN_VEXT_VX_RM(vasubu_vx_w, 4)
+GEN_VEXT_VX_RM(vasubu_vx_d, 8)
 
 /* Vector Single-Width Fractional Multiply with Rounding and Saturation */
 static inline int8_t vsmul8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2710,19 +2718,19 @@ RVVCALL(OPIVV2_RM, vsmul_vv_b, OP_SSS_B, H1, H1, H1, vsmul8)
 RVVCALL(OPIVV2_RM, vsmul_vv_h, OP_SSS_H, H2, H2, H2, vsmul16)
 RVVCALL(OPIVV2_RM, vsmul_vv_w, OP_SSS_W, H4, H4, H4, vsmul32)
 RVVCALL(OPIVV2_RM, vsmul_vv_d, OP_SSS_D, H8, H8, H8, vsmul64)
-GEN_VEXT_VV_RM(vsmul_vv_b)
-GEN_VEXT_VV_RM(vsmul_vv_h)
-GEN_VEXT_VV_RM(vsmul_vv_w)
-GEN_VEXT_VV_RM(vsmul_vv_d)
+GEN_VEXT_VV_RM(vsmul_vv_b, 1)
+GEN_VEXT_VV_RM(vsmul_vv_h, 2)
+GEN_VEXT_VV_RM(vsmul_vv_w, 4)
+GEN_VEXT_VV_RM(vsmul_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vsmul_vx_b, OP_SSS_B, H1, H1, vsmul8)
 RVVCALL(OPIVX2_RM, vsmul_vx_h, OP_SSS_H, H2, H2, vsmul16)
 RVVCALL(OPIVX2_RM, vsmul_vx_w, OP_SSS_W, H4, H4, vsmul32)
 RVVCALL(OPIVX2_RM, vsmul_vx_d, OP_SSS_D, H8, H8, vsmul64)
-GEN_VEXT_VX_RM(vsmul_vx_b)
-GEN_VEXT_VX_RM(vsmul_vx_h)
-GEN_VEXT_VX_RM(vsmul_vx_w)
-GEN_VEXT_VX_RM(vsmul_vx_d)
+GEN_VEXT_VX_RM(vsmul_vx_b, 1)
+GEN_VEXT_VX_RM(vsmul_vx_h, 2)
+GEN_VEXT_VX_RM(vsmul_vx_w, 4)
+GEN_VEXT_VX_RM(vsmul_vx_d, 8)
 
 /* Vector Single-Width Scaling Shift Instructions */
 static inline uint8_t
@@ -2769,19 +2777,19 @@ RVVCALL(OPIVV2_RM, vssrl_vv_b, OP_UUU_B, H1, H1, H1, vssrl8)
 RVVCALL(OPIVV2_RM, vssrl_vv_h, OP_UUU_H, H2, H2, H2, vssrl16)
 RVVCALL(OPIVV2_RM, vssrl_vv_w, OP_UUU_W, H4, H4, H4, vssrl32)
 RVVCALL(OPIVV2_RM, vssrl_vv_d, OP_UUU_D, H8, H8, H8, vssrl64)
-GEN_VEXT_VV_RM(vssrl_vv_b)
-GEN_VEXT_VV_RM(vssrl_vv_h)
-GEN_VEXT_VV_RM(vssrl_vv_w)
-GEN_VEXT_VV_RM(vssrl_vv_d)
+GEN_VEXT_VV_RM(vssrl_vv_b, 1)
+GEN_VEXT_VV_RM(vssrl_vv_h, 2)
+GEN_VEXT_VV_RM(vssrl_vv_w, 4)
+GEN_VEXT_VV_RM(vssrl_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssrl_vx_b, OP_UUU_B, H1, H1, vssrl8)
 RVVCALL(OPIVX2_RM, vssrl_vx_h, OP_UUU_H, H2, H2, vssrl16)
 RVVCALL(OPIVX2_RM, vssrl_vx_w, OP_UUU_W, H4, H4, vssrl32)
 RVVCALL(OPIVX2_RM, vssrl_vx_d, OP_UUU_D, H8, H8, vssrl64)
-GEN_VEXT_VX_RM(vssrl_vx_b)
-GEN_VEXT_VX_RM(vssrl_vx_h)
-GEN_VEXT_VX_RM(vssrl_vx_w)
-GEN_VEXT_VX_RM(vssrl_vx_d)
+GEN_VEXT_VX_RM(vssrl_vx_b, 1)
+GEN_VEXT_VX_RM(vssrl_vx_h, 2)
+GEN_VEXT_VX_RM(vssrl_vx_w, 4)
+GEN_VEXT_VX_RM(vssrl_vx_d, 8)
 
 static inline int8_t
 vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2828,19 +2836,19 @@ RVVCALL(OPIVV2_RM, vssra_vv_b, OP_SSS_B, H1, H1, H1, vssra8)
 RVVCALL(OPIVV2_RM, vssra_vv_h, OP_SSS_H, H2, H2, H2, vssra16)
 RVVCALL(OPIVV2_RM, vssra_vv_w, OP_SSS_W, H4, H4, H4, vssra32)
 RVVCALL(OPIVV2_RM, vssra_vv_d, OP_SSS_D, H8, H8, H8, vssra64)
-GEN_VEXT_VV_RM(vssra_vv_b)
-GEN_VEXT_VV_RM(vssra_vv_h)
-GEN_VEXT_VV_RM(vssra_vv_w)
-GEN_VEXT_VV_RM(vssra_vv_d)
+GEN_VEXT_VV_RM(vssra_vv_b, 1)
+GEN_VEXT_VV_RM(vssra_vv_h, 2)
+GEN_VEXT_VV_RM(vssra_vv_w, 4)
+GEN_VEXT_VV_RM(vssra_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssra_vx_b, OP_SSS_B, H1, H1, vssra8)
 RVVCALL(OPIVX2_RM, vssra_vx_h, OP_SSS_H, H2, H2, vssra16)
 RVVCALL(OPIVX2_RM, vssra_vx_w, OP_SSS_W, H4, H4, vssra32)
 RVVCALL(OPIVX2_RM, vssra_vx_d, OP_SSS_D, H8, H8, vssra64)
-GEN_VEXT_VX_RM(vssra_vx_b)
-GEN_VEXT_VX_RM(vssra_vx_h)
-GEN_VEXT_VX_RM(vssra_vx_w)
-GEN_VEXT_VX_RM(vssra_vx_d)
+GEN_VEXT_VX_RM(vssra_vx_b, 1)
+GEN_VEXT_VX_RM(vssra_vx_h, 2)
+GEN_VEXT_VX_RM(vssra_vx_w, 4)
+GEN_VEXT_VX_RM(vssra_vx_d, 8)
 
 /* Vector Narrowing Fixed-Point Clip Instructions */
 static inline int8_t
@@ -2903,16 +2911,16 @@ vnclip32(CPURISCVState *env, int vxrm, int64_t a, int32_t b)
 RVVCALL(OPIVV2_RM, vnclip_wv_b, NOP_SSS_B, H1, H2, H1, vnclip8)
 RVVCALL(OPIVV2_RM, vnclip_wv_h, NOP_SSS_H, H2, H4, H2, vnclip16)
 RVVCALL(OPIVV2_RM, vnclip_wv_w, NOP_SSS_W, H4, H8, H4, vnclip32)
-GEN_VEXT_VV_RM(vnclip_wv_b)
-GEN_VEXT_VV_RM(vnclip_wv_h)
-GEN_VEXT_VV_RM(vnclip_wv_w)
+GEN_VEXT_VV_RM(vnclip_wv_b, 1)
+GEN_VEXT_VV_RM(vnclip_wv_h, 2)
+GEN_VEXT_VV_RM(vnclip_wv_w, 4)
 
 RVVCALL(OPIVX2_RM, vnclip_wx_b, NOP_SSS_B, H1, H2, vnclip8)
 RVVCALL(OPIVX2_RM, vnclip_wx_h, NOP_SSS_H, H2, H4, vnclip16)
 RVVCALL(OPIVX2_RM, vnclip_wx_w, NOP_SSS_W, H4, H8, vnclip32)
-GEN_VEXT_VX_RM(vnclip_wx_b)
-GEN_VEXT_VX_RM(vnclip_wx_h)
-GEN_VEXT_VX_RM(vnclip_wx_w)
+GEN_VEXT_VX_RM(vnclip_wx_b, 1)
+GEN_VEXT_VX_RM(vnclip_wx_h, 2)
+GEN_VEXT_VX_RM(vnclip_wx_w, 4)
 
 static inline uint8_t
 vnclipu8(CPURISCVState *env, int vxrm, uint16_t a, uint8_t b)
@@ -2965,16 +2973,16 @@ vnclipu32(CPURISCVState *env, int vxrm, uint64_t a, uint32_t b)
 RVVCALL(OPIVV2_RM, vnclipu_wv_b, NOP_UUU_B, H1, H2, H1, vnclipu8)
 RVVCALL(OPIVV2_RM, vnclipu_wv_h, NOP_UUU_H, H2, H4, H2, vnclipu16)
 RVVCALL(OPIVV2_RM, vnclipu_wv_w, NOP_UUU_W, H4, H8, H4, vnclipu32)
-GEN_VEXT_VV_RM(vnclipu_wv_b)
-GEN_VEXT_VV_RM(vnclipu_wv_h)
-GEN_VEXT_VV_RM(vnclipu_wv_w)
+GEN_VEXT_VV_RM(vnclipu_wv_b, 1)
+GEN_VEXT_VV_RM(vnclipu_wv_h, 2)
+GEN_VEXT_VV_RM(vnclipu_wv_w, 4)
 
 RVVCALL(OPIVX2_RM, vnclipu_wx_b, NOP_UUU_B, H1, H2, vnclipu8)
 RVVCALL(OPIVX2_RM, vnclipu_wx_h, NOP_UUU_H, H2, H4, vnclipu16)
 RVVCALL(OPIVX2_RM, vnclipu_wx_w, NOP_UUU_W, H4, H8, vnclipu32)
-GEN_VEXT_VX_RM(vnclipu_wx_b)
-GEN_VEXT_VX_RM(vnclipu_wx_h)
-GEN_VEXT_VX_RM(vnclipu_wx_w)
+GEN_VEXT_VX_RM(vnclipu_wx_b, 1)
+GEN_VEXT_VX_RM(vnclipu_wx_h, 2)
+GEN_VEXT_VX_RM(vnclipu_wx_w, 4)
 
 /*
  *** Vector Float Point Arithmetic Instructions
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 10/14] target/riscv: rvv: Add tail agnostic for vector fix-point arithmetic instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 220 ++++++++++++++++++-----------------
 1 file changed, 114 insertions(+), 106 deletions(-)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 89927a424f..067095e19a 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -2106,10 +2106,12 @@ static inline void
 vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
              CPURISCVState *env,
              uint32_t desc,
-             opivv2_rm_fn *fn)
+             opivv2_rm_fn *fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
 
     switch (env->vxrm) {
     case 0: /* rnu */
@@ -2129,15 +2131,17 @@ vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
                      env, vl, vm, 3, fn);
         break;
     }
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate helpers for fixed point instructions with OPIVV format */
-#define GEN_VEXT_VV_RM(NAME)                                    \
+#define GEN_VEXT_VV_RM(NAME, ESZ)                               \
 void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,     \
                   CPURISCVState *env, uint32_t desc)            \
 {                                                               \
     vext_vv_rm_2(vd, v0, vs1, vs2, env, desc,                   \
-                 do_##NAME);                                    \
+                 do_##NAME, ESZ);                               \
 }
 
 static inline uint8_t saddu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
@@ -2187,10 +2191,10 @@ RVVCALL(OPIVV2_RM, vsaddu_vv_b, OP_UUU_B, H1, H1, H1, saddu8)
 RVVCALL(OPIVV2_RM, vsaddu_vv_h, OP_UUU_H, H2, H2, H2, saddu16)
 RVVCALL(OPIVV2_RM, vsaddu_vv_w, OP_UUU_W, H4, H4, H4, saddu32)
 RVVCALL(OPIVV2_RM, vsaddu_vv_d, OP_UUU_D, H8, H8, H8, saddu64)
-GEN_VEXT_VV_RM(vsaddu_vv_b)
-GEN_VEXT_VV_RM(vsaddu_vv_h)
-GEN_VEXT_VV_RM(vsaddu_vv_w)
-GEN_VEXT_VV_RM(vsaddu_vv_d)
+GEN_VEXT_VV_RM(vsaddu_vv_b, 1)
+GEN_VEXT_VV_RM(vsaddu_vv_h, 2)
+GEN_VEXT_VV_RM(vsaddu_vv_w, 4)
+GEN_VEXT_VV_RM(vsaddu_vv_d, 8)
 
 typedef void opivx2_rm_fn(void *vd, target_long s1, void *vs2, int i,
                           CPURISCVState *env, int vxrm);
@@ -2223,10 +2227,12 @@ static inline void
 vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
              CPURISCVState *env,
              uint32_t desc,
-             opivx2_rm_fn *fn)
+             opivx2_rm_fn *fn, uint32_t esz)
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
 
     switch (env->vxrm) {
     case 0: /* rnu */
@@ -2246,25 +2252,27 @@ vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
                      env, vl, vm, 3, fn);
         break;
     }
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz, total_elems * esz);
 }
 
 /* generate helpers for fixed point instructions with OPIVX format */
-#define GEN_VEXT_VX_RM(NAME)                              \
+#define GEN_VEXT_VX_RM(NAME, ESZ)                         \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
         void *vs2, CPURISCVState *env, uint32_t desc)     \
 {                                                         \
     vext_vx_rm_2(vd, v0, s1, vs2, env, desc,              \
-                 do_##NAME);                              \
+                 do_##NAME, ESZ);                         \
 }
 
 RVVCALL(OPIVX2_RM, vsaddu_vx_b, OP_UUU_B, H1, H1, saddu8)
 RVVCALL(OPIVX2_RM, vsaddu_vx_h, OP_UUU_H, H2, H2, saddu16)
 RVVCALL(OPIVX2_RM, vsaddu_vx_w, OP_UUU_W, H4, H4, saddu32)
 RVVCALL(OPIVX2_RM, vsaddu_vx_d, OP_UUU_D, H8, H8, saddu64)
-GEN_VEXT_VX_RM(vsaddu_vx_b)
-GEN_VEXT_VX_RM(vsaddu_vx_h)
-GEN_VEXT_VX_RM(vsaddu_vx_w)
-GEN_VEXT_VX_RM(vsaddu_vx_d)
+GEN_VEXT_VX_RM(vsaddu_vx_b, 1)
+GEN_VEXT_VX_RM(vsaddu_vx_h, 2)
+GEN_VEXT_VX_RM(vsaddu_vx_w, 4)
+GEN_VEXT_VX_RM(vsaddu_vx_d, 8)
 
 static inline int8_t sadd8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2310,19 +2318,19 @@ RVVCALL(OPIVV2_RM, vsadd_vv_b, OP_SSS_B, H1, H1, H1, sadd8)
 RVVCALL(OPIVV2_RM, vsadd_vv_h, OP_SSS_H, H2, H2, H2, sadd16)
 RVVCALL(OPIVV2_RM, vsadd_vv_w, OP_SSS_W, H4, H4, H4, sadd32)
 RVVCALL(OPIVV2_RM, vsadd_vv_d, OP_SSS_D, H8, H8, H8, sadd64)
-GEN_VEXT_VV_RM(vsadd_vv_b)
-GEN_VEXT_VV_RM(vsadd_vv_h)
-GEN_VEXT_VV_RM(vsadd_vv_w)
-GEN_VEXT_VV_RM(vsadd_vv_d)
+GEN_VEXT_VV_RM(vsadd_vv_b, 1)
+GEN_VEXT_VV_RM(vsadd_vv_h, 2)
+GEN_VEXT_VV_RM(vsadd_vv_w, 4)
+GEN_VEXT_VV_RM(vsadd_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vsadd_vx_b, OP_SSS_B, H1, H1, sadd8)
 RVVCALL(OPIVX2_RM, vsadd_vx_h, OP_SSS_H, H2, H2, sadd16)
 RVVCALL(OPIVX2_RM, vsadd_vx_w, OP_SSS_W, H4, H4, sadd32)
 RVVCALL(OPIVX2_RM, vsadd_vx_d, OP_SSS_D, H8, H8, sadd64)
-GEN_VEXT_VX_RM(vsadd_vx_b)
-GEN_VEXT_VX_RM(vsadd_vx_h)
-GEN_VEXT_VX_RM(vsadd_vx_w)
-GEN_VEXT_VX_RM(vsadd_vx_d)
+GEN_VEXT_VX_RM(vsadd_vx_b, 1)
+GEN_VEXT_VX_RM(vsadd_vx_h, 2)
+GEN_VEXT_VX_RM(vsadd_vx_w, 4)
+GEN_VEXT_VX_RM(vsadd_vx_d, 8)
 
 static inline uint8_t ssubu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
 {
@@ -2371,19 +2379,19 @@ RVVCALL(OPIVV2_RM, vssubu_vv_b, OP_UUU_B, H1, H1, H1, ssubu8)
 RVVCALL(OPIVV2_RM, vssubu_vv_h, OP_UUU_H, H2, H2, H2, ssubu16)
 RVVCALL(OPIVV2_RM, vssubu_vv_w, OP_UUU_W, H4, H4, H4, ssubu32)
 RVVCALL(OPIVV2_RM, vssubu_vv_d, OP_UUU_D, H8, H8, H8, ssubu64)
-GEN_VEXT_VV_RM(vssubu_vv_b)
-GEN_VEXT_VV_RM(vssubu_vv_h)
-GEN_VEXT_VV_RM(vssubu_vv_w)
-GEN_VEXT_VV_RM(vssubu_vv_d)
+GEN_VEXT_VV_RM(vssubu_vv_b, 1)
+GEN_VEXT_VV_RM(vssubu_vv_h, 2)
+GEN_VEXT_VV_RM(vssubu_vv_w, 4)
+GEN_VEXT_VV_RM(vssubu_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssubu_vx_b, OP_UUU_B, H1, H1, ssubu8)
 RVVCALL(OPIVX2_RM, vssubu_vx_h, OP_UUU_H, H2, H2, ssubu16)
 RVVCALL(OPIVX2_RM, vssubu_vx_w, OP_UUU_W, H4, H4, ssubu32)
 RVVCALL(OPIVX2_RM, vssubu_vx_d, OP_UUU_D, H8, H8, ssubu64)
-GEN_VEXT_VX_RM(vssubu_vx_b)
-GEN_VEXT_VX_RM(vssubu_vx_h)
-GEN_VEXT_VX_RM(vssubu_vx_w)
-GEN_VEXT_VX_RM(vssubu_vx_d)
+GEN_VEXT_VX_RM(vssubu_vx_b, 1)
+GEN_VEXT_VX_RM(vssubu_vx_h, 2)
+GEN_VEXT_VX_RM(vssubu_vx_w, 4)
+GEN_VEXT_VX_RM(vssubu_vx_d, 8)
 
 static inline int8_t ssub8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2429,19 +2437,19 @@ RVVCALL(OPIVV2_RM, vssub_vv_b, OP_SSS_B, H1, H1, H1, ssub8)
 RVVCALL(OPIVV2_RM, vssub_vv_h, OP_SSS_H, H2, H2, H2, ssub16)
 RVVCALL(OPIVV2_RM, vssub_vv_w, OP_SSS_W, H4, H4, H4, ssub32)
 RVVCALL(OPIVV2_RM, vssub_vv_d, OP_SSS_D, H8, H8, H8, ssub64)
-GEN_VEXT_VV_RM(vssub_vv_b)
-GEN_VEXT_VV_RM(vssub_vv_h)
-GEN_VEXT_VV_RM(vssub_vv_w)
-GEN_VEXT_VV_RM(vssub_vv_d)
+GEN_VEXT_VV_RM(vssub_vv_b, 1)
+GEN_VEXT_VV_RM(vssub_vv_h, 2)
+GEN_VEXT_VV_RM(vssub_vv_w, 4)
+GEN_VEXT_VV_RM(vssub_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssub_vx_b, OP_SSS_B, H1, H1, ssub8)
 RVVCALL(OPIVX2_RM, vssub_vx_h, OP_SSS_H, H2, H2, ssub16)
 RVVCALL(OPIVX2_RM, vssub_vx_w, OP_SSS_W, H4, H4, ssub32)
 RVVCALL(OPIVX2_RM, vssub_vx_d, OP_SSS_D, H8, H8, ssub64)
-GEN_VEXT_VX_RM(vssub_vx_b)
-GEN_VEXT_VX_RM(vssub_vx_h)
-GEN_VEXT_VX_RM(vssub_vx_w)
-GEN_VEXT_VX_RM(vssub_vx_d)
+GEN_VEXT_VX_RM(vssub_vx_b, 1)
+GEN_VEXT_VX_RM(vssub_vx_h, 2)
+GEN_VEXT_VX_RM(vssub_vx_w, 4)
+GEN_VEXT_VX_RM(vssub_vx_d, 8)
 
 /* Vector Single-Width Averaging Add and Subtract */
 static inline uint8_t get_round(int vxrm, uint64_t v, uint8_t shift)
@@ -2493,19 +2501,19 @@ RVVCALL(OPIVV2_RM, vaadd_vv_b, OP_SSS_B, H1, H1, H1, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_h, OP_SSS_H, H2, H2, H2, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_w, OP_SSS_W, H4, H4, H4, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_d, OP_SSS_D, H8, H8, H8, aadd64)
-GEN_VEXT_VV_RM(vaadd_vv_b)
-GEN_VEXT_VV_RM(vaadd_vv_h)
-GEN_VEXT_VV_RM(vaadd_vv_w)
-GEN_VEXT_VV_RM(vaadd_vv_d)
+GEN_VEXT_VV_RM(vaadd_vv_b, 1)
+GEN_VEXT_VV_RM(vaadd_vv_h, 2)
+GEN_VEXT_VV_RM(vaadd_vv_w, 4)
+GEN_VEXT_VV_RM(vaadd_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vaadd_vx_b, OP_SSS_B, H1, H1, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_h, OP_SSS_H, H2, H2, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_w, OP_SSS_W, H4, H4, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_d, OP_SSS_D, H8, H8, aadd64)
-GEN_VEXT_VX_RM(vaadd_vx_b)
-GEN_VEXT_VX_RM(vaadd_vx_h)
-GEN_VEXT_VX_RM(vaadd_vx_w)
-GEN_VEXT_VX_RM(vaadd_vx_d)
+GEN_VEXT_VX_RM(vaadd_vx_b, 1)
+GEN_VEXT_VX_RM(vaadd_vx_h, 2)
+GEN_VEXT_VX_RM(vaadd_vx_w, 4)
+GEN_VEXT_VX_RM(vaadd_vx_d, 8)
 
 static inline uint32_t aaddu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2530,19 +2538,19 @@ RVVCALL(OPIVV2_RM, vaaddu_vv_b, OP_UUU_B, H1, H1, H1, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_h, OP_UUU_H, H2, H2, H2, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_w, OP_UUU_W, H4, H4, H4, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_d, OP_UUU_D, H8, H8, H8, aaddu64)
-GEN_VEXT_VV_RM(vaaddu_vv_b)
-GEN_VEXT_VV_RM(vaaddu_vv_h)
-GEN_VEXT_VV_RM(vaaddu_vv_w)
-GEN_VEXT_VV_RM(vaaddu_vv_d)
+GEN_VEXT_VV_RM(vaaddu_vv_b, 1)
+GEN_VEXT_VV_RM(vaaddu_vv_h, 2)
+GEN_VEXT_VV_RM(vaaddu_vv_w, 4)
+GEN_VEXT_VV_RM(vaaddu_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vaaddu_vx_b, OP_UUU_B, H1, H1, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_h, OP_UUU_H, H2, H2, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_w, OP_UUU_W, H4, H4, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_d, OP_UUU_D, H8, H8, aaddu64)
-GEN_VEXT_VX_RM(vaaddu_vx_b)
-GEN_VEXT_VX_RM(vaaddu_vx_h)
-GEN_VEXT_VX_RM(vaaddu_vx_w)
-GEN_VEXT_VX_RM(vaaddu_vx_d)
+GEN_VEXT_VX_RM(vaaddu_vx_b, 1)
+GEN_VEXT_VX_RM(vaaddu_vx_h, 2)
+GEN_VEXT_VX_RM(vaaddu_vx_w, 4)
+GEN_VEXT_VX_RM(vaaddu_vx_d, 8)
 
 static inline int32_t asub32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
 {
@@ -2566,19 +2574,19 @@ RVVCALL(OPIVV2_RM, vasub_vv_b, OP_SSS_B, H1, H1, H1, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_h, OP_SSS_H, H2, H2, H2, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_w, OP_SSS_W, H4, H4, H4, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_d, OP_SSS_D, H8, H8, H8, asub64)
-GEN_VEXT_VV_RM(vasub_vv_b)
-GEN_VEXT_VV_RM(vasub_vv_h)
-GEN_VEXT_VV_RM(vasub_vv_w)
-GEN_VEXT_VV_RM(vasub_vv_d)
+GEN_VEXT_VV_RM(vasub_vv_b, 1)
+GEN_VEXT_VV_RM(vasub_vv_h, 2)
+GEN_VEXT_VV_RM(vasub_vv_w, 4)
+GEN_VEXT_VV_RM(vasub_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vasub_vx_b, OP_SSS_B, H1, H1, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_h, OP_SSS_H, H2, H2, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_w, OP_SSS_W, H4, H4, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_d, OP_SSS_D, H8, H8, asub64)
-GEN_VEXT_VX_RM(vasub_vx_b)
-GEN_VEXT_VX_RM(vasub_vx_h)
-GEN_VEXT_VX_RM(vasub_vx_w)
-GEN_VEXT_VX_RM(vasub_vx_d)
+GEN_VEXT_VX_RM(vasub_vx_b, 1)
+GEN_VEXT_VX_RM(vasub_vx_h, 2)
+GEN_VEXT_VX_RM(vasub_vx_w, 4)
+GEN_VEXT_VX_RM(vasub_vx_d, 8)
 
 static inline uint32_t asubu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2603,19 +2611,19 @@ RVVCALL(OPIVV2_RM, vasubu_vv_b, OP_UUU_B, H1, H1, H1, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_h, OP_UUU_H, H2, H2, H2, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_w, OP_UUU_W, H4, H4, H4, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_d, OP_UUU_D, H8, H8, H8, asubu64)
-GEN_VEXT_VV_RM(vasubu_vv_b)
-GEN_VEXT_VV_RM(vasubu_vv_h)
-GEN_VEXT_VV_RM(vasubu_vv_w)
-GEN_VEXT_VV_RM(vasubu_vv_d)
+GEN_VEXT_VV_RM(vasubu_vv_b, 1)
+GEN_VEXT_VV_RM(vasubu_vv_h, 2)
+GEN_VEXT_VV_RM(vasubu_vv_w, 4)
+GEN_VEXT_VV_RM(vasubu_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vasubu_vx_b, OP_UUU_B, H1, H1, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_h, OP_UUU_H, H2, H2, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_w, OP_UUU_W, H4, H4, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_d, OP_UUU_D, H8, H8, asubu64)
-GEN_VEXT_VX_RM(vasubu_vx_b)
-GEN_VEXT_VX_RM(vasubu_vx_h)
-GEN_VEXT_VX_RM(vasubu_vx_w)
-GEN_VEXT_VX_RM(vasubu_vx_d)
+GEN_VEXT_VX_RM(vasubu_vx_b, 1)
+GEN_VEXT_VX_RM(vasubu_vx_h, 2)
+GEN_VEXT_VX_RM(vasubu_vx_w, 4)
+GEN_VEXT_VX_RM(vasubu_vx_d, 8)
 
 /* Vector Single-Width Fractional Multiply with Rounding and Saturation */
 static inline int8_t vsmul8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2710,19 +2718,19 @@ RVVCALL(OPIVV2_RM, vsmul_vv_b, OP_SSS_B, H1, H1, H1, vsmul8)
 RVVCALL(OPIVV2_RM, vsmul_vv_h, OP_SSS_H, H2, H2, H2, vsmul16)
 RVVCALL(OPIVV2_RM, vsmul_vv_w, OP_SSS_W, H4, H4, H4, vsmul32)
 RVVCALL(OPIVV2_RM, vsmul_vv_d, OP_SSS_D, H8, H8, H8, vsmul64)
-GEN_VEXT_VV_RM(vsmul_vv_b)
-GEN_VEXT_VV_RM(vsmul_vv_h)
-GEN_VEXT_VV_RM(vsmul_vv_w)
-GEN_VEXT_VV_RM(vsmul_vv_d)
+GEN_VEXT_VV_RM(vsmul_vv_b, 1)
+GEN_VEXT_VV_RM(vsmul_vv_h, 2)
+GEN_VEXT_VV_RM(vsmul_vv_w, 4)
+GEN_VEXT_VV_RM(vsmul_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vsmul_vx_b, OP_SSS_B, H1, H1, vsmul8)
 RVVCALL(OPIVX2_RM, vsmul_vx_h, OP_SSS_H, H2, H2, vsmul16)
 RVVCALL(OPIVX2_RM, vsmul_vx_w, OP_SSS_W, H4, H4, vsmul32)
 RVVCALL(OPIVX2_RM, vsmul_vx_d, OP_SSS_D, H8, H8, vsmul64)
-GEN_VEXT_VX_RM(vsmul_vx_b)
-GEN_VEXT_VX_RM(vsmul_vx_h)
-GEN_VEXT_VX_RM(vsmul_vx_w)
-GEN_VEXT_VX_RM(vsmul_vx_d)
+GEN_VEXT_VX_RM(vsmul_vx_b, 1)
+GEN_VEXT_VX_RM(vsmul_vx_h, 2)
+GEN_VEXT_VX_RM(vsmul_vx_w, 4)
+GEN_VEXT_VX_RM(vsmul_vx_d, 8)
 
 /* Vector Single-Width Scaling Shift Instructions */
 static inline uint8_t
@@ -2769,19 +2777,19 @@ RVVCALL(OPIVV2_RM, vssrl_vv_b, OP_UUU_B, H1, H1, H1, vssrl8)
 RVVCALL(OPIVV2_RM, vssrl_vv_h, OP_UUU_H, H2, H2, H2, vssrl16)
 RVVCALL(OPIVV2_RM, vssrl_vv_w, OP_UUU_W, H4, H4, H4, vssrl32)
 RVVCALL(OPIVV2_RM, vssrl_vv_d, OP_UUU_D, H8, H8, H8, vssrl64)
-GEN_VEXT_VV_RM(vssrl_vv_b)
-GEN_VEXT_VV_RM(vssrl_vv_h)
-GEN_VEXT_VV_RM(vssrl_vv_w)
-GEN_VEXT_VV_RM(vssrl_vv_d)
+GEN_VEXT_VV_RM(vssrl_vv_b, 1)
+GEN_VEXT_VV_RM(vssrl_vv_h, 2)
+GEN_VEXT_VV_RM(vssrl_vv_w, 4)
+GEN_VEXT_VV_RM(vssrl_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssrl_vx_b, OP_UUU_B, H1, H1, vssrl8)
 RVVCALL(OPIVX2_RM, vssrl_vx_h, OP_UUU_H, H2, H2, vssrl16)
 RVVCALL(OPIVX2_RM, vssrl_vx_w, OP_UUU_W, H4, H4, vssrl32)
 RVVCALL(OPIVX2_RM, vssrl_vx_d, OP_UUU_D, H8, H8, vssrl64)
-GEN_VEXT_VX_RM(vssrl_vx_b)
-GEN_VEXT_VX_RM(vssrl_vx_h)
-GEN_VEXT_VX_RM(vssrl_vx_w)
-GEN_VEXT_VX_RM(vssrl_vx_d)
+GEN_VEXT_VX_RM(vssrl_vx_b, 1)
+GEN_VEXT_VX_RM(vssrl_vx_h, 2)
+GEN_VEXT_VX_RM(vssrl_vx_w, 4)
+GEN_VEXT_VX_RM(vssrl_vx_d, 8)
 
 static inline int8_t
 vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2828,19 +2836,19 @@ RVVCALL(OPIVV2_RM, vssra_vv_b, OP_SSS_B, H1, H1, H1, vssra8)
 RVVCALL(OPIVV2_RM, vssra_vv_h, OP_SSS_H, H2, H2, H2, vssra16)
 RVVCALL(OPIVV2_RM, vssra_vv_w, OP_SSS_W, H4, H4, H4, vssra32)
 RVVCALL(OPIVV2_RM, vssra_vv_d, OP_SSS_D, H8, H8, H8, vssra64)
-GEN_VEXT_VV_RM(vssra_vv_b)
-GEN_VEXT_VV_RM(vssra_vv_h)
-GEN_VEXT_VV_RM(vssra_vv_w)
-GEN_VEXT_VV_RM(vssra_vv_d)
+GEN_VEXT_VV_RM(vssra_vv_b, 1)
+GEN_VEXT_VV_RM(vssra_vv_h, 2)
+GEN_VEXT_VV_RM(vssra_vv_w, 4)
+GEN_VEXT_VV_RM(vssra_vv_d, 8)
 
 RVVCALL(OPIVX2_RM, vssra_vx_b, OP_SSS_B, H1, H1, vssra8)
 RVVCALL(OPIVX2_RM, vssra_vx_h, OP_SSS_H, H2, H2, vssra16)
 RVVCALL(OPIVX2_RM, vssra_vx_w, OP_SSS_W, H4, H4, vssra32)
 RVVCALL(OPIVX2_RM, vssra_vx_d, OP_SSS_D, H8, H8, vssra64)
-GEN_VEXT_VX_RM(vssra_vx_b)
-GEN_VEXT_VX_RM(vssra_vx_h)
-GEN_VEXT_VX_RM(vssra_vx_w)
-GEN_VEXT_VX_RM(vssra_vx_d)
+GEN_VEXT_VX_RM(vssra_vx_b, 1)
+GEN_VEXT_VX_RM(vssra_vx_h, 2)
+GEN_VEXT_VX_RM(vssra_vx_w, 4)
+GEN_VEXT_VX_RM(vssra_vx_d, 8)
 
 /* Vector Narrowing Fixed-Point Clip Instructions */
 static inline int8_t
@@ -2903,16 +2911,16 @@ vnclip32(CPURISCVState *env, int vxrm, int64_t a, int32_t b)
 RVVCALL(OPIVV2_RM, vnclip_wv_b, NOP_SSS_B, H1, H2, H1, vnclip8)
 RVVCALL(OPIVV2_RM, vnclip_wv_h, NOP_SSS_H, H2, H4, H2, vnclip16)
 RVVCALL(OPIVV2_RM, vnclip_wv_w, NOP_SSS_W, H4, H8, H4, vnclip32)
-GEN_VEXT_VV_RM(vnclip_wv_b)
-GEN_VEXT_VV_RM(vnclip_wv_h)
-GEN_VEXT_VV_RM(vnclip_wv_w)
+GEN_VEXT_VV_RM(vnclip_wv_b, 1)
+GEN_VEXT_VV_RM(vnclip_wv_h, 2)
+GEN_VEXT_VV_RM(vnclip_wv_w, 4)
 
 RVVCALL(OPIVX2_RM, vnclip_wx_b, NOP_SSS_B, H1, H2, vnclip8)
 RVVCALL(OPIVX2_RM, vnclip_wx_h, NOP_SSS_H, H2, H4, vnclip16)
 RVVCALL(OPIVX2_RM, vnclip_wx_w, NOP_SSS_W, H4, H8, vnclip32)
-GEN_VEXT_VX_RM(vnclip_wx_b)
-GEN_VEXT_VX_RM(vnclip_wx_h)
-GEN_VEXT_VX_RM(vnclip_wx_w)
+GEN_VEXT_VX_RM(vnclip_wx_b, 1)
+GEN_VEXT_VX_RM(vnclip_wx_h, 2)
+GEN_VEXT_VX_RM(vnclip_wx_w, 4)
 
 static inline uint8_t
 vnclipu8(CPURISCVState *env, int vxrm, uint16_t a, uint8_t b)
@@ -2965,16 +2973,16 @@ vnclipu32(CPURISCVState *env, int vxrm, uint64_t a, uint32_t b)
 RVVCALL(OPIVV2_RM, vnclipu_wv_b, NOP_UUU_B, H1, H2, H1, vnclipu8)
 RVVCALL(OPIVV2_RM, vnclipu_wv_h, NOP_UUU_H, H2, H4, H2, vnclipu16)
 RVVCALL(OPIVV2_RM, vnclipu_wv_w, NOP_UUU_W, H4, H8, H4, vnclipu32)
-GEN_VEXT_VV_RM(vnclipu_wv_b)
-GEN_VEXT_VV_RM(vnclipu_wv_h)
-GEN_VEXT_VV_RM(vnclipu_wv_w)
+GEN_VEXT_VV_RM(vnclipu_wv_b, 1)
+GEN_VEXT_VV_RM(vnclipu_wv_h, 2)
+GEN_VEXT_VV_RM(vnclipu_wv_w, 4)
 
 RVVCALL(OPIVX2_RM, vnclipu_wx_b, NOP_UUU_B, H1, H2, vnclipu8)
 RVVCALL(OPIVX2_RM, vnclipu_wx_h, NOP_UUU_H, H2, H4, vnclipu16)
 RVVCALL(OPIVX2_RM, vnclipu_wx_w, NOP_UUU_W, H4, H8, vnclipu32)
-GEN_VEXT_VX_RM(vnclipu_wx_b)
-GEN_VEXT_VX_RM(vnclipu_wx_h)
-GEN_VEXT_VX_RM(vnclipu_wx_w)
+GEN_VEXT_VX_RM(vnclipu_wx_b, 1)
+GEN_VEXT_VX_RM(vnclipu_wx_h, 2)
+GEN_VEXT_VX_RM(vnclipu_wx_w, 4)
 
 /*
  *** Vector Float Point Arithmetic Instructions
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 11/14] target/riscv: rvv: Add tail agnostic for vector floating-point instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07 10:05 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  11 +
 target/riscv/vector_helper.c            | 452 ++++++++++++++----------
 2 files changed, 267 insertions(+), 196 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 24784b6f5f..5bc340f419 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -2391,6 +2391,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
@@ -2473,6 +2474,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)            \
         gen_set_rm(s, RISCV_FRM_DYN);                             \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);            \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);              \
         return opfvf_trans(a->rd, a->rs1, a->rs2, data,           \
                            fns[s->sew - 1], s);                   \
     }                                                             \
@@ -2511,6 +2513,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
                                                                  \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),   \
                            vreg_ofs(s, a->rs1),                  \
                            vreg_ofs(s, a->rs2), cpu_env,         \
@@ -2550,6 +2553,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         gen_set_rm(s, RISCV_FRM_DYN);                            \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
         return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
                            fns[s->sew - 1], s);                  \
     }                                                            \
@@ -2586,6 +2590,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
@@ -2625,6 +2630,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         gen_set_rm(s, RISCV_FRM_DYN);                            \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
         return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
                            fns[s->sew - 1], s);                  \
     }                                                            \
@@ -2708,6 +2714,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs2), cpu_env,
                            s->cfg_ptr->vlen / 8,
@@ -2921,6 +2928,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
@@ -2972,6 +2980,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
@@ -3039,6 +3048,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
@@ -3092,6 +3102,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 067095e19a..d654377fc8 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -2997,13 +2997,17 @@ static void do_##NAME(void *vd, void *vs1, void *vs2, int i,   \
     *((TD *)vd + HD(i)) = OP(s2, s1, &env->fp_status);         \
 }
 
-#define GEN_VEXT_VV_ENV(NAME)                             \
+#define GEN_VEXT_VV_ENV(NAME, ESZ)                        \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     uint32_t vm = vext_vm(desc);                          \
     uint32_t vl = env->vl;                                \
+    uint32_t total_elems =                                \
+        vext_get_total_elems(env_archcpu(env),            \
+                             env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
                                                           \
     for (i = env->vstart; i < vl; i++) {                  \
@@ -3013,14 +3017,18 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
         do_##NAME(vd, vs1, vs2, i, env);                  \
     }                                                     \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
+                                     vl * ESZ,            \
+                                     total_elems * ESZ);  \
 }
 
 RVVCALL(OPFVV2, vfadd_vv_h, OP_UUU_H, H2, H2, H2, float16_add)
 RVVCALL(OPFVV2, vfadd_vv_w, OP_UUU_W, H4, H4, H4, float32_add)
 RVVCALL(OPFVV2, vfadd_vv_d, OP_UUU_D, H8, H8, H8, float64_add)
-GEN_VEXT_VV_ENV(vfadd_vv_h)
-GEN_VEXT_VV_ENV(vfadd_vv_w)
-GEN_VEXT_VV_ENV(vfadd_vv_d)
+GEN_VEXT_VV_ENV(vfadd_vv_h, 2)
+GEN_VEXT_VV_ENV(vfadd_vv_w, 4)
+GEN_VEXT_VV_ENV(vfadd_vv_d, 8)
 
 #define OPFVF2(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)        \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
@@ -3030,13 +3038,17 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
     *((TD *)vd + HD(i)) = OP(s2, (TX1)(T1)s1, &env->fp_status);\
 }
 
-#define GEN_VEXT_VF(NAME)                                 \
+#define GEN_VEXT_VF(NAME, ESZ)                            \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     uint32_t vm = vext_vm(desc);                          \
     uint32_t vl = env->vl;                                \
+    uint32_t total_elems =                                \
+        vext_get_total_elems(env_archcpu(env),            \
+                             env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
                                                           \
     for (i = env->vstart; i < vl; i++) {                  \
@@ -3046,27 +3058,31 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
         do_##NAME(vd, s1, vs2, i, env);                   \
     }                                                     \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
+                                     vl * ESZ,            \
+                                     total_elems * ESZ);  \
 }
 
 RVVCALL(OPFVF2, vfadd_vf_h, OP_UUU_H, H2, H2, float16_add)
 RVVCALL(OPFVF2, vfadd_vf_w, OP_UUU_W, H4, H4, float32_add)
 RVVCALL(OPFVF2, vfadd_vf_d, OP_UUU_D, H8, H8, float64_add)
-GEN_VEXT_VF(vfadd_vf_h)
-GEN_VEXT_VF(vfadd_vf_w)
-GEN_VEXT_VF(vfadd_vf_d)
+GEN_VEXT_VF(vfadd_vf_h, 2)
+GEN_VEXT_VF(vfadd_vf_w, 4)
+GEN_VEXT_VF(vfadd_vf_d, 8)
 
 RVVCALL(OPFVV2, vfsub_vv_h, OP_UUU_H, H2, H2, H2, float16_sub)
 RVVCALL(OPFVV2, vfsub_vv_w, OP_UUU_W, H4, H4, H4, float32_sub)
 RVVCALL(OPFVV2, vfsub_vv_d, OP_UUU_D, H8, H8, H8, float64_sub)
-GEN_VEXT_VV_ENV(vfsub_vv_h)
-GEN_VEXT_VV_ENV(vfsub_vv_w)
-GEN_VEXT_VV_ENV(vfsub_vv_d)
+GEN_VEXT_VV_ENV(vfsub_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsub_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsub_vv_d, 8)
 RVVCALL(OPFVF2, vfsub_vf_h, OP_UUU_H, H2, H2, float16_sub)
 RVVCALL(OPFVF2, vfsub_vf_w, OP_UUU_W, H4, H4, float32_sub)
 RVVCALL(OPFVF2, vfsub_vf_d, OP_UUU_D, H8, H8, float64_sub)
-GEN_VEXT_VF(vfsub_vf_h)
-GEN_VEXT_VF(vfsub_vf_w)
-GEN_VEXT_VF(vfsub_vf_d)
+GEN_VEXT_VF(vfsub_vf_h, 2)
+GEN_VEXT_VF(vfsub_vf_w, 4)
+GEN_VEXT_VF(vfsub_vf_d, 8)
 
 static uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3086,9 +3102,9 @@ static uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrsub_vf_h, OP_UUU_H, H2, H2, float16_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_w, OP_UUU_W, H4, H4, float32_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub)
-GEN_VEXT_VF(vfrsub_vf_h)
-GEN_VEXT_VF(vfrsub_vf_w)
-GEN_VEXT_VF(vfrsub_vf_d)
+GEN_VEXT_VF(vfrsub_vf_h, 2)
+GEN_VEXT_VF(vfrsub_vf_w, 4)
+GEN_VEXT_VF(vfrsub_vf_d, 8)
 
 /* Vector Widening Floating-Point Add/Subtract Instructions */
 static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s)
@@ -3106,12 +3122,12 @@ static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16)
 RVVCALL(OPFVV2, vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32)
-GEN_VEXT_VV_ENV(vfwadd_vv_h)
-GEN_VEXT_VV_ENV(vfwadd_vv_w)
+GEN_VEXT_VV_ENV(vfwadd_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwadd_vv_w, 8)
 RVVCALL(OPFVF2, vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16)
 RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32)
-GEN_VEXT_VF(vfwadd_vf_h)
-GEN_VEXT_VF(vfwadd_vf_w)
+GEN_VEXT_VF(vfwadd_vf_h, 4)
+GEN_VEXT_VF(vfwadd_vf_w, 8)
 
 static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3128,12 +3144,12 @@ static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16)
 RVVCALL(OPFVV2, vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32)
-GEN_VEXT_VV_ENV(vfwsub_vv_h)
-GEN_VEXT_VV_ENV(vfwsub_vv_w)
+GEN_VEXT_VV_ENV(vfwsub_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwsub_vv_w, 8)
 RVVCALL(OPFVF2, vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16)
 RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32)
-GEN_VEXT_VF(vfwsub_vf_h)
-GEN_VEXT_VF(vfwsub_vf_w)
+GEN_VEXT_VF(vfwsub_vf_h, 4)
+GEN_VEXT_VF(vfwsub_vf_w, 8)
 
 static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -3147,12 +3163,12 @@ static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16)
 RVVCALL(OPFVV2, vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32)
-GEN_VEXT_VV_ENV(vfwadd_wv_h)
-GEN_VEXT_VV_ENV(vfwadd_wv_w)
+GEN_VEXT_VV_ENV(vfwadd_wv_h, 4)
+GEN_VEXT_VV_ENV(vfwadd_wv_w, 8)
 RVVCALL(OPFVF2, vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16)
 RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32)
-GEN_VEXT_VF(vfwadd_wf_h)
-GEN_VEXT_VF(vfwadd_wf_w)
+GEN_VEXT_VF(vfwadd_wf_h, 4)
+GEN_VEXT_VF(vfwadd_wf_w, 8)
 
 static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -3166,39 +3182,39 @@ static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16)
 RVVCALL(OPFVV2, vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32)
-GEN_VEXT_VV_ENV(vfwsub_wv_h)
-GEN_VEXT_VV_ENV(vfwsub_wv_w)
+GEN_VEXT_VV_ENV(vfwsub_wv_h, 4)
+GEN_VEXT_VV_ENV(vfwsub_wv_w, 8)
 RVVCALL(OPFVF2, vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16)
 RVVCALL(OPFVF2, vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32)
-GEN_VEXT_VF(vfwsub_wf_h)
-GEN_VEXT_VF(vfwsub_wf_w)
+GEN_VEXT_VF(vfwsub_wf_h, 4)
+GEN_VEXT_VF(vfwsub_wf_w, 8)
 
 /* Vector Single-Width Floating-Point Multiply/Divide Instructions */
 RVVCALL(OPFVV2, vfmul_vv_h, OP_UUU_H, H2, H2, H2, float16_mul)
 RVVCALL(OPFVV2, vfmul_vv_w, OP_UUU_W, H4, H4, H4, float32_mul)
 RVVCALL(OPFVV2, vfmul_vv_d, OP_UUU_D, H8, H8, H8, float64_mul)
-GEN_VEXT_VV_ENV(vfmul_vv_h)
-GEN_VEXT_VV_ENV(vfmul_vv_w)
-GEN_VEXT_VV_ENV(vfmul_vv_d)
+GEN_VEXT_VV_ENV(vfmul_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmul_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmul_vv_d, 8)
 RVVCALL(OPFVF2, vfmul_vf_h, OP_UUU_H, H2, H2, float16_mul)
 RVVCALL(OPFVF2, vfmul_vf_w, OP_UUU_W, H4, H4, float32_mul)
 RVVCALL(OPFVF2, vfmul_vf_d, OP_UUU_D, H8, H8, float64_mul)
-GEN_VEXT_VF(vfmul_vf_h)
-GEN_VEXT_VF(vfmul_vf_w)
-GEN_VEXT_VF(vfmul_vf_d)
+GEN_VEXT_VF(vfmul_vf_h, 2)
+GEN_VEXT_VF(vfmul_vf_w, 4)
+GEN_VEXT_VF(vfmul_vf_d, 8)
 
 RVVCALL(OPFVV2, vfdiv_vv_h, OP_UUU_H, H2, H2, H2, float16_div)
 RVVCALL(OPFVV2, vfdiv_vv_w, OP_UUU_W, H4, H4, H4, float32_div)
 RVVCALL(OPFVV2, vfdiv_vv_d, OP_UUU_D, H8, H8, H8, float64_div)
-GEN_VEXT_VV_ENV(vfdiv_vv_h)
-GEN_VEXT_VV_ENV(vfdiv_vv_w)
-GEN_VEXT_VV_ENV(vfdiv_vv_d)
+GEN_VEXT_VV_ENV(vfdiv_vv_h, 2)
+GEN_VEXT_VV_ENV(vfdiv_vv_w, 4)
+GEN_VEXT_VV_ENV(vfdiv_vv_d, 8)
 RVVCALL(OPFVF2, vfdiv_vf_h, OP_UUU_H, H2, H2, float16_div)
 RVVCALL(OPFVF2, vfdiv_vf_w, OP_UUU_W, H4, H4, float32_div)
 RVVCALL(OPFVF2, vfdiv_vf_d, OP_UUU_D, H8, H8, float64_div)
-GEN_VEXT_VF(vfdiv_vf_h)
-GEN_VEXT_VF(vfdiv_vf_w)
-GEN_VEXT_VF(vfdiv_vf_d)
+GEN_VEXT_VF(vfdiv_vf_h, 2)
+GEN_VEXT_VF(vfdiv_vf_w, 4)
+GEN_VEXT_VF(vfdiv_vf_d, 8)
 
 static uint16_t float16_rdiv(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3218,9 +3234,9 @@ static uint64_t float64_rdiv(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrdiv_vf_h, OP_UUU_H, H2, H2, float16_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_w, OP_UUU_W, H4, H4, float32_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_d, OP_UUU_D, H8, H8, float64_rdiv)
-GEN_VEXT_VF(vfrdiv_vf_h)
-GEN_VEXT_VF(vfrdiv_vf_w)
-GEN_VEXT_VF(vfrdiv_vf_d)
+GEN_VEXT_VF(vfrdiv_vf_h, 2)
+GEN_VEXT_VF(vfrdiv_vf_w, 4)
+GEN_VEXT_VF(vfrdiv_vf_d, 8)
 
 /* Vector Widening Floating-Point Multiply */
 static uint32_t vfwmul16(uint16_t a, uint16_t b, float_status *s)
@@ -3237,12 +3253,12 @@ static uint64_t vfwmul32(uint32_t a, uint32_t b, float_status *s)
 }
 RVVCALL(OPFVV2, vfwmul_vv_h, WOP_UUU_H, H4, H2, H2, vfwmul16)
 RVVCALL(OPFVV2, vfwmul_vv_w, WOP_UUU_W, H8, H4, H4, vfwmul32)
-GEN_VEXT_VV_ENV(vfwmul_vv_h)
-GEN_VEXT_VV_ENV(vfwmul_vv_w)
+GEN_VEXT_VV_ENV(vfwmul_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwmul_vv_w, 8)
 RVVCALL(OPFVF2, vfwmul_vf_h, WOP_UUU_H, H4, H2, vfwmul16)
 RVVCALL(OPFVF2, vfwmul_vf_w, WOP_UUU_W, H8, H4, vfwmul32)
-GEN_VEXT_VF(vfwmul_vf_h)
-GEN_VEXT_VF(vfwmul_vf_w)
+GEN_VEXT_VF(vfwmul_vf_h, 4)
+GEN_VEXT_VF(vfwmul_vf_w, 8)
 
 /* Vector Single-Width Floating-Point Fused Multiply-Add Instructions */
 #define OPFVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)       \
@@ -3273,9 +3289,9 @@ static uint64_t fmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmacc_vv_h, OP_UUU_H, H2, H2, H2, fmacc16)
 RVVCALL(OPFVV3, vfmacc_vv_w, OP_UUU_W, H4, H4, H4, fmacc32)
 RVVCALL(OPFVV3, vfmacc_vv_d, OP_UUU_D, H8, H8, H8, fmacc64)
-GEN_VEXT_VV_ENV(vfmacc_vv_h)
-GEN_VEXT_VV_ENV(vfmacc_vv_w)
-GEN_VEXT_VV_ENV(vfmacc_vv_d)
+GEN_VEXT_VV_ENV(vfmacc_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmacc_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmacc_vv_d, 8)
 
 #define OPFVF3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)           \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
@@ -3289,9 +3305,9 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
 RVVCALL(OPFVF3, vfmacc_vf_h, OP_UUU_H, H2, H2, fmacc16)
 RVVCALL(OPFVF3, vfmacc_vf_w, OP_UUU_W, H4, H4, fmacc32)
 RVVCALL(OPFVF3, vfmacc_vf_d, OP_UUU_D, H8, H8, fmacc64)
-GEN_VEXT_VF(vfmacc_vf_h)
-GEN_VEXT_VF(vfmacc_vf_w)
-GEN_VEXT_VF(vfmacc_vf_d)
+GEN_VEXT_VF(vfmacc_vf_h, 2)
+GEN_VEXT_VF(vfmacc_vf_w, 4)
+GEN_VEXT_VF(vfmacc_vf_d, 8)
 
 static uint16_t fnmacc16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3314,15 +3330,15 @@ static uint64_t fnmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmacc_vv_h, OP_UUU_H, H2, H2, H2, fnmacc16)
 RVVCALL(OPFVV3, vfnmacc_vv_w, OP_UUU_W, H4, H4, H4, fnmacc32)
 RVVCALL(OPFVV3, vfnmacc_vv_d, OP_UUU_D, H8, H8, H8, fnmacc64)
-GEN_VEXT_VV_ENV(vfnmacc_vv_h)
-GEN_VEXT_VV_ENV(vfnmacc_vv_w)
-GEN_VEXT_VV_ENV(vfnmacc_vv_d)
+GEN_VEXT_VV_ENV(vfnmacc_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmacc_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmacc_vv_d, 8)
 RVVCALL(OPFVF3, vfnmacc_vf_h, OP_UUU_H, H2, H2, fnmacc16)
 RVVCALL(OPFVF3, vfnmacc_vf_w, OP_UUU_W, H4, H4, fnmacc32)
 RVVCALL(OPFVF3, vfnmacc_vf_d, OP_UUU_D, H8, H8, fnmacc64)
-GEN_VEXT_VF(vfnmacc_vf_h)
-GEN_VEXT_VF(vfnmacc_vf_w)
-GEN_VEXT_VF(vfnmacc_vf_d)
+GEN_VEXT_VF(vfnmacc_vf_h, 2)
+GEN_VEXT_VF(vfnmacc_vf_w, 4)
+GEN_VEXT_VF(vfnmacc_vf_d, 8)
 
 static uint16_t fmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3342,15 +3358,15 @@ static uint64_t fmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsac_vv_h, OP_UUU_H, H2, H2, H2, fmsac16)
 RVVCALL(OPFVV3, vfmsac_vv_w, OP_UUU_W, H4, H4, H4, fmsac32)
 RVVCALL(OPFVV3, vfmsac_vv_d, OP_UUU_D, H8, H8, H8, fmsac64)
-GEN_VEXT_VV_ENV(vfmsac_vv_h)
-GEN_VEXT_VV_ENV(vfmsac_vv_w)
-GEN_VEXT_VV_ENV(vfmsac_vv_d)
+GEN_VEXT_VV_ENV(vfmsac_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmsac_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmsac_vv_d, 8)
 RVVCALL(OPFVF3, vfmsac_vf_h, OP_UUU_H, H2, H2, fmsac16)
 RVVCALL(OPFVF3, vfmsac_vf_w, OP_UUU_W, H4, H4, fmsac32)
 RVVCALL(OPFVF3, vfmsac_vf_d, OP_UUU_D, H8, H8, fmsac64)
-GEN_VEXT_VF(vfmsac_vf_h)
-GEN_VEXT_VF(vfmsac_vf_w)
-GEN_VEXT_VF(vfmsac_vf_d)
+GEN_VEXT_VF(vfmsac_vf_h, 2)
+GEN_VEXT_VF(vfmsac_vf_w, 4)
+GEN_VEXT_VF(vfmsac_vf_d, 8)
 
 static uint16_t fnmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3370,15 +3386,15 @@ static uint64_t fnmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsac_vv_h, OP_UUU_H, H2, H2, H2, fnmsac16)
 RVVCALL(OPFVV3, vfnmsac_vv_w, OP_UUU_W, H4, H4, H4, fnmsac32)
 RVVCALL(OPFVV3, vfnmsac_vv_d, OP_UUU_D, H8, H8, H8, fnmsac64)
-GEN_VEXT_VV_ENV(vfnmsac_vv_h)
-GEN_VEXT_VV_ENV(vfnmsac_vv_w)
-GEN_VEXT_VV_ENV(vfnmsac_vv_d)
+GEN_VEXT_VV_ENV(vfnmsac_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmsac_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmsac_vv_d, 8)
 RVVCALL(OPFVF3, vfnmsac_vf_h, OP_UUU_H, H2, H2, fnmsac16)
 RVVCALL(OPFVF3, vfnmsac_vf_w, OP_UUU_W, H4, H4, fnmsac32)
 RVVCALL(OPFVF3, vfnmsac_vf_d, OP_UUU_D, H8, H8, fnmsac64)
-GEN_VEXT_VF(vfnmsac_vf_h)
-GEN_VEXT_VF(vfnmsac_vf_w)
-GEN_VEXT_VF(vfnmsac_vf_d)
+GEN_VEXT_VF(vfnmsac_vf_h, 2)
+GEN_VEXT_VF(vfnmsac_vf_w, 4)
+GEN_VEXT_VF(vfnmsac_vf_d, 8)
 
 static uint16_t fmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3398,15 +3414,15 @@ static uint64_t fmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmadd_vv_h, OP_UUU_H, H2, H2, H2, fmadd16)
 RVVCALL(OPFVV3, vfmadd_vv_w, OP_UUU_W, H4, H4, H4, fmadd32)
 RVVCALL(OPFVV3, vfmadd_vv_d, OP_UUU_D, H8, H8, H8, fmadd64)
-GEN_VEXT_VV_ENV(vfmadd_vv_h)
-GEN_VEXT_VV_ENV(vfmadd_vv_w)
-GEN_VEXT_VV_ENV(vfmadd_vv_d)
+GEN_VEXT_VV_ENV(vfmadd_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmadd_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmadd_vv_d, 8)
 RVVCALL(OPFVF3, vfmadd_vf_h, OP_UUU_H, H2, H2, fmadd16)
 RVVCALL(OPFVF3, vfmadd_vf_w, OP_UUU_W, H4, H4, fmadd32)
 RVVCALL(OPFVF3, vfmadd_vf_d, OP_UUU_D, H8, H8, fmadd64)
-GEN_VEXT_VF(vfmadd_vf_h)
-GEN_VEXT_VF(vfmadd_vf_w)
-GEN_VEXT_VF(vfmadd_vf_d)
+GEN_VEXT_VF(vfmadd_vf_h, 2)
+GEN_VEXT_VF(vfmadd_vf_w, 4)
+GEN_VEXT_VF(vfmadd_vf_d, 8)
 
 static uint16_t fnmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3429,15 +3445,15 @@ static uint64_t fnmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmadd_vv_h, OP_UUU_H, H2, H2, H2, fnmadd16)
 RVVCALL(OPFVV3, vfnmadd_vv_w, OP_UUU_W, H4, H4, H4, fnmadd32)
 RVVCALL(OPFVV3, vfnmadd_vv_d, OP_UUU_D, H8, H8, H8, fnmadd64)
-GEN_VEXT_VV_ENV(vfnmadd_vv_h)
-GEN_VEXT_VV_ENV(vfnmadd_vv_w)
-GEN_VEXT_VV_ENV(vfnmadd_vv_d)
+GEN_VEXT_VV_ENV(vfnmadd_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmadd_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmadd_vv_d, 8)
 RVVCALL(OPFVF3, vfnmadd_vf_h, OP_UUU_H, H2, H2, fnmadd16)
 RVVCALL(OPFVF3, vfnmadd_vf_w, OP_UUU_W, H4, H4, fnmadd32)
 RVVCALL(OPFVF3, vfnmadd_vf_d, OP_UUU_D, H8, H8, fnmadd64)
-GEN_VEXT_VF(vfnmadd_vf_h)
-GEN_VEXT_VF(vfnmadd_vf_w)
-GEN_VEXT_VF(vfnmadd_vf_d)
+GEN_VEXT_VF(vfnmadd_vf_h, 2)
+GEN_VEXT_VF(vfnmadd_vf_w, 4)
+GEN_VEXT_VF(vfnmadd_vf_d, 8)
 
 static uint16_t fmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3457,15 +3473,15 @@ static uint64_t fmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsub_vv_h, OP_UUU_H, H2, H2, H2, fmsub16)
 RVVCALL(OPFVV3, vfmsub_vv_w, OP_UUU_W, H4, H4, H4, fmsub32)
 RVVCALL(OPFVV3, vfmsub_vv_d, OP_UUU_D, H8, H8, H8, fmsub64)
-GEN_VEXT_VV_ENV(vfmsub_vv_h)
-GEN_VEXT_VV_ENV(vfmsub_vv_w)
-GEN_VEXT_VV_ENV(vfmsub_vv_d)
+GEN_VEXT_VV_ENV(vfmsub_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmsub_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmsub_vv_d, 8)
 RVVCALL(OPFVF3, vfmsub_vf_h, OP_UUU_H, H2, H2, fmsub16)
 RVVCALL(OPFVF3, vfmsub_vf_w, OP_UUU_W, H4, H4, fmsub32)
 RVVCALL(OPFVF3, vfmsub_vf_d, OP_UUU_D, H8, H8, fmsub64)
-GEN_VEXT_VF(vfmsub_vf_h)
-GEN_VEXT_VF(vfmsub_vf_w)
-GEN_VEXT_VF(vfmsub_vf_d)
+GEN_VEXT_VF(vfmsub_vf_h, 2)
+GEN_VEXT_VF(vfmsub_vf_w, 4)
+GEN_VEXT_VF(vfmsub_vf_d, 8)
 
 static uint16_t fnmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3485,15 +3501,15 @@ static uint64_t fnmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsub_vv_h, OP_UUU_H, H2, H2, H2, fnmsub16)
 RVVCALL(OPFVV3, vfnmsub_vv_w, OP_UUU_W, H4, H4, H4, fnmsub32)
 RVVCALL(OPFVV3, vfnmsub_vv_d, OP_UUU_D, H8, H8, H8, fnmsub64)
-GEN_VEXT_VV_ENV(vfnmsub_vv_h)
-GEN_VEXT_VV_ENV(vfnmsub_vv_w)
-GEN_VEXT_VV_ENV(vfnmsub_vv_d)
+GEN_VEXT_VV_ENV(vfnmsub_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmsub_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmsub_vv_d, 8)
 RVVCALL(OPFVF3, vfnmsub_vf_h, OP_UUU_H, H2, H2, fnmsub16)
 RVVCALL(OPFVF3, vfnmsub_vf_w, OP_UUU_W, H4, H4, fnmsub32)
 RVVCALL(OPFVF3, vfnmsub_vf_d, OP_UUU_D, H8, H8, fnmsub64)
-GEN_VEXT_VF(vfnmsub_vf_h)
-GEN_VEXT_VF(vfnmsub_vf_w)
-GEN_VEXT_VF(vfnmsub_vf_d)
+GEN_VEXT_VF(vfnmsub_vf_h, 2)
+GEN_VEXT_VF(vfnmsub_vf_w, 4)
+GEN_VEXT_VF(vfnmsub_vf_d, 8)
 
 /* Vector Widening Floating-Point Fused Multiply-Add Instructions */
 static uint32_t fwmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
@@ -3510,12 +3526,12 @@ static uint64_t fwmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwmacc16)
 RVVCALL(OPFVV3, vfwmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwmacc32)
-GEN_VEXT_VV_ENV(vfwmacc_vv_h)
-GEN_VEXT_VV_ENV(vfwmacc_vv_w)
+GEN_VEXT_VV_ENV(vfwmacc_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwmacc_vv_w, 8)
 RVVCALL(OPFVF3, vfwmacc_vf_h, WOP_UUU_H, H4, H2, fwmacc16)
 RVVCALL(OPFVF3, vfwmacc_vf_w, WOP_UUU_W, H8, H4, fwmacc32)
-GEN_VEXT_VF(vfwmacc_vf_h)
-GEN_VEXT_VF(vfwmacc_vf_w)
+GEN_VEXT_VF(vfwmacc_vf_h, 4)
+GEN_VEXT_VF(vfwmacc_vf_w, 8)
 
 static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3533,12 +3549,12 @@ static uint64_t fwnmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwnmacc16)
 RVVCALL(OPFVV3, vfwnmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwnmacc32)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_h)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_w)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_w, 8)
 RVVCALL(OPFVF3, vfwnmacc_vf_h, WOP_UUU_H, H4, H2, fwnmacc16)
 RVVCALL(OPFVF3, vfwnmacc_vf_w, WOP_UUU_W, H8, H4, fwnmacc32)
-GEN_VEXT_VF(vfwnmacc_vf_h)
-GEN_VEXT_VF(vfwnmacc_vf_w)
+GEN_VEXT_VF(vfwnmacc_vf_h, 4)
+GEN_VEXT_VF(vfwnmacc_vf_w, 8)
 
 static uint32_t fwmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3556,12 +3572,12 @@ static uint64_t fwmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwmsac16)
 RVVCALL(OPFVV3, vfwmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwmsac32)
-GEN_VEXT_VV_ENV(vfwmsac_vv_h)
-GEN_VEXT_VV_ENV(vfwmsac_vv_w)
+GEN_VEXT_VV_ENV(vfwmsac_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwmsac_vv_w, 8)
 RVVCALL(OPFVF3, vfwmsac_vf_h, WOP_UUU_H, H4, H2, fwmsac16)
 RVVCALL(OPFVF3, vfwmsac_vf_w, WOP_UUU_W, H8, H4, fwmsac32)
-GEN_VEXT_VF(vfwmsac_vf_h)
-GEN_VEXT_VF(vfwmsac_vf_w)
+GEN_VEXT_VF(vfwmsac_vf_h, 4)
+GEN_VEXT_VF(vfwmsac_vf_w, 8)
 
 static uint32_t fwnmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3579,12 +3595,12 @@ static uint64_t fwnmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwnmsac16)
 RVVCALL(OPFVV3, vfwnmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwnmsac32)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_h)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_w)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_w, 8)
 RVVCALL(OPFVF3, vfwnmsac_vf_h, WOP_UUU_H, H4, H2, fwnmsac16)
 RVVCALL(OPFVF3, vfwnmsac_vf_w, WOP_UUU_W, H8, H4, fwnmsac32)
-GEN_VEXT_VF(vfwnmsac_vf_h)
-GEN_VEXT_VF(vfwnmsac_vf_w)
+GEN_VEXT_VF(vfwnmsac_vf_h, 4)
+GEN_VEXT_VF(vfwnmsac_vf_w, 8)
 
 /* Vector Floating-Point Square-Root Instruction */
 /* (TD, T2, TX2) */
@@ -3600,12 +3616,16 @@ static void do_##NAME(void *vd, void *vs2, int i,      \
     *((TD *)vd + HD(i)) = OP(s2, &env->fp_status);     \
 }
 
-#define GEN_VEXT_V_ENV(NAME)                           \
+#define GEN_VEXT_V_ENV(NAME, ESZ)                      \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         CPURISCVState *env, uint32_t desc)             \
 {                                                      \
     uint32_t vm = vext_vm(desc);                       \
     uint32_t vl = env->vl;                             \
+    uint32_t total_elems =                             \
+        vext_get_total_elems(env_archcpu(env),         \
+                            env->vtype);               \
+    uint32_t vta = vext_vta(desc);                     \
     uint32_t i;                                        \
                                                        \
     if (vl == 0) {                                     \
@@ -3618,14 +3638,19 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         do_##NAME(vd, vs2, i, env);                    \
     }                                                  \
     env->vstart = 0;                                   \
+    /* set tail elements to 1s */                      \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,      \
+                                     vl * ESZ,         \
+                                     total_elems *     \
+                                     ESZ);             \
 }
 
 RVVCALL(OPFVV1, vfsqrt_v_h, OP_UU_H, H2, H2, float16_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_w, OP_UU_W, H4, H4, float32_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_d, OP_UU_D, H8, H8, float64_sqrt)
-GEN_VEXT_V_ENV(vfsqrt_v_h)
-GEN_VEXT_V_ENV(vfsqrt_v_w)
-GEN_VEXT_V_ENV(vfsqrt_v_d)
+GEN_VEXT_V_ENV(vfsqrt_v_h, 2)
+GEN_VEXT_V_ENV(vfsqrt_v_w, 4)
+GEN_VEXT_V_ENV(vfsqrt_v_d, 8)
 
 /*
  * Vector Floating-Point Reciprocal Square-Root Estimate Instruction
@@ -3805,9 +3830,9 @@ static float64 frsqrt7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrsqrt7_v_h, OP_UU_H, H2, H2, frsqrt7_h)
 RVVCALL(OPFVV1, vfrsqrt7_v_w, OP_UU_W, H4, H4, frsqrt7_s)
 RVVCALL(OPFVV1, vfrsqrt7_v_d, OP_UU_D, H8, H8, frsqrt7_d)
-GEN_VEXT_V_ENV(vfrsqrt7_v_h)
-GEN_VEXT_V_ENV(vfrsqrt7_v_w)
-GEN_VEXT_V_ENV(vfrsqrt7_v_d)
+GEN_VEXT_V_ENV(vfrsqrt7_v_h, 2)
+GEN_VEXT_V_ENV(vfrsqrt7_v_w, 4)
+GEN_VEXT_V_ENV(vfrsqrt7_v_d, 8)
 
 /*
  * Vector Floating-Point Reciprocal Estimate Instruction
@@ -3996,36 +4021,36 @@ static float64 frec7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrec7_v_h, OP_UU_H, H2, H2, frec7_h)
 RVVCALL(OPFVV1, vfrec7_v_w, OP_UU_W, H4, H4, frec7_s)
 RVVCALL(OPFVV1, vfrec7_v_d, OP_UU_D, H8, H8, frec7_d)
-GEN_VEXT_V_ENV(vfrec7_v_h)
-GEN_VEXT_V_ENV(vfrec7_v_w)
-GEN_VEXT_V_ENV(vfrec7_v_d)
+GEN_VEXT_V_ENV(vfrec7_v_h, 2)
+GEN_VEXT_V_ENV(vfrec7_v_w, 4)
+GEN_VEXT_V_ENV(vfrec7_v_d, 8)
 
 /* Vector Floating-Point MIN/MAX Instructions */
 RVVCALL(OPFVV2, vfmin_vv_h, OP_UUU_H, H2, H2, H2, float16_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_w, OP_UUU_W, H4, H4, H4, float32_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_d, OP_UUU_D, H8, H8, H8, float64_minimum_number)
-GEN_VEXT_VV_ENV(vfmin_vv_h)
-GEN_VEXT_VV_ENV(vfmin_vv_w)
-GEN_VEXT_VV_ENV(vfmin_vv_d)
+GEN_VEXT_VV_ENV(vfmin_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmin_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmin_vv_d, 8)
 RVVCALL(OPFVF2, vfmin_vf_h, OP_UUU_H, H2, H2, float16_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_w, OP_UUU_W, H4, H4, float32_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_d, OP_UUU_D, H8, H8, float64_minimum_number)
-GEN_VEXT_VF(vfmin_vf_h)
-GEN_VEXT_VF(vfmin_vf_w)
-GEN_VEXT_VF(vfmin_vf_d)
+GEN_VEXT_VF(vfmin_vf_h, 2)
+GEN_VEXT_VF(vfmin_vf_w, 4)
+GEN_VEXT_VF(vfmin_vf_d, 8)
 
 RVVCALL(OPFVV2, vfmax_vv_h, OP_UUU_H, H2, H2, H2, float16_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_w, OP_UUU_W, H4, H4, H4, float32_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_d, OP_UUU_D, H8, H8, H8, float64_maximum_number)
-GEN_VEXT_VV_ENV(vfmax_vv_h)
-GEN_VEXT_VV_ENV(vfmax_vv_w)
-GEN_VEXT_VV_ENV(vfmax_vv_d)
+GEN_VEXT_VV_ENV(vfmax_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmax_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmax_vv_d, 8)
 RVVCALL(OPFVF2, vfmax_vf_h, OP_UUU_H, H2, H2, float16_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_w, OP_UUU_W, H4, H4, float32_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_d, OP_UUU_D, H8, H8, float64_maximum_number)
-GEN_VEXT_VF(vfmax_vf_h)
-GEN_VEXT_VF(vfmax_vf_w)
-GEN_VEXT_VF(vfmax_vf_d)
+GEN_VEXT_VF(vfmax_vf_h, 2)
+GEN_VEXT_VF(vfmax_vf_w, 4)
+GEN_VEXT_VF(vfmax_vf_d, 8)
 
 /* Vector Floating-Point Sign-Injection Instructions */
 static uint16_t fsgnj16(uint16_t a, uint16_t b, float_status *s)
@@ -4046,15 +4071,15 @@ static uint64_t fsgnj64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnj_vv_h, OP_UUU_H, H2, H2, H2, fsgnj16)
 RVVCALL(OPFVV2, vfsgnj_vv_w, OP_UUU_W, H4, H4, H4, fsgnj32)
 RVVCALL(OPFVV2, vfsgnj_vv_d, OP_UUU_D, H8, H8, H8, fsgnj64)
-GEN_VEXT_VV_ENV(vfsgnj_vv_h)
-GEN_VEXT_VV_ENV(vfsgnj_vv_w)
-GEN_VEXT_VV_ENV(vfsgnj_vv_d)
+GEN_VEXT_VV_ENV(vfsgnj_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsgnj_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsgnj_vv_d, 8)
 RVVCALL(OPFVF2, vfsgnj_vf_h, OP_UUU_H, H2, H2, fsgnj16)
 RVVCALL(OPFVF2, vfsgnj_vf_w, OP_UUU_W, H4, H4, fsgnj32)
 RVVCALL(OPFVF2, vfsgnj_vf_d, OP_UUU_D, H8, H8, fsgnj64)
-GEN_VEXT_VF(vfsgnj_vf_h)
-GEN_VEXT_VF(vfsgnj_vf_w)
-GEN_VEXT_VF(vfsgnj_vf_d)
+GEN_VEXT_VF(vfsgnj_vf_h, 2)
+GEN_VEXT_VF(vfsgnj_vf_w, 4)
+GEN_VEXT_VF(vfsgnj_vf_d, 8)
 
 static uint16_t fsgnjn16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -4074,15 +4099,15 @@ static uint64_t fsgnjn64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjn_vv_h, OP_UUU_H, H2, H2, H2, fsgnjn16)
 RVVCALL(OPFVV2, vfsgnjn_vv_w, OP_UUU_W, H4, H4, H4, fsgnjn32)
 RVVCALL(OPFVV2, vfsgnjn_vv_d, OP_UUU_D, H8, H8, H8, fsgnjn64)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_h)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_w)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_d)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_d, 8)
 RVVCALL(OPFVF2, vfsgnjn_vf_h, OP_UUU_H, H2, H2, fsgnjn16)
 RVVCALL(OPFVF2, vfsgnjn_vf_w, OP_UUU_W, H4, H4, fsgnjn32)
 RVVCALL(OPFVF2, vfsgnjn_vf_d, OP_UUU_D, H8, H8, fsgnjn64)
-GEN_VEXT_VF(vfsgnjn_vf_h)
-GEN_VEXT_VF(vfsgnjn_vf_w)
-GEN_VEXT_VF(vfsgnjn_vf_d)
+GEN_VEXT_VF(vfsgnjn_vf_h, 2)
+GEN_VEXT_VF(vfsgnjn_vf_w, 4)
+GEN_VEXT_VF(vfsgnjn_vf_d, 8)
 
 static uint16_t fsgnjx16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -4102,15 +4127,15 @@ static uint64_t fsgnjx64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjx_vv_h, OP_UUU_H, H2, H2, H2, fsgnjx16)
 RVVCALL(OPFVV2, vfsgnjx_vv_w, OP_UUU_W, H4, H4, H4, fsgnjx32)
 RVVCALL(OPFVV2, vfsgnjx_vv_d, OP_UUU_D, H8, H8, H8, fsgnjx64)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_h)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_w)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_d)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_d, 8)
 RVVCALL(OPFVF2, vfsgnjx_vf_h, OP_UUU_H, H2, H2, fsgnjx16)
 RVVCALL(OPFVF2, vfsgnjx_vf_w, OP_UUU_W, H4, H4, fsgnjx32)
 RVVCALL(OPFVF2, vfsgnjx_vf_d, OP_UUU_D, H8, H8, fsgnjx64)
-GEN_VEXT_VF(vfsgnjx_vf_h)
-GEN_VEXT_VF(vfsgnjx_vf_w)
-GEN_VEXT_VF(vfsgnjx_vf_d)
+GEN_VEXT_VF(vfsgnjx_vf_h, 2)
+GEN_VEXT_VF(vfsgnjx_vf_w, 4)
+GEN_VEXT_VF(vfsgnjx_vf_d, 8)
 
 /* Vector Floating-Point Compare Instructions */
 #define GEN_VEXT_CMP_VV_ENV(NAME, ETYPE, H, DO_OP)            \
@@ -4119,6 +4144,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
 {                                                             \
     uint32_t vm = vext_vm(desc);                              \
     uint32_t vl = env->vl;                                    \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -4131,6 +4159,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
                            DO_OP(s2, s1, &env->fp_status));   \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    if (vta) {                                                \
+        for (; i < total_elems; i++) {                        \
+            vext_set_elem_mask(vd, i, 1);                     \
+        }                                                     \
+    }                                                         \
 }
 
 GEN_VEXT_CMP_VV_ENV(vmfeq_vv_h, uint16_t, H2, float16_eq_quiet)
@@ -4143,6 +4177,9 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
 {                                                                   \
     uint32_t vm = vext_vm(desc);                                    \
     uint32_t vl = env->vl;                                          \
+    uint32_t total_elems =                                          \
+        vext_get_total_elems(env_archcpu(env), env->vtype);         \
+    uint32_t vta = vext_vta(desc);                                  \
     uint32_t i;                                                     \
                                                                     \
     for (i = env->vstart; i < vl; i++) {                            \
@@ -4154,6 +4191,12 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
                            DO_OP(s2, (ETYPE)s1, &env->fp_status));  \
     }                                                               \
     env->vstart = 0;                                                \
+    /* set tail elements to 1s */                                   \
+    if (vta) {                                                      \
+        for (; i < total_elems; i++) {                              \
+            vext_set_elem_mask(vd, i, 1);                           \
+        }                                                           \
+    }                                                               \
 }
 
 GEN_VEXT_CMP_VF(vmfeq_vf_h, uint16_t, H2, float16_eq_quiet)
@@ -4254,12 +4297,16 @@ static void do_##NAME(void *vd, void *vs2, int i)      \
     *((TD *)vd + HD(i)) = OP(s2);                      \
 }
 
-#define GEN_VEXT_V(NAME)                               \
+#define GEN_VEXT_V(NAME, ESZ)                          \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
                   CPURISCVState *env, uint32_t desc)   \
 {                                                      \
     uint32_t vm = vext_vm(desc);                       \
     uint32_t vl = env->vl;                             \
+    uint32_t total_elems =                             \
+        vext_get_total_elems(env_archcpu(env),         \
+                            env->vtype);               \
+    uint32_t vta = vext_vta(desc);                     \
     uint32_t i;                                        \
                                                        \
     for (i = env->vstart; i < vl; i++) {               \
@@ -4269,6 +4316,11 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         do_##NAME(vd, vs2, i);                         \
     }                                                  \
     env->vstart = 0;                                   \
+    /* set tail elements to 1s */                      \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,      \
+                                     vl * ESZ,         \
+                                     total_elems *     \
+                                     ESZ);             \
 }
 
 target_ulong fclass_h(uint64_t frs1)
@@ -4331,17 +4383,22 @@ target_ulong fclass_d(uint64_t frs1)
 RVVCALL(OPIVV1, vfclass_v_h, OP_UU_H, H2, H2, fclass_h)
 RVVCALL(OPIVV1, vfclass_v_w, OP_UU_W, H4, H4, fclass_s)
 RVVCALL(OPIVV1, vfclass_v_d, OP_UU_D, H8, H8, fclass_d)
-GEN_VEXT_V(vfclass_v_h)
-GEN_VEXT_V(vfclass_v_w)
-GEN_VEXT_V(vfclass_v_d)
+GEN_VEXT_V(vfclass_v_h, 2)
+GEN_VEXT_V(vfclass_v_w, 4)
+GEN_VEXT_V(vfclass_v_d, 8)
 
 /* Vector Floating-Point Merge Instruction */
+
 #define GEN_VFMERGE_VF(NAME, ETYPE, H)                        \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
     uint32_t vm = vext_vm(desc);                              \
     uint32_t vl = env->vl;                                    \
+    uint32_t esz = sizeof(ETYPE);                             \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -4350,6 +4407,9 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
           = (!vm && !vext_elem_mask(v0, i) ? s2 : s1);        \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,   \
+                                     total_elems * esz);      \
 }
 
 GEN_VFMERGE_VF(vfmerge_vfm_h, int16_t, H2)
@@ -4361,33 +4421,33 @@ GEN_VFMERGE_VF(vfmerge_vfm_d, int64_t, H8)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_h, OP_UU_H, H2, H2, float16_to_uint16)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_w, OP_UU_W, H4, H4, float32_to_uint32)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_d, OP_UU_D, H8, H8, float64_to_uint64)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_h)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_w)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_d)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_d, 8)
 
 /* vfcvt.x.f.v vd, vs2, vm # Convert float to signed integer. */
 RVVCALL(OPFVV1, vfcvt_x_f_v_h, OP_UU_H, H2, H2, float16_to_int16)
 RVVCALL(OPFVV1, vfcvt_x_f_v_w, OP_UU_W, H4, H4, float32_to_int32)
 RVVCALL(OPFVV1, vfcvt_x_f_v_d, OP_UU_D, H8, H8, float64_to_int64)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_h)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_w)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_d)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_d, 8)
 
 /* vfcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_xu_v_h, OP_UU_H, H2, H2, uint16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_w, OP_UU_W, H4, H4, uint32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_d, OP_UU_D, H8, H8, uint64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_h)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_w)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_d)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_d, 8)
 
 /* vfcvt.f.x.v vd, vs2, vm # Convert integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_x_v_h, OP_UU_H, H2, H2, int16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_x_v_w, OP_UU_W, H4, H4, int32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_x_v_d, OP_UU_D, H8, H8, int64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_h)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_w)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_d)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_d, 8)
 
 /* Widening Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4397,30 +4457,30 @@ GEN_VEXT_V_ENV(vfcvt_f_x_v_d)
 /* vfwcvt.xu.f.v vd, vs2, vm # Convert float to double-width unsigned integer.*/
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_h, WOP_UU_H, H4, H2, float16_to_uint32)
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_w, WOP_UU_W, H8, H4, float32_to_uint64)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w, 8)
 
 /* vfwcvt.x.f.v vd, vs2, vm # Convert float to double-width signed integer. */
 RVVCALL(OPFVV1, vfwcvt_x_f_v_h, WOP_UU_H, H4, H2, float16_to_int32)
 RVVCALL(OPFVV1, vfwcvt_x_f_v_w, WOP_UU_W, H8, H4, float32_to_int64)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_h)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_w)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_w, 8)
 
 /* vfwcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to double-width float */
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_b, WOP_UU_B, H2, H1, uint8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_h, WOP_UU_H, H4, H2, uint16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_w, WOP_UU_W, H8, H4, uint32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b, 2)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w, 8)
 
 /* vfwcvt.f.x.v vd, vs2, vm # Convert integer to double-width float. */
 RVVCALL(OPFVV1, vfwcvt_f_x_v_b, WOP_UU_B, H2, H1, int8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_h, WOP_UU_H, H4, H2, int16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_w, WOP_UU_W, H8, H4, int32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_b)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_h)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_w)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_b, 2)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_w, 8)
 
 /*
  * vfwcvt.f.f.v vd, vs2, vm
@@ -4433,8 +4493,8 @@ static uint32_t vfwcvtffv16(uint16_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfwcvt_f_f_v_h, WOP_UU_H, H4, H2, vfwcvtffv16)
 RVVCALL(OPFVV1, vfwcvt_f_f_v_w, WOP_UU_W, H8, H4, float32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_h)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_w)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_w, 8)
 
 /* Narrowing Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4445,29 +4505,29 @@ GEN_VEXT_V_ENV(vfwcvt_f_f_v_w)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_b, NOP_UU_B, H1, H2, float16_to_uint8)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_h, NOP_UU_H, H2, H4, float32_to_uint16)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_w, NOP_UU_W, H4, H8, float64_to_uint32)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_b)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_h)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_w)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_b, 1)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_w, 4)
 
 /* vfncvt.x.f.v vd, vs2, vm # Convert double-width float to signed integer. */
 RVVCALL(OPFVV1, vfncvt_x_f_w_b, NOP_UU_B, H1, H2, float16_to_int8)
 RVVCALL(OPFVV1, vfncvt_x_f_w_h, NOP_UU_H, H2, H4, float32_to_int16)
 RVVCALL(OPFVV1, vfncvt_x_f_w_w, NOP_UU_W, H4, H8, float64_to_int32)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_b)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_h)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_w)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_b, 1)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_w, 4)
 
 /* vfncvt.f.xu.v vd, vs2, vm # Convert double-width unsigned integer to float */
 RVVCALL(OPFVV1, vfncvt_f_xu_w_h, NOP_UU_H, H2, H4, uint32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_xu_w_w, NOP_UU_W, H4, H8, uint64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_h)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_w)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_w, 4)
 
 /* vfncvt.f.x.v vd, vs2, vm # Convert double-width integer to float. */
 RVVCALL(OPFVV1, vfncvt_f_x_w_h, NOP_UU_H, H2, H4, int32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_x_w_w, NOP_UU_W, H4, H8, int64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_h)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_w)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_w, 4)
 
 /* vfncvt.f.f.v vd, vs2, vm # Convert double float to single-width float. */
 static uint16_t vfncvtffv16(uint32_t a, float_status *s)
@@ -4477,8 +4537,8 @@ static uint16_t vfncvtffv16(uint32_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfncvt_f_f_w_h, NOP_UU_H, H2, H4, vfncvtffv16)
 RVVCALL(OPFVV1, vfncvt_f_f_w_w, NOP_UU_W, H4, H8, float64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_h)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_w)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_w, 4)
 
 /*
  *** Vector Reduction Operations
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 11/14] target/riscv: rvv: Add tail agnostic for vector floating-point instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  11 +
 target/riscv/vector_helper.c            | 452 ++++++++++++++----------
 2 files changed, 267 insertions(+), 196 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 24784b6f5f..5bc340f419 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -2391,6 +2391,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
@@ -2473,6 +2474,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)            \
         gen_set_rm(s, RISCV_FRM_DYN);                             \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);            \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);              \
         return opfvf_trans(a->rd, a->rs1, a->rs2, data,           \
                            fns[s->sew - 1], s);                   \
     }                                                             \
@@ -2511,6 +2513,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
                                                                  \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),   \
                            vreg_ofs(s, a->rs1),                  \
                            vreg_ofs(s, a->rs2), cpu_env,         \
@@ -2550,6 +2553,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         gen_set_rm(s, RISCV_FRM_DYN);                            \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
         return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
                            fns[s->sew - 1], s);                  \
     }                                                            \
@@ -2586,6 +2590,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
@@ -2625,6 +2630,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         gen_set_rm(s, RISCV_FRM_DYN);                            \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
         return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
                            fns[s->sew - 1], s);                  \
     }                                                            \
@@ -2708,6 +2714,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs2), cpu_env,
                            s->cfg_ptr->vlen / 8,
@@ -2921,6 +2928,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
@@ -2972,6 +2980,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
@@ -3039,6 +3048,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
@@ -3092,6 +3102,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
                            s->cfg_ptr->vlen / 8,                   \
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 067095e19a..d654377fc8 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -2997,13 +2997,17 @@ static void do_##NAME(void *vd, void *vs1, void *vs2, int i,   \
     *((TD *)vd + HD(i)) = OP(s2, s1, &env->fp_status);         \
 }
 
-#define GEN_VEXT_VV_ENV(NAME)                             \
+#define GEN_VEXT_VV_ENV(NAME, ESZ)                        \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     uint32_t vm = vext_vm(desc);                          \
     uint32_t vl = env->vl;                                \
+    uint32_t total_elems =                                \
+        vext_get_total_elems(env_archcpu(env),            \
+                             env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
                                                           \
     for (i = env->vstart; i < vl; i++) {                  \
@@ -3013,14 +3017,18 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
         do_##NAME(vd, vs1, vs2, i, env);                  \
     }                                                     \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
+                                     vl * ESZ,            \
+                                     total_elems * ESZ);  \
 }
 
 RVVCALL(OPFVV2, vfadd_vv_h, OP_UUU_H, H2, H2, H2, float16_add)
 RVVCALL(OPFVV2, vfadd_vv_w, OP_UUU_W, H4, H4, H4, float32_add)
 RVVCALL(OPFVV2, vfadd_vv_d, OP_UUU_D, H8, H8, H8, float64_add)
-GEN_VEXT_VV_ENV(vfadd_vv_h)
-GEN_VEXT_VV_ENV(vfadd_vv_w)
-GEN_VEXT_VV_ENV(vfadd_vv_d)
+GEN_VEXT_VV_ENV(vfadd_vv_h, 2)
+GEN_VEXT_VV_ENV(vfadd_vv_w, 4)
+GEN_VEXT_VV_ENV(vfadd_vv_d, 8)
 
 #define OPFVF2(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)        \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
@@ -3030,13 +3038,17 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
     *((TD *)vd + HD(i)) = OP(s2, (TX1)(T1)s1, &env->fp_status);\
 }
 
-#define GEN_VEXT_VF(NAME)                                 \
+#define GEN_VEXT_VF(NAME, ESZ)                            \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
     uint32_t vm = vext_vm(desc);                          \
     uint32_t vl = env->vl;                                \
+    uint32_t total_elems =                                \
+        vext_get_total_elems(env_archcpu(env),            \
+                             env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
                                                           \
     for (i = env->vstart; i < vl; i++) {                  \
@@ -3046,27 +3058,31 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
         do_##NAME(vd, s1, vs2, i, env);                   \
     }                                                     \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
+                                     vl * ESZ,            \
+                                     total_elems * ESZ);  \
 }
 
 RVVCALL(OPFVF2, vfadd_vf_h, OP_UUU_H, H2, H2, float16_add)
 RVVCALL(OPFVF2, vfadd_vf_w, OP_UUU_W, H4, H4, float32_add)
 RVVCALL(OPFVF2, vfadd_vf_d, OP_UUU_D, H8, H8, float64_add)
-GEN_VEXT_VF(vfadd_vf_h)
-GEN_VEXT_VF(vfadd_vf_w)
-GEN_VEXT_VF(vfadd_vf_d)
+GEN_VEXT_VF(vfadd_vf_h, 2)
+GEN_VEXT_VF(vfadd_vf_w, 4)
+GEN_VEXT_VF(vfadd_vf_d, 8)
 
 RVVCALL(OPFVV2, vfsub_vv_h, OP_UUU_H, H2, H2, H2, float16_sub)
 RVVCALL(OPFVV2, vfsub_vv_w, OP_UUU_W, H4, H4, H4, float32_sub)
 RVVCALL(OPFVV2, vfsub_vv_d, OP_UUU_D, H8, H8, H8, float64_sub)
-GEN_VEXT_VV_ENV(vfsub_vv_h)
-GEN_VEXT_VV_ENV(vfsub_vv_w)
-GEN_VEXT_VV_ENV(vfsub_vv_d)
+GEN_VEXT_VV_ENV(vfsub_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsub_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsub_vv_d, 8)
 RVVCALL(OPFVF2, vfsub_vf_h, OP_UUU_H, H2, H2, float16_sub)
 RVVCALL(OPFVF2, vfsub_vf_w, OP_UUU_W, H4, H4, float32_sub)
 RVVCALL(OPFVF2, vfsub_vf_d, OP_UUU_D, H8, H8, float64_sub)
-GEN_VEXT_VF(vfsub_vf_h)
-GEN_VEXT_VF(vfsub_vf_w)
-GEN_VEXT_VF(vfsub_vf_d)
+GEN_VEXT_VF(vfsub_vf_h, 2)
+GEN_VEXT_VF(vfsub_vf_w, 4)
+GEN_VEXT_VF(vfsub_vf_d, 8)
 
 static uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3086,9 +3102,9 @@ static uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrsub_vf_h, OP_UUU_H, H2, H2, float16_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_w, OP_UUU_W, H4, H4, float32_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub)
-GEN_VEXT_VF(vfrsub_vf_h)
-GEN_VEXT_VF(vfrsub_vf_w)
-GEN_VEXT_VF(vfrsub_vf_d)
+GEN_VEXT_VF(vfrsub_vf_h, 2)
+GEN_VEXT_VF(vfrsub_vf_w, 4)
+GEN_VEXT_VF(vfrsub_vf_d, 8)
 
 /* Vector Widening Floating-Point Add/Subtract Instructions */
 static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s)
@@ -3106,12 +3122,12 @@ static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16)
 RVVCALL(OPFVV2, vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32)
-GEN_VEXT_VV_ENV(vfwadd_vv_h)
-GEN_VEXT_VV_ENV(vfwadd_vv_w)
+GEN_VEXT_VV_ENV(vfwadd_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwadd_vv_w, 8)
 RVVCALL(OPFVF2, vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16)
 RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32)
-GEN_VEXT_VF(vfwadd_vf_h)
-GEN_VEXT_VF(vfwadd_vf_w)
+GEN_VEXT_VF(vfwadd_vf_h, 4)
+GEN_VEXT_VF(vfwadd_vf_w, 8)
 
 static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3128,12 +3144,12 @@ static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16)
 RVVCALL(OPFVV2, vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32)
-GEN_VEXT_VV_ENV(vfwsub_vv_h)
-GEN_VEXT_VV_ENV(vfwsub_vv_w)
+GEN_VEXT_VV_ENV(vfwsub_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwsub_vv_w, 8)
 RVVCALL(OPFVF2, vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16)
 RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32)
-GEN_VEXT_VF(vfwsub_vf_h)
-GEN_VEXT_VF(vfwsub_vf_w)
+GEN_VEXT_VF(vfwsub_vf_h, 4)
+GEN_VEXT_VF(vfwsub_vf_w, 8)
 
 static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -3147,12 +3163,12 @@ static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16)
 RVVCALL(OPFVV2, vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32)
-GEN_VEXT_VV_ENV(vfwadd_wv_h)
-GEN_VEXT_VV_ENV(vfwadd_wv_w)
+GEN_VEXT_VV_ENV(vfwadd_wv_h, 4)
+GEN_VEXT_VV_ENV(vfwadd_wv_w, 8)
 RVVCALL(OPFVF2, vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16)
 RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32)
-GEN_VEXT_VF(vfwadd_wf_h)
-GEN_VEXT_VF(vfwadd_wf_w)
+GEN_VEXT_VF(vfwadd_wf_h, 4)
+GEN_VEXT_VF(vfwadd_wf_w, 8)
 
 static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -3166,39 +3182,39 @@ static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16)
 RVVCALL(OPFVV2, vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32)
-GEN_VEXT_VV_ENV(vfwsub_wv_h)
-GEN_VEXT_VV_ENV(vfwsub_wv_w)
+GEN_VEXT_VV_ENV(vfwsub_wv_h, 4)
+GEN_VEXT_VV_ENV(vfwsub_wv_w, 8)
 RVVCALL(OPFVF2, vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16)
 RVVCALL(OPFVF2, vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32)
-GEN_VEXT_VF(vfwsub_wf_h)
-GEN_VEXT_VF(vfwsub_wf_w)
+GEN_VEXT_VF(vfwsub_wf_h, 4)
+GEN_VEXT_VF(vfwsub_wf_w, 8)
 
 /* Vector Single-Width Floating-Point Multiply/Divide Instructions */
 RVVCALL(OPFVV2, vfmul_vv_h, OP_UUU_H, H2, H2, H2, float16_mul)
 RVVCALL(OPFVV2, vfmul_vv_w, OP_UUU_W, H4, H4, H4, float32_mul)
 RVVCALL(OPFVV2, vfmul_vv_d, OP_UUU_D, H8, H8, H8, float64_mul)
-GEN_VEXT_VV_ENV(vfmul_vv_h)
-GEN_VEXT_VV_ENV(vfmul_vv_w)
-GEN_VEXT_VV_ENV(vfmul_vv_d)
+GEN_VEXT_VV_ENV(vfmul_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmul_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmul_vv_d, 8)
 RVVCALL(OPFVF2, vfmul_vf_h, OP_UUU_H, H2, H2, float16_mul)
 RVVCALL(OPFVF2, vfmul_vf_w, OP_UUU_W, H4, H4, float32_mul)
 RVVCALL(OPFVF2, vfmul_vf_d, OP_UUU_D, H8, H8, float64_mul)
-GEN_VEXT_VF(vfmul_vf_h)
-GEN_VEXT_VF(vfmul_vf_w)
-GEN_VEXT_VF(vfmul_vf_d)
+GEN_VEXT_VF(vfmul_vf_h, 2)
+GEN_VEXT_VF(vfmul_vf_w, 4)
+GEN_VEXT_VF(vfmul_vf_d, 8)
 
 RVVCALL(OPFVV2, vfdiv_vv_h, OP_UUU_H, H2, H2, H2, float16_div)
 RVVCALL(OPFVV2, vfdiv_vv_w, OP_UUU_W, H4, H4, H4, float32_div)
 RVVCALL(OPFVV2, vfdiv_vv_d, OP_UUU_D, H8, H8, H8, float64_div)
-GEN_VEXT_VV_ENV(vfdiv_vv_h)
-GEN_VEXT_VV_ENV(vfdiv_vv_w)
-GEN_VEXT_VV_ENV(vfdiv_vv_d)
+GEN_VEXT_VV_ENV(vfdiv_vv_h, 2)
+GEN_VEXT_VV_ENV(vfdiv_vv_w, 4)
+GEN_VEXT_VV_ENV(vfdiv_vv_d, 8)
 RVVCALL(OPFVF2, vfdiv_vf_h, OP_UUU_H, H2, H2, float16_div)
 RVVCALL(OPFVF2, vfdiv_vf_w, OP_UUU_W, H4, H4, float32_div)
 RVVCALL(OPFVF2, vfdiv_vf_d, OP_UUU_D, H8, H8, float64_div)
-GEN_VEXT_VF(vfdiv_vf_h)
-GEN_VEXT_VF(vfdiv_vf_w)
-GEN_VEXT_VF(vfdiv_vf_d)
+GEN_VEXT_VF(vfdiv_vf_h, 2)
+GEN_VEXT_VF(vfdiv_vf_w, 4)
+GEN_VEXT_VF(vfdiv_vf_d, 8)
 
 static uint16_t float16_rdiv(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3218,9 +3234,9 @@ static uint64_t float64_rdiv(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrdiv_vf_h, OP_UUU_H, H2, H2, float16_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_w, OP_UUU_W, H4, H4, float32_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_d, OP_UUU_D, H8, H8, float64_rdiv)
-GEN_VEXT_VF(vfrdiv_vf_h)
-GEN_VEXT_VF(vfrdiv_vf_w)
-GEN_VEXT_VF(vfrdiv_vf_d)
+GEN_VEXT_VF(vfrdiv_vf_h, 2)
+GEN_VEXT_VF(vfrdiv_vf_w, 4)
+GEN_VEXT_VF(vfrdiv_vf_d, 8)
 
 /* Vector Widening Floating-Point Multiply */
 static uint32_t vfwmul16(uint16_t a, uint16_t b, float_status *s)
@@ -3237,12 +3253,12 @@ static uint64_t vfwmul32(uint32_t a, uint32_t b, float_status *s)
 }
 RVVCALL(OPFVV2, vfwmul_vv_h, WOP_UUU_H, H4, H2, H2, vfwmul16)
 RVVCALL(OPFVV2, vfwmul_vv_w, WOP_UUU_W, H8, H4, H4, vfwmul32)
-GEN_VEXT_VV_ENV(vfwmul_vv_h)
-GEN_VEXT_VV_ENV(vfwmul_vv_w)
+GEN_VEXT_VV_ENV(vfwmul_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwmul_vv_w, 8)
 RVVCALL(OPFVF2, vfwmul_vf_h, WOP_UUU_H, H4, H2, vfwmul16)
 RVVCALL(OPFVF2, vfwmul_vf_w, WOP_UUU_W, H8, H4, vfwmul32)
-GEN_VEXT_VF(vfwmul_vf_h)
-GEN_VEXT_VF(vfwmul_vf_w)
+GEN_VEXT_VF(vfwmul_vf_h, 4)
+GEN_VEXT_VF(vfwmul_vf_w, 8)
 
 /* Vector Single-Width Floating-Point Fused Multiply-Add Instructions */
 #define OPFVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)       \
@@ -3273,9 +3289,9 @@ static uint64_t fmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmacc_vv_h, OP_UUU_H, H2, H2, H2, fmacc16)
 RVVCALL(OPFVV3, vfmacc_vv_w, OP_UUU_W, H4, H4, H4, fmacc32)
 RVVCALL(OPFVV3, vfmacc_vv_d, OP_UUU_D, H8, H8, H8, fmacc64)
-GEN_VEXT_VV_ENV(vfmacc_vv_h)
-GEN_VEXT_VV_ENV(vfmacc_vv_w)
-GEN_VEXT_VV_ENV(vfmacc_vv_d)
+GEN_VEXT_VV_ENV(vfmacc_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmacc_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmacc_vv_d, 8)
 
 #define OPFVF3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)           \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
@@ -3289,9 +3305,9 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
 RVVCALL(OPFVF3, vfmacc_vf_h, OP_UUU_H, H2, H2, fmacc16)
 RVVCALL(OPFVF3, vfmacc_vf_w, OP_UUU_W, H4, H4, fmacc32)
 RVVCALL(OPFVF3, vfmacc_vf_d, OP_UUU_D, H8, H8, fmacc64)
-GEN_VEXT_VF(vfmacc_vf_h)
-GEN_VEXT_VF(vfmacc_vf_w)
-GEN_VEXT_VF(vfmacc_vf_d)
+GEN_VEXT_VF(vfmacc_vf_h, 2)
+GEN_VEXT_VF(vfmacc_vf_w, 4)
+GEN_VEXT_VF(vfmacc_vf_d, 8)
 
 static uint16_t fnmacc16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3314,15 +3330,15 @@ static uint64_t fnmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmacc_vv_h, OP_UUU_H, H2, H2, H2, fnmacc16)
 RVVCALL(OPFVV3, vfnmacc_vv_w, OP_UUU_W, H4, H4, H4, fnmacc32)
 RVVCALL(OPFVV3, vfnmacc_vv_d, OP_UUU_D, H8, H8, H8, fnmacc64)
-GEN_VEXT_VV_ENV(vfnmacc_vv_h)
-GEN_VEXT_VV_ENV(vfnmacc_vv_w)
-GEN_VEXT_VV_ENV(vfnmacc_vv_d)
+GEN_VEXT_VV_ENV(vfnmacc_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmacc_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmacc_vv_d, 8)
 RVVCALL(OPFVF3, vfnmacc_vf_h, OP_UUU_H, H2, H2, fnmacc16)
 RVVCALL(OPFVF3, vfnmacc_vf_w, OP_UUU_W, H4, H4, fnmacc32)
 RVVCALL(OPFVF3, vfnmacc_vf_d, OP_UUU_D, H8, H8, fnmacc64)
-GEN_VEXT_VF(vfnmacc_vf_h)
-GEN_VEXT_VF(vfnmacc_vf_w)
-GEN_VEXT_VF(vfnmacc_vf_d)
+GEN_VEXT_VF(vfnmacc_vf_h, 2)
+GEN_VEXT_VF(vfnmacc_vf_w, 4)
+GEN_VEXT_VF(vfnmacc_vf_d, 8)
 
 static uint16_t fmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3342,15 +3358,15 @@ static uint64_t fmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsac_vv_h, OP_UUU_H, H2, H2, H2, fmsac16)
 RVVCALL(OPFVV3, vfmsac_vv_w, OP_UUU_W, H4, H4, H4, fmsac32)
 RVVCALL(OPFVV3, vfmsac_vv_d, OP_UUU_D, H8, H8, H8, fmsac64)
-GEN_VEXT_VV_ENV(vfmsac_vv_h)
-GEN_VEXT_VV_ENV(vfmsac_vv_w)
-GEN_VEXT_VV_ENV(vfmsac_vv_d)
+GEN_VEXT_VV_ENV(vfmsac_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmsac_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmsac_vv_d, 8)
 RVVCALL(OPFVF3, vfmsac_vf_h, OP_UUU_H, H2, H2, fmsac16)
 RVVCALL(OPFVF3, vfmsac_vf_w, OP_UUU_W, H4, H4, fmsac32)
 RVVCALL(OPFVF3, vfmsac_vf_d, OP_UUU_D, H8, H8, fmsac64)
-GEN_VEXT_VF(vfmsac_vf_h)
-GEN_VEXT_VF(vfmsac_vf_w)
-GEN_VEXT_VF(vfmsac_vf_d)
+GEN_VEXT_VF(vfmsac_vf_h, 2)
+GEN_VEXT_VF(vfmsac_vf_w, 4)
+GEN_VEXT_VF(vfmsac_vf_d, 8)
 
 static uint16_t fnmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3370,15 +3386,15 @@ static uint64_t fnmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsac_vv_h, OP_UUU_H, H2, H2, H2, fnmsac16)
 RVVCALL(OPFVV3, vfnmsac_vv_w, OP_UUU_W, H4, H4, H4, fnmsac32)
 RVVCALL(OPFVV3, vfnmsac_vv_d, OP_UUU_D, H8, H8, H8, fnmsac64)
-GEN_VEXT_VV_ENV(vfnmsac_vv_h)
-GEN_VEXT_VV_ENV(vfnmsac_vv_w)
-GEN_VEXT_VV_ENV(vfnmsac_vv_d)
+GEN_VEXT_VV_ENV(vfnmsac_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmsac_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmsac_vv_d, 8)
 RVVCALL(OPFVF3, vfnmsac_vf_h, OP_UUU_H, H2, H2, fnmsac16)
 RVVCALL(OPFVF3, vfnmsac_vf_w, OP_UUU_W, H4, H4, fnmsac32)
 RVVCALL(OPFVF3, vfnmsac_vf_d, OP_UUU_D, H8, H8, fnmsac64)
-GEN_VEXT_VF(vfnmsac_vf_h)
-GEN_VEXT_VF(vfnmsac_vf_w)
-GEN_VEXT_VF(vfnmsac_vf_d)
+GEN_VEXT_VF(vfnmsac_vf_h, 2)
+GEN_VEXT_VF(vfnmsac_vf_w, 4)
+GEN_VEXT_VF(vfnmsac_vf_d, 8)
 
 static uint16_t fmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3398,15 +3414,15 @@ static uint64_t fmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmadd_vv_h, OP_UUU_H, H2, H2, H2, fmadd16)
 RVVCALL(OPFVV3, vfmadd_vv_w, OP_UUU_W, H4, H4, H4, fmadd32)
 RVVCALL(OPFVV3, vfmadd_vv_d, OP_UUU_D, H8, H8, H8, fmadd64)
-GEN_VEXT_VV_ENV(vfmadd_vv_h)
-GEN_VEXT_VV_ENV(vfmadd_vv_w)
-GEN_VEXT_VV_ENV(vfmadd_vv_d)
+GEN_VEXT_VV_ENV(vfmadd_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmadd_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmadd_vv_d, 8)
 RVVCALL(OPFVF3, vfmadd_vf_h, OP_UUU_H, H2, H2, fmadd16)
 RVVCALL(OPFVF3, vfmadd_vf_w, OP_UUU_W, H4, H4, fmadd32)
 RVVCALL(OPFVF3, vfmadd_vf_d, OP_UUU_D, H8, H8, fmadd64)
-GEN_VEXT_VF(vfmadd_vf_h)
-GEN_VEXT_VF(vfmadd_vf_w)
-GEN_VEXT_VF(vfmadd_vf_d)
+GEN_VEXT_VF(vfmadd_vf_h, 2)
+GEN_VEXT_VF(vfmadd_vf_w, 4)
+GEN_VEXT_VF(vfmadd_vf_d, 8)
 
 static uint16_t fnmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3429,15 +3445,15 @@ static uint64_t fnmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmadd_vv_h, OP_UUU_H, H2, H2, H2, fnmadd16)
 RVVCALL(OPFVV3, vfnmadd_vv_w, OP_UUU_W, H4, H4, H4, fnmadd32)
 RVVCALL(OPFVV3, vfnmadd_vv_d, OP_UUU_D, H8, H8, H8, fnmadd64)
-GEN_VEXT_VV_ENV(vfnmadd_vv_h)
-GEN_VEXT_VV_ENV(vfnmadd_vv_w)
-GEN_VEXT_VV_ENV(vfnmadd_vv_d)
+GEN_VEXT_VV_ENV(vfnmadd_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmadd_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmadd_vv_d, 8)
 RVVCALL(OPFVF3, vfnmadd_vf_h, OP_UUU_H, H2, H2, fnmadd16)
 RVVCALL(OPFVF3, vfnmadd_vf_w, OP_UUU_W, H4, H4, fnmadd32)
 RVVCALL(OPFVF3, vfnmadd_vf_d, OP_UUU_D, H8, H8, fnmadd64)
-GEN_VEXT_VF(vfnmadd_vf_h)
-GEN_VEXT_VF(vfnmadd_vf_w)
-GEN_VEXT_VF(vfnmadd_vf_d)
+GEN_VEXT_VF(vfnmadd_vf_h, 2)
+GEN_VEXT_VF(vfnmadd_vf_w, 4)
+GEN_VEXT_VF(vfnmadd_vf_d, 8)
 
 static uint16_t fmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3457,15 +3473,15 @@ static uint64_t fmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsub_vv_h, OP_UUU_H, H2, H2, H2, fmsub16)
 RVVCALL(OPFVV3, vfmsub_vv_w, OP_UUU_W, H4, H4, H4, fmsub32)
 RVVCALL(OPFVV3, vfmsub_vv_d, OP_UUU_D, H8, H8, H8, fmsub64)
-GEN_VEXT_VV_ENV(vfmsub_vv_h)
-GEN_VEXT_VV_ENV(vfmsub_vv_w)
-GEN_VEXT_VV_ENV(vfmsub_vv_d)
+GEN_VEXT_VV_ENV(vfmsub_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmsub_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmsub_vv_d, 8)
 RVVCALL(OPFVF3, vfmsub_vf_h, OP_UUU_H, H2, H2, fmsub16)
 RVVCALL(OPFVF3, vfmsub_vf_w, OP_UUU_W, H4, H4, fmsub32)
 RVVCALL(OPFVF3, vfmsub_vf_d, OP_UUU_D, H8, H8, fmsub64)
-GEN_VEXT_VF(vfmsub_vf_h)
-GEN_VEXT_VF(vfmsub_vf_w)
-GEN_VEXT_VF(vfmsub_vf_d)
+GEN_VEXT_VF(vfmsub_vf_h, 2)
+GEN_VEXT_VF(vfmsub_vf_w, 4)
+GEN_VEXT_VF(vfmsub_vf_d, 8)
 
 static uint16_t fnmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3485,15 +3501,15 @@ static uint64_t fnmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsub_vv_h, OP_UUU_H, H2, H2, H2, fnmsub16)
 RVVCALL(OPFVV3, vfnmsub_vv_w, OP_UUU_W, H4, H4, H4, fnmsub32)
 RVVCALL(OPFVV3, vfnmsub_vv_d, OP_UUU_D, H8, H8, H8, fnmsub64)
-GEN_VEXT_VV_ENV(vfnmsub_vv_h)
-GEN_VEXT_VV_ENV(vfnmsub_vv_w)
-GEN_VEXT_VV_ENV(vfnmsub_vv_d)
+GEN_VEXT_VV_ENV(vfnmsub_vv_h, 2)
+GEN_VEXT_VV_ENV(vfnmsub_vv_w, 4)
+GEN_VEXT_VV_ENV(vfnmsub_vv_d, 8)
 RVVCALL(OPFVF3, vfnmsub_vf_h, OP_UUU_H, H2, H2, fnmsub16)
 RVVCALL(OPFVF3, vfnmsub_vf_w, OP_UUU_W, H4, H4, fnmsub32)
 RVVCALL(OPFVF3, vfnmsub_vf_d, OP_UUU_D, H8, H8, fnmsub64)
-GEN_VEXT_VF(vfnmsub_vf_h)
-GEN_VEXT_VF(vfnmsub_vf_w)
-GEN_VEXT_VF(vfnmsub_vf_d)
+GEN_VEXT_VF(vfnmsub_vf_h, 2)
+GEN_VEXT_VF(vfnmsub_vf_w, 4)
+GEN_VEXT_VF(vfnmsub_vf_d, 8)
 
 /* Vector Widening Floating-Point Fused Multiply-Add Instructions */
 static uint32_t fwmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
@@ -3510,12 +3526,12 @@ static uint64_t fwmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwmacc16)
 RVVCALL(OPFVV3, vfwmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwmacc32)
-GEN_VEXT_VV_ENV(vfwmacc_vv_h)
-GEN_VEXT_VV_ENV(vfwmacc_vv_w)
+GEN_VEXT_VV_ENV(vfwmacc_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwmacc_vv_w, 8)
 RVVCALL(OPFVF3, vfwmacc_vf_h, WOP_UUU_H, H4, H2, fwmacc16)
 RVVCALL(OPFVF3, vfwmacc_vf_w, WOP_UUU_W, H8, H4, fwmacc32)
-GEN_VEXT_VF(vfwmacc_vf_h)
-GEN_VEXT_VF(vfwmacc_vf_w)
+GEN_VEXT_VF(vfwmacc_vf_h, 4)
+GEN_VEXT_VF(vfwmacc_vf_w, 8)
 
 static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3533,12 +3549,12 @@ static uint64_t fwnmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwnmacc16)
 RVVCALL(OPFVV3, vfwnmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwnmacc32)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_h)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_w)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_w, 8)
 RVVCALL(OPFVF3, vfwnmacc_vf_h, WOP_UUU_H, H4, H2, fwnmacc16)
 RVVCALL(OPFVF3, vfwnmacc_vf_w, WOP_UUU_W, H8, H4, fwnmacc32)
-GEN_VEXT_VF(vfwnmacc_vf_h)
-GEN_VEXT_VF(vfwnmacc_vf_w)
+GEN_VEXT_VF(vfwnmacc_vf_h, 4)
+GEN_VEXT_VF(vfwnmacc_vf_w, 8)
 
 static uint32_t fwmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3556,12 +3572,12 @@ static uint64_t fwmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwmsac16)
 RVVCALL(OPFVV3, vfwmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwmsac32)
-GEN_VEXT_VV_ENV(vfwmsac_vv_h)
-GEN_VEXT_VV_ENV(vfwmsac_vv_w)
+GEN_VEXT_VV_ENV(vfwmsac_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwmsac_vv_w, 8)
 RVVCALL(OPFVF3, vfwmsac_vf_h, WOP_UUU_H, H4, H2, fwmsac16)
 RVVCALL(OPFVF3, vfwmsac_vf_w, WOP_UUU_W, H8, H4, fwmsac32)
-GEN_VEXT_VF(vfwmsac_vf_h)
-GEN_VEXT_VF(vfwmsac_vf_w)
+GEN_VEXT_VF(vfwmsac_vf_h, 4)
+GEN_VEXT_VF(vfwmsac_vf_w, 8)
 
 static uint32_t fwnmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3579,12 +3595,12 @@ static uint64_t fwnmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwnmsac16)
 RVVCALL(OPFVV3, vfwnmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwnmsac32)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_h)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_w)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_h, 4)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_w, 8)
 RVVCALL(OPFVF3, vfwnmsac_vf_h, WOP_UUU_H, H4, H2, fwnmsac16)
 RVVCALL(OPFVF3, vfwnmsac_vf_w, WOP_UUU_W, H8, H4, fwnmsac32)
-GEN_VEXT_VF(vfwnmsac_vf_h)
-GEN_VEXT_VF(vfwnmsac_vf_w)
+GEN_VEXT_VF(vfwnmsac_vf_h, 4)
+GEN_VEXT_VF(vfwnmsac_vf_w, 8)
 
 /* Vector Floating-Point Square-Root Instruction */
 /* (TD, T2, TX2) */
@@ -3600,12 +3616,16 @@ static void do_##NAME(void *vd, void *vs2, int i,      \
     *((TD *)vd + HD(i)) = OP(s2, &env->fp_status);     \
 }
 
-#define GEN_VEXT_V_ENV(NAME)                           \
+#define GEN_VEXT_V_ENV(NAME, ESZ)                      \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         CPURISCVState *env, uint32_t desc)             \
 {                                                      \
     uint32_t vm = vext_vm(desc);                       \
     uint32_t vl = env->vl;                             \
+    uint32_t total_elems =                             \
+        vext_get_total_elems(env_archcpu(env),         \
+                            env->vtype);               \
+    uint32_t vta = vext_vta(desc);                     \
     uint32_t i;                                        \
                                                        \
     if (vl == 0) {                                     \
@@ -3618,14 +3638,19 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         do_##NAME(vd, vs2, i, env);                    \
     }                                                  \
     env->vstart = 0;                                   \
+    /* set tail elements to 1s */                      \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,      \
+                                     vl * ESZ,         \
+                                     total_elems *     \
+                                     ESZ);             \
 }
 
 RVVCALL(OPFVV1, vfsqrt_v_h, OP_UU_H, H2, H2, float16_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_w, OP_UU_W, H4, H4, float32_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_d, OP_UU_D, H8, H8, float64_sqrt)
-GEN_VEXT_V_ENV(vfsqrt_v_h)
-GEN_VEXT_V_ENV(vfsqrt_v_w)
-GEN_VEXT_V_ENV(vfsqrt_v_d)
+GEN_VEXT_V_ENV(vfsqrt_v_h, 2)
+GEN_VEXT_V_ENV(vfsqrt_v_w, 4)
+GEN_VEXT_V_ENV(vfsqrt_v_d, 8)
 
 /*
  * Vector Floating-Point Reciprocal Square-Root Estimate Instruction
@@ -3805,9 +3830,9 @@ static float64 frsqrt7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrsqrt7_v_h, OP_UU_H, H2, H2, frsqrt7_h)
 RVVCALL(OPFVV1, vfrsqrt7_v_w, OP_UU_W, H4, H4, frsqrt7_s)
 RVVCALL(OPFVV1, vfrsqrt7_v_d, OP_UU_D, H8, H8, frsqrt7_d)
-GEN_VEXT_V_ENV(vfrsqrt7_v_h)
-GEN_VEXT_V_ENV(vfrsqrt7_v_w)
-GEN_VEXT_V_ENV(vfrsqrt7_v_d)
+GEN_VEXT_V_ENV(vfrsqrt7_v_h, 2)
+GEN_VEXT_V_ENV(vfrsqrt7_v_w, 4)
+GEN_VEXT_V_ENV(vfrsqrt7_v_d, 8)
 
 /*
  * Vector Floating-Point Reciprocal Estimate Instruction
@@ -3996,36 +4021,36 @@ static float64 frec7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrec7_v_h, OP_UU_H, H2, H2, frec7_h)
 RVVCALL(OPFVV1, vfrec7_v_w, OP_UU_W, H4, H4, frec7_s)
 RVVCALL(OPFVV1, vfrec7_v_d, OP_UU_D, H8, H8, frec7_d)
-GEN_VEXT_V_ENV(vfrec7_v_h)
-GEN_VEXT_V_ENV(vfrec7_v_w)
-GEN_VEXT_V_ENV(vfrec7_v_d)
+GEN_VEXT_V_ENV(vfrec7_v_h, 2)
+GEN_VEXT_V_ENV(vfrec7_v_w, 4)
+GEN_VEXT_V_ENV(vfrec7_v_d, 8)
 
 /* Vector Floating-Point MIN/MAX Instructions */
 RVVCALL(OPFVV2, vfmin_vv_h, OP_UUU_H, H2, H2, H2, float16_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_w, OP_UUU_W, H4, H4, H4, float32_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_d, OP_UUU_D, H8, H8, H8, float64_minimum_number)
-GEN_VEXT_VV_ENV(vfmin_vv_h)
-GEN_VEXT_VV_ENV(vfmin_vv_w)
-GEN_VEXT_VV_ENV(vfmin_vv_d)
+GEN_VEXT_VV_ENV(vfmin_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmin_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmin_vv_d, 8)
 RVVCALL(OPFVF2, vfmin_vf_h, OP_UUU_H, H2, H2, float16_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_w, OP_UUU_W, H4, H4, float32_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_d, OP_UUU_D, H8, H8, float64_minimum_number)
-GEN_VEXT_VF(vfmin_vf_h)
-GEN_VEXT_VF(vfmin_vf_w)
-GEN_VEXT_VF(vfmin_vf_d)
+GEN_VEXT_VF(vfmin_vf_h, 2)
+GEN_VEXT_VF(vfmin_vf_w, 4)
+GEN_VEXT_VF(vfmin_vf_d, 8)
 
 RVVCALL(OPFVV2, vfmax_vv_h, OP_UUU_H, H2, H2, H2, float16_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_w, OP_UUU_W, H4, H4, H4, float32_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_d, OP_UUU_D, H8, H8, H8, float64_maximum_number)
-GEN_VEXT_VV_ENV(vfmax_vv_h)
-GEN_VEXT_VV_ENV(vfmax_vv_w)
-GEN_VEXT_VV_ENV(vfmax_vv_d)
+GEN_VEXT_VV_ENV(vfmax_vv_h, 2)
+GEN_VEXT_VV_ENV(vfmax_vv_w, 4)
+GEN_VEXT_VV_ENV(vfmax_vv_d, 8)
 RVVCALL(OPFVF2, vfmax_vf_h, OP_UUU_H, H2, H2, float16_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_w, OP_UUU_W, H4, H4, float32_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_d, OP_UUU_D, H8, H8, float64_maximum_number)
-GEN_VEXT_VF(vfmax_vf_h)
-GEN_VEXT_VF(vfmax_vf_w)
-GEN_VEXT_VF(vfmax_vf_d)
+GEN_VEXT_VF(vfmax_vf_h, 2)
+GEN_VEXT_VF(vfmax_vf_w, 4)
+GEN_VEXT_VF(vfmax_vf_d, 8)
 
 /* Vector Floating-Point Sign-Injection Instructions */
 static uint16_t fsgnj16(uint16_t a, uint16_t b, float_status *s)
@@ -4046,15 +4071,15 @@ static uint64_t fsgnj64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnj_vv_h, OP_UUU_H, H2, H2, H2, fsgnj16)
 RVVCALL(OPFVV2, vfsgnj_vv_w, OP_UUU_W, H4, H4, H4, fsgnj32)
 RVVCALL(OPFVV2, vfsgnj_vv_d, OP_UUU_D, H8, H8, H8, fsgnj64)
-GEN_VEXT_VV_ENV(vfsgnj_vv_h)
-GEN_VEXT_VV_ENV(vfsgnj_vv_w)
-GEN_VEXT_VV_ENV(vfsgnj_vv_d)
+GEN_VEXT_VV_ENV(vfsgnj_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsgnj_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsgnj_vv_d, 8)
 RVVCALL(OPFVF2, vfsgnj_vf_h, OP_UUU_H, H2, H2, fsgnj16)
 RVVCALL(OPFVF2, vfsgnj_vf_w, OP_UUU_W, H4, H4, fsgnj32)
 RVVCALL(OPFVF2, vfsgnj_vf_d, OP_UUU_D, H8, H8, fsgnj64)
-GEN_VEXT_VF(vfsgnj_vf_h)
-GEN_VEXT_VF(vfsgnj_vf_w)
-GEN_VEXT_VF(vfsgnj_vf_d)
+GEN_VEXT_VF(vfsgnj_vf_h, 2)
+GEN_VEXT_VF(vfsgnj_vf_w, 4)
+GEN_VEXT_VF(vfsgnj_vf_d, 8)
 
 static uint16_t fsgnjn16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -4074,15 +4099,15 @@ static uint64_t fsgnjn64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjn_vv_h, OP_UUU_H, H2, H2, H2, fsgnjn16)
 RVVCALL(OPFVV2, vfsgnjn_vv_w, OP_UUU_W, H4, H4, H4, fsgnjn32)
 RVVCALL(OPFVV2, vfsgnjn_vv_d, OP_UUU_D, H8, H8, H8, fsgnjn64)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_h)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_w)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_d)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_d, 8)
 RVVCALL(OPFVF2, vfsgnjn_vf_h, OP_UUU_H, H2, H2, fsgnjn16)
 RVVCALL(OPFVF2, vfsgnjn_vf_w, OP_UUU_W, H4, H4, fsgnjn32)
 RVVCALL(OPFVF2, vfsgnjn_vf_d, OP_UUU_D, H8, H8, fsgnjn64)
-GEN_VEXT_VF(vfsgnjn_vf_h)
-GEN_VEXT_VF(vfsgnjn_vf_w)
-GEN_VEXT_VF(vfsgnjn_vf_d)
+GEN_VEXT_VF(vfsgnjn_vf_h, 2)
+GEN_VEXT_VF(vfsgnjn_vf_w, 4)
+GEN_VEXT_VF(vfsgnjn_vf_d, 8)
 
 static uint16_t fsgnjx16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -4102,15 +4127,15 @@ static uint64_t fsgnjx64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjx_vv_h, OP_UUU_H, H2, H2, H2, fsgnjx16)
 RVVCALL(OPFVV2, vfsgnjx_vv_w, OP_UUU_W, H4, H4, H4, fsgnjx32)
 RVVCALL(OPFVV2, vfsgnjx_vv_d, OP_UUU_D, H8, H8, H8, fsgnjx64)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_h)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_w)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_d)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_h, 2)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_w, 4)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_d, 8)
 RVVCALL(OPFVF2, vfsgnjx_vf_h, OP_UUU_H, H2, H2, fsgnjx16)
 RVVCALL(OPFVF2, vfsgnjx_vf_w, OP_UUU_W, H4, H4, fsgnjx32)
 RVVCALL(OPFVF2, vfsgnjx_vf_d, OP_UUU_D, H8, H8, fsgnjx64)
-GEN_VEXT_VF(vfsgnjx_vf_h)
-GEN_VEXT_VF(vfsgnjx_vf_w)
-GEN_VEXT_VF(vfsgnjx_vf_d)
+GEN_VEXT_VF(vfsgnjx_vf_h, 2)
+GEN_VEXT_VF(vfsgnjx_vf_w, 4)
+GEN_VEXT_VF(vfsgnjx_vf_d, 8)
 
 /* Vector Floating-Point Compare Instructions */
 #define GEN_VEXT_CMP_VV_ENV(NAME, ETYPE, H, DO_OP)            \
@@ -4119,6 +4144,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
 {                                                             \
     uint32_t vm = vext_vm(desc);                              \
     uint32_t vl = env->vl;                                    \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -4131,6 +4159,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
                            DO_OP(s2, s1, &env->fp_status));   \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    if (vta) {                                                \
+        for (; i < total_elems; i++) {                        \
+            vext_set_elem_mask(vd, i, 1);                     \
+        }                                                     \
+    }                                                         \
 }
 
 GEN_VEXT_CMP_VV_ENV(vmfeq_vv_h, uint16_t, H2, float16_eq_quiet)
@@ -4143,6 +4177,9 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
 {                                                                   \
     uint32_t vm = vext_vm(desc);                                    \
     uint32_t vl = env->vl;                                          \
+    uint32_t total_elems =                                          \
+        vext_get_total_elems(env_archcpu(env), env->vtype);         \
+    uint32_t vta = vext_vta(desc);                                  \
     uint32_t i;                                                     \
                                                                     \
     for (i = env->vstart; i < vl; i++) {                            \
@@ -4154,6 +4191,12 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
                            DO_OP(s2, (ETYPE)s1, &env->fp_status));  \
     }                                                               \
     env->vstart = 0;                                                \
+    /* set tail elements to 1s */                                   \
+    if (vta) {                                                      \
+        for (; i < total_elems; i++) {                              \
+            vext_set_elem_mask(vd, i, 1);                           \
+        }                                                           \
+    }                                                               \
 }
 
 GEN_VEXT_CMP_VF(vmfeq_vf_h, uint16_t, H2, float16_eq_quiet)
@@ -4254,12 +4297,16 @@ static void do_##NAME(void *vd, void *vs2, int i)      \
     *((TD *)vd + HD(i)) = OP(s2);                      \
 }
 
-#define GEN_VEXT_V(NAME)                               \
+#define GEN_VEXT_V(NAME, ESZ)                          \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
                   CPURISCVState *env, uint32_t desc)   \
 {                                                      \
     uint32_t vm = vext_vm(desc);                       \
     uint32_t vl = env->vl;                             \
+    uint32_t total_elems =                             \
+        vext_get_total_elems(env_archcpu(env),         \
+                            env->vtype);               \
+    uint32_t vta = vext_vta(desc);                     \
     uint32_t i;                                        \
                                                        \
     for (i = env->vstart; i < vl; i++) {               \
@@ -4269,6 +4316,11 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         do_##NAME(vd, vs2, i);                         \
     }                                                  \
     env->vstart = 0;                                   \
+    /* set tail elements to 1s */                      \
+    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,      \
+                                     vl * ESZ,         \
+                                     total_elems *     \
+                                     ESZ);             \
 }
 
 target_ulong fclass_h(uint64_t frs1)
@@ -4331,17 +4383,22 @@ target_ulong fclass_d(uint64_t frs1)
 RVVCALL(OPIVV1, vfclass_v_h, OP_UU_H, H2, H2, fclass_h)
 RVVCALL(OPIVV1, vfclass_v_w, OP_UU_W, H4, H4, fclass_s)
 RVVCALL(OPIVV1, vfclass_v_d, OP_UU_D, H8, H8, fclass_d)
-GEN_VEXT_V(vfclass_v_h)
-GEN_VEXT_V(vfclass_v_w)
-GEN_VEXT_V(vfclass_v_d)
+GEN_VEXT_V(vfclass_v_h, 2)
+GEN_VEXT_V(vfclass_v_w, 4)
+GEN_VEXT_V(vfclass_v_d, 8)
 
 /* Vector Floating-Point Merge Instruction */
+
 #define GEN_VFMERGE_VF(NAME, ETYPE, H)                        \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
     uint32_t vm = vext_vm(desc);                              \
     uint32_t vl = env->vl;                                    \
+    uint32_t esz = sizeof(ETYPE);                             \
+    uint32_t total_elems =                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);   \
+    uint32_t vta = vext_vta(desc);                            \
     uint32_t i;                                               \
                                                               \
     for (i = env->vstart; i < vl; i++) {                      \
@@ -4350,6 +4407,9 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
           = (!vm && !vext_elem_mask(v0, i) ? s2 : s1);        \
     }                                                         \
     env->vstart = 0;                                          \
+    /* set tail elements to 1s */                             \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,   \
+                                     total_elems * esz);      \
 }
 
 GEN_VFMERGE_VF(vfmerge_vfm_h, int16_t, H2)
@@ -4361,33 +4421,33 @@ GEN_VFMERGE_VF(vfmerge_vfm_d, int64_t, H8)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_h, OP_UU_H, H2, H2, float16_to_uint16)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_w, OP_UU_W, H4, H4, float32_to_uint32)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_d, OP_UU_D, H8, H8, float64_to_uint64)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_h)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_w)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_d)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_d, 8)
 
 /* vfcvt.x.f.v vd, vs2, vm # Convert float to signed integer. */
 RVVCALL(OPFVV1, vfcvt_x_f_v_h, OP_UU_H, H2, H2, float16_to_int16)
 RVVCALL(OPFVV1, vfcvt_x_f_v_w, OP_UU_W, H4, H4, float32_to_int32)
 RVVCALL(OPFVV1, vfcvt_x_f_v_d, OP_UU_D, H8, H8, float64_to_int64)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_h)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_w)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_d)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_d, 8)
 
 /* vfcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_xu_v_h, OP_UU_H, H2, H2, uint16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_w, OP_UU_W, H4, H4, uint32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_d, OP_UU_D, H8, H8, uint64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_h)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_w)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_d)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_d, 8)
 
 /* vfcvt.f.x.v vd, vs2, vm # Convert integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_x_v_h, OP_UU_H, H2, H2, int16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_x_v_w, OP_UU_W, H4, H4, int32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_x_v_d, OP_UU_D, H8, H8, int64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_h)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_w)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_d)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_h, 2)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_w, 4)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_d, 8)
 
 /* Widening Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4397,30 +4457,30 @@ GEN_VEXT_V_ENV(vfcvt_f_x_v_d)
 /* vfwcvt.xu.f.v vd, vs2, vm # Convert float to double-width unsigned integer.*/
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_h, WOP_UU_H, H4, H2, float16_to_uint32)
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_w, WOP_UU_W, H8, H4, float32_to_uint64)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w, 8)
 
 /* vfwcvt.x.f.v vd, vs2, vm # Convert float to double-width signed integer. */
 RVVCALL(OPFVV1, vfwcvt_x_f_v_h, WOP_UU_H, H4, H2, float16_to_int32)
 RVVCALL(OPFVV1, vfwcvt_x_f_v_w, WOP_UU_W, H8, H4, float32_to_int64)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_h)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_w)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_w, 8)
 
 /* vfwcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to double-width float */
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_b, WOP_UU_B, H2, H1, uint8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_h, WOP_UU_H, H4, H2, uint16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_w, WOP_UU_W, H8, H4, uint32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b, 2)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w, 8)
 
 /* vfwcvt.f.x.v vd, vs2, vm # Convert integer to double-width float. */
 RVVCALL(OPFVV1, vfwcvt_f_x_v_b, WOP_UU_B, H2, H1, int8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_h, WOP_UU_H, H4, H2, int16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_w, WOP_UU_W, H8, H4, int32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_b)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_h)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_w)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_b, 2)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_w, 8)
 
 /*
  * vfwcvt.f.f.v vd, vs2, vm
@@ -4433,8 +4493,8 @@ static uint32_t vfwcvtffv16(uint16_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfwcvt_f_f_v_h, WOP_UU_H, H4, H2, vfwcvtffv16)
 RVVCALL(OPFVV1, vfwcvt_f_f_v_w, WOP_UU_W, H8, H4, float32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_h)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_w)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_h, 4)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_w, 8)
 
 /* Narrowing Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4445,29 +4505,29 @@ GEN_VEXT_V_ENV(vfwcvt_f_f_v_w)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_b, NOP_UU_B, H1, H2, float16_to_uint8)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_h, NOP_UU_H, H2, H4, float32_to_uint16)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_w, NOP_UU_W, H4, H8, float64_to_uint32)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_b)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_h)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_w)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_b, 1)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_w, 4)
 
 /* vfncvt.x.f.v vd, vs2, vm # Convert double-width float to signed integer. */
 RVVCALL(OPFVV1, vfncvt_x_f_w_b, NOP_UU_B, H1, H2, float16_to_int8)
 RVVCALL(OPFVV1, vfncvt_x_f_w_h, NOP_UU_H, H2, H4, float32_to_int16)
 RVVCALL(OPFVV1, vfncvt_x_f_w_w, NOP_UU_W, H4, H8, float64_to_int32)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_b)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_h)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_w)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_b, 1)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_w, 4)
 
 /* vfncvt.f.xu.v vd, vs2, vm # Convert double-width unsigned integer to float */
 RVVCALL(OPFVV1, vfncvt_f_xu_w_h, NOP_UU_H, H2, H4, uint32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_xu_w_w, NOP_UU_W, H4, H8, uint64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_h)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_w)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_w, 4)
 
 /* vfncvt.f.x.v vd, vs2, vm # Convert double-width integer to float. */
 RVVCALL(OPFVV1, vfncvt_f_x_w_h, NOP_UU_H, H2, H4, int32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_x_w_w, NOP_UU_W, H4, H8, int64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_h)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_w)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_w, 4)
 
 /* vfncvt.f.f.v vd, vs2, vm # Convert double float to single-width float. */
 static uint16_t vfncvtffv16(uint32_t a, float_status *s)
@@ -4477,8 +4537,8 @@ static uint16_t vfncvtffv16(uint32_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfncvt_f_f_w_h, NOP_UU_H, H2, H4, vfncvtffv16)
 RVVCALL(OPFVV1, vfncvt_f_f_w_w, NOP_UU_W, H4, H8, float64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_h)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_w)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_h, 2)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_w, 4)
 
 /*
  *** Vector Reduction Operations
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 11/14] target/riscv: rvv: Add tail agnostic for vector floating-point instructions
  2022-03-24 19:03   ` ~eopxd
@ 2022-03-25 10:42     ` Weiwei Li
  -1 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:42 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, Bin Meng, Alistair Francis, eop Chen,
	Palmer Dabbelt


在 2022/3/7 下午6:05, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc |  11 +
>   target/riscv/vector_helper.c            | 452 ++++++++++++++----------
>   2 files changed, 267 insertions(+), 196 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index 24784b6f5f..5bc340f419 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -2391,6 +2391,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs1),                    \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
> @@ -2473,6 +2474,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)            \
>           gen_set_rm(s, RISCV_FRM_DYN);                             \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);            \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);              \
>           return opfvf_trans(a->rd, a->rs1, a->rs2, data,           \
>                              fns[s->sew - 1], s);                   \
>       }                                                             \
> @@ -2511,6 +2513,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
>                                                                    \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);               \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),   \
>                              vreg_ofs(s, a->rs1),                  \
>                              vreg_ofs(s, a->rs2), cpu_env,         \
> @@ -2550,6 +2553,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
>           gen_set_rm(s, RISCV_FRM_DYN);                            \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);               \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
>           return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
>                              fns[s->sew - 1], s);                  \
>       }                                                            \
> @@ -2586,6 +2590,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs1),                    \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
> @@ -2625,6 +2630,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
>           gen_set_rm(s, RISCV_FRM_DYN);                            \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);               \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
>           return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
>                              fns[s->sew - 1], s);                  \
>       }                                                            \
> @@ -2708,6 +2714,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
>   
>           data = FIELD_DP32(data, VDATA, VM, a->vm);
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                              vreg_ofs(s, a->rs2), cpu_env,
>                              s->cfg_ptr->vlen / 8,
> @@ -2921,6 +2928,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> @@ -2972,6 +2980,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>           tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> @@ -3039,6 +3048,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> @@ -3092,6 +3102,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>           tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index 067095e19a..d654377fc8 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -2997,13 +2997,17 @@ static void do_##NAME(void *vd, void *vs1, void *vs2, int i,   \
>       *((TD *)vd + HD(i)) = OP(s2, s1, &env->fp_status);         \
>   }
>   
> -#define GEN_VEXT_VV_ENV(NAME)                             \
> +#define GEN_VEXT_VV_ENV(NAME, ESZ)                        \
>   void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
>                     void *vs2, CPURISCVState *env,          \
>                     uint32_t desc)                          \
>   {                                                         \
>       uint32_t vm = vext_vm(desc);                          \
>       uint32_t vl = env->vl;                                \
> +    uint32_t total_elems =                                \
> +        vext_get_total_elems(env_archcpu(env),            \
> +                             env->vtype);                 \
> +    uint32_t vta = vext_vta(desc);                        \
>       uint32_t i;                                           \
>                                                             \
>       for (i = env->vstart; i < vl; i++) {                  \
> @@ -3013,14 +3017,18 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
>           do_##NAME(vd, vs1, vs2, i, env);                  \
>       }                                                     \
>       env->vstart = 0;                                      \
> +    /* set tail elements to 1s */                         \
> +    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
> +                                     vl * ESZ,            \
> +                                     total_elems * ESZ);  \
>   }
>   
>   RVVCALL(OPFVV2, vfadd_vv_h, OP_UUU_H, H2, H2, H2, float16_add)
>   RVVCALL(OPFVV2, vfadd_vv_w, OP_UUU_W, H4, H4, H4, float32_add)
>   RVVCALL(OPFVV2, vfadd_vv_d, OP_UUU_D, H8, H8, H8, float64_add)
> -GEN_VEXT_VV_ENV(vfadd_vv_h)
> -GEN_VEXT_VV_ENV(vfadd_vv_w)
> -GEN_VEXT_VV_ENV(vfadd_vv_d)
> +GEN_VEXT_VV_ENV(vfadd_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfadd_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfadd_vv_d, 8)
>   
>   #define OPFVF2(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)        \
>   static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
> @@ -3030,13 +3038,17 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
>       *((TD *)vd + HD(i)) = OP(s2, (TX1)(T1)s1, &env->fp_status);\
>   }
>   
> -#define GEN_VEXT_VF(NAME)                                 \
> +#define GEN_VEXT_VF(NAME, ESZ)                            \
>   void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
>                     void *vs2, CPURISCVState *env,          \
>                     uint32_t desc)                          \
>   {                                                         \
>       uint32_t vm = vext_vm(desc);                          \
>       uint32_t vl = env->vl;                                \
> +    uint32_t total_elems =                                \
> +        vext_get_total_elems(env_archcpu(env),            \
> +                             env->vtype);                 \
> +    uint32_t vta = vext_vta(desc);                        \
>       uint32_t i;                                           \
>                                                             \
>       for (i = env->vstart; i < vl; i++) {                  \
> @@ -3046,27 +3058,31 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
>           do_##NAME(vd, s1, vs2, i, env);                   \
>       }                                                     \
>       env->vstart = 0;                                      \
> +    /* set tail elements to 1s */                         \
> +    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
> +                                     vl * ESZ,            \
> +                                     total_elems * ESZ);  \
>   }
>   
>   RVVCALL(OPFVF2, vfadd_vf_h, OP_UUU_H, H2, H2, float16_add)
>   RVVCALL(OPFVF2, vfadd_vf_w, OP_UUU_W, H4, H4, float32_add)
>   RVVCALL(OPFVF2, vfadd_vf_d, OP_UUU_D, H8, H8, float64_add)
> -GEN_VEXT_VF(vfadd_vf_h)
> -GEN_VEXT_VF(vfadd_vf_w)
> -GEN_VEXT_VF(vfadd_vf_d)
> +GEN_VEXT_VF(vfadd_vf_h, 2)
> +GEN_VEXT_VF(vfadd_vf_w, 4)
> +GEN_VEXT_VF(vfadd_vf_d, 8)
>   
>   RVVCALL(OPFVV2, vfsub_vv_h, OP_UUU_H, H2, H2, H2, float16_sub)
>   RVVCALL(OPFVV2, vfsub_vv_w, OP_UUU_W, H4, H4, H4, float32_sub)
>   RVVCALL(OPFVV2, vfsub_vv_d, OP_UUU_D, H8, H8, H8, float64_sub)
> -GEN_VEXT_VV_ENV(vfsub_vv_h)
> -GEN_VEXT_VV_ENV(vfsub_vv_w)
> -GEN_VEXT_VV_ENV(vfsub_vv_d)
> +GEN_VEXT_VV_ENV(vfsub_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsub_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsub_vv_d, 8)
>   RVVCALL(OPFVF2, vfsub_vf_h, OP_UUU_H, H2, H2, float16_sub)
>   RVVCALL(OPFVF2, vfsub_vf_w, OP_UUU_W, H4, H4, float32_sub)
>   RVVCALL(OPFVF2, vfsub_vf_d, OP_UUU_D, H8, H8, float64_sub)
> -GEN_VEXT_VF(vfsub_vf_h)
> -GEN_VEXT_VF(vfsub_vf_w)
> -GEN_VEXT_VF(vfsub_vf_d)
> +GEN_VEXT_VF(vfsub_vf_h, 2)
> +GEN_VEXT_VF(vfsub_vf_w, 4)
> +GEN_VEXT_VF(vfsub_vf_d, 8)
>   
>   static uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -3086,9 +3102,9 @@ static uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVF2, vfrsub_vf_h, OP_UUU_H, H2, H2, float16_rsub)
>   RVVCALL(OPFVF2, vfrsub_vf_w, OP_UUU_W, H4, H4, float32_rsub)
>   RVVCALL(OPFVF2, vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub)
> -GEN_VEXT_VF(vfrsub_vf_h)
> -GEN_VEXT_VF(vfrsub_vf_w)
> -GEN_VEXT_VF(vfrsub_vf_d)
> +GEN_VEXT_VF(vfrsub_vf_h, 2)
> +GEN_VEXT_VF(vfrsub_vf_w, 4)
> +GEN_VEXT_VF(vfrsub_vf_d, 8)
>   
>   /* Vector Widening Floating-Point Add/Subtract Instructions */
>   static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s)
> @@ -3106,12 +3122,12 @@ static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16)
>   RVVCALL(OPFVV2, vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32)
> -GEN_VEXT_VV_ENV(vfwadd_vv_h)
> -GEN_VEXT_VV_ENV(vfwadd_vv_w)
> +GEN_VEXT_VV_ENV(vfwadd_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwadd_vv_w, 8)
>   RVVCALL(OPFVF2, vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16)
>   RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32)
> -GEN_VEXT_VF(vfwadd_vf_h)
> -GEN_VEXT_VF(vfwadd_vf_w)
> +GEN_VEXT_VF(vfwadd_vf_h, 4)
> +GEN_VEXT_VF(vfwadd_vf_w, 8)
>   
>   static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -3128,12 +3144,12 @@ static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16)
>   RVVCALL(OPFVV2, vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32)
> -GEN_VEXT_VV_ENV(vfwsub_vv_h)
> -GEN_VEXT_VV_ENV(vfwsub_vv_w)
> +GEN_VEXT_VV_ENV(vfwsub_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwsub_vv_w, 8)
>   RVVCALL(OPFVF2, vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16)
>   RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32)
> -GEN_VEXT_VF(vfwsub_vf_h)
> -GEN_VEXT_VF(vfwsub_vf_w)
> +GEN_VEXT_VF(vfwsub_vf_h, 4)
> +GEN_VEXT_VF(vfwsub_vf_w, 8)
>   
>   static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s)
>   {
> @@ -3147,12 +3163,12 @@ static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16)
>   RVVCALL(OPFVV2, vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32)
> -GEN_VEXT_VV_ENV(vfwadd_wv_h)
> -GEN_VEXT_VV_ENV(vfwadd_wv_w)
> +GEN_VEXT_VV_ENV(vfwadd_wv_h, 4)
> +GEN_VEXT_VV_ENV(vfwadd_wv_w, 8)
>   RVVCALL(OPFVF2, vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16)
>   RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32)
> -GEN_VEXT_VF(vfwadd_wf_h)
> -GEN_VEXT_VF(vfwadd_wf_w)
> +GEN_VEXT_VF(vfwadd_wf_h, 4)
> +GEN_VEXT_VF(vfwadd_wf_w, 8)
>   
>   static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s)
>   {
> @@ -3166,39 +3182,39 @@ static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16)
>   RVVCALL(OPFVV2, vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32)
> -GEN_VEXT_VV_ENV(vfwsub_wv_h)
> -GEN_VEXT_VV_ENV(vfwsub_wv_w)
> +GEN_VEXT_VV_ENV(vfwsub_wv_h, 4)
> +GEN_VEXT_VV_ENV(vfwsub_wv_w, 8)
>   RVVCALL(OPFVF2, vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16)
>   RVVCALL(OPFVF2, vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32)
> -GEN_VEXT_VF(vfwsub_wf_h)
> -GEN_VEXT_VF(vfwsub_wf_w)
> +GEN_VEXT_VF(vfwsub_wf_h, 4)
> +GEN_VEXT_VF(vfwsub_wf_w, 8)
>   
>   /* Vector Single-Width Floating-Point Multiply/Divide Instructions */
>   RVVCALL(OPFVV2, vfmul_vv_h, OP_UUU_H, H2, H2, H2, float16_mul)
>   RVVCALL(OPFVV2, vfmul_vv_w, OP_UUU_W, H4, H4, H4, float32_mul)
>   RVVCALL(OPFVV2, vfmul_vv_d, OP_UUU_D, H8, H8, H8, float64_mul)
> -GEN_VEXT_VV_ENV(vfmul_vv_h)
> -GEN_VEXT_VV_ENV(vfmul_vv_w)
> -GEN_VEXT_VV_ENV(vfmul_vv_d)
> +GEN_VEXT_VV_ENV(vfmul_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmul_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmul_vv_d, 8)
>   RVVCALL(OPFVF2, vfmul_vf_h, OP_UUU_H, H2, H2, float16_mul)
>   RVVCALL(OPFVF2, vfmul_vf_w, OP_UUU_W, H4, H4, float32_mul)
>   RVVCALL(OPFVF2, vfmul_vf_d, OP_UUU_D, H8, H8, float64_mul)
> -GEN_VEXT_VF(vfmul_vf_h)
> -GEN_VEXT_VF(vfmul_vf_w)
> -GEN_VEXT_VF(vfmul_vf_d)
> +GEN_VEXT_VF(vfmul_vf_h, 2)
> +GEN_VEXT_VF(vfmul_vf_w, 4)
> +GEN_VEXT_VF(vfmul_vf_d, 8)
>   
>   RVVCALL(OPFVV2, vfdiv_vv_h, OP_UUU_H, H2, H2, H2, float16_div)
>   RVVCALL(OPFVV2, vfdiv_vv_w, OP_UUU_W, H4, H4, H4, float32_div)
>   RVVCALL(OPFVV2, vfdiv_vv_d, OP_UUU_D, H8, H8, H8, float64_div)
> -GEN_VEXT_VV_ENV(vfdiv_vv_h)
> -GEN_VEXT_VV_ENV(vfdiv_vv_w)
> -GEN_VEXT_VV_ENV(vfdiv_vv_d)
> +GEN_VEXT_VV_ENV(vfdiv_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfdiv_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfdiv_vv_d, 8)
>   RVVCALL(OPFVF2, vfdiv_vf_h, OP_UUU_H, H2, H2, float16_div)
>   RVVCALL(OPFVF2, vfdiv_vf_w, OP_UUU_W, H4, H4, float32_div)
>   RVVCALL(OPFVF2, vfdiv_vf_d, OP_UUU_D, H8, H8, float64_div)
> -GEN_VEXT_VF(vfdiv_vf_h)
> -GEN_VEXT_VF(vfdiv_vf_w)
> -GEN_VEXT_VF(vfdiv_vf_d)
> +GEN_VEXT_VF(vfdiv_vf_h, 2)
> +GEN_VEXT_VF(vfdiv_vf_w, 4)
> +GEN_VEXT_VF(vfdiv_vf_d, 8)
>   
>   static uint16_t float16_rdiv(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -3218,9 +3234,9 @@ static uint64_t float64_rdiv(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVF2, vfrdiv_vf_h, OP_UUU_H, H2, H2, float16_rdiv)
>   RVVCALL(OPFVF2, vfrdiv_vf_w, OP_UUU_W, H4, H4, float32_rdiv)
>   RVVCALL(OPFVF2, vfrdiv_vf_d, OP_UUU_D, H8, H8, float64_rdiv)
> -GEN_VEXT_VF(vfrdiv_vf_h)
> -GEN_VEXT_VF(vfrdiv_vf_w)
> -GEN_VEXT_VF(vfrdiv_vf_d)
> +GEN_VEXT_VF(vfrdiv_vf_h, 2)
> +GEN_VEXT_VF(vfrdiv_vf_w, 4)
> +GEN_VEXT_VF(vfrdiv_vf_d, 8)
>   
>   /* Vector Widening Floating-Point Multiply */
>   static uint32_t vfwmul16(uint16_t a, uint16_t b, float_status *s)
> @@ -3237,12 +3253,12 @@ static uint64_t vfwmul32(uint32_t a, uint32_t b, float_status *s)
>   }
>   RVVCALL(OPFVV2, vfwmul_vv_h, WOP_UUU_H, H4, H2, H2, vfwmul16)
>   RVVCALL(OPFVV2, vfwmul_vv_w, WOP_UUU_W, H8, H4, H4, vfwmul32)
> -GEN_VEXT_VV_ENV(vfwmul_vv_h)
> -GEN_VEXT_VV_ENV(vfwmul_vv_w)
> +GEN_VEXT_VV_ENV(vfwmul_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwmul_vv_w, 8)
>   RVVCALL(OPFVF2, vfwmul_vf_h, WOP_UUU_H, H4, H2, vfwmul16)
>   RVVCALL(OPFVF2, vfwmul_vf_w, WOP_UUU_W, H8, H4, vfwmul32)
> -GEN_VEXT_VF(vfwmul_vf_h)
> -GEN_VEXT_VF(vfwmul_vf_w)
> +GEN_VEXT_VF(vfwmul_vf_h, 4)
> +GEN_VEXT_VF(vfwmul_vf_w, 8)
>   
>   /* Vector Single-Width Floating-Point Fused Multiply-Add Instructions */
>   #define OPFVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)       \
> @@ -3273,9 +3289,9 @@ static uint64_t fmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmacc_vv_h, OP_UUU_H, H2, H2, H2, fmacc16)
>   RVVCALL(OPFVV3, vfmacc_vv_w, OP_UUU_W, H4, H4, H4, fmacc32)
>   RVVCALL(OPFVV3, vfmacc_vv_d, OP_UUU_D, H8, H8, H8, fmacc64)
> -GEN_VEXT_VV_ENV(vfmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfmacc_vv_w)
> -GEN_VEXT_VV_ENV(vfmacc_vv_d)
> +GEN_VEXT_VV_ENV(vfmacc_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmacc_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmacc_vv_d, 8)
>   
>   #define OPFVF3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)           \
>   static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
> @@ -3289,9 +3305,9 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
>   RVVCALL(OPFVF3, vfmacc_vf_h, OP_UUU_H, H2, H2, fmacc16)
>   RVVCALL(OPFVF3, vfmacc_vf_w, OP_UUU_W, H4, H4, fmacc32)
>   RVVCALL(OPFVF3, vfmacc_vf_d, OP_UUU_D, H8, H8, fmacc64)
> -GEN_VEXT_VF(vfmacc_vf_h)
> -GEN_VEXT_VF(vfmacc_vf_w)
> -GEN_VEXT_VF(vfmacc_vf_d)
> +GEN_VEXT_VF(vfmacc_vf_h, 2)
> +GEN_VEXT_VF(vfmacc_vf_w, 4)
> +GEN_VEXT_VF(vfmacc_vf_d, 8)
>   
>   static uint16_t fnmacc16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3314,15 +3330,15 @@ static uint64_t fnmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmacc_vv_h, OP_UUU_H, H2, H2, H2, fnmacc16)
>   RVVCALL(OPFVV3, vfnmacc_vv_w, OP_UUU_W, H4, H4, H4, fnmacc32)
>   RVVCALL(OPFVV3, vfnmacc_vv_d, OP_UUU_D, H8, H8, H8, fnmacc64)
> -GEN_VEXT_VV_ENV(vfnmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfnmacc_vv_w)
> -GEN_VEXT_VV_ENV(vfnmacc_vv_d)
> +GEN_VEXT_VV_ENV(vfnmacc_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmacc_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmacc_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmacc_vf_h, OP_UUU_H, H2, H2, fnmacc16)
>   RVVCALL(OPFVF3, vfnmacc_vf_w, OP_UUU_W, H4, H4, fnmacc32)
>   RVVCALL(OPFVF3, vfnmacc_vf_d, OP_UUU_D, H8, H8, fnmacc64)
> -GEN_VEXT_VF(vfnmacc_vf_h)
> -GEN_VEXT_VF(vfnmacc_vf_w)
> -GEN_VEXT_VF(vfnmacc_vf_d)
> +GEN_VEXT_VF(vfnmacc_vf_h, 2)
> +GEN_VEXT_VF(vfnmacc_vf_w, 4)
> +GEN_VEXT_VF(vfnmacc_vf_d, 8)
>   
>   static uint16_t fmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3342,15 +3358,15 @@ static uint64_t fmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmsac_vv_h, OP_UUU_H, H2, H2, H2, fmsac16)
>   RVVCALL(OPFVV3, vfmsac_vv_w, OP_UUU_W, H4, H4, H4, fmsac32)
>   RVVCALL(OPFVV3, vfmsac_vv_d, OP_UUU_D, H8, H8, H8, fmsac64)
> -GEN_VEXT_VV_ENV(vfmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfmsac_vv_w)
> -GEN_VEXT_VV_ENV(vfmsac_vv_d)
> +GEN_VEXT_VV_ENV(vfmsac_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmsac_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmsac_vv_d, 8)
>   RVVCALL(OPFVF3, vfmsac_vf_h, OP_UUU_H, H2, H2, fmsac16)
>   RVVCALL(OPFVF3, vfmsac_vf_w, OP_UUU_W, H4, H4, fmsac32)
>   RVVCALL(OPFVF3, vfmsac_vf_d, OP_UUU_D, H8, H8, fmsac64)
> -GEN_VEXT_VF(vfmsac_vf_h)
> -GEN_VEXT_VF(vfmsac_vf_w)
> -GEN_VEXT_VF(vfmsac_vf_d)
> +GEN_VEXT_VF(vfmsac_vf_h, 2)
> +GEN_VEXT_VF(vfmsac_vf_w, 4)
> +GEN_VEXT_VF(vfmsac_vf_d, 8)
>   
>   static uint16_t fnmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3370,15 +3386,15 @@ static uint64_t fnmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmsac_vv_h, OP_UUU_H, H2, H2, H2, fnmsac16)
>   RVVCALL(OPFVV3, vfnmsac_vv_w, OP_UUU_W, H4, H4, H4, fnmsac32)
>   RVVCALL(OPFVV3, vfnmsac_vv_d, OP_UUU_D, H8, H8, H8, fnmsac64)
> -GEN_VEXT_VV_ENV(vfnmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfnmsac_vv_w)
> -GEN_VEXT_VV_ENV(vfnmsac_vv_d)
> +GEN_VEXT_VV_ENV(vfnmsac_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmsac_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmsac_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmsac_vf_h, OP_UUU_H, H2, H2, fnmsac16)
>   RVVCALL(OPFVF3, vfnmsac_vf_w, OP_UUU_W, H4, H4, fnmsac32)
>   RVVCALL(OPFVF3, vfnmsac_vf_d, OP_UUU_D, H8, H8, fnmsac64)
> -GEN_VEXT_VF(vfnmsac_vf_h)
> -GEN_VEXT_VF(vfnmsac_vf_w)
> -GEN_VEXT_VF(vfnmsac_vf_d)
> +GEN_VEXT_VF(vfnmsac_vf_h, 2)
> +GEN_VEXT_VF(vfnmsac_vf_w, 4)
> +GEN_VEXT_VF(vfnmsac_vf_d, 8)
>   
>   static uint16_t fmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3398,15 +3414,15 @@ static uint64_t fmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmadd_vv_h, OP_UUU_H, H2, H2, H2, fmadd16)
>   RVVCALL(OPFVV3, vfmadd_vv_w, OP_UUU_W, H4, H4, H4, fmadd32)
>   RVVCALL(OPFVV3, vfmadd_vv_d, OP_UUU_D, H8, H8, H8, fmadd64)
> -GEN_VEXT_VV_ENV(vfmadd_vv_h)
> -GEN_VEXT_VV_ENV(vfmadd_vv_w)
> -GEN_VEXT_VV_ENV(vfmadd_vv_d)
> +GEN_VEXT_VV_ENV(vfmadd_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmadd_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmadd_vv_d, 8)
>   RVVCALL(OPFVF3, vfmadd_vf_h, OP_UUU_H, H2, H2, fmadd16)
>   RVVCALL(OPFVF3, vfmadd_vf_w, OP_UUU_W, H4, H4, fmadd32)
>   RVVCALL(OPFVF3, vfmadd_vf_d, OP_UUU_D, H8, H8, fmadd64)
> -GEN_VEXT_VF(vfmadd_vf_h)
> -GEN_VEXT_VF(vfmadd_vf_w)
> -GEN_VEXT_VF(vfmadd_vf_d)
> +GEN_VEXT_VF(vfmadd_vf_h, 2)
> +GEN_VEXT_VF(vfmadd_vf_w, 4)
> +GEN_VEXT_VF(vfmadd_vf_d, 8)
>   
>   static uint16_t fnmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3429,15 +3445,15 @@ static uint64_t fnmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmadd_vv_h, OP_UUU_H, H2, H2, H2, fnmadd16)
>   RVVCALL(OPFVV3, vfnmadd_vv_w, OP_UUU_W, H4, H4, H4, fnmadd32)
>   RVVCALL(OPFVV3, vfnmadd_vv_d, OP_UUU_D, H8, H8, H8, fnmadd64)
> -GEN_VEXT_VV_ENV(vfnmadd_vv_h)
> -GEN_VEXT_VV_ENV(vfnmadd_vv_w)
> -GEN_VEXT_VV_ENV(vfnmadd_vv_d)
> +GEN_VEXT_VV_ENV(vfnmadd_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmadd_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmadd_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmadd_vf_h, OP_UUU_H, H2, H2, fnmadd16)
>   RVVCALL(OPFVF3, vfnmadd_vf_w, OP_UUU_W, H4, H4, fnmadd32)
>   RVVCALL(OPFVF3, vfnmadd_vf_d, OP_UUU_D, H8, H8, fnmadd64)
> -GEN_VEXT_VF(vfnmadd_vf_h)
> -GEN_VEXT_VF(vfnmadd_vf_w)
> -GEN_VEXT_VF(vfnmadd_vf_d)
> +GEN_VEXT_VF(vfnmadd_vf_h, 2)
> +GEN_VEXT_VF(vfnmadd_vf_w, 4)
> +GEN_VEXT_VF(vfnmadd_vf_d, 8)
>   
>   static uint16_t fmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3457,15 +3473,15 @@ static uint64_t fmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmsub_vv_h, OP_UUU_H, H2, H2, H2, fmsub16)
>   RVVCALL(OPFVV3, vfmsub_vv_w, OP_UUU_W, H4, H4, H4, fmsub32)
>   RVVCALL(OPFVV3, vfmsub_vv_d, OP_UUU_D, H8, H8, H8, fmsub64)
> -GEN_VEXT_VV_ENV(vfmsub_vv_h)
> -GEN_VEXT_VV_ENV(vfmsub_vv_w)
> -GEN_VEXT_VV_ENV(vfmsub_vv_d)
> +GEN_VEXT_VV_ENV(vfmsub_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmsub_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmsub_vv_d, 8)
>   RVVCALL(OPFVF3, vfmsub_vf_h, OP_UUU_H, H2, H2, fmsub16)
>   RVVCALL(OPFVF3, vfmsub_vf_w, OP_UUU_W, H4, H4, fmsub32)
>   RVVCALL(OPFVF3, vfmsub_vf_d, OP_UUU_D, H8, H8, fmsub64)
> -GEN_VEXT_VF(vfmsub_vf_h)
> -GEN_VEXT_VF(vfmsub_vf_w)
> -GEN_VEXT_VF(vfmsub_vf_d)
> +GEN_VEXT_VF(vfmsub_vf_h, 2)
> +GEN_VEXT_VF(vfmsub_vf_w, 4)
> +GEN_VEXT_VF(vfmsub_vf_d, 8)
>   
>   static uint16_t fnmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3485,15 +3501,15 @@ static uint64_t fnmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmsub_vv_h, OP_UUU_H, H2, H2, H2, fnmsub16)
>   RVVCALL(OPFVV3, vfnmsub_vv_w, OP_UUU_W, H4, H4, H4, fnmsub32)
>   RVVCALL(OPFVV3, vfnmsub_vv_d, OP_UUU_D, H8, H8, H8, fnmsub64)
> -GEN_VEXT_VV_ENV(vfnmsub_vv_h)
> -GEN_VEXT_VV_ENV(vfnmsub_vv_w)
> -GEN_VEXT_VV_ENV(vfnmsub_vv_d)
> +GEN_VEXT_VV_ENV(vfnmsub_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmsub_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmsub_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmsub_vf_h, OP_UUU_H, H2, H2, fnmsub16)
>   RVVCALL(OPFVF3, vfnmsub_vf_w, OP_UUU_W, H4, H4, fnmsub32)
>   RVVCALL(OPFVF3, vfnmsub_vf_d, OP_UUU_D, H8, H8, fnmsub64)
> -GEN_VEXT_VF(vfnmsub_vf_h)
> -GEN_VEXT_VF(vfnmsub_vf_w)
> -GEN_VEXT_VF(vfnmsub_vf_d)
> +GEN_VEXT_VF(vfnmsub_vf_h, 2)
> +GEN_VEXT_VF(vfnmsub_vf_w, 4)
> +GEN_VEXT_VF(vfnmsub_vf_d, 8)
>   
>   /* Vector Widening Floating-Point Fused Multiply-Add Instructions */
>   static uint32_t fwmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
> @@ -3510,12 +3526,12 @@ static uint64_t fwmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwmacc16)
>   RVVCALL(OPFVV3, vfwmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwmacc32)
> -GEN_VEXT_VV_ENV(vfwmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfwmacc_vv_w)
> +GEN_VEXT_VV_ENV(vfwmacc_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwmacc_vv_w, 8)
>   RVVCALL(OPFVF3, vfwmacc_vf_h, WOP_UUU_H, H4, H2, fwmacc16)
>   RVVCALL(OPFVF3, vfwmacc_vf_w, WOP_UUU_W, H8, H4, fwmacc32)
> -GEN_VEXT_VF(vfwmacc_vf_h)
> -GEN_VEXT_VF(vfwmacc_vf_w)
> +GEN_VEXT_VF(vfwmacc_vf_h, 4)
> +GEN_VEXT_VF(vfwmacc_vf_w, 8)
>   
>   static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
>   {
> @@ -3533,12 +3549,12 @@ static uint64_t fwnmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwnmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwnmacc16)
>   RVVCALL(OPFVV3, vfwnmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwnmacc32)
> -GEN_VEXT_VV_ENV(vfwnmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfwnmacc_vv_w)
> +GEN_VEXT_VV_ENV(vfwnmacc_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwnmacc_vv_w, 8)
>   RVVCALL(OPFVF3, vfwnmacc_vf_h, WOP_UUU_H, H4, H2, fwnmacc16)
>   RVVCALL(OPFVF3, vfwnmacc_vf_w, WOP_UUU_W, H8, H4, fwnmacc32)
> -GEN_VEXT_VF(vfwnmacc_vf_h)
> -GEN_VEXT_VF(vfwnmacc_vf_w)
> +GEN_VEXT_VF(vfwnmacc_vf_h, 4)
> +GEN_VEXT_VF(vfwnmacc_vf_w, 8)
>   
>   static uint32_t fwmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
>   {
> @@ -3556,12 +3572,12 @@ static uint64_t fwmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwmsac16)
>   RVVCALL(OPFVV3, vfwmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwmsac32)
> -GEN_VEXT_VV_ENV(vfwmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfwmsac_vv_w)
> +GEN_VEXT_VV_ENV(vfwmsac_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwmsac_vv_w, 8)
>   RVVCALL(OPFVF3, vfwmsac_vf_h, WOP_UUU_H, H4, H2, fwmsac16)
>   RVVCALL(OPFVF3, vfwmsac_vf_w, WOP_UUU_W, H8, H4, fwmsac32)
> -GEN_VEXT_VF(vfwmsac_vf_h)
> -GEN_VEXT_VF(vfwmsac_vf_w)
> +GEN_VEXT_VF(vfwmsac_vf_h, 4)
> +GEN_VEXT_VF(vfwmsac_vf_w, 8)
>   
>   static uint32_t fwnmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
>   {
> @@ -3579,12 +3595,12 @@ static uint64_t fwnmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwnmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwnmsac16)
>   RVVCALL(OPFVV3, vfwnmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwnmsac32)
> -GEN_VEXT_VV_ENV(vfwnmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfwnmsac_vv_w)
> +GEN_VEXT_VV_ENV(vfwnmsac_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwnmsac_vv_w, 8)
>   RVVCALL(OPFVF3, vfwnmsac_vf_h, WOP_UUU_H, H4, H2, fwnmsac16)
>   RVVCALL(OPFVF3, vfwnmsac_vf_w, WOP_UUU_W, H8, H4, fwnmsac32)
> -GEN_VEXT_VF(vfwnmsac_vf_h)
> -GEN_VEXT_VF(vfwnmsac_vf_w)
> +GEN_VEXT_VF(vfwnmsac_vf_h, 4)
> +GEN_VEXT_VF(vfwnmsac_vf_w, 8)
>   
>   /* Vector Floating-Point Square-Root Instruction */
>   /* (TD, T2, TX2) */
> @@ -3600,12 +3616,16 @@ static void do_##NAME(void *vd, void *vs2, int i,      \
>       *((TD *)vd + HD(i)) = OP(s2, &env->fp_status);     \
>   }
>   
> -#define GEN_VEXT_V_ENV(NAME)                           \
> +#define GEN_VEXT_V_ENV(NAME, ESZ)                      \
>   void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
>           CPURISCVState *env, uint32_t desc)             \
>   {                                                      \
>       uint32_t vm = vext_vm(desc);                       \
>       uint32_t vl = env->vl;                             \
> +    uint32_t total_elems =                             \
> +        vext_get_total_elems(env_archcpu(env),         \
> +                            env->vtype);               \
> +    uint32_t vta = vext_vta(desc);                     \
>       uint32_t i;                                        \
>                                                          \
>       if (vl == 0) {                                     \
> @@ -3618,14 +3638,19 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
>           do_##NAME(vd, vs2, i, env);                    \
>       }                                                  \
>       env->vstart = 0;                                   \
> +    /* set tail elements to 1s */                      \
> +    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,      \
> +                                     vl * ESZ,         \
> +                                     total_elems *     \
> +                                     ESZ);             \
>   }
>   
>   RVVCALL(OPFVV1, vfsqrt_v_h, OP_UU_H, H2, H2, float16_sqrt)
>   RVVCALL(OPFVV1, vfsqrt_v_w, OP_UU_W, H4, H4, float32_sqrt)
>   RVVCALL(OPFVV1, vfsqrt_v_d, OP_UU_D, H8, H8, float64_sqrt)
> -GEN_VEXT_V_ENV(vfsqrt_v_h)
> -GEN_VEXT_V_ENV(vfsqrt_v_w)
> -GEN_VEXT_V_ENV(vfsqrt_v_d)
> +GEN_VEXT_V_ENV(vfsqrt_v_h, 2)
> +GEN_VEXT_V_ENV(vfsqrt_v_w, 4)
> +GEN_VEXT_V_ENV(vfsqrt_v_d, 8)
>   
>   /*
>    * Vector Floating-Point Reciprocal Square-Root Estimate Instruction
> @@ -3805,9 +3830,9 @@ static float64 frsqrt7_d(float64 f, float_status *s)
>   RVVCALL(OPFVV1, vfrsqrt7_v_h, OP_UU_H, H2, H2, frsqrt7_h)
>   RVVCALL(OPFVV1, vfrsqrt7_v_w, OP_UU_W, H4, H4, frsqrt7_s)
>   RVVCALL(OPFVV1, vfrsqrt7_v_d, OP_UU_D, H8, H8, frsqrt7_d)
> -GEN_VEXT_V_ENV(vfrsqrt7_v_h)
> -GEN_VEXT_V_ENV(vfrsqrt7_v_w)
> -GEN_VEXT_V_ENV(vfrsqrt7_v_d)
> +GEN_VEXT_V_ENV(vfrsqrt7_v_h, 2)
> +GEN_VEXT_V_ENV(vfrsqrt7_v_w, 4)
> +GEN_VEXT_V_ENV(vfrsqrt7_v_d, 8)
>   
>   /*
>    * Vector Floating-Point Reciprocal Estimate Instruction
> @@ -3996,36 +4021,36 @@ static float64 frec7_d(float64 f, float_status *s)
>   RVVCALL(OPFVV1, vfrec7_v_h, OP_UU_H, H2, H2, frec7_h)
>   RVVCALL(OPFVV1, vfrec7_v_w, OP_UU_W, H4, H4, frec7_s)
>   RVVCALL(OPFVV1, vfrec7_v_d, OP_UU_D, H8, H8, frec7_d)
> -GEN_VEXT_V_ENV(vfrec7_v_h)
> -GEN_VEXT_V_ENV(vfrec7_v_w)
> -GEN_VEXT_V_ENV(vfrec7_v_d)
> +GEN_VEXT_V_ENV(vfrec7_v_h, 2)
> +GEN_VEXT_V_ENV(vfrec7_v_w, 4)
> +GEN_VEXT_V_ENV(vfrec7_v_d, 8)
>   
>   /* Vector Floating-Point MIN/MAX Instructions */
>   RVVCALL(OPFVV2, vfmin_vv_h, OP_UUU_H, H2, H2, H2, float16_minimum_number)
>   RVVCALL(OPFVV2, vfmin_vv_w, OP_UUU_W, H4, H4, H4, float32_minimum_number)
>   RVVCALL(OPFVV2, vfmin_vv_d, OP_UUU_D, H8, H8, H8, float64_minimum_number)
> -GEN_VEXT_VV_ENV(vfmin_vv_h)
> -GEN_VEXT_VV_ENV(vfmin_vv_w)
> -GEN_VEXT_VV_ENV(vfmin_vv_d)
> +GEN_VEXT_VV_ENV(vfmin_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmin_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmin_vv_d, 8)
>   RVVCALL(OPFVF2, vfmin_vf_h, OP_UUU_H, H2, H2, float16_minimum_number)
>   RVVCALL(OPFVF2, vfmin_vf_w, OP_UUU_W, H4, H4, float32_minimum_number)
>   RVVCALL(OPFVF2, vfmin_vf_d, OP_UUU_D, H8, H8, float64_minimum_number)
> -GEN_VEXT_VF(vfmin_vf_h)
> -GEN_VEXT_VF(vfmin_vf_w)
> -GEN_VEXT_VF(vfmin_vf_d)
> +GEN_VEXT_VF(vfmin_vf_h, 2)
> +GEN_VEXT_VF(vfmin_vf_w, 4)
> +GEN_VEXT_VF(vfmin_vf_d, 8)
>   
>   RVVCALL(OPFVV2, vfmax_vv_h, OP_UUU_H, H2, H2, H2, float16_maximum_number)
>   RVVCALL(OPFVV2, vfmax_vv_w, OP_UUU_W, H4, H4, H4, float32_maximum_number)
>   RVVCALL(OPFVV2, vfmax_vv_d, OP_UUU_D, H8, H8, H8, float64_maximum_number)
> -GEN_VEXT_VV_ENV(vfmax_vv_h)
> -GEN_VEXT_VV_ENV(vfmax_vv_w)
> -GEN_VEXT_VV_ENV(vfmax_vv_d)
> +GEN_VEXT_VV_ENV(vfmax_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmax_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmax_vv_d, 8)
>   RVVCALL(OPFVF2, vfmax_vf_h, OP_UUU_H, H2, H2, float16_maximum_number)
>   RVVCALL(OPFVF2, vfmax_vf_w, OP_UUU_W, H4, H4, float32_maximum_number)
>   RVVCALL(OPFVF2, vfmax_vf_d, OP_UUU_D, H8, H8, float64_maximum_number)
> -GEN_VEXT_VF(vfmax_vf_h)
> -GEN_VEXT_VF(vfmax_vf_w)
> -GEN_VEXT_VF(vfmax_vf_d)
> +GEN_VEXT_VF(vfmax_vf_h, 2)
> +GEN_VEXT_VF(vfmax_vf_w, 4)
> +GEN_VEXT_VF(vfmax_vf_d, 8)
>   
>   /* Vector Floating-Point Sign-Injection Instructions */
>   static uint16_t fsgnj16(uint16_t a, uint16_t b, float_status *s)
> @@ -4046,15 +4071,15 @@ static uint64_t fsgnj64(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVV2, vfsgnj_vv_h, OP_UUU_H, H2, H2, H2, fsgnj16)
>   RVVCALL(OPFVV2, vfsgnj_vv_w, OP_UUU_W, H4, H4, H4, fsgnj32)
>   RVVCALL(OPFVV2, vfsgnj_vv_d, OP_UUU_D, H8, H8, H8, fsgnj64)
> -GEN_VEXT_VV_ENV(vfsgnj_vv_h)
> -GEN_VEXT_VV_ENV(vfsgnj_vv_w)
> -GEN_VEXT_VV_ENV(vfsgnj_vv_d)
> +GEN_VEXT_VV_ENV(vfsgnj_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsgnj_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsgnj_vv_d, 8)
>   RVVCALL(OPFVF2, vfsgnj_vf_h, OP_UUU_H, H2, H2, fsgnj16)
>   RVVCALL(OPFVF2, vfsgnj_vf_w, OP_UUU_W, H4, H4, fsgnj32)
>   RVVCALL(OPFVF2, vfsgnj_vf_d, OP_UUU_D, H8, H8, fsgnj64)
> -GEN_VEXT_VF(vfsgnj_vf_h)
> -GEN_VEXT_VF(vfsgnj_vf_w)
> -GEN_VEXT_VF(vfsgnj_vf_d)
> +GEN_VEXT_VF(vfsgnj_vf_h, 2)
> +GEN_VEXT_VF(vfsgnj_vf_w, 4)
> +GEN_VEXT_VF(vfsgnj_vf_d, 8)
>   
>   static uint16_t fsgnjn16(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -4074,15 +4099,15 @@ static uint64_t fsgnjn64(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVV2, vfsgnjn_vv_h, OP_UUU_H, H2, H2, H2, fsgnjn16)
>   RVVCALL(OPFVV2, vfsgnjn_vv_w, OP_UUU_W, H4, H4, H4, fsgnjn32)
>   RVVCALL(OPFVV2, vfsgnjn_vv_d, OP_UUU_D, H8, H8, H8, fsgnjn64)
> -GEN_VEXT_VV_ENV(vfsgnjn_vv_h)
> -GEN_VEXT_VV_ENV(vfsgnjn_vv_w)
> -GEN_VEXT_VV_ENV(vfsgnjn_vv_d)
> +GEN_VEXT_VV_ENV(vfsgnjn_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsgnjn_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsgnjn_vv_d, 8)
>   RVVCALL(OPFVF2, vfsgnjn_vf_h, OP_UUU_H, H2, H2, fsgnjn16)
>   RVVCALL(OPFVF2, vfsgnjn_vf_w, OP_UUU_W, H4, H4, fsgnjn32)
>   RVVCALL(OPFVF2, vfsgnjn_vf_d, OP_UUU_D, H8, H8, fsgnjn64)
> -GEN_VEXT_VF(vfsgnjn_vf_h)
> -GEN_VEXT_VF(vfsgnjn_vf_w)
> -GEN_VEXT_VF(vfsgnjn_vf_d)
> +GEN_VEXT_VF(vfsgnjn_vf_h, 2)
> +GEN_VEXT_VF(vfsgnjn_vf_w, 4)
> +GEN_VEXT_VF(vfsgnjn_vf_d, 8)
>   
>   static uint16_t fsgnjx16(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -4102,15 +4127,15 @@ static uint64_t fsgnjx64(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVV2, vfsgnjx_vv_h, OP_UUU_H, H2, H2, H2, fsgnjx16)
>   RVVCALL(OPFVV2, vfsgnjx_vv_w, OP_UUU_W, H4, H4, H4, fsgnjx32)
>   RVVCALL(OPFVV2, vfsgnjx_vv_d, OP_UUU_D, H8, H8, H8, fsgnjx64)
> -GEN_VEXT_VV_ENV(vfsgnjx_vv_h)
> -GEN_VEXT_VV_ENV(vfsgnjx_vv_w)
> -GEN_VEXT_VV_ENV(vfsgnjx_vv_d)
> +GEN_VEXT_VV_ENV(vfsgnjx_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsgnjx_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsgnjx_vv_d, 8)
>   RVVCALL(OPFVF2, vfsgnjx_vf_h, OP_UUU_H, H2, H2, fsgnjx16)
>   RVVCALL(OPFVF2, vfsgnjx_vf_w, OP_UUU_W, H4, H4, fsgnjx32)
>   RVVCALL(OPFVF2, vfsgnjx_vf_d, OP_UUU_D, H8, H8, fsgnjx64)
> -GEN_VEXT_VF(vfsgnjx_vf_h)
> -GEN_VEXT_VF(vfsgnjx_vf_w)
> -GEN_VEXT_VF(vfsgnjx_vf_d)
> +GEN_VEXT_VF(vfsgnjx_vf_h, 2)
> +GEN_VEXT_VF(vfsgnjx_vf_w, 4)
> +GEN_VEXT_VF(vfsgnjx_vf_d, 8)
>   
>   /* Vector Floating-Point Compare Instructions */
>   #define GEN_VEXT_CMP_VV_ENV(NAME, ETYPE, H, DO_OP)            \
> @@ -4119,6 +4144,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
>   {                                                             \
>       uint32_t vm = vext_vm(desc);                              \
>       uint32_t vl = env->vl;                                    \
> +    uint32_t total_elems =                                    \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);   \
> +    uint32_t vta = vext_vta(desc);                            \
>       uint32_t i;                                               \
>                                                                 \
>       for (i = env->vstart; i < vl; i++) {                      \
> @@ -4131,6 +4159,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
>                              DO_OP(s2, s1, &env->fp_status));   \
>       }                                                         \
>       env->vstart = 0;                                          \
> +    /* set tail elements to 1s */                             \
> +    if (vta) {                                                \
> +        for (; i < total_elems; i++) {                        \
> +            vext_set_elem_mask(vd, i, 1);                     \
> +        }                                                     \
> +    }                                                         \
>   }
>   
>   GEN_VEXT_CMP_VV_ENV(vmfeq_vv_h, uint16_t, H2, float16_eq_quiet)
> @@ -4143,6 +4177,9 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
>   {                                                                   \
>       uint32_t vm = vext_vm(desc);                                    \
>       uint32_t vl = env->vl;                                          \
> +    uint32_t total_elems =                                          \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);         \
> +    uint32_t vta = vext_vta(desc);                                  \
>       uint32_t i;                                                     \
>                                                                       \
>       for (i = env->vstart; i < vl; i++) {                            \
> @@ -4154,6 +4191,12 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
>                              DO_OP(s2, (ETYPE)s1, &env->fp_status));  \
>       }                                                               \
>       env->vstart = 0;                                                \
> +    /* set tail elements to 1s */                                   \
> +    if (vta) {                                                      \
> +        for (; i < total_elems; i++) {                              \
> +            vext_set_elem_mask(vd, i, 1);                           \
> +        }                                                           \
> +    }                                                               \
>   }
>   
>   GEN_VEXT_CMP_VF(vmfeq_vf_h, uint16_t, H2, float16_eq_quiet)
> @@ -4254,12 +4297,16 @@ static void do_##NAME(void *vd, void *vs2, int i)      \
>       *((TD *)vd + HD(i)) = OP(s2);                      \
>   }
>   
> -#define GEN_VEXT_V(NAME)                               \
> +#define GEN_VEXT_V(NAME, ESZ)                          \
>   void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
>                     CPURISCVState *env, uint32_t desc)   \
>   {                                                      \
>       uint32_t vm = vext_vm(desc);                       \
>       uint32_t vl = env->vl;                             \
> +    uint32_t total_elems =                             \
> +        vext_get_total_elems(env_archcpu(env),         \
> +                            env->vtype);               \
It's  better to aligned here too.

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 11/14] target/riscv: rvv: Add tail agnostic for vector floating-point instructions
@ 2022-03-25 10:42     ` Weiwei Li
  0 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:42 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt


在 2022/3/7 下午6:05, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc |  11 +
>   target/riscv/vector_helper.c            | 452 ++++++++++++++----------
>   2 files changed, 267 insertions(+), 196 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index 24784b6f5f..5bc340f419 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -2391,6 +2391,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs1),                    \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
> @@ -2473,6 +2474,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)            \
>           gen_set_rm(s, RISCV_FRM_DYN);                             \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);            \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);              \
>           return opfvf_trans(a->rd, a->rs1, a->rs2, data,           \
>                              fns[s->sew - 1], s);                   \
>       }                                                             \
> @@ -2511,6 +2513,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
>                                                                    \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);               \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),   \
>                              vreg_ofs(s, a->rs1),                  \
>                              vreg_ofs(s, a->rs2), cpu_env,         \
> @@ -2550,6 +2553,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
>           gen_set_rm(s, RISCV_FRM_DYN);                            \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);               \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
>           return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
>                              fns[s->sew - 1], s);                  \
>       }                                                            \
> @@ -2586,6 +2590,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs1),                    \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
> @@ -2625,6 +2630,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
>           gen_set_rm(s, RISCV_FRM_DYN);                            \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);               \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);             \
>           return opfvf_trans(a->rd, a->rs1, a->rs2, data,          \
>                              fns[s->sew - 1], s);                  \
>       }                                                            \
> @@ -2708,6 +2714,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
>   
>           data = FIELD_DP32(data, VDATA, VM, a->vm);
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                              vreg_ofs(s, a->rs2), cpu_env,
>                              s->cfg_ptr->vlen / 8,
> @@ -2921,6 +2928,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> @@ -2972,6 +2980,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>           tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> @@ -3039,6 +3048,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> @@ -3092,6 +3102,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
>           tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
>                                                                      \
>           data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
>           tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
>                              vreg_ofs(s, a->rs2), cpu_env,           \
>                              s->cfg_ptr->vlen / 8,                   \
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index 067095e19a..d654377fc8 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -2997,13 +2997,17 @@ static void do_##NAME(void *vd, void *vs1, void *vs2, int i,   \
>       *((TD *)vd + HD(i)) = OP(s2, s1, &env->fp_status);         \
>   }
>   
> -#define GEN_VEXT_VV_ENV(NAME)                             \
> +#define GEN_VEXT_VV_ENV(NAME, ESZ)                        \
>   void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
>                     void *vs2, CPURISCVState *env,          \
>                     uint32_t desc)                          \
>   {                                                         \
>       uint32_t vm = vext_vm(desc);                          \
>       uint32_t vl = env->vl;                                \
> +    uint32_t total_elems =                                \
> +        vext_get_total_elems(env_archcpu(env),            \
> +                             env->vtype);                 \
> +    uint32_t vta = vext_vta(desc);                        \
>       uint32_t i;                                           \
>                                                             \
>       for (i = env->vstart; i < vl; i++) {                  \
> @@ -3013,14 +3017,18 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
>           do_##NAME(vd, vs1, vs2, i, env);                  \
>       }                                                     \
>       env->vstart = 0;                                      \
> +    /* set tail elements to 1s */                         \
> +    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
> +                                     vl * ESZ,            \
> +                                     total_elems * ESZ);  \
>   }
>   
>   RVVCALL(OPFVV2, vfadd_vv_h, OP_UUU_H, H2, H2, H2, float16_add)
>   RVVCALL(OPFVV2, vfadd_vv_w, OP_UUU_W, H4, H4, H4, float32_add)
>   RVVCALL(OPFVV2, vfadd_vv_d, OP_UUU_D, H8, H8, H8, float64_add)
> -GEN_VEXT_VV_ENV(vfadd_vv_h)
> -GEN_VEXT_VV_ENV(vfadd_vv_w)
> -GEN_VEXT_VV_ENV(vfadd_vv_d)
> +GEN_VEXT_VV_ENV(vfadd_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfadd_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfadd_vv_d, 8)
>   
>   #define OPFVF2(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)        \
>   static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
> @@ -3030,13 +3038,17 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
>       *((TD *)vd + HD(i)) = OP(s2, (TX1)(T1)s1, &env->fp_status);\
>   }
>   
> -#define GEN_VEXT_VF(NAME)                                 \
> +#define GEN_VEXT_VF(NAME, ESZ)                            \
>   void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
>                     void *vs2, CPURISCVState *env,          \
>                     uint32_t desc)                          \
>   {                                                         \
>       uint32_t vm = vext_vm(desc);                          \
>       uint32_t vl = env->vl;                                \
> +    uint32_t total_elems =                                \
> +        vext_get_total_elems(env_archcpu(env),            \
> +                             env->vtype);                 \
> +    uint32_t vta = vext_vta(desc);                        \
>       uint32_t i;                                           \
>                                                             \
>       for (i = env->vstart; i < vl; i++) {                  \
> @@ -3046,27 +3058,31 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
>           do_##NAME(vd, s1, vs2, i, env);                   \
>       }                                                     \
>       env->vstart = 0;                                      \
> +    /* set tail elements to 1s */                         \
> +    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,         \
> +                                     vl * ESZ,            \
> +                                     total_elems * ESZ);  \
>   }
>   
>   RVVCALL(OPFVF2, vfadd_vf_h, OP_UUU_H, H2, H2, float16_add)
>   RVVCALL(OPFVF2, vfadd_vf_w, OP_UUU_W, H4, H4, float32_add)
>   RVVCALL(OPFVF2, vfadd_vf_d, OP_UUU_D, H8, H8, float64_add)
> -GEN_VEXT_VF(vfadd_vf_h)
> -GEN_VEXT_VF(vfadd_vf_w)
> -GEN_VEXT_VF(vfadd_vf_d)
> +GEN_VEXT_VF(vfadd_vf_h, 2)
> +GEN_VEXT_VF(vfadd_vf_w, 4)
> +GEN_VEXT_VF(vfadd_vf_d, 8)
>   
>   RVVCALL(OPFVV2, vfsub_vv_h, OP_UUU_H, H2, H2, H2, float16_sub)
>   RVVCALL(OPFVV2, vfsub_vv_w, OP_UUU_W, H4, H4, H4, float32_sub)
>   RVVCALL(OPFVV2, vfsub_vv_d, OP_UUU_D, H8, H8, H8, float64_sub)
> -GEN_VEXT_VV_ENV(vfsub_vv_h)
> -GEN_VEXT_VV_ENV(vfsub_vv_w)
> -GEN_VEXT_VV_ENV(vfsub_vv_d)
> +GEN_VEXT_VV_ENV(vfsub_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsub_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsub_vv_d, 8)
>   RVVCALL(OPFVF2, vfsub_vf_h, OP_UUU_H, H2, H2, float16_sub)
>   RVVCALL(OPFVF2, vfsub_vf_w, OP_UUU_W, H4, H4, float32_sub)
>   RVVCALL(OPFVF2, vfsub_vf_d, OP_UUU_D, H8, H8, float64_sub)
> -GEN_VEXT_VF(vfsub_vf_h)
> -GEN_VEXT_VF(vfsub_vf_w)
> -GEN_VEXT_VF(vfsub_vf_d)
> +GEN_VEXT_VF(vfsub_vf_h, 2)
> +GEN_VEXT_VF(vfsub_vf_w, 4)
> +GEN_VEXT_VF(vfsub_vf_d, 8)
>   
>   static uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -3086,9 +3102,9 @@ static uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVF2, vfrsub_vf_h, OP_UUU_H, H2, H2, float16_rsub)
>   RVVCALL(OPFVF2, vfrsub_vf_w, OP_UUU_W, H4, H4, float32_rsub)
>   RVVCALL(OPFVF2, vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub)
> -GEN_VEXT_VF(vfrsub_vf_h)
> -GEN_VEXT_VF(vfrsub_vf_w)
> -GEN_VEXT_VF(vfrsub_vf_d)
> +GEN_VEXT_VF(vfrsub_vf_h, 2)
> +GEN_VEXT_VF(vfrsub_vf_w, 4)
> +GEN_VEXT_VF(vfrsub_vf_d, 8)
>   
>   /* Vector Widening Floating-Point Add/Subtract Instructions */
>   static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s)
> @@ -3106,12 +3122,12 @@ static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16)
>   RVVCALL(OPFVV2, vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32)
> -GEN_VEXT_VV_ENV(vfwadd_vv_h)
> -GEN_VEXT_VV_ENV(vfwadd_vv_w)
> +GEN_VEXT_VV_ENV(vfwadd_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwadd_vv_w, 8)
>   RVVCALL(OPFVF2, vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16)
>   RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32)
> -GEN_VEXT_VF(vfwadd_vf_h)
> -GEN_VEXT_VF(vfwadd_vf_w)
> +GEN_VEXT_VF(vfwadd_vf_h, 4)
> +GEN_VEXT_VF(vfwadd_vf_w, 8)
>   
>   static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -3128,12 +3144,12 @@ static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16)
>   RVVCALL(OPFVV2, vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32)
> -GEN_VEXT_VV_ENV(vfwsub_vv_h)
> -GEN_VEXT_VV_ENV(vfwsub_vv_w)
> +GEN_VEXT_VV_ENV(vfwsub_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwsub_vv_w, 8)
>   RVVCALL(OPFVF2, vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16)
>   RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32)
> -GEN_VEXT_VF(vfwsub_vf_h)
> -GEN_VEXT_VF(vfwsub_vf_w)
> +GEN_VEXT_VF(vfwsub_vf_h, 4)
> +GEN_VEXT_VF(vfwsub_vf_w, 8)
>   
>   static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s)
>   {
> @@ -3147,12 +3163,12 @@ static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16)
>   RVVCALL(OPFVV2, vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32)
> -GEN_VEXT_VV_ENV(vfwadd_wv_h)
> -GEN_VEXT_VV_ENV(vfwadd_wv_w)
> +GEN_VEXT_VV_ENV(vfwadd_wv_h, 4)
> +GEN_VEXT_VV_ENV(vfwadd_wv_w, 8)
>   RVVCALL(OPFVF2, vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16)
>   RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32)
> -GEN_VEXT_VF(vfwadd_wf_h)
> -GEN_VEXT_VF(vfwadd_wf_w)
> +GEN_VEXT_VF(vfwadd_wf_h, 4)
> +GEN_VEXT_VF(vfwadd_wf_w, 8)
>   
>   static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s)
>   {
> @@ -3166,39 +3182,39 @@ static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s)
>   
>   RVVCALL(OPFVV2, vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16)
>   RVVCALL(OPFVV2, vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32)
> -GEN_VEXT_VV_ENV(vfwsub_wv_h)
> -GEN_VEXT_VV_ENV(vfwsub_wv_w)
> +GEN_VEXT_VV_ENV(vfwsub_wv_h, 4)
> +GEN_VEXT_VV_ENV(vfwsub_wv_w, 8)
>   RVVCALL(OPFVF2, vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16)
>   RVVCALL(OPFVF2, vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32)
> -GEN_VEXT_VF(vfwsub_wf_h)
> -GEN_VEXT_VF(vfwsub_wf_w)
> +GEN_VEXT_VF(vfwsub_wf_h, 4)
> +GEN_VEXT_VF(vfwsub_wf_w, 8)
>   
>   /* Vector Single-Width Floating-Point Multiply/Divide Instructions */
>   RVVCALL(OPFVV2, vfmul_vv_h, OP_UUU_H, H2, H2, H2, float16_mul)
>   RVVCALL(OPFVV2, vfmul_vv_w, OP_UUU_W, H4, H4, H4, float32_mul)
>   RVVCALL(OPFVV2, vfmul_vv_d, OP_UUU_D, H8, H8, H8, float64_mul)
> -GEN_VEXT_VV_ENV(vfmul_vv_h)
> -GEN_VEXT_VV_ENV(vfmul_vv_w)
> -GEN_VEXT_VV_ENV(vfmul_vv_d)
> +GEN_VEXT_VV_ENV(vfmul_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmul_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmul_vv_d, 8)
>   RVVCALL(OPFVF2, vfmul_vf_h, OP_UUU_H, H2, H2, float16_mul)
>   RVVCALL(OPFVF2, vfmul_vf_w, OP_UUU_W, H4, H4, float32_mul)
>   RVVCALL(OPFVF2, vfmul_vf_d, OP_UUU_D, H8, H8, float64_mul)
> -GEN_VEXT_VF(vfmul_vf_h)
> -GEN_VEXT_VF(vfmul_vf_w)
> -GEN_VEXT_VF(vfmul_vf_d)
> +GEN_VEXT_VF(vfmul_vf_h, 2)
> +GEN_VEXT_VF(vfmul_vf_w, 4)
> +GEN_VEXT_VF(vfmul_vf_d, 8)
>   
>   RVVCALL(OPFVV2, vfdiv_vv_h, OP_UUU_H, H2, H2, H2, float16_div)
>   RVVCALL(OPFVV2, vfdiv_vv_w, OP_UUU_W, H4, H4, H4, float32_div)
>   RVVCALL(OPFVV2, vfdiv_vv_d, OP_UUU_D, H8, H8, H8, float64_div)
> -GEN_VEXT_VV_ENV(vfdiv_vv_h)
> -GEN_VEXT_VV_ENV(vfdiv_vv_w)
> -GEN_VEXT_VV_ENV(vfdiv_vv_d)
> +GEN_VEXT_VV_ENV(vfdiv_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfdiv_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfdiv_vv_d, 8)
>   RVVCALL(OPFVF2, vfdiv_vf_h, OP_UUU_H, H2, H2, float16_div)
>   RVVCALL(OPFVF2, vfdiv_vf_w, OP_UUU_W, H4, H4, float32_div)
>   RVVCALL(OPFVF2, vfdiv_vf_d, OP_UUU_D, H8, H8, float64_div)
> -GEN_VEXT_VF(vfdiv_vf_h)
> -GEN_VEXT_VF(vfdiv_vf_w)
> -GEN_VEXT_VF(vfdiv_vf_d)
> +GEN_VEXT_VF(vfdiv_vf_h, 2)
> +GEN_VEXT_VF(vfdiv_vf_w, 4)
> +GEN_VEXT_VF(vfdiv_vf_d, 8)
>   
>   static uint16_t float16_rdiv(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -3218,9 +3234,9 @@ static uint64_t float64_rdiv(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVF2, vfrdiv_vf_h, OP_UUU_H, H2, H2, float16_rdiv)
>   RVVCALL(OPFVF2, vfrdiv_vf_w, OP_UUU_W, H4, H4, float32_rdiv)
>   RVVCALL(OPFVF2, vfrdiv_vf_d, OP_UUU_D, H8, H8, float64_rdiv)
> -GEN_VEXT_VF(vfrdiv_vf_h)
> -GEN_VEXT_VF(vfrdiv_vf_w)
> -GEN_VEXT_VF(vfrdiv_vf_d)
> +GEN_VEXT_VF(vfrdiv_vf_h, 2)
> +GEN_VEXT_VF(vfrdiv_vf_w, 4)
> +GEN_VEXT_VF(vfrdiv_vf_d, 8)
>   
>   /* Vector Widening Floating-Point Multiply */
>   static uint32_t vfwmul16(uint16_t a, uint16_t b, float_status *s)
> @@ -3237,12 +3253,12 @@ static uint64_t vfwmul32(uint32_t a, uint32_t b, float_status *s)
>   }
>   RVVCALL(OPFVV2, vfwmul_vv_h, WOP_UUU_H, H4, H2, H2, vfwmul16)
>   RVVCALL(OPFVV2, vfwmul_vv_w, WOP_UUU_W, H8, H4, H4, vfwmul32)
> -GEN_VEXT_VV_ENV(vfwmul_vv_h)
> -GEN_VEXT_VV_ENV(vfwmul_vv_w)
> +GEN_VEXT_VV_ENV(vfwmul_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwmul_vv_w, 8)
>   RVVCALL(OPFVF2, vfwmul_vf_h, WOP_UUU_H, H4, H2, vfwmul16)
>   RVVCALL(OPFVF2, vfwmul_vf_w, WOP_UUU_W, H8, H4, vfwmul32)
> -GEN_VEXT_VF(vfwmul_vf_h)
> -GEN_VEXT_VF(vfwmul_vf_w)
> +GEN_VEXT_VF(vfwmul_vf_h, 4)
> +GEN_VEXT_VF(vfwmul_vf_w, 8)
>   
>   /* Vector Single-Width Floating-Point Fused Multiply-Add Instructions */
>   #define OPFVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)       \
> @@ -3273,9 +3289,9 @@ static uint64_t fmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmacc_vv_h, OP_UUU_H, H2, H2, H2, fmacc16)
>   RVVCALL(OPFVV3, vfmacc_vv_w, OP_UUU_W, H4, H4, H4, fmacc32)
>   RVVCALL(OPFVV3, vfmacc_vv_d, OP_UUU_D, H8, H8, H8, fmacc64)
> -GEN_VEXT_VV_ENV(vfmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfmacc_vv_w)
> -GEN_VEXT_VV_ENV(vfmacc_vv_d)
> +GEN_VEXT_VV_ENV(vfmacc_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmacc_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmacc_vv_d, 8)
>   
>   #define OPFVF3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)           \
>   static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
> @@ -3289,9 +3305,9 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
>   RVVCALL(OPFVF3, vfmacc_vf_h, OP_UUU_H, H2, H2, fmacc16)
>   RVVCALL(OPFVF3, vfmacc_vf_w, OP_UUU_W, H4, H4, fmacc32)
>   RVVCALL(OPFVF3, vfmacc_vf_d, OP_UUU_D, H8, H8, fmacc64)
> -GEN_VEXT_VF(vfmacc_vf_h)
> -GEN_VEXT_VF(vfmacc_vf_w)
> -GEN_VEXT_VF(vfmacc_vf_d)
> +GEN_VEXT_VF(vfmacc_vf_h, 2)
> +GEN_VEXT_VF(vfmacc_vf_w, 4)
> +GEN_VEXT_VF(vfmacc_vf_d, 8)
>   
>   static uint16_t fnmacc16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3314,15 +3330,15 @@ static uint64_t fnmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmacc_vv_h, OP_UUU_H, H2, H2, H2, fnmacc16)
>   RVVCALL(OPFVV3, vfnmacc_vv_w, OP_UUU_W, H4, H4, H4, fnmacc32)
>   RVVCALL(OPFVV3, vfnmacc_vv_d, OP_UUU_D, H8, H8, H8, fnmacc64)
> -GEN_VEXT_VV_ENV(vfnmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfnmacc_vv_w)
> -GEN_VEXT_VV_ENV(vfnmacc_vv_d)
> +GEN_VEXT_VV_ENV(vfnmacc_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmacc_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmacc_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmacc_vf_h, OP_UUU_H, H2, H2, fnmacc16)
>   RVVCALL(OPFVF3, vfnmacc_vf_w, OP_UUU_W, H4, H4, fnmacc32)
>   RVVCALL(OPFVF3, vfnmacc_vf_d, OP_UUU_D, H8, H8, fnmacc64)
> -GEN_VEXT_VF(vfnmacc_vf_h)
> -GEN_VEXT_VF(vfnmacc_vf_w)
> -GEN_VEXT_VF(vfnmacc_vf_d)
> +GEN_VEXT_VF(vfnmacc_vf_h, 2)
> +GEN_VEXT_VF(vfnmacc_vf_w, 4)
> +GEN_VEXT_VF(vfnmacc_vf_d, 8)
>   
>   static uint16_t fmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3342,15 +3358,15 @@ static uint64_t fmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmsac_vv_h, OP_UUU_H, H2, H2, H2, fmsac16)
>   RVVCALL(OPFVV3, vfmsac_vv_w, OP_UUU_W, H4, H4, H4, fmsac32)
>   RVVCALL(OPFVV3, vfmsac_vv_d, OP_UUU_D, H8, H8, H8, fmsac64)
> -GEN_VEXT_VV_ENV(vfmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfmsac_vv_w)
> -GEN_VEXT_VV_ENV(vfmsac_vv_d)
> +GEN_VEXT_VV_ENV(vfmsac_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmsac_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmsac_vv_d, 8)
>   RVVCALL(OPFVF3, vfmsac_vf_h, OP_UUU_H, H2, H2, fmsac16)
>   RVVCALL(OPFVF3, vfmsac_vf_w, OP_UUU_W, H4, H4, fmsac32)
>   RVVCALL(OPFVF3, vfmsac_vf_d, OP_UUU_D, H8, H8, fmsac64)
> -GEN_VEXT_VF(vfmsac_vf_h)
> -GEN_VEXT_VF(vfmsac_vf_w)
> -GEN_VEXT_VF(vfmsac_vf_d)
> +GEN_VEXT_VF(vfmsac_vf_h, 2)
> +GEN_VEXT_VF(vfmsac_vf_w, 4)
> +GEN_VEXT_VF(vfmsac_vf_d, 8)
>   
>   static uint16_t fnmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3370,15 +3386,15 @@ static uint64_t fnmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmsac_vv_h, OP_UUU_H, H2, H2, H2, fnmsac16)
>   RVVCALL(OPFVV3, vfnmsac_vv_w, OP_UUU_W, H4, H4, H4, fnmsac32)
>   RVVCALL(OPFVV3, vfnmsac_vv_d, OP_UUU_D, H8, H8, H8, fnmsac64)
> -GEN_VEXT_VV_ENV(vfnmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfnmsac_vv_w)
> -GEN_VEXT_VV_ENV(vfnmsac_vv_d)
> +GEN_VEXT_VV_ENV(vfnmsac_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmsac_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmsac_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmsac_vf_h, OP_UUU_H, H2, H2, fnmsac16)
>   RVVCALL(OPFVF3, vfnmsac_vf_w, OP_UUU_W, H4, H4, fnmsac32)
>   RVVCALL(OPFVF3, vfnmsac_vf_d, OP_UUU_D, H8, H8, fnmsac64)
> -GEN_VEXT_VF(vfnmsac_vf_h)
> -GEN_VEXT_VF(vfnmsac_vf_w)
> -GEN_VEXT_VF(vfnmsac_vf_d)
> +GEN_VEXT_VF(vfnmsac_vf_h, 2)
> +GEN_VEXT_VF(vfnmsac_vf_w, 4)
> +GEN_VEXT_VF(vfnmsac_vf_d, 8)
>   
>   static uint16_t fmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3398,15 +3414,15 @@ static uint64_t fmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmadd_vv_h, OP_UUU_H, H2, H2, H2, fmadd16)
>   RVVCALL(OPFVV3, vfmadd_vv_w, OP_UUU_W, H4, H4, H4, fmadd32)
>   RVVCALL(OPFVV3, vfmadd_vv_d, OP_UUU_D, H8, H8, H8, fmadd64)
> -GEN_VEXT_VV_ENV(vfmadd_vv_h)
> -GEN_VEXT_VV_ENV(vfmadd_vv_w)
> -GEN_VEXT_VV_ENV(vfmadd_vv_d)
> +GEN_VEXT_VV_ENV(vfmadd_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmadd_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmadd_vv_d, 8)
>   RVVCALL(OPFVF3, vfmadd_vf_h, OP_UUU_H, H2, H2, fmadd16)
>   RVVCALL(OPFVF3, vfmadd_vf_w, OP_UUU_W, H4, H4, fmadd32)
>   RVVCALL(OPFVF3, vfmadd_vf_d, OP_UUU_D, H8, H8, fmadd64)
> -GEN_VEXT_VF(vfmadd_vf_h)
> -GEN_VEXT_VF(vfmadd_vf_w)
> -GEN_VEXT_VF(vfmadd_vf_d)
> +GEN_VEXT_VF(vfmadd_vf_h, 2)
> +GEN_VEXT_VF(vfmadd_vf_w, 4)
> +GEN_VEXT_VF(vfmadd_vf_d, 8)
>   
>   static uint16_t fnmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3429,15 +3445,15 @@ static uint64_t fnmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmadd_vv_h, OP_UUU_H, H2, H2, H2, fnmadd16)
>   RVVCALL(OPFVV3, vfnmadd_vv_w, OP_UUU_W, H4, H4, H4, fnmadd32)
>   RVVCALL(OPFVV3, vfnmadd_vv_d, OP_UUU_D, H8, H8, H8, fnmadd64)
> -GEN_VEXT_VV_ENV(vfnmadd_vv_h)
> -GEN_VEXT_VV_ENV(vfnmadd_vv_w)
> -GEN_VEXT_VV_ENV(vfnmadd_vv_d)
> +GEN_VEXT_VV_ENV(vfnmadd_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmadd_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmadd_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmadd_vf_h, OP_UUU_H, H2, H2, fnmadd16)
>   RVVCALL(OPFVF3, vfnmadd_vf_w, OP_UUU_W, H4, H4, fnmadd32)
>   RVVCALL(OPFVF3, vfnmadd_vf_d, OP_UUU_D, H8, H8, fnmadd64)
> -GEN_VEXT_VF(vfnmadd_vf_h)
> -GEN_VEXT_VF(vfnmadd_vf_w)
> -GEN_VEXT_VF(vfnmadd_vf_d)
> +GEN_VEXT_VF(vfnmadd_vf_h, 2)
> +GEN_VEXT_VF(vfnmadd_vf_w, 4)
> +GEN_VEXT_VF(vfnmadd_vf_d, 8)
>   
>   static uint16_t fmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3457,15 +3473,15 @@ static uint64_t fmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfmsub_vv_h, OP_UUU_H, H2, H2, H2, fmsub16)
>   RVVCALL(OPFVV3, vfmsub_vv_w, OP_UUU_W, H4, H4, H4, fmsub32)
>   RVVCALL(OPFVV3, vfmsub_vv_d, OP_UUU_D, H8, H8, H8, fmsub64)
> -GEN_VEXT_VV_ENV(vfmsub_vv_h)
> -GEN_VEXT_VV_ENV(vfmsub_vv_w)
> -GEN_VEXT_VV_ENV(vfmsub_vv_d)
> +GEN_VEXT_VV_ENV(vfmsub_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmsub_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmsub_vv_d, 8)
>   RVVCALL(OPFVF3, vfmsub_vf_h, OP_UUU_H, H2, H2, fmsub16)
>   RVVCALL(OPFVF3, vfmsub_vf_w, OP_UUU_W, H4, H4, fmsub32)
>   RVVCALL(OPFVF3, vfmsub_vf_d, OP_UUU_D, H8, H8, fmsub64)
> -GEN_VEXT_VF(vfmsub_vf_h)
> -GEN_VEXT_VF(vfmsub_vf_w)
> -GEN_VEXT_VF(vfmsub_vf_d)
> +GEN_VEXT_VF(vfmsub_vf_h, 2)
> +GEN_VEXT_VF(vfmsub_vf_w, 4)
> +GEN_VEXT_VF(vfmsub_vf_d, 8)
>   
>   static uint16_t fnmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
>   {
> @@ -3485,15 +3501,15 @@ static uint64_t fnmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
>   RVVCALL(OPFVV3, vfnmsub_vv_h, OP_UUU_H, H2, H2, H2, fnmsub16)
>   RVVCALL(OPFVV3, vfnmsub_vv_w, OP_UUU_W, H4, H4, H4, fnmsub32)
>   RVVCALL(OPFVV3, vfnmsub_vv_d, OP_UUU_D, H8, H8, H8, fnmsub64)
> -GEN_VEXT_VV_ENV(vfnmsub_vv_h)
> -GEN_VEXT_VV_ENV(vfnmsub_vv_w)
> -GEN_VEXT_VV_ENV(vfnmsub_vv_d)
> +GEN_VEXT_VV_ENV(vfnmsub_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfnmsub_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfnmsub_vv_d, 8)
>   RVVCALL(OPFVF3, vfnmsub_vf_h, OP_UUU_H, H2, H2, fnmsub16)
>   RVVCALL(OPFVF3, vfnmsub_vf_w, OP_UUU_W, H4, H4, fnmsub32)
>   RVVCALL(OPFVF3, vfnmsub_vf_d, OP_UUU_D, H8, H8, fnmsub64)
> -GEN_VEXT_VF(vfnmsub_vf_h)
> -GEN_VEXT_VF(vfnmsub_vf_w)
> -GEN_VEXT_VF(vfnmsub_vf_d)
> +GEN_VEXT_VF(vfnmsub_vf_h, 2)
> +GEN_VEXT_VF(vfnmsub_vf_w, 4)
> +GEN_VEXT_VF(vfnmsub_vf_d, 8)
>   
>   /* Vector Widening Floating-Point Fused Multiply-Add Instructions */
>   static uint32_t fwmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
> @@ -3510,12 +3526,12 @@ static uint64_t fwmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwmacc16)
>   RVVCALL(OPFVV3, vfwmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwmacc32)
> -GEN_VEXT_VV_ENV(vfwmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfwmacc_vv_w)
> +GEN_VEXT_VV_ENV(vfwmacc_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwmacc_vv_w, 8)
>   RVVCALL(OPFVF3, vfwmacc_vf_h, WOP_UUU_H, H4, H2, fwmacc16)
>   RVVCALL(OPFVF3, vfwmacc_vf_w, WOP_UUU_W, H8, H4, fwmacc32)
> -GEN_VEXT_VF(vfwmacc_vf_h)
> -GEN_VEXT_VF(vfwmacc_vf_w)
> +GEN_VEXT_VF(vfwmacc_vf_h, 4)
> +GEN_VEXT_VF(vfwmacc_vf_w, 8)
>   
>   static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
>   {
> @@ -3533,12 +3549,12 @@ static uint64_t fwnmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwnmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwnmacc16)
>   RVVCALL(OPFVV3, vfwnmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwnmacc32)
> -GEN_VEXT_VV_ENV(vfwnmacc_vv_h)
> -GEN_VEXT_VV_ENV(vfwnmacc_vv_w)
> +GEN_VEXT_VV_ENV(vfwnmacc_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwnmacc_vv_w, 8)
>   RVVCALL(OPFVF3, vfwnmacc_vf_h, WOP_UUU_H, H4, H2, fwnmacc16)
>   RVVCALL(OPFVF3, vfwnmacc_vf_w, WOP_UUU_W, H8, H4, fwnmacc32)
> -GEN_VEXT_VF(vfwnmacc_vf_h)
> -GEN_VEXT_VF(vfwnmacc_vf_w)
> +GEN_VEXT_VF(vfwnmacc_vf_h, 4)
> +GEN_VEXT_VF(vfwnmacc_vf_w, 8)
>   
>   static uint32_t fwmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
>   {
> @@ -3556,12 +3572,12 @@ static uint64_t fwmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwmsac16)
>   RVVCALL(OPFVV3, vfwmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwmsac32)
> -GEN_VEXT_VV_ENV(vfwmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfwmsac_vv_w)
> +GEN_VEXT_VV_ENV(vfwmsac_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwmsac_vv_w, 8)
>   RVVCALL(OPFVF3, vfwmsac_vf_h, WOP_UUU_H, H4, H2, fwmsac16)
>   RVVCALL(OPFVF3, vfwmsac_vf_w, WOP_UUU_W, H8, H4, fwmsac32)
> -GEN_VEXT_VF(vfwmsac_vf_h)
> -GEN_VEXT_VF(vfwmsac_vf_w)
> +GEN_VEXT_VF(vfwmsac_vf_h, 4)
> +GEN_VEXT_VF(vfwmsac_vf_w, 8)
>   
>   static uint32_t fwnmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
>   {
> @@ -3579,12 +3595,12 @@ static uint64_t fwnmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
>   
>   RVVCALL(OPFVV3, vfwnmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwnmsac16)
>   RVVCALL(OPFVV3, vfwnmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwnmsac32)
> -GEN_VEXT_VV_ENV(vfwnmsac_vv_h)
> -GEN_VEXT_VV_ENV(vfwnmsac_vv_w)
> +GEN_VEXT_VV_ENV(vfwnmsac_vv_h, 4)
> +GEN_VEXT_VV_ENV(vfwnmsac_vv_w, 8)
>   RVVCALL(OPFVF3, vfwnmsac_vf_h, WOP_UUU_H, H4, H2, fwnmsac16)
>   RVVCALL(OPFVF3, vfwnmsac_vf_w, WOP_UUU_W, H8, H4, fwnmsac32)
> -GEN_VEXT_VF(vfwnmsac_vf_h)
> -GEN_VEXT_VF(vfwnmsac_vf_w)
> +GEN_VEXT_VF(vfwnmsac_vf_h, 4)
> +GEN_VEXT_VF(vfwnmsac_vf_w, 8)
>   
>   /* Vector Floating-Point Square-Root Instruction */
>   /* (TD, T2, TX2) */
> @@ -3600,12 +3616,16 @@ static void do_##NAME(void *vd, void *vs2, int i,      \
>       *((TD *)vd + HD(i)) = OP(s2, &env->fp_status);     \
>   }
>   
> -#define GEN_VEXT_V_ENV(NAME)                           \
> +#define GEN_VEXT_V_ENV(NAME, ESZ)                      \
>   void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
>           CPURISCVState *env, uint32_t desc)             \
>   {                                                      \
>       uint32_t vm = vext_vm(desc);                       \
>       uint32_t vl = env->vl;                             \
> +    uint32_t total_elems =                             \
> +        vext_get_total_elems(env_archcpu(env),         \
> +                            env->vtype);               \
> +    uint32_t vta = vext_vta(desc);                     \
>       uint32_t i;                                        \
>                                                          \
>       if (vl == 0) {                                     \
> @@ -3618,14 +3638,19 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
>           do_##NAME(vd, vs2, i, env);                    \
>       }                                                  \
>       env->vstart = 0;                                   \
> +    /* set tail elements to 1s */                      \
> +    vext_set_elems_1s_fns[ctzl(ESZ)](vd, vta, vl,      \
> +                                     vl * ESZ,         \
> +                                     total_elems *     \
> +                                     ESZ);             \
>   }
>   
>   RVVCALL(OPFVV1, vfsqrt_v_h, OP_UU_H, H2, H2, float16_sqrt)
>   RVVCALL(OPFVV1, vfsqrt_v_w, OP_UU_W, H4, H4, float32_sqrt)
>   RVVCALL(OPFVV1, vfsqrt_v_d, OP_UU_D, H8, H8, float64_sqrt)
> -GEN_VEXT_V_ENV(vfsqrt_v_h)
> -GEN_VEXT_V_ENV(vfsqrt_v_w)
> -GEN_VEXT_V_ENV(vfsqrt_v_d)
> +GEN_VEXT_V_ENV(vfsqrt_v_h, 2)
> +GEN_VEXT_V_ENV(vfsqrt_v_w, 4)
> +GEN_VEXT_V_ENV(vfsqrt_v_d, 8)
>   
>   /*
>    * Vector Floating-Point Reciprocal Square-Root Estimate Instruction
> @@ -3805,9 +3830,9 @@ static float64 frsqrt7_d(float64 f, float_status *s)
>   RVVCALL(OPFVV1, vfrsqrt7_v_h, OP_UU_H, H2, H2, frsqrt7_h)
>   RVVCALL(OPFVV1, vfrsqrt7_v_w, OP_UU_W, H4, H4, frsqrt7_s)
>   RVVCALL(OPFVV1, vfrsqrt7_v_d, OP_UU_D, H8, H8, frsqrt7_d)
> -GEN_VEXT_V_ENV(vfrsqrt7_v_h)
> -GEN_VEXT_V_ENV(vfrsqrt7_v_w)
> -GEN_VEXT_V_ENV(vfrsqrt7_v_d)
> +GEN_VEXT_V_ENV(vfrsqrt7_v_h, 2)
> +GEN_VEXT_V_ENV(vfrsqrt7_v_w, 4)
> +GEN_VEXT_V_ENV(vfrsqrt7_v_d, 8)
>   
>   /*
>    * Vector Floating-Point Reciprocal Estimate Instruction
> @@ -3996,36 +4021,36 @@ static float64 frec7_d(float64 f, float_status *s)
>   RVVCALL(OPFVV1, vfrec7_v_h, OP_UU_H, H2, H2, frec7_h)
>   RVVCALL(OPFVV1, vfrec7_v_w, OP_UU_W, H4, H4, frec7_s)
>   RVVCALL(OPFVV1, vfrec7_v_d, OP_UU_D, H8, H8, frec7_d)
> -GEN_VEXT_V_ENV(vfrec7_v_h)
> -GEN_VEXT_V_ENV(vfrec7_v_w)
> -GEN_VEXT_V_ENV(vfrec7_v_d)
> +GEN_VEXT_V_ENV(vfrec7_v_h, 2)
> +GEN_VEXT_V_ENV(vfrec7_v_w, 4)
> +GEN_VEXT_V_ENV(vfrec7_v_d, 8)
>   
>   /* Vector Floating-Point MIN/MAX Instructions */
>   RVVCALL(OPFVV2, vfmin_vv_h, OP_UUU_H, H2, H2, H2, float16_minimum_number)
>   RVVCALL(OPFVV2, vfmin_vv_w, OP_UUU_W, H4, H4, H4, float32_minimum_number)
>   RVVCALL(OPFVV2, vfmin_vv_d, OP_UUU_D, H8, H8, H8, float64_minimum_number)
> -GEN_VEXT_VV_ENV(vfmin_vv_h)
> -GEN_VEXT_VV_ENV(vfmin_vv_w)
> -GEN_VEXT_VV_ENV(vfmin_vv_d)
> +GEN_VEXT_VV_ENV(vfmin_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmin_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmin_vv_d, 8)
>   RVVCALL(OPFVF2, vfmin_vf_h, OP_UUU_H, H2, H2, float16_minimum_number)
>   RVVCALL(OPFVF2, vfmin_vf_w, OP_UUU_W, H4, H4, float32_minimum_number)
>   RVVCALL(OPFVF2, vfmin_vf_d, OP_UUU_D, H8, H8, float64_minimum_number)
> -GEN_VEXT_VF(vfmin_vf_h)
> -GEN_VEXT_VF(vfmin_vf_w)
> -GEN_VEXT_VF(vfmin_vf_d)
> +GEN_VEXT_VF(vfmin_vf_h, 2)
> +GEN_VEXT_VF(vfmin_vf_w, 4)
> +GEN_VEXT_VF(vfmin_vf_d, 8)
>   
>   RVVCALL(OPFVV2, vfmax_vv_h, OP_UUU_H, H2, H2, H2, float16_maximum_number)
>   RVVCALL(OPFVV2, vfmax_vv_w, OP_UUU_W, H4, H4, H4, float32_maximum_number)
>   RVVCALL(OPFVV2, vfmax_vv_d, OP_UUU_D, H8, H8, H8, float64_maximum_number)
> -GEN_VEXT_VV_ENV(vfmax_vv_h)
> -GEN_VEXT_VV_ENV(vfmax_vv_w)
> -GEN_VEXT_VV_ENV(vfmax_vv_d)
> +GEN_VEXT_VV_ENV(vfmax_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfmax_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfmax_vv_d, 8)
>   RVVCALL(OPFVF2, vfmax_vf_h, OP_UUU_H, H2, H2, float16_maximum_number)
>   RVVCALL(OPFVF2, vfmax_vf_w, OP_UUU_W, H4, H4, float32_maximum_number)
>   RVVCALL(OPFVF2, vfmax_vf_d, OP_UUU_D, H8, H8, float64_maximum_number)
> -GEN_VEXT_VF(vfmax_vf_h)
> -GEN_VEXT_VF(vfmax_vf_w)
> -GEN_VEXT_VF(vfmax_vf_d)
> +GEN_VEXT_VF(vfmax_vf_h, 2)
> +GEN_VEXT_VF(vfmax_vf_w, 4)
> +GEN_VEXT_VF(vfmax_vf_d, 8)
>   
>   /* Vector Floating-Point Sign-Injection Instructions */
>   static uint16_t fsgnj16(uint16_t a, uint16_t b, float_status *s)
> @@ -4046,15 +4071,15 @@ static uint64_t fsgnj64(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVV2, vfsgnj_vv_h, OP_UUU_H, H2, H2, H2, fsgnj16)
>   RVVCALL(OPFVV2, vfsgnj_vv_w, OP_UUU_W, H4, H4, H4, fsgnj32)
>   RVVCALL(OPFVV2, vfsgnj_vv_d, OP_UUU_D, H8, H8, H8, fsgnj64)
> -GEN_VEXT_VV_ENV(vfsgnj_vv_h)
> -GEN_VEXT_VV_ENV(vfsgnj_vv_w)
> -GEN_VEXT_VV_ENV(vfsgnj_vv_d)
> +GEN_VEXT_VV_ENV(vfsgnj_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsgnj_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsgnj_vv_d, 8)
>   RVVCALL(OPFVF2, vfsgnj_vf_h, OP_UUU_H, H2, H2, fsgnj16)
>   RVVCALL(OPFVF2, vfsgnj_vf_w, OP_UUU_W, H4, H4, fsgnj32)
>   RVVCALL(OPFVF2, vfsgnj_vf_d, OP_UUU_D, H8, H8, fsgnj64)
> -GEN_VEXT_VF(vfsgnj_vf_h)
> -GEN_VEXT_VF(vfsgnj_vf_w)
> -GEN_VEXT_VF(vfsgnj_vf_d)
> +GEN_VEXT_VF(vfsgnj_vf_h, 2)
> +GEN_VEXT_VF(vfsgnj_vf_w, 4)
> +GEN_VEXT_VF(vfsgnj_vf_d, 8)
>   
>   static uint16_t fsgnjn16(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -4074,15 +4099,15 @@ static uint64_t fsgnjn64(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVV2, vfsgnjn_vv_h, OP_UUU_H, H2, H2, H2, fsgnjn16)
>   RVVCALL(OPFVV2, vfsgnjn_vv_w, OP_UUU_W, H4, H4, H4, fsgnjn32)
>   RVVCALL(OPFVV2, vfsgnjn_vv_d, OP_UUU_D, H8, H8, H8, fsgnjn64)
> -GEN_VEXT_VV_ENV(vfsgnjn_vv_h)
> -GEN_VEXT_VV_ENV(vfsgnjn_vv_w)
> -GEN_VEXT_VV_ENV(vfsgnjn_vv_d)
> +GEN_VEXT_VV_ENV(vfsgnjn_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsgnjn_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsgnjn_vv_d, 8)
>   RVVCALL(OPFVF2, vfsgnjn_vf_h, OP_UUU_H, H2, H2, fsgnjn16)
>   RVVCALL(OPFVF2, vfsgnjn_vf_w, OP_UUU_W, H4, H4, fsgnjn32)
>   RVVCALL(OPFVF2, vfsgnjn_vf_d, OP_UUU_D, H8, H8, fsgnjn64)
> -GEN_VEXT_VF(vfsgnjn_vf_h)
> -GEN_VEXT_VF(vfsgnjn_vf_w)
> -GEN_VEXT_VF(vfsgnjn_vf_d)
> +GEN_VEXT_VF(vfsgnjn_vf_h, 2)
> +GEN_VEXT_VF(vfsgnjn_vf_w, 4)
> +GEN_VEXT_VF(vfsgnjn_vf_d, 8)
>   
>   static uint16_t fsgnjx16(uint16_t a, uint16_t b, float_status *s)
>   {
> @@ -4102,15 +4127,15 @@ static uint64_t fsgnjx64(uint64_t a, uint64_t b, float_status *s)
>   RVVCALL(OPFVV2, vfsgnjx_vv_h, OP_UUU_H, H2, H2, H2, fsgnjx16)
>   RVVCALL(OPFVV2, vfsgnjx_vv_w, OP_UUU_W, H4, H4, H4, fsgnjx32)
>   RVVCALL(OPFVV2, vfsgnjx_vv_d, OP_UUU_D, H8, H8, H8, fsgnjx64)
> -GEN_VEXT_VV_ENV(vfsgnjx_vv_h)
> -GEN_VEXT_VV_ENV(vfsgnjx_vv_w)
> -GEN_VEXT_VV_ENV(vfsgnjx_vv_d)
> +GEN_VEXT_VV_ENV(vfsgnjx_vv_h, 2)
> +GEN_VEXT_VV_ENV(vfsgnjx_vv_w, 4)
> +GEN_VEXT_VV_ENV(vfsgnjx_vv_d, 8)
>   RVVCALL(OPFVF2, vfsgnjx_vf_h, OP_UUU_H, H2, H2, fsgnjx16)
>   RVVCALL(OPFVF2, vfsgnjx_vf_w, OP_UUU_W, H4, H4, fsgnjx32)
>   RVVCALL(OPFVF2, vfsgnjx_vf_d, OP_UUU_D, H8, H8, fsgnjx64)
> -GEN_VEXT_VF(vfsgnjx_vf_h)
> -GEN_VEXT_VF(vfsgnjx_vf_w)
> -GEN_VEXT_VF(vfsgnjx_vf_d)
> +GEN_VEXT_VF(vfsgnjx_vf_h, 2)
> +GEN_VEXT_VF(vfsgnjx_vf_w, 4)
> +GEN_VEXT_VF(vfsgnjx_vf_d, 8)
>   
>   /* Vector Floating-Point Compare Instructions */
>   #define GEN_VEXT_CMP_VV_ENV(NAME, ETYPE, H, DO_OP)            \
> @@ -4119,6 +4144,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
>   {                                                             \
>       uint32_t vm = vext_vm(desc);                              \
>       uint32_t vl = env->vl;                                    \
> +    uint32_t total_elems =                                    \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);   \
> +    uint32_t vta = vext_vta(desc);                            \
>       uint32_t i;                                               \
>                                                                 \
>       for (i = env->vstart; i < vl; i++) {                      \
> @@ -4131,6 +4159,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,   \
>                              DO_OP(s2, s1, &env->fp_status));   \
>       }                                                         \
>       env->vstart = 0;                                          \
> +    /* set tail elements to 1s */                             \
> +    if (vta) {                                                \
> +        for (; i < total_elems; i++) {                        \
> +            vext_set_elem_mask(vd, i, 1);                     \
> +        }                                                     \
> +    }                                                         \
>   }
>   
>   GEN_VEXT_CMP_VV_ENV(vmfeq_vv_h, uint16_t, H2, float16_eq_quiet)
> @@ -4143,6 +4177,9 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
>   {                                                                   \
>       uint32_t vm = vext_vm(desc);                                    \
>       uint32_t vl = env->vl;                                          \
> +    uint32_t total_elems =                                          \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);         \
> +    uint32_t vta = vext_vta(desc);                                  \
>       uint32_t i;                                                     \
>                                                                       \
>       for (i = env->vstart; i < vl; i++) {                            \
> @@ -4154,6 +4191,12 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2,       \
>                              DO_OP(s2, (ETYPE)s1, &env->fp_status));  \
>       }                                                               \
>       env->vstart = 0;                                                \
> +    /* set tail elements to 1s */                                   \
> +    if (vta) {                                                      \
> +        for (; i < total_elems; i++) {                              \
> +            vext_set_elem_mask(vd, i, 1);                           \
> +        }                                                           \
> +    }                                                               \
>   }
>   
>   GEN_VEXT_CMP_VF(vmfeq_vf_h, uint16_t, H2, float16_eq_quiet)
> @@ -4254,12 +4297,16 @@ static void do_##NAME(void *vd, void *vs2, int i)      \
>       *((TD *)vd + HD(i)) = OP(s2);                      \
>   }
>   
> -#define GEN_VEXT_V(NAME)                               \
> +#define GEN_VEXT_V(NAME, ESZ)                          \
>   void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
>                     CPURISCVState *env, uint32_t desc)   \
>   {                                                      \
>       uint32_t vm = vext_vm(desc);                       \
>       uint32_t vl = env->vl;                             \
> +    uint32_t total_elems =                             \
> +        vext_get_total_elems(env_archcpu(env),         \
> +                            env->vtype);               \
It's  better to aligned here too.

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 12/14] target/riscv: rvv: Add tail agnostic for vector reduction instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07 12:21 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index d654377fc8..3d19b9d141 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4550,6 +4550,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
 {                                                         \
     uint32_t vm = vext_vm(desc);                          \
     uint32_t vl = env->vl;                                \
+    uint32_t esz = sizeof(TD);                            \
+    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;     \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
     TD s1 =  *((TD *)vs1 + HD(0));                        \
                                                           \
@@ -4562,6 +4565,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
     }                                                     \
     *((TD *)vd + HD(0)) = s1;                             \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz,     \
+                                     vlenb);              \
 }
 
 /* vd[0] = sum(vs1[0], vs2[*]) */
@@ -4667,6 +4673,9 @@ void HELPER(vfwredsum_vs_h)(void *vd, void *v0, void *vs1,
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t esz = sizeof(uint32_t);
+    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
     uint32_t s1 =  *((uint32_t *)vs1 + H4(0));
 
@@ -4680,6 +4689,8 @@ void HELPER(vfwredsum_vs_h)(void *vd, void *v0, void *vs1,
     }
     *((uint32_t *)vd + H4(0)) = s1;
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz, vlenb);
 }
 
 void HELPER(vfwredsum_vs_w)(void *vd, void *v0, void *vs1,
@@ -4687,6 +4698,9 @@ void HELPER(vfwredsum_vs_w)(void *vd, void *v0, void *vs1,
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t esz = sizeof(uint64_t);
+    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
     uint64_t s1 =  *((uint64_t *)vs1);
 
@@ -4700,6 +4714,8 @@ void HELPER(vfwredsum_vs_w)(void *vd, void *v0, void *vs1,
     }
     *((uint64_t *)vd) = s1;
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz, vlenb);
 }
 
 /*
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 12/14] target/riscv: rvv: Add tail agnostic for vector reduction instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index d654377fc8..3d19b9d141 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4550,6 +4550,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
 {                                                         \
     uint32_t vm = vext_vm(desc);                          \
     uint32_t vl = env->vl;                                \
+    uint32_t esz = sizeof(TD);                            \
+    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;     \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
     TD s1 =  *((TD *)vs1 + HD(0));                        \
                                                           \
@@ -4562,6 +4565,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
     }                                                     \
     *((TD *)vd + HD(0)) = s1;                             \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz,     \
+                                     vlenb);              \
 }
 
 /* vd[0] = sum(vs1[0], vs2[*]) */
@@ -4667,6 +4673,9 @@ void HELPER(vfwredsum_vs_h)(void *vd, void *v0, void *vs1,
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t esz = sizeof(uint32_t);
+    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
     uint32_t s1 =  *((uint32_t *)vs1 + H4(0));
 
@@ -4680,6 +4689,8 @@ void HELPER(vfwredsum_vs_h)(void *vd, void *v0, void *vs1,
     }
     *((uint32_t *)vd + H4(0)) = s1;
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz, vlenb);
 }
 
 void HELPER(vfwredsum_vs_w)(void *vd, void *v0, void *vs1,
@@ -4687,6 +4698,9 @@ void HELPER(vfwredsum_vs_w)(void *vd, void *v0, void *vs1,
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t esz = sizeof(uint64_t);
+    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;
+    uint32_t vta = vext_vta(desc);
     uint32_t i;
     uint64_t s1 =  *((uint64_t *)vs1);
 
@@ -4700,6 +4714,8 @@ void HELPER(vfwredsum_vs_w)(void *vd, void *v0, void *vs1,
     }
     *((uint64_t *)vd) = s1;
     env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz, vlenb);
 }
 
 /*
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 12/14] target/riscv: rvv: Add tail agnostic for vector reduction instructions
  2022-03-24 19:03   ` ~eopxd
@ 2022-03-25 10:47     ` Weiwei Li
  -1 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:47 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, Bin Meng, Alistair Francis, eop Chen,
	Palmer Dabbelt

[-- Attachment #1: Type: text/plain, Size: 1177 bytes --]


在 2022/3/7 下午8:21, ~eopxd 写道:
> +    uint32_t vta = vext_vta(desc);                        \
>       uint32_t i;                                           \
>       TD s1 =  *((TD *)vs1 + HD(0));                        \
>                                                             \
> @@ -4562,6 +4565,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
>       }                                                     \
>       *((TD *)vd + HD(0)) = s1;                             \
>       env->vstart = 0;                                      \
> +    /* set tail elements to 1s */                         \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz,     \
> +                                     vlenb);              \
>   }
>   
>   /* vd[0] = sum(vs1[0], vs2[*]) */
> @@ -4667,6 +4673,9 @@ void HELPER(vfwredsum_vs_h)(void *vd, void *v0, void *vs1,
>   {
>       uint32_t vm = vext_vm(desc);
>       uint32_t vl = env->vl;
> +    uint32_t esz = sizeof(uint32_t);
> +    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;

vlenb  also can get from maxsz field of desc,  that is

uint32_tvlenb= simd_maxsz(desc)

Regards,

Weiwei Li


[-- Attachment #2: Type: text/html, Size: 2070 bytes --]

^ permalink raw reply	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 12/14] target/riscv: rvv: Add tail agnostic for vector reduction instructions
@ 2022-03-25 10:47     ` Weiwei Li
  0 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:47 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

[-- Attachment #1: Type: text/plain, Size: 1177 bytes --]


在 2022/3/7 下午8:21, ~eopxd 写道:
> +    uint32_t vta = vext_vta(desc);                        \
>       uint32_t i;                                           \
>       TD s1 =  *((TD *)vs1 + HD(0));                        \
>                                                             \
> @@ -4562,6 +4565,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
>       }                                                     \
>       *((TD *)vd + HD(0)) = s1;                             \
>       env->vstart = 0;                                      \
> +    /* set tail elements to 1s */                         \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, 1, esz,     \
> +                                     vlenb);              \
>   }
>   
>   /* vd[0] = sum(vs1[0], vs2[*]) */
> @@ -4667,6 +4673,9 @@ void HELPER(vfwredsum_vs_h)(void *vd, void *v0, void *vs1,
>   {
>       uint32_t vm = vext_vm(desc);
>       uint32_t vl = env->vl;
> +    uint32_t esz = sizeof(uint32_t);
> +    uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;

vlenb  also can get from maxsz field of desc,  that is

uint32_tvlenb= simd_maxsz(desc)

Regards,

Weiwei Li


[-- Attachment #2: Type: text/html, Size: 2070 bytes --]

^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 13/14] target/riscv: rvv: Add tail agnostic for vector mask instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07 15:26 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  4 ++++
 target/riscv/vector_helper.c            | 32 +++++++++++++++++++++++++
 2 files changed, 36 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 5bc340f419..8b24570e22 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -3191,6 +3191,7 @@ static bool trans_##NAME(DisasContext *s, arg_r *a)                \
         tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
@@ -3295,6 +3296,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd),                     \
                            vreg_ofs(s, 0), vreg_ofs(s, a->rs2),    \
                            cpu_env, s->cfg_ptr->vlen / 8,          \
@@ -3332,6 +3334,7 @@ static bool trans_viota_m(DisasContext *s, arg_viota_m *a)
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         static gen_helper_gvec_3_ptr * const fns[4] = {
             gen_helper_viota_m_b, gen_helper_viota_m_h,
             gen_helper_viota_m_w, gen_helper_viota_m_d,
@@ -3361,6 +3364,7 @@ static bool trans_vid_v(DisasContext *s, arg_vid_v *a)
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         static gen_helper_gvec_2_ptr * const fns[4] = {
             gen_helper_vid_v_b, gen_helper_vid_v_h,
             gen_helper_vid_v_w, gen_helper_vid_v_d,
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 3d19b9d141..0670489679 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4728,6 +4728,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   uint32_t desc)                          \
 {                                                         \
     uint32_t vl = env->vl;                                \
+    uint32_t total_elems =                                \
+        vext_get_total_elems(env_archcpu(env),            \
+                             env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
     int a, b;                                             \
                                                           \
@@ -4737,6 +4741,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
         vext_set_elem_mask(vd, i, OP(b, a));              \
     }                                                     \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    if (vta) {                                            \
+        for (; i < total_elems; i++) {                    \
+            vext_set_elem_mask(vd, i, 1);                 \
+        }                                                 \
+    }                                                     \
 }
 
 #define DO_NAND(N, M)  (!(N & M))
@@ -4804,6 +4814,8 @@ static void vmsetm(void *vd, void *v0, void *vs2, CPURISCVState *env,
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
     int i;
     bool first_mask_bit = false;
 
@@ -4832,6 +4844,12 @@ static void vmsetm(void *vd, void *v0, void *vs2, CPURISCVState *env,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    if (vta) {
+        for (; i < total_elems; i++) {
+            vext_set_elem_mask(vd, i, 1);
+        }
+    }
 }
 
 void HELPER(vmsbf_m)(void *vd, void *v0, void *vs2, CPURISCVState *env,
@@ -4859,6 +4877,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2, CPURISCVState *env,      \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint32_t sum = 0;                                                     \
     int i;                                                                \
                                                                           \
@@ -4872,6 +4894,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2, CPURISCVState *env,      \
         }                                                                 \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 GEN_VEXT_VIOTA_M(viota_m_b, uint8_t,  H1)
@@ -4885,6 +4910,10 @@ void HELPER(NAME)(void *vd, void *v0, CPURISCVState *env, uint32_t desc)  \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     int i;                                                                \
                                                                           \
     for (i = env->vstart; i < vl; i++) {                                  \
@@ -4894,6 +4923,9 @@ void HELPER(NAME)(void *vd, void *v0, CPURISCVState *env, uint32_t desc)  \
         *((ETYPE *)vd + H(i)) = i;                                        \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz); \
 }
 
 GEN_VEXT_VID_V(vid_v_b, uint8_t,  H1)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 13/14] target/riscv: rvv: Add tail agnostic for vector mask instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc |  4 ++++
 target/riscv/vector_helper.c            | 32 +++++++++++++++++++++++++
 2 files changed, 36 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 5bc340f419..8b24570e22 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -3191,6 +3191,7 @@ static bool trans_##NAME(DisasContext *s, arg_r *a)                \
         tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
@@ -3295,6 +3296,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);               \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd),                     \
                            vreg_ofs(s, 0), vreg_ofs(s, a->rs2),    \
                            cpu_env, s->cfg_ptr->vlen / 8,          \
@@ -3332,6 +3334,7 @@ static bool trans_viota_m(DisasContext *s, arg_viota_m *a)
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         static gen_helper_gvec_3_ptr * const fns[4] = {
             gen_helper_viota_m_b, gen_helper_viota_m_h,
             gen_helper_viota_m_w, gen_helper_viota_m_d,
@@ -3361,6 +3364,7 @@ static bool trans_vid_v(DisasContext *s, arg_vid_v *a)
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         static gen_helper_gvec_2_ptr * const fns[4] = {
             gen_helper_vid_v_b, gen_helper_vid_v_h,
             gen_helper_vid_v_w, gen_helper_vid_v_d,
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 3d19b9d141..0670489679 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4728,6 +4728,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   uint32_t desc)                          \
 {                                                         \
     uint32_t vl = env->vl;                                \
+    uint32_t total_elems =                                \
+        vext_get_total_elems(env_archcpu(env),            \
+                             env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                        \
     uint32_t i;                                           \
     int a, b;                                             \
                                                           \
@@ -4737,6 +4741,12 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
         vext_set_elem_mask(vd, i, OP(b, a));              \
     }                                                     \
     env->vstart = 0;                                      \
+    /* set tail elements to 1s */                         \
+    if (vta) {                                            \
+        for (; i < total_elems; i++) {                    \
+            vext_set_elem_mask(vd, i, 1);                 \
+        }                                                 \
+    }                                                     \
 }
 
 #define DO_NAND(N, M)  (!(N & M))
@@ -4804,6 +4814,8 @@ static void vmsetm(void *vd, void *v0, void *vs2, CPURISCVState *env,
 {
     uint32_t vm = vext_vm(desc);
     uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env_archcpu(env), env->vtype);
+    uint32_t vta = vext_vta(desc);
     int i;
     bool first_mask_bit = false;
 
@@ -4832,6 +4844,12 @@ static void vmsetm(void *vd, void *v0, void *vs2, CPURISCVState *env,
         }
     }
     env->vstart = 0;
+    /* set tail elements to 1s */
+    if (vta) {
+        for (; i < total_elems; i++) {
+            vext_set_elem_mask(vd, i, 1);
+        }
+    }
 }
 
 void HELPER(vmsbf_m)(void *vd, void *v0, void *vs2, CPURISCVState *env,
@@ -4859,6 +4877,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2, CPURISCVState *env,      \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint32_t sum = 0;                                                     \
     int i;                                                                \
                                                                           \
@@ -4872,6 +4894,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2, CPURISCVState *env,      \
         }                                                                 \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 GEN_VEXT_VIOTA_M(viota_m_b, uint8_t,  H1)
@@ -4885,6 +4910,10 @@ void HELPER(NAME)(void *vd, void *v0, CPURISCVState *env, uint32_t desc)  \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     int i;                                                                \
                                                                           \
     for (i = env->vstart; i < vl; i++) {                                  \
@@ -4894,6 +4923,9 @@ void HELPER(NAME)(void *vd, void *v0, CPURISCVState *env, uint32_t desc)  \
         *((ETYPE *)vd + H(i)) = i;                                        \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz); \
 }
 
 GEN_VEXT_VID_V(vid_v_b, uint8_t,  H1)
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 14/14] target/riscv: rvv: Add tail agnostic for vector permutation instructions
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-07 15:59 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 20 ++++++++
 target/riscv/vector_helper.c            | 68 ++++++++++++++++++++++---
 2 files changed, 82 insertions(+), 6 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 8b24570e22..f037d1875c 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -3724,6 +3724,15 @@ static bool trans_vrgather_vx(DisasContext *s, arg_rmrr *a)
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         int scale = s->lmul - (s->sew + 3);
         int vlmax = scale < 0 ?
                        s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
@@ -3757,6 +3766,15 @@ static bool trans_vrgather_vi(DisasContext *s, arg_rmrr *a)
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         int scale = s->lmul - (s->sew + 3);
         int vlmax = scale < 0 ?
                        s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
@@ -3809,6 +3827,7 @@ static bool trans_vcompress_vm(DisasContext *s, arg_r *a)
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
 
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
                            cpu_env, s->cfg_ptr->vlen / 8,
@@ -3914,6 +3933,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
     }
 
     data = FIELD_DP32(data, VDATA, VM, a->vm);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
 
     tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                        vreg_ofs(s, a->rs2), cpu_env,
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 0670489679..6e13d6bdcf 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4944,6 +4944,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     target_ulong offset = s1, i_min, i;                                   \
                                                                           \
     i_min = MAX(env->vstart, offset);                                     \
@@ -4953,6 +4957,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
         }                                                                 \
         *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i - offset));          \
     }                                                                     \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vslideup.vx vd, vs2, rs1, vm # vd[i+rs1] = vs2[i] */
@@ -4965,12 +4972,16 @@ GEN_VEXT_VSLIDEUP_VX(vslideup_vx_d, uint64_t, H8)
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
-    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(ETYPE)));           \
+    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(ETYPE)));        \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     target_ulong i_max, i;                                                \
                                                                           \
-    i_max = MAX(MIN(s1 < vlmax ? vlmax - s1 : 0, vl), env->vstart);       \
+    i_max = MAX(MIN(s1 < max_elem ? max_elem - s1 : 0, vl), env->vstart); \
     for (i = env->vstart; i < i_max; ++i) {                               \
         if (vm || vext_elem_mask(v0, i)) {                                \
             *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i + s1));          \
@@ -4984,6 +4995,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
     }                                                                     \
                                                                           \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vslidedown.vx vd, vs2, rs1, vm # vd[i] = vs2[i+rs1] */
@@ -4999,6 +5013,10 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
     typedef uint##BITWIDTH##_t ETYPE;                                       \
     uint32_t vm = vext_vm(desc);                                            \
     uint32_t vl = env->vl;                                                  \
+    uint32_t esz = sizeof(ETYPE);                                           \
+    uint32_t total_elems =                                                  \
+        vext_get_total_elems(env_archcpu(env), env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                                          \
     uint32_t i;                                                             \
                                                                             \
     for (i = env->vstart; i < vl; i++) {                                    \
@@ -5012,6 +5030,9 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
         }                                                                   \
     }                                                                       \
     env->vstart = 0;                                                        \
+    /* set tail elements to 1s */                                           \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                 \
+                                     total_elems * esz);                    \
 }
 
 GEN_VEXT_VSLIE1UP(8,  H1)
@@ -5039,6 +5060,10 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
     typedef uint##BITWIDTH##_t ETYPE;                                         \
     uint32_t vm = vext_vm(desc);                                              \
     uint32_t vl = env->vl;                                                    \
+    uint32_t esz = sizeof(ETYPE);                                             \
+    uint32_t total_elems =                                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);                   \
+    uint32_t vta = vext_vta(desc);                                            \
     uint32_t i;                                                               \
                                                                               \
     for (i = env->vstart; i < vl; i++) {                                      \
@@ -5052,6 +5077,9 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
         }                                                                     \
     }                                                                         \
     env->vstart = 0;                                                          \
+    /* set tail elements to 1s */                                             \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                   \
+                                     total_elems * esz);                      \
 }
 
 GEN_VEXT_VSLIDE1DOWN(8,  H1)
@@ -5102,9 +5130,13 @@ GEN_VEXT_VFSLIDE1DOWN_VF(vfslide1down_vf_d, 64)
 void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
-    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(TS2)));             \
+    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(TS2)));          \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(TS2);                                           \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint64_t index;                                                       \
     uint32_t i;                                                           \
                                                                           \
@@ -5113,13 +5145,16 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
             continue;                                                     \
         }                                                                 \
         index = *((TS1 *)vs1 + HS1(i));                                   \
-        if (index >= vlmax) {                                             \
+        if (index >= max_elem) {                                          \
             *((TS2 *)vd + HS2(i)) = 0;                                    \
         } else {                                                          \
             *((TS2 *)vd + HS2(i)) = *((TS2 *)vs2 + HS2(index));           \
         }                                                                 \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vd[i] = (vs1[i] >= VLMAX) ? 0 : vs2[vs1[i]]; */
@@ -5137,9 +5172,13 @@ GEN_VEXT_VRGATHER_VV(vrgatherei16_vv_d, uint16_t, uint64_t, H2, H8)
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
-    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(ETYPE)));           \
+    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(ETYPE)));        \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint64_t index = s1;                                                  \
     uint32_t i;                                                           \
                                                                           \
@@ -5147,13 +5186,16 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
         if (!vm && !vext_elem_mask(v0, i)) {                              \
             continue;                                                     \
         }                                                                 \
-        if (index >= vlmax) {                                             \
+        if (index >= max_elem) {                                          \
             *((ETYPE *)vd + H(i)) = 0;                                    \
         } else {                                                          \
             *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(index));           \
         }                                                                 \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vd[i] = (x[rs1] >= VLMAX) ? 0 : vs2[rs1] */
@@ -5168,6 +5210,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint32_t num = 0, i;                                                  \
                                                                           \
     for (i = env->vstart; i < vl; i++) {                                  \
@@ -5178,6 +5224,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
         num++;                                                            \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* Compress into vd elements of vs2 where vs1 is enabled */
@@ -5214,6 +5263,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,                 \
 {                                                                \
     uint32_t vl = env->vl;                                       \
     uint32_t vm = vext_vm(desc);                                 \
+    uint32_t esz = sizeof(ETYPE);                                \
+    uint32_t total_elems =                                       \
+        vext_get_total_elems(env_archcpu(env), env->vtype);      \
+    uint32_t vta = vext_vta(desc);                               \
     uint32_t i;                                                  \
                                                                  \
     for (i = env->vstart; i < vl; i++) {                         \
@@ -5223,6 +5276,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,                 \
         *((ETYPE *)vd + HD(i)) = *((DTYPE *)vs2 + HS1(i));       \
     }                                                            \
     env->vstart = 0;                                             \
+    /* set tail elements to 1s */                                \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,      \
+                                     total_elems * esz);         \
 }
 
 GEN_VEXT_INT_EXT(vzext_vf2_h, uint16_t, uint8_t,  H2, H1)
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 14/14] target/riscv: rvv: Add tail agnostic for vector permutation instructions
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 20 ++++++++
 target/riscv/vector_helper.c            | 68 ++++++++++++++++++++++---
 2 files changed, 82 insertions(+), 6 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 8b24570e22..f037d1875c 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -3724,6 +3724,15 @@ static bool trans_vrgather_vx(DisasContext *s, arg_rmrr *a)
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         int scale = s->lmul - (s->sew + 3);
         int vlmax = scale < 0 ?
                        s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
@@ -3757,6 +3766,15 @@ static bool trans_vrgather_vi(DisasContext *s, arg_rmrr *a)
     }
 
     if (a->vm && s->vl_eq_vlmax) {
+        if (s->vta && s->lmul < 0) {
+            /* tail elements may pass vlmax when lmul < 0
+             * set tail elements to 1s
+             */
+            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
+            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
+                             vreg_ofs(s, a->rd), -1,
+                             vlenb, vlenb);
+        }
         int scale = s->lmul - (s->sew + 3);
         int vlmax = scale < 0 ?
                        s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
@@ -3809,6 +3827,7 @@ static bool trans_vcompress_vm(DisasContext *s, arg_r *a)
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
 
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
+        data = FIELD_DP32(data, VDATA, VTA, s->vta);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
                            cpu_env, s->cfg_ptr->vlen / 8,
@@ -3914,6 +3933,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
     }
 
     data = FIELD_DP32(data, VDATA, VM, a->vm);
+    data = FIELD_DP32(data, VDATA, VTA, s->vta);
 
     tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                        vreg_ofs(s, a->rs2), cpu_env,
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 0670489679..6e13d6bdcf 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4944,6 +4944,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
 {                                                                         \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     target_ulong offset = s1, i_min, i;                                   \
                                                                           \
     i_min = MAX(env->vstart, offset);                                     \
@@ -4953,6 +4957,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
         }                                                                 \
         *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i - offset));          \
     }                                                                     \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vslideup.vx vd, vs2, rs1, vm # vd[i+rs1] = vs2[i] */
@@ -4965,12 +4972,16 @@ GEN_VEXT_VSLIDEUP_VX(vslideup_vx_d, uint64_t, H8)
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
-    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(ETYPE)));           \
+    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(ETYPE)));        \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     target_ulong i_max, i;                                                \
                                                                           \
-    i_max = MAX(MIN(s1 < vlmax ? vlmax - s1 : 0, vl), env->vstart);       \
+    i_max = MAX(MIN(s1 < max_elem ? max_elem - s1 : 0, vl), env->vstart); \
     for (i = env->vstart; i < i_max; ++i) {                               \
         if (vm || vext_elem_mask(v0, i)) {                                \
             *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i + s1));          \
@@ -4984,6 +4995,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
     }                                                                     \
                                                                           \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vslidedown.vx vd, vs2, rs1, vm # vd[i] = vs2[i+rs1] */
@@ -4999,6 +5013,10 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
     typedef uint##BITWIDTH##_t ETYPE;                                       \
     uint32_t vm = vext_vm(desc);                                            \
     uint32_t vl = env->vl;                                                  \
+    uint32_t esz = sizeof(ETYPE);                                           \
+    uint32_t total_elems =                                                  \
+        vext_get_total_elems(env_archcpu(env), env->vtype);                 \
+    uint32_t vta = vext_vta(desc);                                          \
     uint32_t i;                                                             \
                                                                             \
     for (i = env->vstart; i < vl; i++) {                                    \
@@ -5012,6 +5030,9 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
         }                                                                   \
     }                                                                       \
     env->vstart = 0;                                                        \
+    /* set tail elements to 1s */                                           \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                 \
+                                     total_elems * esz);                    \
 }
 
 GEN_VEXT_VSLIE1UP(8,  H1)
@@ -5039,6 +5060,10 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
     typedef uint##BITWIDTH##_t ETYPE;                                         \
     uint32_t vm = vext_vm(desc);                                              \
     uint32_t vl = env->vl;                                                    \
+    uint32_t esz = sizeof(ETYPE);                                             \
+    uint32_t total_elems =                                                    \
+        vext_get_total_elems(env_archcpu(env), env->vtype);                   \
+    uint32_t vta = vext_vta(desc);                                            \
     uint32_t i;                                                               \
                                                                               \
     for (i = env->vstart; i < vl; i++) {                                      \
@@ -5052,6 +5077,9 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
         }                                                                     \
     }                                                                         \
     env->vstart = 0;                                                          \
+    /* set tail elements to 1s */                                             \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                   \
+                                     total_elems * esz);                      \
 }
 
 GEN_VEXT_VSLIDE1DOWN(8,  H1)
@@ -5102,9 +5130,13 @@ GEN_VEXT_VFSLIDE1DOWN_VF(vfslide1down_vf_d, 64)
 void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
-    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(TS2)));             \
+    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(TS2)));          \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(TS2);                                           \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint64_t index;                                                       \
     uint32_t i;                                                           \
                                                                           \
@@ -5113,13 +5145,16 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
             continue;                                                     \
         }                                                                 \
         index = *((TS1 *)vs1 + HS1(i));                                   \
-        if (index >= vlmax) {                                             \
+        if (index >= max_elem) {                                          \
             *((TS2 *)vd + HS2(i)) = 0;                                    \
         } else {                                                          \
             *((TS2 *)vd + HS2(i)) = *((TS2 *)vs2 + HS2(index));           \
         }                                                                 \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vd[i] = (vs1[i] >= VLMAX) ? 0 : vs2[vs1[i]]; */
@@ -5137,9 +5172,13 @@ GEN_VEXT_VRGATHER_VV(vrgatherei16_vv_d, uint16_t, uint64_t, H2, H8)
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
-    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(ETYPE)));           \
+    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(ETYPE)));        \
     uint32_t vm = vext_vm(desc);                                          \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint64_t index = s1;                                                  \
     uint32_t i;                                                           \
                                                                           \
@@ -5147,13 +5186,16 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
         if (!vm && !vext_elem_mask(v0, i)) {                              \
             continue;                                                     \
         }                                                                 \
-        if (index >= vlmax) {                                             \
+        if (index >= max_elem) {                                          \
             *((ETYPE *)vd + H(i)) = 0;                                    \
         } else {                                                          \
             *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(index));           \
         }                                                                 \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* vd[i] = (x[rs1] >= VLMAX) ? 0 : vs2[rs1] */
@@ -5168,6 +5210,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
                   CPURISCVState *env, uint32_t desc)                      \
 {                                                                         \
     uint32_t vl = env->vl;                                                \
+    uint32_t esz = sizeof(ETYPE);                                         \
+    uint32_t total_elems =                                                \
+        vext_get_total_elems(env_archcpu(env), env->vtype);               \
+    uint32_t vta = vext_vta(desc);                                        \
     uint32_t num = 0, i;                                                  \
                                                                           \
     for (i = env->vstart; i < vl; i++) {                                  \
@@ -5178,6 +5224,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
         num++;                                                            \
     }                                                                     \
     env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
+                                     total_elems * esz);                  \
 }
 
 /* Compress into vd elements of vs2 where vs1 is enabled */
@@ -5214,6 +5263,10 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,                 \
 {                                                                \
     uint32_t vl = env->vl;                                       \
     uint32_t vm = vext_vm(desc);                                 \
+    uint32_t esz = sizeof(ETYPE);                                \
+    uint32_t total_elems =                                       \
+        vext_get_total_elems(env_archcpu(env), env->vtype);      \
+    uint32_t vta = vext_vta(desc);                               \
     uint32_t i;                                                  \
                                                                  \
     for (i = env->vstart; i < vl; i++) {                         \
@@ -5223,6 +5276,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,                 \
         *((ETYPE *)vd + HD(i)) = *((DTYPE *)vs2 + HS1(i));       \
     }                                                            \
     env->vstart = 0;                                             \
+    /* set tail elements to 1s */                                \
+    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,      \
+                                     total_elems * esz);         \
 }
 
 GEN_VEXT_INT_EXT(vzext_vf2_h, uint16_t, uint8_t,  H2, H1)
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 14/14] target/riscv: rvv: Add tail agnostic for vector permutation instructions
  2022-03-24 19:03   ` ~eopxd
@ 2022-03-25 10:51     ` Weiwei Li
  -1 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:51 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, Bin Meng, Alistair Francis, eop Chen,
	Palmer Dabbelt


在 2022/3/7 下午11:59, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc | 20 ++++++++
>   target/riscv/vector_helper.c            | 68 ++++++++++++++++++++++---
>   2 files changed, 82 insertions(+), 6 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index 8b24570e22..f037d1875c 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -3724,6 +3724,15 @@ static bool trans_vrgather_vx(DisasContext *s, arg_rmrr *a)
>       }
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */
> +            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                             vreg_ofs(s, a->rd), -1,
> +                             vlenb, vlenb);
> +        }
>           int scale = s->lmul - (s->sew + 3);
>           int vlmax = scale < 0 ?
>                          s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
> @@ -3757,6 +3766,15 @@ static bool trans_vrgather_vi(DisasContext *s, arg_rmrr *a)
>       }
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */
> +            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                             vreg_ofs(s, a->rd), -1,
> +                             vlenb, vlenb);
> +        }
>           int scale = s->lmul - (s->sew + 3);
>           int vlmax = scale < 0 ?
>                          s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
> @@ -3809,6 +3827,7 @@ static bool trans_vcompress_vm(DisasContext *s, arg_r *a)
>           tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
>   
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                              vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
>                              cpu_env, s->cfg_ptr->vlen / 8,
> @@ -3914,6 +3933,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
>       }
>   
>       data = FIELD_DP32(data, VDATA, VM, a->vm);
> +    data = FIELD_DP32(data, VDATA, VTA, s->vta);
>   
>       tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                          vreg_ofs(s, a->rs2), cpu_env,
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index 0670489679..6e13d6bdcf 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -4944,6 +4944,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>   {                                                                         \
>       uint32_t vm = vext_vm(desc);                                          \
>       uint32_t vl = env->vl;                                                \
> +    uint32_t esz = sizeof(ETYPE);                                         \
> +    uint32_t total_elems =                                                \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);               \
> +    uint32_t vta = vext_vta(desc);                                        \
>       target_ulong offset = s1, i_min, i;                                   \
>                                                                             \
>       i_min = MAX(env->vstart, offset);                                     \
> @@ -4953,6 +4957,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>           }                                                                 \
>           *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i - offset));          \
>       }                                                                     \
> +    /* set tail elements to 1s */                                         \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
> +                                     total_elems * esz);                  \
>   }
>   
>   /* vslideup.vx vd, vs2, rs1, vm # vd[i+rs1] = vs2[i] */
> @@ -4965,12 +4972,16 @@ GEN_VEXT_VSLIDEUP_VX(vslideup_vx_d, uint64_t, H8)
>   void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>                     CPURISCVState *env, uint32_t desc)                      \
>   {                                                                         \
> -    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(ETYPE)));           \
> +    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(ETYPE)));        \

This change seems unnecessary. It's truely vlmax here.

>       uint32_t vm = vext_vm(desc);                                          \
>       uint32_t vl = env->vl;                                                \
> +    uint32_t esz = sizeof(ETYPE);                                         \
> +    uint32_t total_elems =                                                \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);               \
> +    uint32_t vta = vext_vta(desc);                                        \
>       target_ulong i_max, i;                                                \
>                                                                             \
> -    i_max = MAX(MIN(s1 < vlmax ? vlmax - s1 : 0, vl), env->vstart);       \
> +    i_max = MAX(MIN(s1 < max_elem ? max_elem - s1 : 0, vl), env->vstart); \
>       for (i = env->vstart; i < i_max; ++i) {                               \
>           if (vm || vext_elem_mask(v0, i)) {                                \
>               *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i + s1));          \
> @@ -4984,6 +4995,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>       }                                                                     \
>                                                                             \
>       env->vstart = 0;                                                      \
> +    /* set tail elements to 1s */                                         \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
> +                                     total_elems * esz);                  \
>   }
>   
>   /* vslidedown.vx vd, vs2, rs1, vm # vd[i] = vs2[i+rs1] */
> @@ -4999,6 +5013,10 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>       typedef uint##BITWIDTH##_t ETYPE;                                       \
>       uint32_t vm = vext_vm(desc);                                            \
>       uint32_t vl = env->vl;                                                  \
> +    uint32_t esz = sizeof(ETYPE);                                           \
> +    uint32_t total_elems =                                                  \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);                 \
> +    uint32_t vta = vext_vta(desc);                                          \
>       uint32_t i;                                                             \
>                                                                               \
>       for (i = env->vstart; i < vl; i++) {                                    \
> @@ -5012,6 +5030,9 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>           }                                                                   \
>       }                                                                       \
>       env->vstart = 0;                                                        \
> +    /* set tail elements to 1s */                                           \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                 \
> +                                     total_elems * esz);                    \
>   }
>   
>   GEN_VEXT_VSLIE1UP(8,  H1)
> @@ -5039,6 +5060,10 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>       typedef uint##BITWIDTH##_t ETYPE;                                         \
>       uint32_t vm = vext_vm(desc);                                              \
>       uint32_t vl = env->vl;                                                    \
> +    uint32_t esz = sizeof(ETYPE);                                             \
> +    uint32_t total_elems =                                                    \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);                   \
> +    uint32_t vta = vext_vta(desc);                                            \
>       uint32_t i;                                                               \
>                                                                                 \
>       for (i = env->vstart; i < vl; i++) {                                      \
> @@ -5052,6 +5077,9 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>           }                                                                     \
>       }                                                                         \
>       env->vstart = 0;                                                          \
> +    /* set tail elements to 1s */                                             \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                   \
> +                                     total_elems * esz);                      \
>   }
>   
>   GEN_VEXT_VSLIDE1DOWN(8,  H1)
> @@ -5102,9 +5130,13 @@ GEN_VEXT_VFSLIDE1DOWN_VF(vfslide1down_vf_d, 64)
>   void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
>                     CPURISCVState *env, uint32_t desc)                      \
>   {                                                                         \
> -    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(TS2)));             \
> +    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(TS2)));          \
similar to above.

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* Re: [PATCH qemu v4 14/14] target/riscv: rvv: Add tail agnostic for vector permutation instructions
@ 2022-03-25 10:51     ` Weiwei Li
  0 siblings, 0 replies; 42+ messages in thread
From: Weiwei Li @ 2022-03-25 10:51 UTC (permalink / raw)
  To: ~eopxd, qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt


在 2022/3/7 下午11:59, ~eopxd 写道:
> From: eopXD <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc | 20 ++++++++
>   target/riscv/vector_helper.c            | 68 ++++++++++++++++++++++---
>   2 files changed, 82 insertions(+), 6 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
> index 8b24570e22..f037d1875c 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -3724,6 +3724,15 @@ static bool trans_vrgather_vx(DisasContext *s, arg_rmrr *a)
>       }
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */
> +            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                             vreg_ofs(s, a->rd), -1,
> +                             vlenb, vlenb);
> +        }
>           int scale = s->lmul - (s->sew + 3);
>           int vlmax = scale < 0 ?
>                          s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
> @@ -3757,6 +3766,15 @@ static bool trans_vrgather_vi(DisasContext *s, arg_rmrr *a)
>       }
>   
>       if (a->vm && s->vl_eq_vlmax) {
> +        if (s->vta && s->lmul < 0) {
> +            /* tail elements may pass vlmax when lmul < 0
> +             * set tail elements to 1s
> +             */
> +            uint32_t vlenb = s->cfg_ptr->vlen >> 3;
> +            tcg_gen_gvec_ori(s->sew, vreg_ofs(s, a->rd),
> +                             vreg_ofs(s, a->rd), -1,
> +                             vlenb, vlenb);
> +        }
>           int scale = s->lmul - (s->sew + 3);
>           int vlmax = scale < 0 ?
>                          s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
> @@ -3809,6 +3827,7 @@ static bool trans_vcompress_vm(DisasContext *s, arg_r *a)
>           tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
>   
>           data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
> +        data = FIELD_DP32(data, VDATA, VTA, s->vta);
>           tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                              vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
>                              cpu_env, s->cfg_ptr->vlen / 8,
> @@ -3914,6 +3933,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
>       }
>   
>       data = FIELD_DP32(data, VDATA, VM, a->vm);
> +    data = FIELD_DP32(data, VDATA, VTA, s->vta);
>   
>       tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
>                          vreg_ofs(s, a->rs2), cpu_env,
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index 0670489679..6e13d6bdcf 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -4944,6 +4944,10 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>   {                                                                         \
>       uint32_t vm = vext_vm(desc);                                          \
>       uint32_t vl = env->vl;                                                \
> +    uint32_t esz = sizeof(ETYPE);                                         \
> +    uint32_t total_elems =                                                \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);               \
> +    uint32_t vta = vext_vta(desc);                                        \
>       target_ulong offset = s1, i_min, i;                                   \
>                                                                             \
>       i_min = MAX(env->vstart, offset);                                     \
> @@ -4953,6 +4957,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>           }                                                                 \
>           *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i - offset));          \
>       }                                                                     \
> +    /* set tail elements to 1s */                                         \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
> +                                     total_elems * esz);                  \
>   }
>   
>   /* vslideup.vx vd, vs2, rs1, vm # vd[i+rs1] = vs2[i] */
> @@ -4965,12 +4972,16 @@ GEN_VEXT_VSLIDEUP_VX(vslideup_vx_d, uint64_t, H8)
>   void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>                     CPURISCVState *env, uint32_t desc)                      \
>   {                                                                         \
> -    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(ETYPE)));           \
> +    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(ETYPE)));        \

This change seems unnecessary. It's truely vlmax here.

>       uint32_t vm = vext_vm(desc);                                          \
>       uint32_t vl = env->vl;                                                \
> +    uint32_t esz = sizeof(ETYPE);                                         \
> +    uint32_t total_elems =                                                \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);               \
> +    uint32_t vta = vext_vta(desc);                                        \
>       target_ulong i_max, i;                                                \
>                                                                             \
> -    i_max = MAX(MIN(s1 < vlmax ? vlmax - s1 : 0, vl), env->vstart);       \
> +    i_max = MAX(MIN(s1 < max_elem ? max_elem - s1 : 0, vl), env->vstart); \
>       for (i = env->vstart; i < i_max; ++i) {                               \
>           if (vm || vext_elem_mask(v0, i)) {                                \
>               *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i + s1));          \
> @@ -4984,6 +4995,9 @@ void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
>       }                                                                     \
>                                                                             \
>       env->vstart = 0;                                                      \
> +    /* set tail elements to 1s */                                         \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,               \
> +                                     total_elems * esz);                  \
>   }
>   
>   /* vslidedown.vx vd, vs2, rs1, vm # vd[i] = vs2[i+rs1] */
> @@ -4999,6 +5013,10 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>       typedef uint##BITWIDTH##_t ETYPE;                                       \
>       uint32_t vm = vext_vm(desc);                                            \
>       uint32_t vl = env->vl;                                                  \
> +    uint32_t esz = sizeof(ETYPE);                                           \
> +    uint32_t total_elems =                                                  \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);                 \
> +    uint32_t vta = vext_vta(desc);                                          \
>       uint32_t i;                                                             \
>                                                                               \
>       for (i = env->vstart; i < vl; i++) {                                    \
> @@ -5012,6 +5030,9 @@ static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>           }                                                                   \
>       }                                                                       \
>       env->vstart = 0;                                                        \
> +    /* set tail elements to 1s */                                           \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                 \
> +                                     total_elems * esz);                    \
>   }
>   
>   GEN_VEXT_VSLIE1UP(8,  H1)
> @@ -5039,6 +5060,10 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>       typedef uint##BITWIDTH##_t ETYPE;                                         \
>       uint32_t vm = vext_vm(desc);                                              \
>       uint32_t vl = env->vl;                                                    \
> +    uint32_t esz = sizeof(ETYPE);                                             \
> +    uint32_t total_elems =                                                    \
> +        vext_get_total_elems(env_archcpu(env), env->vtype);                   \
> +    uint32_t vta = vext_vta(desc);                                            \
>       uint32_t i;                                                               \
>                                                                                 \
>       for (i = env->vstart; i < vl; i++) {                                      \
> @@ -5052,6 +5077,9 @@ static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
>           }                                                                     \
>       }                                                                         \
>       env->vstart = 0;                                                          \
> +    /* set tail elements to 1s */                                             \
> +    vext_set_elems_1s_fns[ctzl(esz)](vd, vta, vl, vl * esz,                   \
> +                                     total_elems * esz);                      \
>   }
>   
>   GEN_VEXT_VSLIDE1DOWN(8,  H1)
> @@ -5102,9 +5130,13 @@ GEN_VEXT_VFSLIDE1DOWN_VF(vfslide1down_vf_d, 64)
>   void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,               \
>                     CPURISCVState *env, uint32_t desc)                      \
>   {                                                                         \
> -    uint32_t vlmax = vext_max_elems(desc, ctzl(sizeof(TS2)));             \
> +    uint32_t max_elem = vext_max_elems(desc, ctzl(sizeof(TS2)));          \
similar to above.

Regards,

Weiwei Li



^ permalink raw reply	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 02/14] target/riscv: rvv: Rename ambiguous esz
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-24 19:03   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-09  8:34 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

No functional change intended in this commit.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 76 ++++++++++++++++++------------------
 1 file changed, 38 insertions(+), 38 deletions(-)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index e94caf1a3c..d0452a7756 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -125,9 +125,9 @@ static inline int32_t vext_lmul(uint32_t desc)
 /*
  * Get the maximum number of elements can be operated.
  *
- * esz: log2 of element size in bytes.
+ * log2_esz: log2 of element size in bytes.
  */
-static inline uint32_t vext_max_elems(uint32_t desc, uint32_t esz)
+static inline uint32_t vext_max_elems(uint32_t desc, uint32_t log2_esz)
 {
     /*
      * As simd_desc support at most 2048 bytes, the max vlen is 1024 bits.
@@ -136,7 +136,7 @@ static inline uint32_t vext_max_elems(uint32_t desc, uint32_t esz)
     uint32_t vlenb = simd_maxsz(desc);
 
     /* Return VLMAX */
-    int scale = vext_lmul(desc) - esz;
+    int scale = vext_lmul(desc) - log2_esz;
     return scale < 0 ? vlenb >> -scale : vlenb << scale;
 }
 
@@ -231,11 +231,11 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
                  target_ulong stride, CPURISCVState *env,
                  uint32_t desc, uint32_t vm,
                  vext_ldst_elem_fn *ldst_elem,
-                 uint32_t esz, uintptr_t ra, MMUAccessType access_type)
+                 uint32_t log2_esz, uintptr_t ra, MMUAccessType access_type)
 {
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
 
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
         if (!vm && !vext_elem_mask(v0, i)) {
@@ -244,7 +244,7 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
 
         k = 0;
         while (k < nf) {
-            target_ulong addr = base + stride * i + (k << esz);
+            target_ulong addr = base + stride * i + (k << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -289,18 +289,18 @@ GEN_VEXT_ST_STRIDE(vsse64_v, int64_t, ste_d)
 /* unmasked unit-stride load and store operation*/
 static void
 vext_ldst_us(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
-             vext_ldst_elem_fn *ldst_elem, uint32_t esz, uint32_t evl,
+             vext_ldst_elem_fn *ldst_elem, uint32_t log2_esz, uint32_t evl,
              uintptr_t ra, MMUAccessType access_type)
 {
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < evl; i++, env->vstart++) {
         k = 0;
         while (k < nf) {
-            target_ulong addr = base + ((i * nf + k) << esz);
+            target_ulong addr = base + ((i * nf + k) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -399,12 +399,12 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
                 void *vs2, CPURISCVState *env, uint32_t desc,
                 vext_get_index_addr get_index_addr,
                 vext_ldst_elem_fn *ldst_elem,
-                uint32_t esz, uintptr_t ra, MMUAccessType access_type)
+                uint32_t log2_esz, uintptr_t ra, MMUAccessType access_type)
 {
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
@@ -414,7 +414,7 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
 
         k = 0;
         while (k < nf) {
-            abi_ptr addr = get_index_addr(base, i, vs2) + (k << esz);
+            abi_ptr addr = get_index_addr(base, i, vs2) + (k << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -480,13 +480,13 @@ static inline void
 vext_ldff(void *vd, void *v0, target_ulong base,
           CPURISCVState *env, uint32_t desc,
           vext_ldst_elem_fn *ldst_elem,
-          uint32_t esz, uintptr_t ra)
+          uint32_t log2_esz, uintptr_t ra)
 {
     void *host;
     uint32_t i, k, vl = 0;
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
     target_ulong addr, offset, remain;
 
     /* probe every access*/
@@ -494,12 +494,12 @@ vext_ldff(void *vd, void *v0, target_ulong base,
         if (!vm && !vext_elem_mask(v0, i)) {
             continue;
         }
-        addr = adjust_addr(env, base + i * (nf << esz));
+        addr = adjust_addr(env, base + i * (nf << log2_esz));
         if (i == 0) {
-            probe_pages(env, addr, nf << esz, ra, MMU_DATA_LOAD);
+            probe_pages(env, addr, nf << log2_esz, ra, MMU_DATA_LOAD);
         } else {
             /* if it triggers an exception, no need to check watchpoint */
-            remain = nf << esz;
+            remain = nf << log2_esz;
             while (remain > 0) {
                 offset = -(addr | TARGET_PAGE_MASK);
                 host = tlb_vaddr_to_host(env, addr, MMU_DATA_LOAD,
@@ -536,7 +536,7 @@ ProbeSuccess:
             continue;
         }
         while (k < nf) {
-            target_ulong addr = base + ((i * nf + k) << esz);
+            target_ulong addr = base + ((i * nf + k) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -576,13 +576,13 @@ GEN_VEXT_LDFF(vle64ff_v, int64_t, lde_d)
  */
 static void
 vext_ldst_whole(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
-                vext_ldst_elem_fn *ldst_elem, uint32_t esz, uintptr_t ra,
+                vext_ldst_elem_fn *ldst_elem, uint32_t log2_esz, uintptr_t ra,
                 MMUAccessType access_type)
 {
     uint32_t i, k, off, pos;
     uint32_t nf = vext_nf(desc);
     uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;
-    uint32_t max_elems = vlenb >> esz;
+    uint32_t max_elems = vlenb >> log2_esz;
 
     k = env->vstart / max_elems;
     off = env->vstart % max_elems;
@@ -590,7 +590,7 @@ vext_ldst_whole(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
     if (off) {
         /* load/store rest of elements of current segment pointed by vstart */
         for (pos = off; pos < max_elems; pos++, env->vstart++) {
-            target_ulong addr = base + ((pos + k * max_elems) << esz);
+            target_ulong addr = base + ((pos + k * max_elems) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), pos + k * max_elems, vd, ra);
         }
         k++;
@@ -599,7 +599,7 @@ vext_ldst_whole(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
     /* load/store elements for rest of segments */
     for (; k < nf; k++) {
         for (i = 0; i < max_elems; i++, env->vstart++) {
-            target_ulong addr = base + ((i + k * max_elems) << esz);
+            target_ulong addr = base + ((i + k * max_elems) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
         }
     }
@@ -4691,11 +4691,11 @@ GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_h, uint16_t, H2)
 GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_w, uint32_t, H4)
 GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_d, uint64_t, H8)
 
-#define GEN_VEXT_VSLIE1UP(ESZ, H)                                           \
-static void vslide1up_##ESZ(void *vd, void *v0, target_ulong s1, void *vs2, \
-                     CPURISCVState *env, uint32_t desc)                     \
+#define GEN_VEXT_VSLIE1UP(BITWIDTH, H)                                      \
+static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
+                     void *vs2, CPURISCVState *env, uint32_t desc)          \
 {                                                                           \
-    typedef uint##ESZ##_t ETYPE;                                            \
+    typedef uint##BITWIDTH##_t ETYPE;                                       \
     uint32_t vm = vext_vm(desc);                                            \
     uint32_t vl = env->vl;                                                  \
     uint32_t i;                                                             \
@@ -4718,11 +4718,11 @@ GEN_VEXT_VSLIE1UP(16, H2)
 GEN_VEXT_VSLIE1UP(32, H4)
 GEN_VEXT_VSLIE1UP(64, H8)
 
-#define GEN_VEXT_VSLIDE1UP_VX(NAME, ESZ)                          \
+#define GEN_VEXT_VSLIDE1UP_VX(NAME, BITWIDTH)                     \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)              \
 {                                                                 \
-    vslide1up_##ESZ(vd, v0, s1, vs2, env, desc);                  \
+    vslide1up_##BITWIDTH(vd, v0, s1, vs2, env, desc);             \
 }
 
 /* vslide1up.vx vd, vs2, rs1, vm # vd[0]=x[rs1], vd[i+1] = vs2[i] */
@@ -4731,11 +4731,11 @@ GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_h, 16)
 GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_w, 32)
 GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_d, 64)
 
-#define GEN_VEXT_VSLIDE1DOWN(ESZ, H)                                          \
-static void vslide1down_##ESZ(void *vd, void *v0, target_ulong s1, void *vs2, \
-                       CPURISCVState *env, uint32_t desc)                     \
+#define GEN_VEXT_VSLIDE1DOWN(BITWIDTH, H)                                     \
+static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
+                       void *vs2, CPURISCVState *env, uint32_t desc)          \
 {                                                                             \
-    typedef uint##ESZ##_t ETYPE;                                              \
+    typedef uint##BITWIDTH##_t ETYPE;                                         \
     uint32_t vm = vext_vm(desc);                                              \
     uint32_t vl = env->vl;                                                    \
     uint32_t i;                                                               \
@@ -4758,11 +4758,11 @@ GEN_VEXT_VSLIDE1DOWN(16, H2)
 GEN_VEXT_VSLIDE1DOWN(32, H4)
 GEN_VEXT_VSLIDE1DOWN(64, H8)
 
-#define GEN_VEXT_VSLIDE1DOWN_VX(NAME, ESZ)                        \
+#define GEN_VEXT_VSLIDE1DOWN_VX(NAME, BITWIDTH)                   \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)              \
 {                                                                 \
-    vslide1down_##ESZ(vd, v0, s1, vs2, env, desc);                \
+    vslide1down_##BITWIDTH(vd, v0, s1, vs2, env, desc);           \
 }
 
 /* vslide1down.vx vd, vs2, rs1, vm # vd[i] = vs2[i+1], vd[vl-1]=x[rs1] */
@@ -4772,11 +4772,11 @@ GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_w, 32)
 GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_d, 64)
 
 /* Vector Floating-Point Slide Instructions */
-#define GEN_VEXT_VFSLIDE1UP_VF(NAME, ESZ)                     \
+#define GEN_VEXT_VFSLIDE1UP_VF(NAME, BITWIDTH)                \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
-    vslide1up_##ESZ(vd, v0, s1, vs2, env, desc);              \
+    vslide1up_##BITWIDTH(vd, v0, s1, vs2, env, desc);         \
 }
 
 /* vfslide1up.vf vd, vs2, rs1, vm # vd[0]=f[rs1], vd[i+1] = vs2[i] */
@@ -4784,11 +4784,11 @@ GEN_VEXT_VFSLIDE1UP_VF(vfslide1up_vf_h, 16)
 GEN_VEXT_VFSLIDE1UP_VF(vfslide1up_vf_w, 32)
 GEN_VEXT_VFSLIDE1UP_VF(vfslide1up_vf_d, 64)
 
-#define GEN_VEXT_VFSLIDE1DOWN_VF(NAME, ESZ)                   \
+#define GEN_VEXT_VFSLIDE1DOWN_VF(NAME, BITWIDTH)              \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
-    vslide1down_##ESZ(vd, v0, s1, vs2, env, desc);            \
+    vslide1down_##BITWIDTH(vd, v0, s1, vs2, env, desc);       \
 }
 
 /* vfslide1down.vf vd, vs2, rs1, vm # vd[i] = vs2[i+1], vd[vl-1]=f[rs1] */
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 02/14] target/riscv: rvv: Rename ambiguous esz
@ 2022-03-24 19:03   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-24 19:03 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

No functional change intended in this commit.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 76 ++++++++++++++++++------------------
 1 file changed, 38 insertions(+), 38 deletions(-)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index e94caf1a3c..d0452a7756 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -125,9 +125,9 @@ static inline int32_t vext_lmul(uint32_t desc)
 /*
  * Get the maximum number of elements can be operated.
  *
- * esz: log2 of element size in bytes.
+ * log2_esz: log2 of element size in bytes.
  */
-static inline uint32_t vext_max_elems(uint32_t desc, uint32_t esz)
+static inline uint32_t vext_max_elems(uint32_t desc, uint32_t log2_esz)
 {
     /*
      * As simd_desc support at most 2048 bytes, the max vlen is 1024 bits.
@@ -136,7 +136,7 @@ static inline uint32_t vext_max_elems(uint32_t desc, uint32_t esz)
     uint32_t vlenb = simd_maxsz(desc);
 
     /* Return VLMAX */
-    int scale = vext_lmul(desc) - esz;
+    int scale = vext_lmul(desc) - log2_esz;
     return scale < 0 ? vlenb >> -scale : vlenb << scale;
 }
 
@@ -231,11 +231,11 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
                  target_ulong stride, CPURISCVState *env,
                  uint32_t desc, uint32_t vm,
                  vext_ldst_elem_fn *ldst_elem,
-                 uint32_t esz, uintptr_t ra, MMUAccessType access_type)
+                 uint32_t log2_esz, uintptr_t ra, MMUAccessType access_type)
 {
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
 
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
         if (!vm && !vext_elem_mask(v0, i)) {
@@ -244,7 +244,7 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
 
         k = 0;
         while (k < nf) {
-            target_ulong addr = base + stride * i + (k << esz);
+            target_ulong addr = base + stride * i + (k << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -289,18 +289,18 @@ GEN_VEXT_ST_STRIDE(vsse64_v, int64_t, ste_d)
 /* unmasked unit-stride load and store operation*/
 static void
 vext_ldst_us(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
-             vext_ldst_elem_fn *ldst_elem, uint32_t esz, uint32_t evl,
+             vext_ldst_elem_fn *ldst_elem, uint32_t log2_esz, uint32_t evl,
              uintptr_t ra, MMUAccessType access_type)
 {
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < evl; i++, env->vstart++) {
         k = 0;
         while (k < nf) {
-            target_ulong addr = base + ((i * nf + k) << esz);
+            target_ulong addr = base + ((i * nf + k) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -399,12 +399,12 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
                 void *vs2, CPURISCVState *env, uint32_t desc,
                 vext_get_index_addr get_index_addr,
                 vext_ldst_elem_fn *ldst_elem,
-                uint32_t esz, uintptr_t ra, MMUAccessType access_type)
+                uint32_t log2_esz, uintptr_t ra, MMUAccessType access_type)
 {
     uint32_t i, k;
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
 
     /* load bytes from guest memory */
     for (i = env->vstart; i < env->vl; i++, env->vstart++) {
@@ -414,7 +414,7 @@ vext_ldst_index(void *vd, void *v0, target_ulong base,
 
         k = 0;
         while (k < nf) {
-            abi_ptr addr = get_index_addr(base, i, vs2) + (k << esz);
+            abi_ptr addr = get_index_addr(base, i, vs2) + (k << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -480,13 +480,13 @@ static inline void
 vext_ldff(void *vd, void *v0, target_ulong base,
           CPURISCVState *env, uint32_t desc,
           vext_ldst_elem_fn *ldst_elem,
-          uint32_t esz, uintptr_t ra)
+          uint32_t log2_esz, uintptr_t ra)
 {
     void *host;
     uint32_t i, k, vl = 0;
     uint32_t nf = vext_nf(desc);
     uint32_t vm = vext_vm(desc);
-    uint32_t max_elems = vext_max_elems(desc, esz);
+    uint32_t max_elems = vext_max_elems(desc, log2_esz);
     target_ulong addr, offset, remain;
 
     /* probe every access*/
@@ -494,12 +494,12 @@ vext_ldff(void *vd, void *v0, target_ulong base,
         if (!vm && !vext_elem_mask(v0, i)) {
             continue;
         }
-        addr = adjust_addr(env, base + i * (nf << esz));
+        addr = adjust_addr(env, base + i * (nf << log2_esz));
         if (i == 0) {
-            probe_pages(env, addr, nf << esz, ra, MMU_DATA_LOAD);
+            probe_pages(env, addr, nf << log2_esz, ra, MMU_DATA_LOAD);
         } else {
             /* if it triggers an exception, no need to check watchpoint */
-            remain = nf << esz;
+            remain = nf << log2_esz;
             while (remain > 0) {
                 offset = -(addr | TARGET_PAGE_MASK);
                 host = tlb_vaddr_to_host(env, addr, MMU_DATA_LOAD,
@@ -536,7 +536,7 @@ ProbeSuccess:
             continue;
         }
         while (k < nf) {
-            target_ulong addr = base + ((i * nf + k) << esz);
+            target_ulong addr = base + ((i * nf + k) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
             k++;
         }
@@ -576,13 +576,13 @@ GEN_VEXT_LDFF(vle64ff_v, int64_t, lde_d)
  */
 static void
 vext_ldst_whole(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
-                vext_ldst_elem_fn *ldst_elem, uint32_t esz, uintptr_t ra,
+                vext_ldst_elem_fn *ldst_elem, uint32_t log2_esz, uintptr_t ra,
                 MMUAccessType access_type)
 {
     uint32_t i, k, off, pos;
     uint32_t nf = vext_nf(desc);
     uint32_t vlenb = env_archcpu(env)->cfg.vlen >> 3;
-    uint32_t max_elems = vlenb >> esz;
+    uint32_t max_elems = vlenb >> log2_esz;
 
     k = env->vstart / max_elems;
     off = env->vstart % max_elems;
@@ -590,7 +590,7 @@ vext_ldst_whole(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
     if (off) {
         /* load/store rest of elements of current segment pointed by vstart */
         for (pos = off; pos < max_elems; pos++, env->vstart++) {
-            target_ulong addr = base + ((pos + k * max_elems) << esz);
+            target_ulong addr = base + ((pos + k * max_elems) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), pos + k * max_elems, vd, ra);
         }
         k++;
@@ -599,7 +599,7 @@ vext_ldst_whole(void *vd, target_ulong base, CPURISCVState *env, uint32_t desc,
     /* load/store elements for rest of segments */
     for (; k < nf; k++) {
         for (i = 0; i < max_elems; i++, env->vstart++) {
-            target_ulong addr = base + ((i + k * max_elems) << esz);
+            target_ulong addr = base + ((i + k * max_elems) << log2_esz);
             ldst_elem(env, adjust_addr(env, addr), i + k * max_elems, vd, ra);
         }
     }
@@ -4691,11 +4691,11 @@ GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_h, uint16_t, H2)
 GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_w, uint32_t, H4)
 GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_d, uint64_t, H8)
 
-#define GEN_VEXT_VSLIE1UP(ESZ, H)                                           \
-static void vslide1up_##ESZ(void *vd, void *v0, target_ulong s1, void *vs2, \
-                     CPURISCVState *env, uint32_t desc)                     \
+#define GEN_VEXT_VSLIE1UP(BITWIDTH, H)                                      \
+static void vslide1up_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
+                     void *vs2, CPURISCVState *env, uint32_t desc)          \
 {                                                                           \
-    typedef uint##ESZ##_t ETYPE;                                            \
+    typedef uint##BITWIDTH##_t ETYPE;                                       \
     uint32_t vm = vext_vm(desc);                                            \
     uint32_t vl = env->vl;                                                  \
     uint32_t i;                                                             \
@@ -4718,11 +4718,11 @@ GEN_VEXT_VSLIE1UP(16, H2)
 GEN_VEXT_VSLIE1UP(32, H4)
 GEN_VEXT_VSLIE1UP(64, H8)
 
-#define GEN_VEXT_VSLIDE1UP_VX(NAME, ESZ)                          \
+#define GEN_VEXT_VSLIDE1UP_VX(NAME, BITWIDTH)                     \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)              \
 {                                                                 \
-    vslide1up_##ESZ(vd, v0, s1, vs2, env, desc);                  \
+    vslide1up_##BITWIDTH(vd, v0, s1, vs2, env, desc);             \
 }
 
 /* vslide1up.vx vd, vs2, rs1, vm # vd[0]=x[rs1], vd[i+1] = vs2[i] */
@@ -4731,11 +4731,11 @@ GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_h, 16)
 GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_w, 32)
 GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_d, 64)
 
-#define GEN_VEXT_VSLIDE1DOWN(ESZ, H)                                          \
-static void vslide1down_##ESZ(void *vd, void *v0, target_ulong s1, void *vs2, \
-                       CPURISCVState *env, uint32_t desc)                     \
+#define GEN_VEXT_VSLIDE1DOWN(BITWIDTH, H)                                     \
+static void vslide1down_##BITWIDTH(void *vd, void *v0, target_ulong s1,       \
+                       void *vs2, CPURISCVState *env, uint32_t desc)          \
 {                                                                             \
-    typedef uint##ESZ##_t ETYPE;                                              \
+    typedef uint##BITWIDTH##_t ETYPE;                                         \
     uint32_t vm = vext_vm(desc);                                              \
     uint32_t vl = env->vl;                                                    \
     uint32_t i;                                                               \
@@ -4758,11 +4758,11 @@ GEN_VEXT_VSLIDE1DOWN(16, H2)
 GEN_VEXT_VSLIDE1DOWN(32, H4)
 GEN_VEXT_VSLIDE1DOWN(64, H8)
 
-#define GEN_VEXT_VSLIDE1DOWN_VX(NAME, ESZ)                        \
+#define GEN_VEXT_VSLIDE1DOWN_VX(NAME, BITWIDTH)                   \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)              \
 {                                                                 \
-    vslide1down_##ESZ(vd, v0, s1, vs2, env, desc);                \
+    vslide1down_##BITWIDTH(vd, v0, s1, vs2, env, desc);           \
 }
 
 /* vslide1down.vx vd, vs2, rs1, vm # vd[i] = vs2[i+1], vd[vl-1]=x[rs1] */
@@ -4772,11 +4772,11 @@ GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_w, 32)
 GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_d, 64)
 
 /* Vector Floating-Point Slide Instructions */
-#define GEN_VEXT_VFSLIDE1UP_VF(NAME, ESZ)                     \
+#define GEN_VEXT_VFSLIDE1UP_VF(NAME, BITWIDTH)                \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
-    vslide1up_##ESZ(vd, v0, s1, vs2, env, desc);              \
+    vslide1up_##BITWIDTH(vd, v0, s1, vs2, env, desc);         \
 }
 
 /* vfslide1up.vf vd, vs2, rs1, vm # vd[0]=f[rs1], vd[i+1] = vs2[i] */
@@ -4784,11 +4784,11 @@ GEN_VEXT_VFSLIDE1UP_VF(vfslide1up_vf_h, 16)
 GEN_VEXT_VFSLIDE1UP_VF(vfslide1up_vf_w, 32)
 GEN_VEXT_VFSLIDE1UP_VF(vfslide1up_vf_d, 64)
 
-#define GEN_VEXT_VFSLIDE1DOWN_VF(NAME, ESZ)                   \
+#define GEN_VEXT_VFSLIDE1DOWN_VF(NAME, BITWIDTH)              \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1, void *vs2, \
                   CPURISCVState *env, uint32_t desc)          \
 {                                                             \
-    vslide1down_##ESZ(vd, v0, s1, vs2, env, desc);            \
+    vslide1down_##BITWIDTH(vd, v0, s1, vs2, env, desc);       \
 }
 
 /* vfslide1down.vf vd, vs2, rs1, vm # vd[i] = vs2[i+1], vd[vl-1]=f[rs1] */
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 03/14] target/riscv: rvv: Early exit when vstart >= vl
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-12  6:28   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-12  6:28 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

According to v-spec (section 5.4):
When vstart ≥ vl, there are no body elements, and no elements are
updated in any destination vector register group, including that
no tail elements are updated with agnostic values.

vmsbf.m, vmsif.m, vmsof.m, viota.m, vcompress instructions themselves
require vstart to be zero. So they don't need the early exit.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 27 +++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 275fded6e4..57953923d5 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -652,6 +652,7 @@ static bool ldst_us_trans(uint32_t vd, uint32_t rs1, uint32_t data,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -818,6 +819,7 @@ static bool ldst_stride_trans(uint32_t vd, uint32_t rs1, uint32_t rs2,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -925,6 +927,7 @@ static bool ldst_index_trans(uint32_t vd, uint32_t rs1, uint32_t vs2,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1067,6 +1070,7 @@ static bool ldff_trans(uint32_t vd, uint32_t rs1, uint32_t data,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1216,6 +1220,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
     }
 
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     if (a->vm && s->vl_eq_vlmax) {
         gvec_fn(s->sew, vreg_ofs(s, a->rd),
@@ -1263,6 +1268,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1427,6 +1433,7 @@ static bool opivi_trans(uint32_t vd, uint32_t imm, uint32_t vs2, uint32_t vm,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1513,6 +1520,7 @@ static bool do_opivv_widen(DisasContext *s, arg_rmrr *a,
         uint32_t data = 0;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -1593,6 +1601,7 @@ static bool do_opiwv_widen(DisasContext *s, arg_rmrr *a,
         uint32_t data = 0;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -1670,6 +1679,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         };                                                         \
         TCGLabel *over = gen_new_label();                          \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -1851,6 +1861,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         };                                                         \
         TCGLabel *over = gen_new_label();                          \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2061,6 +2072,7 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
             };
             TCGLabel *over = gen_new_label();
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+            tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
             tcg_gen_gvec_2_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, a->rs1),
                                cpu_env, s->cfg_ptr->vlen / 8,
@@ -2084,6 +2096,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
         TCGv s1;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         s1 = get_gpr(s, a->rs1, EXT_SIGN);
 
@@ -2139,6 +2152,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
             };
             TCGLabel *over = gen_new_label();
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+            tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
             s1 = tcg_constant_i64(simm);
             dest = tcg_temp_new_ptr();
@@ -2291,6 +2305,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, RISCV_FRM_DYN);                              \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2321,6 +2336,7 @@ static bool opfvf_trans(uint32_t vd, uint32_t rs1, uint32_t vs2,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -2409,6 +2425,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         TCGLabel *over = gen_new_label();                        \
         gen_set_rm(s, RISCV_FRM_DYN);                            \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);        \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);\
                                                                  \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
@@ -2483,6 +2500,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, RISCV_FRM_DYN);                              \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2604,6 +2622,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
         TCGLabel *over = gen_new_label();
         gen_set_rm(s, rm);
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -2717,6 +2736,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
             };
             TCGLabel *over = gen_new_label();
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+            tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
             t1 = tcg_temp_new_i64();
             /* NaN-box f[rs1] */
@@ -2805,6 +2825,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, FRM);                                        \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2856,6 +2877,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, RISCV_FRM_DYN);                              \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
@@ -2921,6 +2943,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, FRM);                                        \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2974,6 +2997,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, FRM);                                        \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
@@ -3061,6 +3085,7 @@ static bool trans_##NAME(DisasContext *s, arg_r *a)                \
         gen_helper_gvec_4_ptr *fn = gen_helper_##NAME;             \
         TCGLabel *over = gen_new_label();                          \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
@@ -3229,6 +3254,7 @@ static bool trans_vid_v(DisasContext *s, arg_vid_v *a)
         uint32_t data = 0;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -3746,6 +3772,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
     gen_helper_gvec_3_ptr *fn;
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     static gen_helper_gvec_3_ptr * const fns[6][4] = {
         {
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 03/14] target/riscv: rvv: Early exit when vstart >= vl
@ 2022-03-12  6:28   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-12  6:28 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

According to v-spec (section 5.4):
When vstart ≥ vl, there are no body elements, and no elements are
updated in any destination vector register group, including that
no tail elements are updated with agnostic values.

vmsbf.m, vmsif.m, vmsof.m, viota.m, vcompress instructions themselves
require vstart to be zero. So they don't need the early exit.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 27 +++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 275fded6e4..57953923d5 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -652,6 +652,7 @@ static bool ldst_us_trans(uint32_t vd, uint32_t rs1, uint32_t data,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -818,6 +819,7 @@ static bool ldst_stride_trans(uint32_t vd, uint32_t rs1, uint32_t rs2,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -925,6 +927,7 @@ static bool ldst_index_trans(uint32_t vd, uint32_t rs1, uint32_t vs2,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1067,6 +1070,7 @@ static bool ldff_trans(uint32_t vd, uint32_t rs1, uint32_t data,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1216,6 +1220,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
     }
 
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     if (a->vm && s->vl_eq_vlmax) {
         gvec_fn(s->sew, vreg_ofs(s, a->rd),
@@ -1263,6 +1268,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1427,6 +1433,7 @@ static bool opivi_trans(uint32_t vd, uint32_t imm, uint32_t vs2, uint32_t vm,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -1513,6 +1520,7 @@ static bool do_opivv_widen(DisasContext *s, arg_rmrr *a,
         uint32_t data = 0;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -1593,6 +1601,7 @@ static bool do_opiwv_widen(DisasContext *s, arg_rmrr *a,
         uint32_t data = 0;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -1670,6 +1679,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         };                                                         \
         TCGLabel *over = gen_new_label();                          \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -1851,6 +1861,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         };                                                         \
         TCGLabel *over = gen_new_label();                          \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2061,6 +2072,7 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
             };
             TCGLabel *over = gen_new_label();
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+            tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
             tcg_gen_gvec_2_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, a->rs1),
                                cpu_env, s->cfg_ptr->vlen / 8,
@@ -2084,6 +2096,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
         TCGv s1;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         s1 = get_gpr(s, a->rs1, EXT_SIGN);
 
@@ -2139,6 +2152,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
             };
             TCGLabel *over = gen_new_label();
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+            tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
             s1 = tcg_constant_i64(simm);
             dest = tcg_temp_new_ptr();
@@ -2291,6 +2305,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, RISCV_FRM_DYN);                              \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2321,6 +2336,7 @@ static bool opfvf_trans(uint32_t vd, uint32_t rs1, uint32_t vs2,
 
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
@@ -2409,6 +2425,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         TCGLabel *over = gen_new_label();                        \
         gen_set_rm(s, RISCV_FRM_DYN);                            \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);        \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);\
                                                                  \
         data = FIELD_DP32(data, VDATA, VM, a->vm);               \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);           \
@@ -2483,6 +2500,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, RISCV_FRM_DYN);                              \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2604,6 +2622,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
         TCGLabel *over = gen_new_label();
         gen_set_rm(s, rm);
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -2717,6 +2736,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
             };
             TCGLabel *over = gen_new_label();
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+            tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
             t1 = tcg_temp_new_i64();
             /* NaN-box f[rs1] */
@@ -2805,6 +2825,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, FRM);                                        \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2856,6 +2877,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, RISCV_FRM_DYN);                              \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
@@ -2921,6 +2943,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, FRM);                                        \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
@@ -2974,6 +2997,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         TCGLabel *over = gen_new_label();                          \
         gen_set_rm(s, FRM);                                        \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
@@ -3061,6 +3085,7 @@ static bool trans_##NAME(DisasContext *s, arg_r *a)                \
         gen_helper_gvec_4_ptr *fn = gen_helper_##NAME;             \
         TCGLabel *over = gen_new_label();                          \
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
                                                                    \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
@@ -3229,6 +3254,7 @@ static bool trans_vid_v(DisasContext *s, arg_vid_v *a)
         uint32_t data = 0;
         TCGLabel *over = gen_new_label();
         tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+        tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
         data = FIELD_DP32(data, VDATA, VM, a->vm);
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
@@ -3746,6 +3772,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
     gen_helper_gvec_3_ptr *fn;
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
 
     static gen_helper_gvec_3_ptr * const fns[6][4] = {
         {
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 01/14] target/riscv: rvv: Prune redundant ESZ, DSZ parameter passed
  2022-03-24 19:03 ` ~eopxd
@ 2022-03-14  7:38   ` ~eopxd
  -1 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-14  7:38 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: WeiWei Li, Frank Chang, eop Chen, Bin Meng, Alistair Francis,
	Palmer Dabbelt

From: eopXD <eop.chen@sifive.com>

No functional change intended in this commit.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 1132 +++++++++++++++++-----------------
 1 file changed, 565 insertions(+), 567 deletions(-)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 3bd4aac9c9..e94caf1a3c 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -710,7 +710,6 @@ RVVCALL(OPIVV2, vsub_vv_d, OP_SSS_D, H8, H8, H8, DO_SUB)
 
 static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       uint32_t esz, uint32_t dsz,
                        opivv2_fn *fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -727,23 +726,23 @@ static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
 }
 
 /* generate the helpers for OPIVV */
-#define GEN_VEXT_VV(NAME, ESZ, DSZ)                       \
+#define GEN_VEXT_VV(NAME)                                 \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
-    do_vext_vv(vd, v0, vs1, vs2, env, desc, ESZ, DSZ,     \
+    do_vext_vv(vd, v0, vs1, vs2, env, desc,               \
                do_##NAME);                                \
 }
 
-GEN_VEXT_VV(vadd_vv_b, 1, 1)
-GEN_VEXT_VV(vadd_vv_h, 2, 2)
-GEN_VEXT_VV(vadd_vv_w, 4, 4)
-GEN_VEXT_VV(vadd_vv_d, 8, 8)
-GEN_VEXT_VV(vsub_vv_b, 1, 1)
-GEN_VEXT_VV(vsub_vv_h, 2, 2)
-GEN_VEXT_VV(vsub_vv_w, 4, 4)
-GEN_VEXT_VV(vsub_vv_d, 8, 8)
+GEN_VEXT_VV(vadd_vv_b)
+GEN_VEXT_VV(vadd_vv_h)
+GEN_VEXT_VV(vadd_vv_w)
+GEN_VEXT_VV(vadd_vv_d)
+GEN_VEXT_VV(vsub_vv_b)
+GEN_VEXT_VV(vsub_vv_h)
+GEN_VEXT_VV(vsub_vv_w)
+GEN_VEXT_VV(vsub_vv_d)
 
 typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
 
@@ -773,7 +772,6 @@ RVVCALL(OPIVX2, vrsub_vx_d, OP_SSS_D, H8, H8, DO_RSUB)
 
 static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       uint32_t esz, uint32_t dsz,
                        opivx2_fn fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -790,27 +788,27 @@ static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
 }
 
 /* generate the helpers for OPIVX */
-#define GEN_VEXT_VX(NAME, ESZ, DSZ)                       \
+#define GEN_VEXT_VX(NAME)                                 \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
-    do_vext_vx(vd, v0, s1, vs2, env, desc, ESZ, DSZ,      \
+    do_vext_vx(vd, v0, s1, vs2, env, desc,                \
                do_##NAME);                                \
 }
 
-GEN_VEXT_VX(vadd_vx_b, 1, 1)
-GEN_VEXT_VX(vadd_vx_h, 2, 2)
-GEN_VEXT_VX(vadd_vx_w, 4, 4)
-GEN_VEXT_VX(vadd_vx_d, 8, 8)
-GEN_VEXT_VX(vsub_vx_b, 1, 1)
-GEN_VEXT_VX(vsub_vx_h, 2, 2)
-GEN_VEXT_VX(vsub_vx_w, 4, 4)
-GEN_VEXT_VX(vsub_vx_d, 8, 8)
-GEN_VEXT_VX(vrsub_vx_b, 1, 1)
-GEN_VEXT_VX(vrsub_vx_h, 2, 2)
-GEN_VEXT_VX(vrsub_vx_w, 4, 4)
-GEN_VEXT_VX(vrsub_vx_d, 8, 8)
+GEN_VEXT_VX(vadd_vx_b)
+GEN_VEXT_VX(vadd_vx_h)
+GEN_VEXT_VX(vadd_vx_w)
+GEN_VEXT_VX(vadd_vx_d)
+GEN_VEXT_VX(vsub_vx_b)
+GEN_VEXT_VX(vsub_vx_h)
+GEN_VEXT_VX(vsub_vx_w)
+GEN_VEXT_VX(vsub_vx_d)
+GEN_VEXT_VX(vrsub_vx_b)
+GEN_VEXT_VX(vrsub_vx_h)
+GEN_VEXT_VX(vrsub_vx_w)
+GEN_VEXT_VX(vrsub_vx_d)
 
 void HELPER(vec_rsubs8)(void *d, void *a, uint64_t b, uint32_t desc)
 {
@@ -889,30 +887,30 @@ RVVCALL(OPIVV2, vwadd_wv_w, WOP_WSSS_W, H8, H4, H4, DO_ADD)
 RVVCALL(OPIVV2, vwsub_wv_b, WOP_WSSS_B, H2, H1, H1, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_h, WOP_WSSS_H, H4, H2, H2, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_w, WOP_WSSS_W, H8, H4, H4, DO_SUB)
-GEN_VEXT_VV(vwaddu_vv_b, 1, 2)
-GEN_VEXT_VV(vwaddu_vv_h, 2, 4)
-GEN_VEXT_VV(vwaddu_vv_w, 4, 8)
-GEN_VEXT_VV(vwsubu_vv_b, 1, 2)
-GEN_VEXT_VV(vwsubu_vv_h, 2, 4)
-GEN_VEXT_VV(vwsubu_vv_w, 4, 8)
-GEN_VEXT_VV(vwadd_vv_b, 1, 2)
-GEN_VEXT_VV(vwadd_vv_h, 2, 4)
-GEN_VEXT_VV(vwadd_vv_w, 4, 8)
-GEN_VEXT_VV(vwsub_vv_b, 1, 2)
-GEN_VEXT_VV(vwsub_vv_h, 2, 4)
-GEN_VEXT_VV(vwsub_vv_w, 4, 8)
-GEN_VEXT_VV(vwaddu_wv_b, 1, 2)
-GEN_VEXT_VV(vwaddu_wv_h, 2, 4)
-GEN_VEXT_VV(vwaddu_wv_w, 4, 8)
-GEN_VEXT_VV(vwsubu_wv_b, 1, 2)
-GEN_VEXT_VV(vwsubu_wv_h, 2, 4)
-GEN_VEXT_VV(vwsubu_wv_w, 4, 8)
-GEN_VEXT_VV(vwadd_wv_b, 1, 2)
-GEN_VEXT_VV(vwadd_wv_h, 2, 4)
-GEN_VEXT_VV(vwadd_wv_w, 4, 8)
-GEN_VEXT_VV(vwsub_wv_b, 1, 2)
-GEN_VEXT_VV(vwsub_wv_h, 2, 4)
-GEN_VEXT_VV(vwsub_wv_w, 4, 8)
+GEN_VEXT_VV(vwaddu_vv_b)
+GEN_VEXT_VV(vwaddu_vv_h)
+GEN_VEXT_VV(vwaddu_vv_w)
+GEN_VEXT_VV(vwsubu_vv_b)
+GEN_VEXT_VV(vwsubu_vv_h)
+GEN_VEXT_VV(vwsubu_vv_w)
+GEN_VEXT_VV(vwadd_vv_b)
+GEN_VEXT_VV(vwadd_vv_h)
+GEN_VEXT_VV(vwadd_vv_w)
+GEN_VEXT_VV(vwsub_vv_b)
+GEN_VEXT_VV(vwsub_vv_h)
+GEN_VEXT_VV(vwsub_vv_w)
+GEN_VEXT_VV(vwaddu_wv_b)
+GEN_VEXT_VV(vwaddu_wv_h)
+GEN_VEXT_VV(vwaddu_wv_w)
+GEN_VEXT_VV(vwsubu_wv_b)
+GEN_VEXT_VV(vwsubu_wv_h)
+GEN_VEXT_VV(vwsubu_wv_w)
+GEN_VEXT_VV(vwadd_wv_b)
+GEN_VEXT_VV(vwadd_wv_h)
+GEN_VEXT_VV(vwadd_wv_w)
+GEN_VEXT_VV(vwsub_wv_b)
+GEN_VEXT_VV(vwsub_wv_h)
+GEN_VEXT_VV(vwsub_wv_w)
 
 RVVCALL(OPIVX2, vwaddu_vx_b, WOP_UUU_B, H2, H1, DO_ADD)
 RVVCALL(OPIVX2, vwaddu_vx_h, WOP_UUU_H, H4, H2, DO_ADD)
@@ -938,30 +936,30 @@ RVVCALL(OPIVX2, vwadd_wx_w, WOP_WSSS_W, H8, H4, DO_ADD)
 RVVCALL(OPIVX2, vwsub_wx_b, WOP_WSSS_B, H2, H1, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_h, WOP_WSSS_H, H4, H2, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_w, WOP_WSSS_W, H8, H4, DO_SUB)
-GEN_VEXT_VX(vwaddu_vx_b, 1, 2)
-GEN_VEXT_VX(vwaddu_vx_h, 2, 4)
-GEN_VEXT_VX(vwaddu_vx_w, 4, 8)
-GEN_VEXT_VX(vwsubu_vx_b, 1, 2)
-GEN_VEXT_VX(vwsubu_vx_h, 2, 4)
-GEN_VEXT_VX(vwsubu_vx_w, 4, 8)
-GEN_VEXT_VX(vwadd_vx_b, 1, 2)
-GEN_VEXT_VX(vwadd_vx_h, 2, 4)
-GEN_VEXT_VX(vwadd_vx_w, 4, 8)
-GEN_VEXT_VX(vwsub_vx_b, 1, 2)
-GEN_VEXT_VX(vwsub_vx_h, 2, 4)
-GEN_VEXT_VX(vwsub_vx_w, 4, 8)
-GEN_VEXT_VX(vwaddu_wx_b, 1, 2)
-GEN_VEXT_VX(vwaddu_wx_h, 2, 4)
-GEN_VEXT_VX(vwaddu_wx_w, 4, 8)
-GEN_VEXT_VX(vwsubu_wx_b, 1, 2)
-GEN_VEXT_VX(vwsubu_wx_h, 2, 4)
-GEN_VEXT_VX(vwsubu_wx_w, 4, 8)
-GEN_VEXT_VX(vwadd_wx_b, 1, 2)
-GEN_VEXT_VX(vwadd_wx_h, 2, 4)
-GEN_VEXT_VX(vwadd_wx_w, 4, 8)
-GEN_VEXT_VX(vwsub_wx_b, 1, 2)
-GEN_VEXT_VX(vwsub_wx_h, 2, 4)
-GEN_VEXT_VX(vwsub_wx_w, 4, 8)
+GEN_VEXT_VX(vwaddu_vx_b)
+GEN_VEXT_VX(vwaddu_vx_h)
+GEN_VEXT_VX(vwaddu_vx_w)
+GEN_VEXT_VX(vwsubu_vx_b)
+GEN_VEXT_VX(vwsubu_vx_h)
+GEN_VEXT_VX(vwsubu_vx_w)
+GEN_VEXT_VX(vwadd_vx_b)
+GEN_VEXT_VX(vwadd_vx_h)
+GEN_VEXT_VX(vwadd_vx_w)
+GEN_VEXT_VX(vwsub_vx_b)
+GEN_VEXT_VX(vwsub_vx_h)
+GEN_VEXT_VX(vwsub_vx_w)
+GEN_VEXT_VX(vwaddu_wx_b)
+GEN_VEXT_VX(vwaddu_wx_h)
+GEN_VEXT_VX(vwaddu_wx_w)
+GEN_VEXT_VX(vwsubu_wx_b)
+GEN_VEXT_VX(vwsubu_wx_h)
+GEN_VEXT_VX(vwsubu_wx_w)
+GEN_VEXT_VX(vwadd_wx_b)
+GEN_VEXT_VX(vwadd_wx_h)
+GEN_VEXT_VX(vwadd_wx_w)
+GEN_VEXT_VX(vwsub_wx_b)
+GEN_VEXT_VX(vwsub_wx_h)
+GEN_VEXT_VX(vwsub_wx_w)
 
 /* Vector Integer Add-with-Carry / Subtract-with-Borrow Instructions */
 #define DO_VADC(N, M, C) (N + M + C)
@@ -1091,18 +1089,18 @@ RVVCALL(OPIVV2, vxor_vv_b, OP_SSS_B, H1, H1, H1, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_h, OP_SSS_H, H2, H2, H2, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_w, OP_SSS_W, H4, H4, H4, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_d, OP_SSS_D, H8, H8, H8, DO_XOR)
-GEN_VEXT_VV(vand_vv_b, 1, 1)
-GEN_VEXT_VV(vand_vv_h, 2, 2)
-GEN_VEXT_VV(vand_vv_w, 4, 4)
-GEN_VEXT_VV(vand_vv_d, 8, 8)
-GEN_VEXT_VV(vor_vv_b, 1, 1)
-GEN_VEXT_VV(vor_vv_h, 2, 2)
-GEN_VEXT_VV(vor_vv_w, 4, 4)
-GEN_VEXT_VV(vor_vv_d, 8, 8)
-GEN_VEXT_VV(vxor_vv_b, 1, 1)
-GEN_VEXT_VV(vxor_vv_h, 2, 2)
-GEN_VEXT_VV(vxor_vv_w, 4, 4)
-GEN_VEXT_VV(vxor_vv_d, 8, 8)
+GEN_VEXT_VV(vand_vv_b)
+GEN_VEXT_VV(vand_vv_h)
+GEN_VEXT_VV(vand_vv_w)
+GEN_VEXT_VV(vand_vv_d)
+GEN_VEXT_VV(vor_vv_b)
+GEN_VEXT_VV(vor_vv_h)
+GEN_VEXT_VV(vor_vv_w)
+GEN_VEXT_VV(vor_vv_d)
+GEN_VEXT_VV(vxor_vv_b)
+GEN_VEXT_VV(vxor_vv_h)
+GEN_VEXT_VV(vxor_vv_w)
+GEN_VEXT_VV(vxor_vv_d)
 
 RVVCALL(OPIVX2, vand_vx_b, OP_SSS_B, H1, H1, DO_AND)
 RVVCALL(OPIVX2, vand_vx_h, OP_SSS_H, H2, H2, DO_AND)
@@ -1116,18 +1114,18 @@ RVVCALL(OPIVX2, vxor_vx_b, OP_SSS_B, H1, H1, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_h, OP_SSS_H, H2, H2, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_w, OP_SSS_W, H4, H4, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_d, OP_SSS_D, H8, H8, DO_XOR)
-GEN_VEXT_VX(vand_vx_b, 1, 1)
-GEN_VEXT_VX(vand_vx_h, 2, 2)
-GEN_VEXT_VX(vand_vx_w, 4, 4)
-GEN_VEXT_VX(vand_vx_d, 8, 8)
-GEN_VEXT_VX(vor_vx_b, 1, 1)
-GEN_VEXT_VX(vor_vx_h, 2, 2)
-GEN_VEXT_VX(vor_vx_w, 4, 4)
-GEN_VEXT_VX(vor_vx_d, 8, 8)
-GEN_VEXT_VX(vxor_vx_b, 1, 1)
-GEN_VEXT_VX(vxor_vx_h, 2, 2)
-GEN_VEXT_VX(vxor_vx_w, 4, 4)
-GEN_VEXT_VX(vxor_vx_d, 8, 8)
+GEN_VEXT_VX(vand_vx_b)
+GEN_VEXT_VX(vand_vx_h)
+GEN_VEXT_VX(vand_vx_w)
+GEN_VEXT_VX(vand_vx_d)
+GEN_VEXT_VX(vor_vx_b)
+GEN_VEXT_VX(vor_vx_h)
+GEN_VEXT_VX(vor_vx_w)
+GEN_VEXT_VX(vor_vx_d)
+GEN_VEXT_VX(vxor_vx_b)
+GEN_VEXT_VX(vxor_vx_h)
+GEN_VEXT_VX(vxor_vx_w)
+GEN_VEXT_VX(vxor_vx_d)
 
 /* Vector Single-Width Bit Shift Instructions */
 #define DO_SLL(N, M)  (N << (M))
@@ -1348,22 +1346,22 @@ RVVCALL(OPIVV2, vmax_vv_b, OP_SSS_B, H1, H1, H1, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_h, OP_SSS_H, H2, H2, H2, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_w, OP_SSS_W, H4, H4, H4, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_d, OP_SSS_D, H8, H8, H8, DO_MAX)
-GEN_VEXT_VV(vminu_vv_b, 1, 1)
-GEN_VEXT_VV(vminu_vv_h, 2, 2)
-GEN_VEXT_VV(vminu_vv_w, 4, 4)
-GEN_VEXT_VV(vminu_vv_d, 8, 8)
-GEN_VEXT_VV(vmin_vv_b, 1, 1)
-GEN_VEXT_VV(vmin_vv_h, 2, 2)
-GEN_VEXT_VV(vmin_vv_w, 4, 4)
-GEN_VEXT_VV(vmin_vv_d, 8, 8)
-GEN_VEXT_VV(vmaxu_vv_b, 1, 1)
-GEN_VEXT_VV(vmaxu_vv_h, 2, 2)
-GEN_VEXT_VV(vmaxu_vv_w, 4, 4)
-GEN_VEXT_VV(vmaxu_vv_d, 8, 8)
-GEN_VEXT_VV(vmax_vv_b, 1, 1)
-GEN_VEXT_VV(vmax_vv_h, 2, 2)
-GEN_VEXT_VV(vmax_vv_w, 4, 4)
-GEN_VEXT_VV(vmax_vv_d, 8, 8)
+GEN_VEXT_VV(vminu_vv_b)
+GEN_VEXT_VV(vminu_vv_h)
+GEN_VEXT_VV(vminu_vv_w)
+GEN_VEXT_VV(vminu_vv_d)
+GEN_VEXT_VV(vmin_vv_b)
+GEN_VEXT_VV(vmin_vv_h)
+GEN_VEXT_VV(vmin_vv_w)
+GEN_VEXT_VV(vmin_vv_d)
+GEN_VEXT_VV(vmaxu_vv_b)
+GEN_VEXT_VV(vmaxu_vv_h)
+GEN_VEXT_VV(vmaxu_vv_w)
+GEN_VEXT_VV(vmaxu_vv_d)
+GEN_VEXT_VV(vmax_vv_b)
+GEN_VEXT_VV(vmax_vv_h)
+GEN_VEXT_VV(vmax_vv_w)
+GEN_VEXT_VV(vmax_vv_d)
 
 RVVCALL(OPIVX2, vminu_vx_b, OP_UUU_B, H1, H1, DO_MIN)
 RVVCALL(OPIVX2, vminu_vx_h, OP_UUU_H, H2, H2, DO_MIN)
@@ -1381,22 +1379,22 @@ RVVCALL(OPIVX2, vmax_vx_b, OP_SSS_B, H1, H1, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_h, OP_SSS_H, H2, H2, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_w, OP_SSS_W, H4, H4, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_d, OP_SSS_D, H8, H8, DO_MAX)
-GEN_VEXT_VX(vminu_vx_b, 1, 1)
-GEN_VEXT_VX(vminu_vx_h, 2, 2)
-GEN_VEXT_VX(vminu_vx_w, 4, 4)
-GEN_VEXT_VX(vminu_vx_d, 8, 8)
-GEN_VEXT_VX(vmin_vx_b, 1, 1)
-GEN_VEXT_VX(vmin_vx_h, 2, 2)
-GEN_VEXT_VX(vmin_vx_w, 4, 4)
-GEN_VEXT_VX(vmin_vx_d, 8, 8)
-GEN_VEXT_VX(vmaxu_vx_b, 1, 1)
-GEN_VEXT_VX(vmaxu_vx_h, 2, 2)
-GEN_VEXT_VX(vmaxu_vx_w, 4, 4)
-GEN_VEXT_VX(vmaxu_vx_d, 8, 8)
-GEN_VEXT_VX(vmax_vx_b, 1, 1)
-GEN_VEXT_VX(vmax_vx_h, 2, 2)
-GEN_VEXT_VX(vmax_vx_w, 4, 4)
-GEN_VEXT_VX(vmax_vx_d, 8, 8)
+GEN_VEXT_VX(vminu_vx_b)
+GEN_VEXT_VX(vminu_vx_h)
+GEN_VEXT_VX(vminu_vx_w)
+GEN_VEXT_VX(vminu_vx_d)
+GEN_VEXT_VX(vmin_vx_b)
+GEN_VEXT_VX(vmin_vx_h)
+GEN_VEXT_VX(vmin_vx_w)
+GEN_VEXT_VX(vmin_vx_d)
+GEN_VEXT_VX(vmaxu_vx_b)
+GEN_VEXT_VX(vmaxu_vx_h)
+GEN_VEXT_VX(vmaxu_vx_w)
+GEN_VEXT_VX(vmaxu_vx_d)
+GEN_VEXT_VX(vmax_vx_b)
+GEN_VEXT_VX(vmax_vx_h)
+GEN_VEXT_VX(vmax_vx_w)
+GEN_VEXT_VX(vmax_vx_d)
 
 /* Vector Single-Width Integer Multiply Instructions */
 #define DO_MUL(N, M) (N * M)
@@ -1404,10 +1402,10 @@ RVVCALL(OPIVV2, vmul_vv_b, OP_SSS_B, H1, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_h, OP_SSS_H, H2, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_w, OP_SSS_W, H4, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_d, OP_SSS_D, H8, H8, H8, DO_MUL)
-GEN_VEXT_VV(vmul_vv_b, 1, 1)
-GEN_VEXT_VV(vmul_vv_h, 2, 2)
-GEN_VEXT_VV(vmul_vv_w, 4, 4)
-GEN_VEXT_VV(vmul_vv_d, 8, 8)
+GEN_VEXT_VV(vmul_vv_b)
+GEN_VEXT_VV(vmul_vv_h)
+GEN_VEXT_VV(vmul_vv_w)
+GEN_VEXT_VV(vmul_vv_d)
 
 static int8_t do_mulh_b(int8_t s2, int8_t s1)
 {
@@ -1511,18 +1509,18 @@ RVVCALL(OPIVV2, vmulhsu_vv_b, OP_SUS_B, H1, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVV2, vmulhsu_vv_h, OP_SUS_H, H2, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVV2, vmulhsu_vv_w, OP_SUS_W, H4, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVV2, vmulhsu_vv_d, OP_SUS_D, H8, H8, H8, do_mulhsu_d)
-GEN_VEXT_VV(vmulh_vv_b, 1, 1)
-GEN_VEXT_VV(vmulh_vv_h, 2, 2)
-GEN_VEXT_VV(vmulh_vv_w, 4, 4)
-GEN_VEXT_VV(vmulh_vv_d, 8, 8)
-GEN_VEXT_VV(vmulhu_vv_b, 1, 1)
-GEN_VEXT_VV(vmulhu_vv_h, 2, 2)
-GEN_VEXT_VV(vmulhu_vv_w, 4, 4)
-GEN_VEXT_VV(vmulhu_vv_d, 8, 8)
-GEN_VEXT_VV(vmulhsu_vv_b, 1, 1)
-GEN_VEXT_VV(vmulhsu_vv_h, 2, 2)
-GEN_VEXT_VV(vmulhsu_vv_w, 4, 4)
-GEN_VEXT_VV(vmulhsu_vv_d, 8, 8)
+GEN_VEXT_VV(vmulh_vv_b)
+GEN_VEXT_VV(vmulh_vv_h)
+GEN_VEXT_VV(vmulh_vv_w)
+GEN_VEXT_VV(vmulh_vv_d)
+GEN_VEXT_VV(vmulhu_vv_b)
+GEN_VEXT_VV(vmulhu_vv_h)
+GEN_VEXT_VV(vmulhu_vv_w)
+GEN_VEXT_VV(vmulhu_vv_d)
+GEN_VEXT_VV(vmulhsu_vv_b)
+GEN_VEXT_VV(vmulhsu_vv_h)
+GEN_VEXT_VV(vmulhsu_vv_w)
+GEN_VEXT_VV(vmulhsu_vv_d)
 
 RVVCALL(OPIVX2, vmul_vx_b, OP_SSS_B, H1, H1, DO_MUL)
 RVVCALL(OPIVX2, vmul_vx_h, OP_SSS_H, H2, H2, DO_MUL)
@@ -1540,22 +1538,22 @@ RVVCALL(OPIVX2, vmulhsu_vx_b, OP_SUS_B, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVX2, vmulhsu_vx_h, OP_SUS_H, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVX2, vmulhsu_vx_w, OP_SUS_W, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVX2, vmulhsu_vx_d, OP_SUS_D, H8, H8, do_mulhsu_d)
-GEN_VEXT_VX(vmul_vx_b, 1, 1)
-GEN_VEXT_VX(vmul_vx_h, 2, 2)
-GEN_VEXT_VX(vmul_vx_w, 4, 4)
-GEN_VEXT_VX(vmul_vx_d, 8, 8)
-GEN_VEXT_VX(vmulh_vx_b, 1, 1)
-GEN_VEXT_VX(vmulh_vx_h, 2, 2)
-GEN_VEXT_VX(vmulh_vx_w, 4, 4)
-GEN_VEXT_VX(vmulh_vx_d, 8, 8)
-GEN_VEXT_VX(vmulhu_vx_b, 1, 1)
-GEN_VEXT_VX(vmulhu_vx_h, 2, 2)
-GEN_VEXT_VX(vmulhu_vx_w, 4, 4)
-GEN_VEXT_VX(vmulhu_vx_d, 8, 8)
-GEN_VEXT_VX(vmulhsu_vx_b, 1, 1)
-GEN_VEXT_VX(vmulhsu_vx_h, 2, 2)
-GEN_VEXT_VX(vmulhsu_vx_w, 4, 4)
-GEN_VEXT_VX(vmulhsu_vx_d, 8, 8)
+GEN_VEXT_VX(vmul_vx_b)
+GEN_VEXT_VX(vmul_vx_h)
+GEN_VEXT_VX(vmul_vx_w)
+GEN_VEXT_VX(vmul_vx_d)
+GEN_VEXT_VX(vmulh_vx_b)
+GEN_VEXT_VX(vmulh_vx_h)
+GEN_VEXT_VX(vmulh_vx_w)
+GEN_VEXT_VX(vmulh_vx_d)
+GEN_VEXT_VX(vmulhu_vx_b)
+GEN_VEXT_VX(vmulhu_vx_h)
+GEN_VEXT_VX(vmulhu_vx_w)
+GEN_VEXT_VX(vmulhu_vx_d)
+GEN_VEXT_VX(vmulhsu_vx_b)
+GEN_VEXT_VX(vmulhsu_vx_h)
+GEN_VEXT_VX(vmulhsu_vx_w)
+GEN_VEXT_VX(vmulhsu_vx_d)
 
 /* Vector Integer Divide Instructions */
 #define DO_DIVU(N, M) (unlikely(M == 0) ? (__typeof(N))(-1) : N / M)
@@ -1581,22 +1579,22 @@ RVVCALL(OPIVV2, vrem_vv_b, OP_SSS_B, H1, H1, H1, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_h, OP_SSS_H, H2, H2, H2, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_w, OP_SSS_W, H4, H4, H4, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_d, OP_SSS_D, H8, H8, H8, DO_REM)
-GEN_VEXT_VV(vdivu_vv_b, 1, 1)
-GEN_VEXT_VV(vdivu_vv_h, 2, 2)
-GEN_VEXT_VV(vdivu_vv_w, 4, 4)
-GEN_VEXT_VV(vdivu_vv_d, 8, 8)
-GEN_VEXT_VV(vdiv_vv_b, 1, 1)
-GEN_VEXT_VV(vdiv_vv_h, 2, 2)
-GEN_VEXT_VV(vdiv_vv_w, 4, 4)
-GEN_VEXT_VV(vdiv_vv_d, 8, 8)
-GEN_VEXT_VV(vremu_vv_b, 1, 1)
-GEN_VEXT_VV(vremu_vv_h, 2, 2)
-GEN_VEXT_VV(vremu_vv_w, 4, 4)
-GEN_VEXT_VV(vremu_vv_d, 8, 8)
-GEN_VEXT_VV(vrem_vv_b, 1, 1)
-GEN_VEXT_VV(vrem_vv_h, 2, 2)
-GEN_VEXT_VV(vrem_vv_w, 4, 4)
-GEN_VEXT_VV(vrem_vv_d, 8, 8)
+GEN_VEXT_VV(vdivu_vv_b)
+GEN_VEXT_VV(vdivu_vv_h)
+GEN_VEXT_VV(vdivu_vv_w)
+GEN_VEXT_VV(vdivu_vv_d)
+GEN_VEXT_VV(vdiv_vv_b)
+GEN_VEXT_VV(vdiv_vv_h)
+GEN_VEXT_VV(vdiv_vv_w)
+GEN_VEXT_VV(vdiv_vv_d)
+GEN_VEXT_VV(vremu_vv_b)
+GEN_VEXT_VV(vremu_vv_h)
+GEN_VEXT_VV(vremu_vv_w)
+GEN_VEXT_VV(vremu_vv_d)
+GEN_VEXT_VV(vrem_vv_b)
+GEN_VEXT_VV(vrem_vv_h)
+GEN_VEXT_VV(vrem_vv_w)
+GEN_VEXT_VV(vrem_vv_d)
 
 RVVCALL(OPIVX2, vdivu_vx_b, OP_UUU_B, H1, H1, DO_DIVU)
 RVVCALL(OPIVX2, vdivu_vx_h, OP_UUU_H, H2, H2, DO_DIVU)
@@ -1614,22 +1612,22 @@ RVVCALL(OPIVX2, vrem_vx_b, OP_SSS_B, H1, H1, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_h, OP_SSS_H, H2, H2, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_w, OP_SSS_W, H4, H4, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_d, OP_SSS_D, H8, H8, DO_REM)
-GEN_VEXT_VX(vdivu_vx_b, 1, 1)
-GEN_VEXT_VX(vdivu_vx_h, 2, 2)
-GEN_VEXT_VX(vdivu_vx_w, 4, 4)
-GEN_VEXT_VX(vdivu_vx_d, 8, 8)
-GEN_VEXT_VX(vdiv_vx_b, 1, 1)
-GEN_VEXT_VX(vdiv_vx_h, 2, 2)
-GEN_VEXT_VX(vdiv_vx_w, 4, 4)
-GEN_VEXT_VX(vdiv_vx_d, 8, 8)
-GEN_VEXT_VX(vremu_vx_b, 1, 1)
-GEN_VEXT_VX(vremu_vx_h, 2, 2)
-GEN_VEXT_VX(vremu_vx_w, 4, 4)
-GEN_VEXT_VX(vremu_vx_d, 8, 8)
-GEN_VEXT_VX(vrem_vx_b, 1, 1)
-GEN_VEXT_VX(vrem_vx_h, 2, 2)
-GEN_VEXT_VX(vrem_vx_w, 4, 4)
-GEN_VEXT_VX(vrem_vx_d, 8, 8)
+GEN_VEXT_VX(vdivu_vx_b)
+GEN_VEXT_VX(vdivu_vx_h)
+GEN_VEXT_VX(vdivu_vx_w)
+GEN_VEXT_VX(vdivu_vx_d)
+GEN_VEXT_VX(vdiv_vx_b)
+GEN_VEXT_VX(vdiv_vx_h)
+GEN_VEXT_VX(vdiv_vx_w)
+GEN_VEXT_VX(vdiv_vx_d)
+GEN_VEXT_VX(vremu_vx_b)
+GEN_VEXT_VX(vremu_vx_h)
+GEN_VEXT_VX(vremu_vx_w)
+GEN_VEXT_VX(vremu_vx_d)
+GEN_VEXT_VX(vrem_vx_b)
+GEN_VEXT_VX(vrem_vx_h)
+GEN_VEXT_VX(vrem_vx_w)
+GEN_VEXT_VX(vrem_vx_d)
 
 /* Vector Widening Integer Multiply Instructions */
 RVVCALL(OPIVV2, vwmul_vv_b, WOP_SSS_B, H2, H1, H1, DO_MUL)
@@ -1641,15 +1639,15 @@ RVVCALL(OPIVV2, vwmulu_vv_w, WOP_UUU_W, H8, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_b, WOP_SUS_B, H2, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_h, WOP_SUS_H, H4, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_w, WOP_SUS_W, H8, H4, H4, DO_MUL)
-GEN_VEXT_VV(vwmul_vv_b, 1, 2)
-GEN_VEXT_VV(vwmul_vv_h, 2, 4)
-GEN_VEXT_VV(vwmul_vv_w, 4, 8)
-GEN_VEXT_VV(vwmulu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmulu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmulu_vv_w, 4, 8)
-GEN_VEXT_VV(vwmulsu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmulsu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmulsu_vv_w, 4, 8)
+GEN_VEXT_VV(vwmul_vv_b)
+GEN_VEXT_VV(vwmul_vv_h)
+GEN_VEXT_VV(vwmul_vv_w)
+GEN_VEXT_VV(vwmulu_vv_b)
+GEN_VEXT_VV(vwmulu_vv_h)
+GEN_VEXT_VV(vwmulu_vv_w)
+GEN_VEXT_VV(vwmulsu_vv_b)
+GEN_VEXT_VV(vwmulsu_vv_h)
+GEN_VEXT_VV(vwmulsu_vv_w)
 
 RVVCALL(OPIVX2, vwmul_vx_b, WOP_SSS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmul_vx_h, WOP_SSS_H, H4, H2, DO_MUL)
@@ -1660,15 +1658,15 @@ RVVCALL(OPIVX2, vwmulu_vx_w, WOP_UUU_W, H8, H4, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_b, WOP_SUS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_h, WOP_SUS_H, H4, H2, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_w, WOP_SUS_W, H8, H4, DO_MUL)
-GEN_VEXT_VX(vwmul_vx_b, 1, 2)
-GEN_VEXT_VX(vwmul_vx_h, 2, 4)
-GEN_VEXT_VX(vwmul_vx_w, 4, 8)
-GEN_VEXT_VX(vwmulu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmulu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmulu_vx_w, 4, 8)
-GEN_VEXT_VX(vwmulsu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmulsu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmulsu_vx_w, 4, 8)
+GEN_VEXT_VX(vwmul_vx_b)
+GEN_VEXT_VX(vwmul_vx_h)
+GEN_VEXT_VX(vwmul_vx_w)
+GEN_VEXT_VX(vwmulu_vx_b)
+GEN_VEXT_VX(vwmulu_vx_h)
+GEN_VEXT_VX(vwmulu_vx_w)
+GEN_VEXT_VX(vwmulsu_vx_b)
+GEN_VEXT_VX(vwmulsu_vx_h)
+GEN_VEXT_VX(vwmulsu_vx_w)
 
 /* Vector Single-Width Integer Multiply-Add Instructions */
 #define OPIVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)   \
@@ -1700,22 +1698,22 @@ RVVCALL(OPIVV3, vnmsub_vv_b, OP_SSS_B, H1, H1, H1, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_h, OP_SSS_H, H2, H2, H2, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_w, OP_SSS_W, H4, H4, H4, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_d, OP_SSS_D, H8, H8, H8, DO_NMSUB)
-GEN_VEXT_VV(vmacc_vv_b, 1, 1)
-GEN_VEXT_VV(vmacc_vv_h, 2, 2)
-GEN_VEXT_VV(vmacc_vv_w, 4, 4)
-GEN_VEXT_VV(vmacc_vv_d, 8, 8)
-GEN_VEXT_VV(vnmsac_vv_b, 1, 1)
-GEN_VEXT_VV(vnmsac_vv_h, 2, 2)
-GEN_VEXT_VV(vnmsac_vv_w, 4, 4)
-GEN_VEXT_VV(vnmsac_vv_d, 8, 8)
-GEN_VEXT_VV(vmadd_vv_b, 1, 1)
-GEN_VEXT_VV(vmadd_vv_h, 2, 2)
-GEN_VEXT_VV(vmadd_vv_w, 4, 4)
-GEN_VEXT_VV(vmadd_vv_d, 8, 8)
-GEN_VEXT_VV(vnmsub_vv_b, 1, 1)
-GEN_VEXT_VV(vnmsub_vv_h, 2, 2)
-GEN_VEXT_VV(vnmsub_vv_w, 4, 4)
-GEN_VEXT_VV(vnmsub_vv_d, 8, 8)
+GEN_VEXT_VV(vmacc_vv_b)
+GEN_VEXT_VV(vmacc_vv_h)
+GEN_VEXT_VV(vmacc_vv_w)
+GEN_VEXT_VV(vmacc_vv_d)
+GEN_VEXT_VV(vnmsac_vv_b)
+GEN_VEXT_VV(vnmsac_vv_h)
+GEN_VEXT_VV(vnmsac_vv_w)
+GEN_VEXT_VV(vnmsac_vv_d)
+GEN_VEXT_VV(vmadd_vv_b)
+GEN_VEXT_VV(vmadd_vv_h)
+GEN_VEXT_VV(vmadd_vv_w)
+GEN_VEXT_VV(vmadd_vv_d)
+GEN_VEXT_VV(vnmsub_vv_b)
+GEN_VEXT_VV(vnmsub_vv_h)
+GEN_VEXT_VV(vnmsub_vv_w)
+GEN_VEXT_VV(vnmsub_vv_d)
 
 #define OPIVX3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)             \
 static void do_##NAME(void *vd, target_long s1, void *vs2, int i)   \
@@ -1741,22 +1739,22 @@ RVVCALL(OPIVX3, vnmsub_vx_b, OP_SSS_B, H1, H1, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_h, OP_SSS_H, H2, H2, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_w, OP_SSS_W, H4, H4, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_d, OP_SSS_D, H8, H8, DO_NMSUB)
-GEN_VEXT_VX(vmacc_vx_b, 1, 1)
-GEN_VEXT_VX(vmacc_vx_h, 2, 2)
-GEN_VEXT_VX(vmacc_vx_w, 4, 4)
-GEN_VEXT_VX(vmacc_vx_d, 8, 8)
-GEN_VEXT_VX(vnmsac_vx_b, 1, 1)
-GEN_VEXT_VX(vnmsac_vx_h, 2, 2)
-GEN_VEXT_VX(vnmsac_vx_w, 4, 4)
-GEN_VEXT_VX(vnmsac_vx_d, 8, 8)
-GEN_VEXT_VX(vmadd_vx_b, 1, 1)
-GEN_VEXT_VX(vmadd_vx_h, 2, 2)
-GEN_VEXT_VX(vmadd_vx_w, 4, 4)
-GEN_VEXT_VX(vmadd_vx_d, 8, 8)
-GEN_VEXT_VX(vnmsub_vx_b, 1, 1)
-GEN_VEXT_VX(vnmsub_vx_h, 2, 2)
-GEN_VEXT_VX(vnmsub_vx_w, 4, 4)
-GEN_VEXT_VX(vnmsub_vx_d, 8, 8)
+GEN_VEXT_VX(vmacc_vx_b)
+GEN_VEXT_VX(vmacc_vx_h)
+GEN_VEXT_VX(vmacc_vx_w)
+GEN_VEXT_VX(vmacc_vx_d)
+GEN_VEXT_VX(vnmsac_vx_b)
+GEN_VEXT_VX(vnmsac_vx_h)
+GEN_VEXT_VX(vnmsac_vx_w)
+GEN_VEXT_VX(vnmsac_vx_d)
+GEN_VEXT_VX(vmadd_vx_b)
+GEN_VEXT_VX(vmadd_vx_h)
+GEN_VEXT_VX(vmadd_vx_w)
+GEN_VEXT_VX(vmadd_vx_d)
+GEN_VEXT_VX(vnmsub_vx_b)
+GEN_VEXT_VX(vnmsub_vx_h)
+GEN_VEXT_VX(vnmsub_vx_w)
+GEN_VEXT_VX(vnmsub_vx_d)
 
 /* Vector Widening Integer Multiply-Add Instructions */
 RVVCALL(OPIVV3, vwmaccu_vv_b, WOP_UUU_B, H2, H1, H1, DO_MACC)
@@ -1768,15 +1766,15 @@ RVVCALL(OPIVV3, vwmacc_vv_w, WOP_SSS_W, H8, H4, H4, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_b, WOP_SSU_B, H2, H1, H1, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_h, WOP_SSU_H, H4, H2, H2, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_w, WOP_SSU_W, H8, H4, H4, DO_MACC)
-GEN_VEXT_VV(vwmaccu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmaccu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmaccu_vv_w, 4, 8)
-GEN_VEXT_VV(vwmacc_vv_b, 1, 2)
-GEN_VEXT_VV(vwmacc_vv_h, 2, 4)
-GEN_VEXT_VV(vwmacc_vv_w, 4, 8)
-GEN_VEXT_VV(vwmaccsu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmaccsu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmaccsu_vv_w, 4, 8)
+GEN_VEXT_VV(vwmaccu_vv_b)
+GEN_VEXT_VV(vwmaccu_vv_h)
+GEN_VEXT_VV(vwmaccu_vv_w)
+GEN_VEXT_VV(vwmacc_vv_b)
+GEN_VEXT_VV(vwmacc_vv_h)
+GEN_VEXT_VV(vwmacc_vv_w)
+GEN_VEXT_VV(vwmaccsu_vv_b)
+GEN_VEXT_VV(vwmaccsu_vv_h)
+GEN_VEXT_VV(vwmaccsu_vv_w)
 
 RVVCALL(OPIVX3, vwmaccu_vx_b, WOP_UUU_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccu_vx_h, WOP_UUU_H, H4, H2, DO_MACC)
@@ -1790,18 +1788,18 @@ RVVCALL(OPIVX3, vwmaccsu_vx_w, WOP_SSU_W, H8, H4, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_b, WOP_SUS_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_h, WOP_SUS_H, H4, H2, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_w, WOP_SUS_W, H8, H4, DO_MACC)
-GEN_VEXT_VX(vwmaccu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmaccu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmaccu_vx_w, 4, 8)
-GEN_VEXT_VX(vwmacc_vx_b, 1, 2)
-GEN_VEXT_VX(vwmacc_vx_h, 2, 4)
-GEN_VEXT_VX(vwmacc_vx_w, 4, 8)
-GEN_VEXT_VX(vwmaccsu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmaccsu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmaccsu_vx_w, 4, 8)
-GEN_VEXT_VX(vwmaccus_vx_b, 1, 2)
-GEN_VEXT_VX(vwmaccus_vx_h, 2, 4)
-GEN_VEXT_VX(vwmaccus_vx_w, 4, 8)
+GEN_VEXT_VX(vwmaccu_vx_b)
+GEN_VEXT_VX(vwmaccu_vx_h)
+GEN_VEXT_VX(vwmaccu_vx_w)
+GEN_VEXT_VX(vwmacc_vx_b)
+GEN_VEXT_VX(vwmacc_vx_h)
+GEN_VEXT_VX(vwmacc_vx_w)
+GEN_VEXT_VX(vwmaccsu_vx_b)
+GEN_VEXT_VX(vwmaccsu_vx_h)
+GEN_VEXT_VX(vwmaccsu_vx_w)
+GEN_VEXT_VX(vwmaccus_vx_b)
+GEN_VEXT_VX(vwmaccus_vx_h)
+GEN_VEXT_VX(vwmaccus_vx_w)
 
 /* Vector Integer Merge and Move Instructions */
 #define GEN_VEXT_VMV_VV(NAME, ETYPE, H)                              \
@@ -1922,7 +1920,7 @@ vext_vv_rm_1(void *vd, void *v0, void *vs1, void *vs2,
 static inline void
 vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
              CPURISCVState *env,
-             uint32_t desc, uint32_t esz, uint32_t dsz,
+             uint32_t desc,
              opivv2_rm_fn *fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -1949,11 +1947,11 @@ vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
 }
 
 /* generate helpers for fixed point instructions with OPIVV format */
-#define GEN_VEXT_VV_RM(NAME, ESZ, DSZ)                          \
+#define GEN_VEXT_VV_RM(NAME)                                    \
 void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,     \
                   CPURISCVState *env, uint32_t desc)            \
 {                                                               \
-    vext_vv_rm_2(vd, v0, vs1, vs2, env, desc, ESZ, DSZ,         \
+    vext_vv_rm_2(vd, v0, vs1, vs2, env, desc,                   \
                  do_##NAME);                                    \
 }
 
@@ -2004,10 +2002,10 @@ RVVCALL(OPIVV2_RM, vsaddu_vv_b, OP_UUU_B, H1, H1, H1, saddu8)
 RVVCALL(OPIVV2_RM, vsaddu_vv_h, OP_UUU_H, H2, H2, H2, saddu16)
 RVVCALL(OPIVV2_RM, vsaddu_vv_w, OP_UUU_W, H4, H4, H4, saddu32)
 RVVCALL(OPIVV2_RM, vsaddu_vv_d, OP_UUU_D, H8, H8, H8, saddu64)
-GEN_VEXT_VV_RM(vsaddu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vsaddu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vsaddu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vsaddu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vsaddu_vv_b)
+GEN_VEXT_VV_RM(vsaddu_vv_h)
+GEN_VEXT_VV_RM(vsaddu_vv_w)
+GEN_VEXT_VV_RM(vsaddu_vv_d)
 
 typedef void opivx2_rm_fn(void *vd, target_long s1, void *vs2, int i,
                           CPURISCVState *env, int vxrm);
@@ -2039,7 +2037,7 @@ vext_vx_rm_1(void *vd, void *v0, target_long s1, void *vs2,
 static inline void
 vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
              CPURISCVState *env,
-             uint32_t desc, uint32_t esz, uint32_t dsz,
+             uint32_t desc,
              opivx2_rm_fn *fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -2066,11 +2064,11 @@ vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
 }
 
 /* generate helpers for fixed point instructions with OPIVX format */
-#define GEN_VEXT_VX_RM(NAME, ESZ, DSZ)                    \
+#define GEN_VEXT_VX_RM(NAME)                              \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
         void *vs2, CPURISCVState *env, uint32_t desc)     \
 {                                                         \
-    vext_vx_rm_2(vd, v0, s1, vs2, env, desc, ESZ, DSZ,    \
+    vext_vx_rm_2(vd, v0, s1, vs2, env, desc,              \
                  do_##NAME);                              \
 }
 
@@ -2078,10 +2076,10 @@ RVVCALL(OPIVX2_RM, vsaddu_vx_b, OP_UUU_B, H1, H1, saddu8)
 RVVCALL(OPIVX2_RM, vsaddu_vx_h, OP_UUU_H, H2, H2, saddu16)
 RVVCALL(OPIVX2_RM, vsaddu_vx_w, OP_UUU_W, H4, H4, saddu32)
 RVVCALL(OPIVX2_RM, vsaddu_vx_d, OP_UUU_D, H8, H8, saddu64)
-GEN_VEXT_VX_RM(vsaddu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vsaddu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vsaddu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vsaddu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vsaddu_vx_b)
+GEN_VEXT_VX_RM(vsaddu_vx_h)
+GEN_VEXT_VX_RM(vsaddu_vx_w)
+GEN_VEXT_VX_RM(vsaddu_vx_d)
 
 static inline int8_t sadd8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2127,19 +2125,19 @@ RVVCALL(OPIVV2_RM, vsadd_vv_b, OP_SSS_B, H1, H1, H1, sadd8)
 RVVCALL(OPIVV2_RM, vsadd_vv_h, OP_SSS_H, H2, H2, H2, sadd16)
 RVVCALL(OPIVV2_RM, vsadd_vv_w, OP_SSS_W, H4, H4, H4, sadd32)
 RVVCALL(OPIVV2_RM, vsadd_vv_d, OP_SSS_D, H8, H8, H8, sadd64)
-GEN_VEXT_VV_RM(vsadd_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vsadd_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vsadd_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vsadd_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vsadd_vv_b)
+GEN_VEXT_VV_RM(vsadd_vv_h)
+GEN_VEXT_VV_RM(vsadd_vv_w)
+GEN_VEXT_VV_RM(vsadd_vv_d)
 
 RVVCALL(OPIVX2_RM, vsadd_vx_b, OP_SSS_B, H1, H1, sadd8)
 RVVCALL(OPIVX2_RM, vsadd_vx_h, OP_SSS_H, H2, H2, sadd16)
 RVVCALL(OPIVX2_RM, vsadd_vx_w, OP_SSS_W, H4, H4, sadd32)
 RVVCALL(OPIVX2_RM, vsadd_vx_d, OP_SSS_D, H8, H8, sadd64)
-GEN_VEXT_VX_RM(vsadd_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vsadd_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vsadd_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vsadd_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vsadd_vx_b)
+GEN_VEXT_VX_RM(vsadd_vx_h)
+GEN_VEXT_VX_RM(vsadd_vx_w)
+GEN_VEXT_VX_RM(vsadd_vx_d)
 
 static inline uint8_t ssubu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
 {
@@ -2188,19 +2186,19 @@ RVVCALL(OPIVV2_RM, vssubu_vv_b, OP_UUU_B, H1, H1, H1, ssubu8)
 RVVCALL(OPIVV2_RM, vssubu_vv_h, OP_UUU_H, H2, H2, H2, ssubu16)
 RVVCALL(OPIVV2_RM, vssubu_vv_w, OP_UUU_W, H4, H4, H4, ssubu32)
 RVVCALL(OPIVV2_RM, vssubu_vv_d, OP_UUU_D, H8, H8, H8, ssubu64)
-GEN_VEXT_VV_RM(vssubu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssubu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssubu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssubu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssubu_vv_b)
+GEN_VEXT_VV_RM(vssubu_vv_h)
+GEN_VEXT_VV_RM(vssubu_vv_w)
+GEN_VEXT_VV_RM(vssubu_vv_d)
 
 RVVCALL(OPIVX2_RM, vssubu_vx_b, OP_UUU_B, H1, H1, ssubu8)
 RVVCALL(OPIVX2_RM, vssubu_vx_h, OP_UUU_H, H2, H2, ssubu16)
 RVVCALL(OPIVX2_RM, vssubu_vx_w, OP_UUU_W, H4, H4, ssubu32)
 RVVCALL(OPIVX2_RM, vssubu_vx_d, OP_UUU_D, H8, H8, ssubu64)
-GEN_VEXT_VX_RM(vssubu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssubu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssubu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssubu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssubu_vx_b)
+GEN_VEXT_VX_RM(vssubu_vx_h)
+GEN_VEXT_VX_RM(vssubu_vx_w)
+GEN_VEXT_VX_RM(vssubu_vx_d)
 
 static inline int8_t ssub8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2246,19 +2244,19 @@ RVVCALL(OPIVV2_RM, vssub_vv_b, OP_SSS_B, H1, H1, H1, ssub8)
 RVVCALL(OPIVV2_RM, vssub_vv_h, OP_SSS_H, H2, H2, H2, ssub16)
 RVVCALL(OPIVV2_RM, vssub_vv_w, OP_SSS_W, H4, H4, H4, ssub32)
 RVVCALL(OPIVV2_RM, vssub_vv_d, OP_SSS_D, H8, H8, H8, ssub64)
-GEN_VEXT_VV_RM(vssub_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssub_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssub_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssub_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssub_vv_b)
+GEN_VEXT_VV_RM(vssub_vv_h)
+GEN_VEXT_VV_RM(vssub_vv_w)
+GEN_VEXT_VV_RM(vssub_vv_d)
 
 RVVCALL(OPIVX2_RM, vssub_vx_b, OP_SSS_B, H1, H1, ssub8)
 RVVCALL(OPIVX2_RM, vssub_vx_h, OP_SSS_H, H2, H2, ssub16)
 RVVCALL(OPIVX2_RM, vssub_vx_w, OP_SSS_W, H4, H4, ssub32)
 RVVCALL(OPIVX2_RM, vssub_vx_d, OP_SSS_D, H8, H8, ssub64)
-GEN_VEXT_VX_RM(vssub_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssub_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssub_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssub_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssub_vx_b)
+GEN_VEXT_VX_RM(vssub_vx_h)
+GEN_VEXT_VX_RM(vssub_vx_w)
+GEN_VEXT_VX_RM(vssub_vx_d)
 
 /* Vector Single-Width Averaging Add and Subtract */
 static inline uint8_t get_round(int vxrm, uint64_t v, uint8_t shift)
@@ -2310,19 +2308,19 @@ RVVCALL(OPIVV2_RM, vaadd_vv_b, OP_SSS_B, H1, H1, H1, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_h, OP_SSS_H, H2, H2, H2, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_w, OP_SSS_W, H4, H4, H4, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_d, OP_SSS_D, H8, H8, H8, aadd64)
-GEN_VEXT_VV_RM(vaadd_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vaadd_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vaadd_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vaadd_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vaadd_vv_b)
+GEN_VEXT_VV_RM(vaadd_vv_h)
+GEN_VEXT_VV_RM(vaadd_vv_w)
+GEN_VEXT_VV_RM(vaadd_vv_d)
 
 RVVCALL(OPIVX2_RM, vaadd_vx_b, OP_SSS_B, H1, H1, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_h, OP_SSS_H, H2, H2, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_w, OP_SSS_W, H4, H4, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_d, OP_SSS_D, H8, H8, aadd64)
-GEN_VEXT_VX_RM(vaadd_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vaadd_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vaadd_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vaadd_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vaadd_vx_b)
+GEN_VEXT_VX_RM(vaadd_vx_h)
+GEN_VEXT_VX_RM(vaadd_vx_w)
+GEN_VEXT_VX_RM(vaadd_vx_d)
 
 static inline uint32_t aaddu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2347,19 +2345,19 @@ RVVCALL(OPIVV2_RM, vaaddu_vv_b, OP_UUU_B, H1, H1, H1, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_h, OP_UUU_H, H2, H2, H2, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_w, OP_UUU_W, H4, H4, H4, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_d, OP_UUU_D, H8, H8, H8, aaddu64)
-GEN_VEXT_VV_RM(vaaddu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vaaddu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vaaddu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vaaddu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vaaddu_vv_b)
+GEN_VEXT_VV_RM(vaaddu_vv_h)
+GEN_VEXT_VV_RM(vaaddu_vv_w)
+GEN_VEXT_VV_RM(vaaddu_vv_d)
 
 RVVCALL(OPIVX2_RM, vaaddu_vx_b, OP_UUU_B, H1, H1, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_h, OP_UUU_H, H2, H2, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_w, OP_UUU_W, H4, H4, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_d, OP_UUU_D, H8, H8, aaddu64)
-GEN_VEXT_VX_RM(vaaddu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vaaddu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vaaddu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vaaddu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vaaddu_vx_b)
+GEN_VEXT_VX_RM(vaaddu_vx_h)
+GEN_VEXT_VX_RM(vaaddu_vx_w)
+GEN_VEXT_VX_RM(vaaddu_vx_d)
 
 static inline int32_t asub32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
 {
@@ -2383,19 +2381,19 @@ RVVCALL(OPIVV2_RM, vasub_vv_b, OP_SSS_B, H1, H1, H1, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_h, OP_SSS_H, H2, H2, H2, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_w, OP_SSS_W, H4, H4, H4, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_d, OP_SSS_D, H8, H8, H8, asub64)
-GEN_VEXT_VV_RM(vasub_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vasub_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vasub_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vasub_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vasub_vv_b)
+GEN_VEXT_VV_RM(vasub_vv_h)
+GEN_VEXT_VV_RM(vasub_vv_w)
+GEN_VEXT_VV_RM(vasub_vv_d)
 
 RVVCALL(OPIVX2_RM, vasub_vx_b, OP_SSS_B, H1, H1, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_h, OP_SSS_H, H2, H2, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_w, OP_SSS_W, H4, H4, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_d, OP_SSS_D, H8, H8, asub64)
-GEN_VEXT_VX_RM(vasub_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vasub_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vasub_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vasub_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vasub_vx_b)
+GEN_VEXT_VX_RM(vasub_vx_h)
+GEN_VEXT_VX_RM(vasub_vx_w)
+GEN_VEXT_VX_RM(vasub_vx_d)
 
 static inline uint32_t asubu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2420,19 +2418,19 @@ RVVCALL(OPIVV2_RM, vasubu_vv_b, OP_UUU_B, H1, H1, H1, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_h, OP_UUU_H, H2, H2, H2, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_w, OP_UUU_W, H4, H4, H4, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_d, OP_UUU_D, H8, H8, H8, asubu64)
-GEN_VEXT_VV_RM(vasubu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vasubu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vasubu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vasubu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vasubu_vv_b)
+GEN_VEXT_VV_RM(vasubu_vv_h)
+GEN_VEXT_VV_RM(vasubu_vv_w)
+GEN_VEXT_VV_RM(vasubu_vv_d)
 
 RVVCALL(OPIVX2_RM, vasubu_vx_b, OP_UUU_B, H1, H1, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_h, OP_UUU_H, H2, H2, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_w, OP_UUU_W, H4, H4, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_d, OP_UUU_D, H8, H8, asubu64)
-GEN_VEXT_VX_RM(vasubu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vasubu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vasubu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vasubu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vasubu_vx_b)
+GEN_VEXT_VX_RM(vasubu_vx_h)
+GEN_VEXT_VX_RM(vasubu_vx_w)
+GEN_VEXT_VX_RM(vasubu_vx_d)
 
 /* Vector Single-Width Fractional Multiply with Rounding and Saturation */
 static inline int8_t vsmul8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2527,19 +2525,19 @@ RVVCALL(OPIVV2_RM, vsmul_vv_b, OP_SSS_B, H1, H1, H1, vsmul8)
 RVVCALL(OPIVV2_RM, vsmul_vv_h, OP_SSS_H, H2, H2, H2, vsmul16)
 RVVCALL(OPIVV2_RM, vsmul_vv_w, OP_SSS_W, H4, H4, H4, vsmul32)
 RVVCALL(OPIVV2_RM, vsmul_vv_d, OP_SSS_D, H8, H8, H8, vsmul64)
-GEN_VEXT_VV_RM(vsmul_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vsmul_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vsmul_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vsmul_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vsmul_vv_b)
+GEN_VEXT_VV_RM(vsmul_vv_h)
+GEN_VEXT_VV_RM(vsmul_vv_w)
+GEN_VEXT_VV_RM(vsmul_vv_d)
 
 RVVCALL(OPIVX2_RM, vsmul_vx_b, OP_SSS_B, H1, H1, vsmul8)
 RVVCALL(OPIVX2_RM, vsmul_vx_h, OP_SSS_H, H2, H2, vsmul16)
 RVVCALL(OPIVX2_RM, vsmul_vx_w, OP_SSS_W, H4, H4, vsmul32)
 RVVCALL(OPIVX2_RM, vsmul_vx_d, OP_SSS_D, H8, H8, vsmul64)
-GEN_VEXT_VX_RM(vsmul_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vsmul_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vsmul_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vsmul_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vsmul_vx_b)
+GEN_VEXT_VX_RM(vsmul_vx_h)
+GEN_VEXT_VX_RM(vsmul_vx_w)
+GEN_VEXT_VX_RM(vsmul_vx_d)
 
 /* Vector Single-Width Scaling Shift Instructions */
 static inline uint8_t
@@ -2586,19 +2584,19 @@ RVVCALL(OPIVV2_RM, vssrl_vv_b, OP_UUU_B, H1, H1, H1, vssrl8)
 RVVCALL(OPIVV2_RM, vssrl_vv_h, OP_UUU_H, H2, H2, H2, vssrl16)
 RVVCALL(OPIVV2_RM, vssrl_vv_w, OP_UUU_W, H4, H4, H4, vssrl32)
 RVVCALL(OPIVV2_RM, vssrl_vv_d, OP_UUU_D, H8, H8, H8, vssrl64)
-GEN_VEXT_VV_RM(vssrl_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssrl_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssrl_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssrl_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssrl_vv_b)
+GEN_VEXT_VV_RM(vssrl_vv_h)
+GEN_VEXT_VV_RM(vssrl_vv_w)
+GEN_VEXT_VV_RM(vssrl_vv_d)
 
 RVVCALL(OPIVX2_RM, vssrl_vx_b, OP_UUU_B, H1, H1, vssrl8)
 RVVCALL(OPIVX2_RM, vssrl_vx_h, OP_UUU_H, H2, H2, vssrl16)
 RVVCALL(OPIVX2_RM, vssrl_vx_w, OP_UUU_W, H4, H4, vssrl32)
 RVVCALL(OPIVX2_RM, vssrl_vx_d, OP_UUU_D, H8, H8, vssrl64)
-GEN_VEXT_VX_RM(vssrl_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssrl_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssrl_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssrl_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssrl_vx_b)
+GEN_VEXT_VX_RM(vssrl_vx_h)
+GEN_VEXT_VX_RM(vssrl_vx_w)
+GEN_VEXT_VX_RM(vssrl_vx_d)
 
 static inline int8_t
 vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2645,19 +2643,19 @@ RVVCALL(OPIVV2_RM, vssra_vv_b, OP_SSS_B, H1, H1, H1, vssra8)
 RVVCALL(OPIVV2_RM, vssra_vv_h, OP_SSS_H, H2, H2, H2, vssra16)
 RVVCALL(OPIVV2_RM, vssra_vv_w, OP_SSS_W, H4, H4, H4, vssra32)
 RVVCALL(OPIVV2_RM, vssra_vv_d, OP_SSS_D, H8, H8, H8, vssra64)
-GEN_VEXT_VV_RM(vssra_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssra_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssra_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssra_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssra_vv_b)
+GEN_VEXT_VV_RM(vssra_vv_h)
+GEN_VEXT_VV_RM(vssra_vv_w)
+GEN_VEXT_VV_RM(vssra_vv_d)
 
 RVVCALL(OPIVX2_RM, vssra_vx_b, OP_SSS_B, H1, H1, vssra8)
 RVVCALL(OPIVX2_RM, vssra_vx_h, OP_SSS_H, H2, H2, vssra16)
 RVVCALL(OPIVX2_RM, vssra_vx_w, OP_SSS_W, H4, H4, vssra32)
 RVVCALL(OPIVX2_RM, vssra_vx_d, OP_SSS_D, H8, H8, vssra64)
-GEN_VEXT_VX_RM(vssra_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssra_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssra_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssra_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssra_vx_b)
+GEN_VEXT_VX_RM(vssra_vx_h)
+GEN_VEXT_VX_RM(vssra_vx_w)
+GEN_VEXT_VX_RM(vssra_vx_d)
 
 /* Vector Narrowing Fixed-Point Clip Instructions */
 static inline int8_t
@@ -2720,16 +2718,16 @@ vnclip32(CPURISCVState *env, int vxrm, int64_t a, int32_t b)
 RVVCALL(OPIVV2_RM, vnclip_wv_b, NOP_SSS_B, H1, H2, H1, vnclip8)
 RVVCALL(OPIVV2_RM, vnclip_wv_h, NOP_SSS_H, H2, H4, H2, vnclip16)
 RVVCALL(OPIVV2_RM, vnclip_wv_w, NOP_SSS_W, H4, H8, H4, vnclip32)
-GEN_VEXT_VV_RM(vnclip_wv_b, 1, 1)
-GEN_VEXT_VV_RM(vnclip_wv_h, 2, 2)
-GEN_VEXT_VV_RM(vnclip_wv_w, 4, 4)
+GEN_VEXT_VV_RM(vnclip_wv_b)
+GEN_VEXT_VV_RM(vnclip_wv_h)
+GEN_VEXT_VV_RM(vnclip_wv_w)
 
 RVVCALL(OPIVX2_RM, vnclip_wx_b, NOP_SSS_B, H1, H2, vnclip8)
 RVVCALL(OPIVX2_RM, vnclip_wx_h, NOP_SSS_H, H2, H4, vnclip16)
 RVVCALL(OPIVX2_RM, vnclip_wx_w, NOP_SSS_W, H4, H8, vnclip32)
-GEN_VEXT_VX_RM(vnclip_wx_b, 1, 1)
-GEN_VEXT_VX_RM(vnclip_wx_h, 2, 2)
-GEN_VEXT_VX_RM(vnclip_wx_w, 4, 4)
+GEN_VEXT_VX_RM(vnclip_wx_b)
+GEN_VEXT_VX_RM(vnclip_wx_h)
+GEN_VEXT_VX_RM(vnclip_wx_w)
 
 static inline uint8_t
 vnclipu8(CPURISCVState *env, int vxrm, uint16_t a, uint8_t b)
@@ -2782,16 +2780,16 @@ vnclipu32(CPURISCVState *env, int vxrm, uint64_t a, uint32_t b)
 RVVCALL(OPIVV2_RM, vnclipu_wv_b, NOP_UUU_B, H1, H2, H1, vnclipu8)
 RVVCALL(OPIVV2_RM, vnclipu_wv_h, NOP_UUU_H, H2, H4, H2, vnclipu16)
 RVVCALL(OPIVV2_RM, vnclipu_wv_w, NOP_UUU_W, H4, H8, H4, vnclipu32)
-GEN_VEXT_VV_RM(vnclipu_wv_b, 1, 1)
-GEN_VEXT_VV_RM(vnclipu_wv_h, 2, 2)
-GEN_VEXT_VV_RM(vnclipu_wv_w, 4, 4)
+GEN_VEXT_VV_RM(vnclipu_wv_b)
+GEN_VEXT_VV_RM(vnclipu_wv_h)
+GEN_VEXT_VV_RM(vnclipu_wv_w)
 
 RVVCALL(OPIVX2_RM, vnclipu_wx_b, NOP_UUU_B, H1, H2, vnclipu8)
 RVVCALL(OPIVX2_RM, vnclipu_wx_h, NOP_UUU_H, H2, H4, vnclipu16)
 RVVCALL(OPIVX2_RM, vnclipu_wx_w, NOP_UUU_W, H4, H8, vnclipu32)
-GEN_VEXT_VX_RM(vnclipu_wx_b, 1, 1)
-GEN_VEXT_VX_RM(vnclipu_wx_h, 2, 2)
-GEN_VEXT_VX_RM(vnclipu_wx_w, 4, 4)
+GEN_VEXT_VX_RM(vnclipu_wx_b)
+GEN_VEXT_VX_RM(vnclipu_wx_h)
+GEN_VEXT_VX_RM(vnclipu_wx_w)
 
 /*
  *** Vector Float Point Arithmetic Instructions
@@ -2806,7 +2804,7 @@ static void do_##NAME(void *vd, void *vs1, void *vs2, int i,   \
     *((TD *)vd + HD(i)) = OP(s2, s1, &env->fp_status);         \
 }
 
-#define GEN_VEXT_VV_ENV(NAME, ESZ, DSZ)                   \
+#define GEN_VEXT_VV_ENV(NAME)                             \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
@@ -2827,9 +2825,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
 RVVCALL(OPFVV2, vfadd_vv_h, OP_UUU_H, H2, H2, H2, float16_add)
 RVVCALL(OPFVV2, vfadd_vv_w, OP_UUU_W, H4, H4, H4, float32_add)
 RVVCALL(OPFVV2, vfadd_vv_d, OP_UUU_D, H8, H8, H8, float64_add)
-GEN_VEXT_VV_ENV(vfadd_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfadd_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfadd_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfadd_vv_h)
+GEN_VEXT_VV_ENV(vfadd_vv_w)
+GEN_VEXT_VV_ENV(vfadd_vv_d)
 
 #define OPFVF2(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)        \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
@@ -2839,7 +2837,7 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
     *((TD *)vd + HD(i)) = OP(s2, (TX1)(T1)s1, &env->fp_status);\
 }
 
-#define GEN_VEXT_VF(NAME, ESZ, DSZ)                       \
+#define GEN_VEXT_VF(NAME)                                 \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
@@ -2860,22 +2858,22 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
 RVVCALL(OPFVF2, vfadd_vf_h, OP_UUU_H, H2, H2, float16_add)
 RVVCALL(OPFVF2, vfadd_vf_w, OP_UUU_W, H4, H4, float32_add)
 RVVCALL(OPFVF2, vfadd_vf_d, OP_UUU_D, H8, H8, float64_add)
-GEN_VEXT_VF(vfadd_vf_h, 2, 2)
-GEN_VEXT_VF(vfadd_vf_w, 4, 4)
-GEN_VEXT_VF(vfadd_vf_d, 8, 8)
+GEN_VEXT_VF(vfadd_vf_h)
+GEN_VEXT_VF(vfadd_vf_w)
+GEN_VEXT_VF(vfadd_vf_d)
 
 RVVCALL(OPFVV2, vfsub_vv_h, OP_UUU_H, H2, H2, H2, float16_sub)
 RVVCALL(OPFVV2, vfsub_vv_w, OP_UUU_W, H4, H4, H4, float32_sub)
 RVVCALL(OPFVV2, vfsub_vv_d, OP_UUU_D, H8, H8, H8, float64_sub)
-GEN_VEXT_VV_ENV(vfsub_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsub_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsub_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsub_vv_h)
+GEN_VEXT_VV_ENV(vfsub_vv_w)
+GEN_VEXT_VV_ENV(vfsub_vv_d)
 RVVCALL(OPFVF2, vfsub_vf_h, OP_UUU_H, H2, H2, float16_sub)
 RVVCALL(OPFVF2, vfsub_vf_w, OP_UUU_W, H4, H4, float32_sub)
 RVVCALL(OPFVF2, vfsub_vf_d, OP_UUU_D, H8, H8, float64_sub)
-GEN_VEXT_VF(vfsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfsub_vf_h)
+GEN_VEXT_VF(vfsub_vf_w)
+GEN_VEXT_VF(vfsub_vf_d)
 
 static uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s)
 {
@@ -2895,9 +2893,9 @@ static uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrsub_vf_h, OP_UUU_H, H2, H2, float16_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_w, OP_UUU_W, H4, H4, float32_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub)
-GEN_VEXT_VF(vfrsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfrsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfrsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfrsub_vf_h)
+GEN_VEXT_VF(vfrsub_vf_w)
+GEN_VEXT_VF(vfrsub_vf_d)
 
 /* Vector Widening Floating-Point Add/Subtract Instructions */
 static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s)
@@ -2915,12 +2913,12 @@ static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16)
 RVVCALL(OPFVV2, vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32)
-GEN_VEXT_VV_ENV(vfwadd_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwadd_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwadd_vv_h)
+GEN_VEXT_VV_ENV(vfwadd_vv_w)
 RVVCALL(OPFVF2, vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16)
 RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32)
-GEN_VEXT_VF(vfwadd_vf_h, 2, 4)
-GEN_VEXT_VF(vfwadd_vf_w, 4, 8)
+GEN_VEXT_VF(vfwadd_vf_h)
+GEN_VEXT_VF(vfwadd_vf_w)
 
 static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -2937,12 +2935,12 @@ static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16)
 RVVCALL(OPFVV2, vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32)
-GEN_VEXT_VV_ENV(vfwsub_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwsub_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwsub_vv_h)
+GEN_VEXT_VV_ENV(vfwsub_vv_w)
 RVVCALL(OPFVF2, vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16)
 RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32)
-GEN_VEXT_VF(vfwsub_vf_h, 2, 4)
-GEN_VEXT_VF(vfwsub_vf_w, 4, 8)
+GEN_VEXT_VF(vfwsub_vf_h)
+GEN_VEXT_VF(vfwsub_vf_w)
 
 static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -2956,12 +2954,12 @@ static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16)
 RVVCALL(OPFVV2, vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32)
-GEN_VEXT_VV_ENV(vfwadd_wv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwadd_wv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwadd_wv_h)
+GEN_VEXT_VV_ENV(vfwadd_wv_w)
 RVVCALL(OPFVF2, vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16)
 RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32)
-GEN_VEXT_VF(vfwadd_wf_h, 2, 4)
-GEN_VEXT_VF(vfwadd_wf_w, 4, 8)
+GEN_VEXT_VF(vfwadd_wf_h)
+GEN_VEXT_VF(vfwadd_wf_w)
 
 static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -2975,39 +2973,39 @@ static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16)
 RVVCALL(OPFVV2, vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32)
-GEN_VEXT_VV_ENV(vfwsub_wv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwsub_wv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwsub_wv_h)
+GEN_VEXT_VV_ENV(vfwsub_wv_w)
 RVVCALL(OPFVF2, vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16)
 RVVCALL(OPFVF2, vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32)
-GEN_VEXT_VF(vfwsub_wf_h, 2, 4)
-GEN_VEXT_VF(vfwsub_wf_w, 4, 8)
+GEN_VEXT_VF(vfwsub_wf_h)
+GEN_VEXT_VF(vfwsub_wf_w)
 
 /* Vector Single-Width Floating-Point Multiply/Divide Instructions */
 RVVCALL(OPFVV2, vfmul_vv_h, OP_UUU_H, H2, H2, H2, float16_mul)
 RVVCALL(OPFVV2, vfmul_vv_w, OP_UUU_W, H4, H4, H4, float32_mul)
 RVVCALL(OPFVV2, vfmul_vv_d, OP_UUU_D, H8, H8, H8, float64_mul)
-GEN_VEXT_VV_ENV(vfmul_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmul_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmul_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmul_vv_h)
+GEN_VEXT_VV_ENV(vfmul_vv_w)
+GEN_VEXT_VV_ENV(vfmul_vv_d)
 RVVCALL(OPFVF2, vfmul_vf_h, OP_UUU_H, H2, H2, float16_mul)
 RVVCALL(OPFVF2, vfmul_vf_w, OP_UUU_W, H4, H4, float32_mul)
 RVVCALL(OPFVF2, vfmul_vf_d, OP_UUU_D, H8, H8, float64_mul)
-GEN_VEXT_VF(vfmul_vf_h, 2, 2)
-GEN_VEXT_VF(vfmul_vf_w, 4, 4)
-GEN_VEXT_VF(vfmul_vf_d, 8, 8)
+GEN_VEXT_VF(vfmul_vf_h)
+GEN_VEXT_VF(vfmul_vf_w)
+GEN_VEXT_VF(vfmul_vf_d)
 
 RVVCALL(OPFVV2, vfdiv_vv_h, OP_UUU_H, H2, H2, H2, float16_div)
 RVVCALL(OPFVV2, vfdiv_vv_w, OP_UUU_W, H4, H4, H4, float32_div)
 RVVCALL(OPFVV2, vfdiv_vv_d, OP_UUU_D, H8, H8, H8, float64_div)
-GEN_VEXT_VV_ENV(vfdiv_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfdiv_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfdiv_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfdiv_vv_h)
+GEN_VEXT_VV_ENV(vfdiv_vv_w)
+GEN_VEXT_VV_ENV(vfdiv_vv_d)
 RVVCALL(OPFVF2, vfdiv_vf_h, OP_UUU_H, H2, H2, float16_div)
 RVVCALL(OPFVF2, vfdiv_vf_w, OP_UUU_W, H4, H4, float32_div)
 RVVCALL(OPFVF2, vfdiv_vf_d, OP_UUU_D, H8, H8, float64_div)
-GEN_VEXT_VF(vfdiv_vf_h, 2, 2)
-GEN_VEXT_VF(vfdiv_vf_w, 4, 4)
-GEN_VEXT_VF(vfdiv_vf_d, 8, 8)
+GEN_VEXT_VF(vfdiv_vf_h)
+GEN_VEXT_VF(vfdiv_vf_w)
+GEN_VEXT_VF(vfdiv_vf_d)
 
 static uint16_t float16_rdiv(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3027,9 +3025,9 @@ static uint64_t float64_rdiv(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrdiv_vf_h, OP_UUU_H, H2, H2, float16_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_w, OP_UUU_W, H4, H4, float32_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_d, OP_UUU_D, H8, H8, float64_rdiv)
-GEN_VEXT_VF(vfrdiv_vf_h, 2, 2)
-GEN_VEXT_VF(vfrdiv_vf_w, 4, 4)
-GEN_VEXT_VF(vfrdiv_vf_d, 8, 8)
+GEN_VEXT_VF(vfrdiv_vf_h)
+GEN_VEXT_VF(vfrdiv_vf_w)
+GEN_VEXT_VF(vfrdiv_vf_d)
 
 /* Vector Widening Floating-Point Multiply */
 static uint32_t vfwmul16(uint16_t a, uint16_t b, float_status *s)
@@ -3046,12 +3044,12 @@ static uint64_t vfwmul32(uint32_t a, uint32_t b, float_status *s)
 }
 RVVCALL(OPFVV2, vfwmul_vv_h, WOP_UUU_H, H4, H2, H2, vfwmul16)
 RVVCALL(OPFVV2, vfwmul_vv_w, WOP_UUU_W, H8, H4, H4, vfwmul32)
-GEN_VEXT_VV_ENV(vfwmul_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwmul_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwmul_vv_h)
+GEN_VEXT_VV_ENV(vfwmul_vv_w)
 RVVCALL(OPFVF2, vfwmul_vf_h, WOP_UUU_H, H4, H2, vfwmul16)
 RVVCALL(OPFVF2, vfwmul_vf_w, WOP_UUU_W, H8, H4, vfwmul32)
-GEN_VEXT_VF(vfwmul_vf_h, 2, 4)
-GEN_VEXT_VF(vfwmul_vf_w, 4, 8)
+GEN_VEXT_VF(vfwmul_vf_h)
+GEN_VEXT_VF(vfwmul_vf_w)
 
 /* Vector Single-Width Floating-Point Fused Multiply-Add Instructions */
 #define OPFVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)       \
@@ -3082,9 +3080,9 @@ static uint64_t fmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmacc_vv_h, OP_UUU_H, H2, H2, H2, fmacc16)
 RVVCALL(OPFVV3, vfmacc_vv_w, OP_UUU_W, H4, H4, H4, fmacc32)
 RVVCALL(OPFVV3, vfmacc_vv_d, OP_UUU_D, H8, H8, H8, fmacc64)
-GEN_VEXT_VV_ENV(vfmacc_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmacc_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmacc_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmacc_vv_h)
+GEN_VEXT_VV_ENV(vfmacc_vv_w)
+GEN_VEXT_VV_ENV(vfmacc_vv_d)
 
 #define OPFVF3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)           \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
@@ -3098,9 +3096,9 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
 RVVCALL(OPFVF3, vfmacc_vf_h, OP_UUU_H, H2, H2, fmacc16)
 RVVCALL(OPFVF3, vfmacc_vf_w, OP_UUU_W, H4, H4, fmacc32)
 RVVCALL(OPFVF3, vfmacc_vf_d, OP_UUU_D, H8, H8, fmacc64)
-GEN_VEXT_VF(vfmacc_vf_h, 2, 2)
-GEN_VEXT_VF(vfmacc_vf_w, 4, 4)
-GEN_VEXT_VF(vfmacc_vf_d, 8, 8)
+GEN_VEXT_VF(vfmacc_vf_h)
+GEN_VEXT_VF(vfmacc_vf_w)
+GEN_VEXT_VF(vfmacc_vf_d)
 
 static uint16_t fnmacc16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3123,15 +3121,15 @@ static uint64_t fnmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmacc_vv_h, OP_UUU_H, H2, H2, H2, fnmacc16)
 RVVCALL(OPFVV3, vfnmacc_vv_w, OP_UUU_W, H4, H4, H4, fnmacc32)
 RVVCALL(OPFVV3, vfnmacc_vv_d, OP_UUU_D, H8, H8, H8, fnmacc64)
-GEN_VEXT_VV_ENV(vfnmacc_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmacc_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmacc_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmacc_vv_h)
+GEN_VEXT_VV_ENV(vfnmacc_vv_w)
+GEN_VEXT_VV_ENV(vfnmacc_vv_d)
 RVVCALL(OPFVF3, vfnmacc_vf_h, OP_UUU_H, H2, H2, fnmacc16)
 RVVCALL(OPFVF3, vfnmacc_vf_w, OP_UUU_W, H4, H4, fnmacc32)
 RVVCALL(OPFVF3, vfnmacc_vf_d, OP_UUU_D, H8, H8, fnmacc64)
-GEN_VEXT_VF(vfnmacc_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmacc_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmacc_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmacc_vf_h)
+GEN_VEXT_VF(vfnmacc_vf_w)
+GEN_VEXT_VF(vfnmacc_vf_d)
 
 static uint16_t fmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3151,15 +3149,15 @@ static uint64_t fmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsac_vv_h, OP_UUU_H, H2, H2, H2, fmsac16)
 RVVCALL(OPFVV3, vfmsac_vv_w, OP_UUU_W, H4, H4, H4, fmsac32)
 RVVCALL(OPFVV3, vfmsac_vv_d, OP_UUU_D, H8, H8, H8, fmsac64)
-GEN_VEXT_VV_ENV(vfmsac_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmsac_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmsac_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmsac_vv_h)
+GEN_VEXT_VV_ENV(vfmsac_vv_w)
+GEN_VEXT_VV_ENV(vfmsac_vv_d)
 RVVCALL(OPFVF3, vfmsac_vf_h, OP_UUU_H, H2, H2, fmsac16)
 RVVCALL(OPFVF3, vfmsac_vf_w, OP_UUU_W, H4, H4, fmsac32)
 RVVCALL(OPFVF3, vfmsac_vf_d, OP_UUU_D, H8, H8, fmsac64)
-GEN_VEXT_VF(vfmsac_vf_h, 2, 2)
-GEN_VEXT_VF(vfmsac_vf_w, 4, 4)
-GEN_VEXT_VF(vfmsac_vf_d, 8, 8)
+GEN_VEXT_VF(vfmsac_vf_h)
+GEN_VEXT_VF(vfmsac_vf_w)
+GEN_VEXT_VF(vfmsac_vf_d)
 
 static uint16_t fnmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3179,15 +3177,15 @@ static uint64_t fnmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsac_vv_h, OP_UUU_H, H2, H2, H2, fnmsac16)
 RVVCALL(OPFVV3, vfnmsac_vv_w, OP_UUU_W, H4, H4, H4, fnmsac32)
 RVVCALL(OPFVV3, vfnmsac_vv_d, OP_UUU_D, H8, H8, H8, fnmsac64)
-GEN_VEXT_VV_ENV(vfnmsac_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmsac_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmsac_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmsac_vv_h)
+GEN_VEXT_VV_ENV(vfnmsac_vv_w)
+GEN_VEXT_VV_ENV(vfnmsac_vv_d)
 RVVCALL(OPFVF3, vfnmsac_vf_h, OP_UUU_H, H2, H2, fnmsac16)
 RVVCALL(OPFVF3, vfnmsac_vf_w, OP_UUU_W, H4, H4, fnmsac32)
 RVVCALL(OPFVF3, vfnmsac_vf_d, OP_UUU_D, H8, H8, fnmsac64)
-GEN_VEXT_VF(vfnmsac_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmsac_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmsac_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmsac_vf_h)
+GEN_VEXT_VF(vfnmsac_vf_w)
+GEN_VEXT_VF(vfnmsac_vf_d)
 
 static uint16_t fmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3207,15 +3205,15 @@ static uint64_t fmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmadd_vv_h, OP_UUU_H, H2, H2, H2, fmadd16)
 RVVCALL(OPFVV3, vfmadd_vv_w, OP_UUU_W, H4, H4, H4, fmadd32)
 RVVCALL(OPFVV3, vfmadd_vv_d, OP_UUU_D, H8, H8, H8, fmadd64)
-GEN_VEXT_VV_ENV(vfmadd_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmadd_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmadd_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmadd_vv_h)
+GEN_VEXT_VV_ENV(vfmadd_vv_w)
+GEN_VEXT_VV_ENV(vfmadd_vv_d)
 RVVCALL(OPFVF3, vfmadd_vf_h, OP_UUU_H, H2, H2, fmadd16)
 RVVCALL(OPFVF3, vfmadd_vf_w, OP_UUU_W, H4, H4, fmadd32)
 RVVCALL(OPFVF3, vfmadd_vf_d, OP_UUU_D, H8, H8, fmadd64)
-GEN_VEXT_VF(vfmadd_vf_h, 2, 2)
-GEN_VEXT_VF(vfmadd_vf_w, 4, 4)
-GEN_VEXT_VF(vfmadd_vf_d, 8, 8)
+GEN_VEXT_VF(vfmadd_vf_h)
+GEN_VEXT_VF(vfmadd_vf_w)
+GEN_VEXT_VF(vfmadd_vf_d)
 
 static uint16_t fnmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3238,15 +3236,15 @@ static uint64_t fnmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmadd_vv_h, OP_UUU_H, H2, H2, H2, fnmadd16)
 RVVCALL(OPFVV3, vfnmadd_vv_w, OP_UUU_W, H4, H4, H4, fnmadd32)
 RVVCALL(OPFVV3, vfnmadd_vv_d, OP_UUU_D, H8, H8, H8, fnmadd64)
-GEN_VEXT_VV_ENV(vfnmadd_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmadd_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmadd_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmadd_vv_h)
+GEN_VEXT_VV_ENV(vfnmadd_vv_w)
+GEN_VEXT_VV_ENV(vfnmadd_vv_d)
 RVVCALL(OPFVF3, vfnmadd_vf_h, OP_UUU_H, H2, H2, fnmadd16)
 RVVCALL(OPFVF3, vfnmadd_vf_w, OP_UUU_W, H4, H4, fnmadd32)
 RVVCALL(OPFVF3, vfnmadd_vf_d, OP_UUU_D, H8, H8, fnmadd64)
-GEN_VEXT_VF(vfnmadd_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmadd_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmadd_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmadd_vf_h)
+GEN_VEXT_VF(vfnmadd_vf_w)
+GEN_VEXT_VF(vfnmadd_vf_d)
 
 static uint16_t fmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3266,15 +3264,15 @@ static uint64_t fmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsub_vv_h, OP_UUU_H, H2, H2, H2, fmsub16)
 RVVCALL(OPFVV3, vfmsub_vv_w, OP_UUU_W, H4, H4, H4, fmsub32)
 RVVCALL(OPFVV3, vfmsub_vv_d, OP_UUU_D, H8, H8, H8, fmsub64)
-GEN_VEXT_VV_ENV(vfmsub_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmsub_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmsub_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmsub_vv_h)
+GEN_VEXT_VV_ENV(vfmsub_vv_w)
+GEN_VEXT_VV_ENV(vfmsub_vv_d)
 RVVCALL(OPFVF3, vfmsub_vf_h, OP_UUU_H, H2, H2, fmsub16)
 RVVCALL(OPFVF3, vfmsub_vf_w, OP_UUU_W, H4, H4, fmsub32)
 RVVCALL(OPFVF3, vfmsub_vf_d, OP_UUU_D, H8, H8, fmsub64)
-GEN_VEXT_VF(vfmsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfmsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfmsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfmsub_vf_h)
+GEN_VEXT_VF(vfmsub_vf_w)
+GEN_VEXT_VF(vfmsub_vf_d)
 
 static uint16_t fnmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3294,15 +3292,15 @@ static uint64_t fnmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsub_vv_h, OP_UUU_H, H2, H2, H2, fnmsub16)
 RVVCALL(OPFVV3, vfnmsub_vv_w, OP_UUU_W, H4, H4, H4, fnmsub32)
 RVVCALL(OPFVV3, vfnmsub_vv_d, OP_UUU_D, H8, H8, H8, fnmsub64)
-GEN_VEXT_VV_ENV(vfnmsub_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmsub_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmsub_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmsub_vv_h)
+GEN_VEXT_VV_ENV(vfnmsub_vv_w)
+GEN_VEXT_VV_ENV(vfnmsub_vv_d)
 RVVCALL(OPFVF3, vfnmsub_vf_h, OP_UUU_H, H2, H2, fnmsub16)
 RVVCALL(OPFVF3, vfnmsub_vf_w, OP_UUU_W, H4, H4, fnmsub32)
 RVVCALL(OPFVF3, vfnmsub_vf_d, OP_UUU_D, H8, H8, fnmsub64)
-GEN_VEXT_VF(vfnmsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmsub_vf_h)
+GEN_VEXT_VF(vfnmsub_vf_w)
+GEN_VEXT_VF(vfnmsub_vf_d)
 
 /* Vector Widening Floating-Point Fused Multiply-Add Instructions */
 static uint32_t fwmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
@@ -3319,12 +3317,12 @@ static uint64_t fwmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwmacc16)
 RVVCALL(OPFVV3, vfwmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwmacc32)
-GEN_VEXT_VV_ENV(vfwmacc_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwmacc_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwmacc_vv_h)
+GEN_VEXT_VV_ENV(vfwmacc_vv_w)
 RVVCALL(OPFVF3, vfwmacc_vf_h, WOP_UUU_H, H4, H2, fwmacc16)
 RVVCALL(OPFVF3, vfwmacc_vf_w, WOP_UUU_W, H8, H4, fwmacc32)
-GEN_VEXT_VF(vfwmacc_vf_h, 2, 4)
-GEN_VEXT_VF(vfwmacc_vf_w, 4, 8)
+GEN_VEXT_VF(vfwmacc_vf_h)
+GEN_VEXT_VF(vfwmacc_vf_w)
 
 static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3342,12 +3340,12 @@ static uint64_t fwnmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwnmacc16)
 RVVCALL(OPFVV3, vfwnmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwnmacc32)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_h)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_w)
 RVVCALL(OPFVF3, vfwnmacc_vf_h, WOP_UUU_H, H4, H2, fwnmacc16)
 RVVCALL(OPFVF3, vfwnmacc_vf_w, WOP_UUU_W, H8, H4, fwnmacc32)
-GEN_VEXT_VF(vfwnmacc_vf_h, 2, 4)
-GEN_VEXT_VF(vfwnmacc_vf_w, 4, 8)
+GEN_VEXT_VF(vfwnmacc_vf_h)
+GEN_VEXT_VF(vfwnmacc_vf_w)
 
 static uint32_t fwmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3365,12 +3363,12 @@ static uint64_t fwmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwmsac16)
 RVVCALL(OPFVV3, vfwmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwmsac32)
-GEN_VEXT_VV_ENV(vfwmsac_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwmsac_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwmsac_vv_h)
+GEN_VEXT_VV_ENV(vfwmsac_vv_w)
 RVVCALL(OPFVF3, vfwmsac_vf_h, WOP_UUU_H, H4, H2, fwmsac16)
 RVVCALL(OPFVF3, vfwmsac_vf_w, WOP_UUU_W, H8, H4, fwmsac32)
-GEN_VEXT_VF(vfwmsac_vf_h, 2, 4)
-GEN_VEXT_VF(vfwmsac_vf_w, 4, 8)
+GEN_VEXT_VF(vfwmsac_vf_h)
+GEN_VEXT_VF(vfwmsac_vf_w)
 
 static uint32_t fwnmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3388,12 +3386,12 @@ static uint64_t fwnmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwnmsac16)
 RVVCALL(OPFVV3, vfwnmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwnmsac32)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_h)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_w)
 RVVCALL(OPFVF3, vfwnmsac_vf_h, WOP_UUU_H, H4, H2, fwnmsac16)
 RVVCALL(OPFVF3, vfwnmsac_vf_w, WOP_UUU_W, H8, H4, fwnmsac32)
-GEN_VEXT_VF(vfwnmsac_vf_h, 2, 4)
-GEN_VEXT_VF(vfwnmsac_vf_w, 4, 8)
+GEN_VEXT_VF(vfwnmsac_vf_h)
+GEN_VEXT_VF(vfwnmsac_vf_w)
 
 /* Vector Floating-Point Square-Root Instruction */
 /* (TD, T2, TX2) */
@@ -3409,7 +3407,7 @@ static void do_##NAME(void *vd, void *vs2, int i,      \
     *((TD *)vd + HD(i)) = OP(s2, &env->fp_status);     \
 }
 
-#define GEN_VEXT_V_ENV(NAME, ESZ, DSZ)                 \
+#define GEN_VEXT_V_ENV(NAME)                           \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         CPURISCVState *env, uint32_t desc)             \
 {                                                      \
@@ -3432,9 +3430,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
 RVVCALL(OPFVV1, vfsqrt_v_h, OP_UU_H, H2, H2, float16_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_w, OP_UU_W, H4, H4, float32_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_d, OP_UU_D, H8, H8, float64_sqrt)
-GEN_VEXT_V_ENV(vfsqrt_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfsqrt_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfsqrt_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfsqrt_v_h)
+GEN_VEXT_V_ENV(vfsqrt_v_w)
+GEN_VEXT_V_ENV(vfsqrt_v_d)
 
 /*
  * Vector Floating-Point Reciprocal Square-Root Estimate Instruction
@@ -3614,9 +3612,9 @@ static float64 frsqrt7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrsqrt7_v_h, OP_UU_H, H2, H2, frsqrt7_h)
 RVVCALL(OPFVV1, vfrsqrt7_v_w, OP_UU_W, H4, H4, frsqrt7_s)
 RVVCALL(OPFVV1, vfrsqrt7_v_d, OP_UU_D, H8, H8, frsqrt7_d)
-GEN_VEXT_V_ENV(vfrsqrt7_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfrsqrt7_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfrsqrt7_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfrsqrt7_v_h)
+GEN_VEXT_V_ENV(vfrsqrt7_v_w)
+GEN_VEXT_V_ENV(vfrsqrt7_v_d)
 
 /*
  * Vector Floating-Point Reciprocal Estimate Instruction
@@ -3805,36 +3803,36 @@ static float64 frec7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrec7_v_h, OP_UU_H, H2, H2, frec7_h)
 RVVCALL(OPFVV1, vfrec7_v_w, OP_UU_W, H4, H4, frec7_s)
 RVVCALL(OPFVV1, vfrec7_v_d, OP_UU_D, H8, H8, frec7_d)
-GEN_VEXT_V_ENV(vfrec7_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfrec7_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfrec7_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfrec7_v_h)
+GEN_VEXT_V_ENV(vfrec7_v_w)
+GEN_VEXT_V_ENV(vfrec7_v_d)
 
 /* Vector Floating-Point MIN/MAX Instructions */
 RVVCALL(OPFVV2, vfmin_vv_h, OP_UUU_H, H2, H2, H2, float16_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_w, OP_UUU_W, H4, H4, H4, float32_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_d, OP_UUU_D, H8, H8, H8, float64_minimum_number)
-GEN_VEXT_VV_ENV(vfmin_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmin_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmin_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmin_vv_h)
+GEN_VEXT_VV_ENV(vfmin_vv_w)
+GEN_VEXT_VV_ENV(vfmin_vv_d)
 RVVCALL(OPFVF2, vfmin_vf_h, OP_UUU_H, H2, H2, float16_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_w, OP_UUU_W, H4, H4, float32_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_d, OP_UUU_D, H8, H8, float64_minimum_number)
-GEN_VEXT_VF(vfmin_vf_h, 2, 2)
-GEN_VEXT_VF(vfmin_vf_w, 4, 4)
-GEN_VEXT_VF(vfmin_vf_d, 8, 8)
+GEN_VEXT_VF(vfmin_vf_h)
+GEN_VEXT_VF(vfmin_vf_w)
+GEN_VEXT_VF(vfmin_vf_d)
 
 RVVCALL(OPFVV2, vfmax_vv_h, OP_UUU_H, H2, H2, H2, float16_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_w, OP_UUU_W, H4, H4, H4, float32_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_d, OP_UUU_D, H8, H8, H8, float64_maximum_number)
-GEN_VEXT_VV_ENV(vfmax_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmax_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmax_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmax_vv_h)
+GEN_VEXT_VV_ENV(vfmax_vv_w)
+GEN_VEXT_VV_ENV(vfmax_vv_d)
 RVVCALL(OPFVF2, vfmax_vf_h, OP_UUU_H, H2, H2, float16_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_w, OP_UUU_W, H4, H4, float32_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_d, OP_UUU_D, H8, H8, float64_maximum_number)
-GEN_VEXT_VF(vfmax_vf_h, 2, 2)
-GEN_VEXT_VF(vfmax_vf_w, 4, 4)
-GEN_VEXT_VF(vfmax_vf_d, 8, 8)
+GEN_VEXT_VF(vfmax_vf_h)
+GEN_VEXT_VF(vfmax_vf_w)
+GEN_VEXT_VF(vfmax_vf_d)
 
 /* Vector Floating-Point Sign-Injection Instructions */
 static uint16_t fsgnj16(uint16_t a, uint16_t b, float_status *s)
@@ -3855,15 +3853,15 @@ static uint64_t fsgnj64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnj_vv_h, OP_UUU_H, H2, H2, H2, fsgnj16)
 RVVCALL(OPFVV2, vfsgnj_vv_w, OP_UUU_W, H4, H4, H4, fsgnj32)
 RVVCALL(OPFVV2, vfsgnj_vv_d, OP_UUU_D, H8, H8, H8, fsgnj64)
-GEN_VEXT_VV_ENV(vfsgnj_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsgnj_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsgnj_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsgnj_vv_h)
+GEN_VEXT_VV_ENV(vfsgnj_vv_w)
+GEN_VEXT_VV_ENV(vfsgnj_vv_d)
 RVVCALL(OPFVF2, vfsgnj_vf_h, OP_UUU_H, H2, H2, fsgnj16)
 RVVCALL(OPFVF2, vfsgnj_vf_w, OP_UUU_W, H4, H4, fsgnj32)
 RVVCALL(OPFVF2, vfsgnj_vf_d, OP_UUU_D, H8, H8, fsgnj64)
-GEN_VEXT_VF(vfsgnj_vf_h, 2, 2)
-GEN_VEXT_VF(vfsgnj_vf_w, 4, 4)
-GEN_VEXT_VF(vfsgnj_vf_d, 8, 8)
+GEN_VEXT_VF(vfsgnj_vf_h)
+GEN_VEXT_VF(vfsgnj_vf_w)
+GEN_VEXT_VF(vfsgnj_vf_d)
 
 static uint16_t fsgnjn16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3883,15 +3881,15 @@ static uint64_t fsgnjn64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjn_vv_h, OP_UUU_H, H2, H2, H2, fsgnjn16)
 RVVCALL(OPFVV2, vfsgnjn_vv_w, OP_UUU_W, H4, H4, H4, fsgnjn32)
 RVVCALL(OPFVV2, vfsgnjn_vv_d, OP_UUU_D, H8, H8, H8, fsgnjn64)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_h)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_w)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_d)
 RVVCALL(OPFVF2, vfsgnjn_vf_h, OP_UUU_H, H2, H2, fsgnjn16)
 RVVCALL(OPFVF2, vfsgnjn_vf_w, OP_UUU_W, H4, H4, fsgnjn32)
 RVVCALL(OPFVF2, vfsgnjn_vf_d, OP_UUU_D, H8, H8, fsgnjn64)
-GEN_VEXT_VF(vfsgnjn_vf_h, 2, 2)
-GEN_VEXT_VF(vfsgnjn_vf_w, 4, 4)
-GEN_VEXT_VF(vfsgnjn_vf_d, 8, 8)
+GEN_VEXT_VF(vfsgnjn_vf_h)
+GEN_VEXT_VF(vfsgnjn_vf_w)
+GEN_VEXT_VF(vfsgnjn_vf_d)
 
 static uint16_t fsgnjx16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3911,15 +3909,15 @@ static uint64_t fsgnjx64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjx_vv_h, OP_UUU_H, H2, H2, H2, fsgnjx16)
 RVVCALL(OPFVV2, vfsgnjx_vv_w, OP_UUU_W, H4, H4, H4, fsgnjx32)
 RVVCALL(OPFVV2, vfsgnjx_vv_d, OP_UUU_D, H8, H8, H8, fsgnjx64)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_h)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_w)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_d)
 RVVCALL(OPFVF2, vfsgnjx_vf_h, OP_UUU_H, H2, H2, fsgnjx16)
 RVVCALL(OPFVF2, vfsgnjx_vf_w, OP_UUU_W, H4, H4, fsgnjx32)
 RVVCALL(OPFVF2, vfsgnjx_vf_d, OP_UUU_D, H8, H8, fsgnjx64)
-GEN_VEXT_VF(vfsgnjx_vf_h, 2, 2)
-GEN_VEXT_VF(vfsgnjx_vf_w, 4, 4)
-GEN_VEXT_VF(vfsgnjx_vf_d, 8, 8)
+GEN_VEXT_VF(vfsgnjx_vf_h)
+GEN_VEXT_VF(vfsgnjx_vf_w)
+GEN_VEXT_VF(vfsgnjx_vf_d)
 
 /* Vector Floating-Point Compare Instructions */
 #define GEN_VEXT_CMP_VV_ENV(NAME, ETYPE, H, DO_OP)            \
@@ -4063,7 +4061,7 @@ static void do_##NAME(void *vd, void *vs2, int i)      \
     *((TD *)vd + HD(i)) = OP(s2);                      \
 }
 
-#define GEN_VEXT_V(NAME, ESZ, DSZ)                     \
+#define GEN_VEXT_V(NAME)                               \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
                   CPURISCVState *env, uint32_t desc)   \
 {                                                      \
@@ -4140,9 +4138,9 @@ target_ulong fclass_d(uint64_t frs1)
 RVVCALL(OPIVV1, vfclass_v_h, OP_UU_H, H2, H2, fclass_h)
 RVVCALL(OPIVV1, vfclass_v_w, OP_UU_W, H4, H4, fclass_s)
 RVVCALL(OPIVV1, vfclass_v_d, OP_UU_D, H8, H8, fclass_d)
-GEN_VEXT_V(vfclass_v_h, 2, 2)
-GEN_VEXT_V(vfclass_v_w, 4, 4)
-GEN_VEXT_V(vfclass_v_d, 8, 8)
+GEN_VEXT_V(vfclass_v_h)
+GEN_VEXT_V(vfclass_v_w)
+GEN_VEXT_V(vfclass_v_d)
 
 /* Vector Floating-Point Merge Instruction */
 #define GEN_VFMERGE_VF(NAME, ETYPE, H)                        \
@@ -4170,33 +4168,33 @@ GEN_VFMERGE_VF(vfmerge_vfm_d, int64_t, H8)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_h, OP_UU_H, H2, H2, float16_to_uint16)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_w, OP_UU_W, H4, H4, float32_to_uint32)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_d, OP_UU_D, H8, H8, float64_to_uint64)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_h)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_w)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_d)
 
 /* vfcvt.x.f.v vd, vs2, vm # Convert float to signed integer. */
 RVVCALL(OPFVV1, vfcvt_x_f_v_h, OP_UU_H, H2, H2, float16_to_int16)
 RVVCALL(OPFVV1, vfcvt_x_f_v_w, OP_UU_W, H4, H4, float32_to_int32)
 RVVCALL(OPFVV1, vfcvt_x_f_v_d, OP_UU_D, H8, H8, float64_to_int64)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_h)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_w)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_d)
 
 /* vfcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_xu_v_h, OP_UU_H, H2, H2, uint16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_w, OP_UU_W, H4, H4, uint32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_d, OP_UU_D, H8, H8, uint64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_h)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_w)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_d)
 
 /* vfcvt.f.x.v vd, vs2, vm # Convert integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_x_v_h, OP_UU_H, H2, H2, int16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_x_v_w, OP_UU_W, H4, H4, int32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_x_v_d, OP_UU_D, H8, H8, int64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_h)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_w)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_d)
 
 /* Widening Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4206,30 +4204,30 @@ GEN_VEXT_V_ENV(vfcvt_f_x_v_d, 8, 8)
 /* vfwcvt.xu.f.v vd, vs2, vm # Convert float to double-width unsigned integer.*/
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_h, WOP_UU_H, H4, H2, float16_to_uint32)
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_w, WOP_UU_W, H8, H4, float32_to_uint64)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w)
 
 /* vfwcvt.x.f.v vd, vs2, vm # Convert float to double-width signed integer. */
 RVVCALL(OPFVV1, vfwcvt_x_f_v_h, WOP_UU_H, H4, H2, float16_to_int32)
 RVVCALL(OPFVV1, vfwcvt_x_f_v_w, WOP_UU_W, H8, H4, float32_to_int64)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_h)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_w)
 
 /* vfwcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to double-width float */
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_b, WOP_UU_B, H2, H1, uint8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_h, WOP_UU_H, H4, H2, uint16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_w, WOP_UU_W, H8, H4, uint32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b, 1, 2)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w)
 
 /* vfwcvt.f.x.v vd, vs2, vm # Convert integer to double-width float. */
 RVVCALL(OPFVV1, vfwcvt_f_x_v_b, WOP_UU_B, H2, H1, int8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_h, WOP_UU_H, H4, H2, int16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_w, WOP_UU_W, H8, H4, int32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_b, 1, 2)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_b)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_h)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_w)
 
 /*
  * vfwcvt.f.f.v vd, vs2, vm
@@ -4242,8 +4240,8 @@ static uint32_t vfwcvtffv16(uint16_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfwcvt_f_f_v_h, WOP_UU_H, H4, H2, vfwcvtffv16)
 RVVCALL(OPFVV1, vfwcvt_f_f_v_w, WOP_UU_W, H8, H4, float32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_h)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_w)
 
 /* Narrowing Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4254,29 +4252,29 @@ GEN_VEXT_V_ENV(vfwcvt_f_f_v_w, 4, 8)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_b, NOP_UU_B, H1, H2, float16_to_uint8)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_h, NOP_UU_H, H2, H4, float32_to_uint16)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_w, NOP_UU_W, H4, H8, float64_to_uint32)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_b, 1, 1)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_b)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_h)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_w)
 
 /* vfncvt.x.f.v vd, vs2, vm # Convert double-width float to signed integer. */
 RVVCALL(OPFVV1, vfncvt_x_f_w_b, NOP_UU_B, H1, H2, float16_to_int8)
 RVVCALL(OPFVV1, vfncvt_x_f_w_h, NOP_UU_H, H2, H4, float32_to_int16)
 RVVCALL(OPFVV1, vfncvt_x_f_w_w, NOP_UU_W, H4, H8, float64_to_int32)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_b, 1, 1)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_b)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_h)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_w)
 
 /* vfncvt.f.xu.v vd, vs2, vm # Convert double-width unsigned integer to float */
 RVVCALL(OPFVV1, vfncvt_f_xu_w_h, NOP_UU_H, H2, H4, uint32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_xu_w_w, NOP_UU_W, H4, H8, uint64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_h)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_w)
 
 /* vfncvt.f.x.v vd, vs2, vm # Convert double-width integer to float. */
 RVVCALL(OPFVV1, vfncvt_f_x_w_h, NOP_UU_H, H2, H4, int32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_x_w_w, NOP_UU_W, H4, H8, int64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_h)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_w)
 
 /* vfncvt.f.f.v vd, vs2, vm # Convert double float to single-width float. */
 static uint16_t vfncvtffv16(uint32_t a, float_status *s)
@@ -4286,8 +4284,8 @@ static uint16_t vfncvtffv16(uint32_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfncvt_f_f_w_h, NOP_UU_H, H2, H4, vfncvtffv16)
 RVVCALL(OPFVV1, vfncvt_f_f_w_w, NOP_UU_W, H4, H8, float64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_h)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_w)
 
 /*
  *** Vector Reduction Operations
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

* [PATCH qemu v4 01/14] target/riscv: rvv: Prune redundant ESZ, DSZ parameter passed
@ 2022-03-14  7:38   ` ~eopxd
  0 siblings, 0 replies; 42+ messages in thread
From: ~eopxd @ 2022-03-14  7:38 UTC (permalink / raw)
  To: qemu-devel, qemu-riscv
  Cc: Palmer Dabbelt, Alistair Francis, Bin Meng, Frank Chang,
	WeiWei Li, eop Chen

From: eopXD <eop.chen@sifive.com>

No functional change intended in this commit.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
---
 target/riscv/vector_helper.c | 1132 +++++++++++++++++-----------------
 1 file changed, 565 insertions(+), 567 deletions(-)

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 3bd4aac9c9..e94caf1a3c 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -710,7 +710,6 @@ RVVCALL(OPIVV2, vsub_vv_d, OP_SSS_D, H8, H8, H8, DO_SUB)
 
 static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       uint32_t esz, uint32_t dsz,
                        opivv2_fn *fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -727,23 +726,23 @@ static void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
 }
 
 /* generate the helpers for OPIVV */
-#define GEN_VEXT_VV(NAME, ESZ, DSZ)                       \
+#define GEN_VEXT_VV(NAME)                                 \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
-    do_vext_vv(vd, v0, vs1, vs2, env, desc, ESZ, DSZ,     \
+    do_vext_vv(vd, v0, vs1, vs2, env, desc,               \
                do_##NAME);                                \
 }
 
-GEN_VEXT_VV(vadd_vv_b, 1, 1)
-GEN_VEXT_VV(vadd_vv_h, 2, 2)
-GEN_VEXT_VV(vadd_vv_w, 4, 4)
-GEN_VEXT_VV(vadd_vv_d, 8, 8)
-GEN_VEXT_VV(vsub_vv_b, 1, 1)
-GEN_VEXT_VV(vsub_vv_h, 2, 2)
-GEN_VEXT_VV(vsub_vv_w, 4, 4)
-GEN_VEXT_VV(vsub_vv_d, 8, 8)
+GEN_VEXT_VV(vadd_vv_b)
+GEN_VEXT_VV(vadd_vv_h)
+GEN_VEXT_VV(vadd_vv_w)
+GEN_VEXT_VV(vadd_vv_d)
+GEN_VEXT_VV(vsub_vv_b)
+GEN_VEXT_VV(vsub_vv_h)
+GEN_VEXT_VV(vsub_vv_w)
+GEN_VEXT_VV(vsub_vv_d)
 
 typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
 
@@ -773,7 +772,6 @@ RVVCALL(OPIVX2, vrsub_vx_d, OP_SSS_D, H8, H8, DO_RSUB)
 
 static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
                        CPURISCVState *env, uint32_t desc,
-                       uint32_t esz, uint32_t dsz,
                        opivx2_fn fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -790,27 +788,27 @@ static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
 }
 
 /* generate the helpers for OPIVX */
-#define GEN_VEXT_VX(NAME, ESZ, DSZ)                       \
+#define GEN_VEXT_VX(NAME)                                 \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
 {                                                         \
-    do_vext_vx(vd, v0, s1, vs2, env, desc, ESZ, DSZ,      \
+    do_vext_vx(vd, v0, s1, vs2, env, desc,                \
                do_##NAME);                                \
 }
 
-GEN_VEXT_VX(vadd_vx_b, 1, 1)
-GEN_VEXT_VX(vadd_vx_h, 2, 2)
-GEN_VEXT_VX(vadd_vx_w, 4, 4)
-GEN_VEXT_VX(vadd_vx_d, 8, 8)
-GEN_VEXT_VX(vsub_vx_b, 1, 1)
-GEN_VEXT_VX(vsub_vx_h, 2, 2)
-GEN_VEXT_VX(vsub_vx_w, 4, 4)
-GEN_VEXT_VX(vsub_vx_d, 8, 8)
-GEN_VEXT_VX(vrsub_vx_b, 1, 1)
-GEN_VEXT_VX(vrsub_vx_h, 2, 2)
-GEN_VEXT_VX(vrsub_vx_w, 4, 4)
-GEN_VEXT_VX(vrsub_vx_d, 8, 8)
+GEN_VEXT_VX(vadd_vx_b)
+GEN_VEXT_VX(vadd_vx_h)
+GEN_VEXT_VX(vadd_vx_w)
+GEN_VEXT_VX(vadd_vx_d)
+GEN_VEXT_VX(vsub_vx_b)
+GEN_VEXT_VX(vsub_vx_h)
+GEN_VEXT_VX(vsub_vx_w)
+GEN_VEXT_VX(vsub_vx_d)
+GEN_VEXT_VX(vrsub_vx_b)
+GEN_VEXT_VX(vrsub_vx_h)
+GEN_VEXT_VX(vrsub_vx_w)
+GEN_VEXT_VX(vrsub_vx_d)
 
 void HELPER(vec_rsubs8)(void *d, void *a, uint64_t b, uint32_t desc)
 {
@@ -889,30 +887,30 @@ RVVCALL(OPIVV2, vwadd_wv_w, WOP_WSSS_W, H8, H4, H4, DO_ADD)
 RVVCALL(OPIVV2, vwsub_wv_b, WOP_WSSS_B, H2, H1, H1, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_h, WOP_WSSS_H, H4, H2, H2, DO_SUB)
 RVVCALL(OPIVV2, vwsub_wv_w, WOP_WSSS_W, H8, H4, H4, DO_SUB)
-GEN_VEXT_VV(vwaddu_vv_b, 1, 2)
-GEN_VEXT_VV(vwaddu_vv_h, 2, 4)
-GEN_VEXT_VV(vwaddu_vv_w, 4, 8)
-GEN_VEXT_VV(vwsubu_vv_b, 1, 2)
-GEN_VEXT_VV(vwsubu_vv_h, 2, 4)
-GEN_VEXT_VV(vwsubu_vv_w, 4, 8)
-GEN_VEXT_VV(vwadd_vv_b, 1, 2)
-GEN_VEXT_VV(vwadd_vv_h, 2, 4)
-GEN_VEXT_VV(vwadd_vv_w, 4, 8)
-GEN_VEXT_VV(vwsub_vv_b, 1, 2)
-GEN_VEXT_VV(vwsub_vv_h, 2, 4)
-GEN_VEXT_VV(vwsub_vv_w, 4, 8)
-GEN_VEXT_VV(vwaddu_wv_b, 1, 2)
-GEN_VEXT_VV(vwaddu_wv_h, 2, 4)
-GEN_VEXT_VV(vwaddu_wv_w, 4, 8)
-GEN_VEXT_VV(vwsubu_wv_b, 1, 2)
-GEN_VEXT_VV(vwsubu_wv_h, 2, 4)
-GEN_VEXT_VV(vwsubu_wv_w, 4, 8)
-GEN_VEXT_VV(vwadd_wv_b, 1, 2)
-GEN_VEXT_VV(vwadd_wv_h, 2, 4)
-GEN_VEXT_VV(vwadd_wv_w, 4, 8)
-GEN_VEXT_VV(vwsub_wv_b, 1, 2)
-GEN_VEXT_VV(vwsub_wv_h, 2, 4)
-GEN_VEXT_VV(vwsub_wv_w, 4, 8)
+GEN_VEXT_VV(vwaddu_vv_b)
+GEN_VEXT_VV(vwaddu_vv_h)
+GEN_VEXT_VV(vwaddu_vv_w)
+GEN_VEXT_VV(vwsubu_vv_b)
+GEN_VEXT_VV(vwsubu_vv_h)
+GEN_VEXT_VV(vwsubu_vv_w)
+GEN_VEXT_VV(vwadd_vv_b)
+GEN_VEXT_VV(vwadd_vv_h)
+GEN_VEXT_VV(vwadd_vv_w)
+GEN_VEXT_VV(vwsub_vv_b)
+GEN_VEXT_VV(vwsub_vv_h)
+GEN_VEXT_VV(vwsub_vv_w)
+GEN_VEXT_VV(vwaddu_wv_b)
+GEN_VEXT_VV(vwaddu_wv_h)
+GEN_VEXT_VV(vwaddu_wv_w)
+GEN_VEXT_VV(vwsubu_wv_b)
+GEN_VEXT_VV(vwsubu_wv_h)
+GEN_VEXT_VV(vwsubu_wv_w)
+GEN_VEXT_VV(vwadd_wv_b)
+GEN_VEXT_VV(vwadd_wv_h)
+GEN_VEXT_VV(vwadd_wv_w)
+GEN_VEXT_VV(vwsub_wv_b)
+GEN_VEXT_VV(vwsub_wv_h)
+GEN_VEXT_VV(vwsub_wv_w)
 
 RVVCALL(OPIVX2, vwaddu_vx_b, WOP_UUU_B, H2, H1, DO_ADD)
 RVVCALL(OPIVX2, vwaddu_vx_h, WOP_UUU_H, H4, H2, DO_ADD)
@@ -938,30 +936,30 @@ RVVCALL(OPIVX2, vwadd_wx_w, WOP_WSSS_W, H8, H4, DO_ADD)
 RVVCALL(OPIVX2, vwsub_wx_b, WOP_WSSS_B, H2, H1, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_h, WOP_WSSS_H, H4, H2, DO_SUB)
 RVVCALL(OPIVX2, vwsub_wx_w, WOP_WSSS_W, H8, H4, DO_SUB)
-GEN_VEXT_VX(vwaddu_vx_b, 1, 2)
-GEN_VEXT_VX(vwaddu_vx_h, 2, 4)
-GEN_VEXT_VX(vwaddu_vx_w, 4, 8)
-GEN_VEXT_VX(vwsubu_vx_b, 1, 2)
-GEN_VEXT_VX(vwsubu_vx_h, 2, 4)
-GEN_VEXT_VX(vwsubu_vx_w, 4, 8)
-GEN_VEXT_VX(vwadd_vx_b, 1, 2)
-GEN_VEXT_VX(vwadd_vx_h, 2, 4)
-GEN_VEXT_VX(vwadd_vx_w, 4, 8)
-GEN_VEXT_VX(vwsub_vx_b, 1, 2)
-GEN_VEXT_VX(vwsub_vx_h, 2, 4)
-GEN_VEXT_VX(vwsub_vx_w, 4, 8)
-GEN_VEXT_VX(vwaddu_wx_b, 1, 2)
-GEN_VEXT_VX(vwaddu_wx_h, 2, 4)
-GEN_VEXT_VX(vwaddu_wx_w, 4, 8)
-GEN_VEXT_VX(vwsubu_wx_b, 1, 2)
-GEN_VEXT_VX(vwsubu_wx_h, 2, 4)
-GEN_VEXT_VX(vwsubu_wx_w, 4, 8)
-GEN_VEXT_VX(vwadd_wx_b, 1, 2)
-GEN_VEXT_VX(vwadd_wx_h, 2, 4)
-GEN_VEXT_VX(vwadd_wx_w, 4, 8)
-GEN_VEXT_VX(vwsub_wx_b, 1, 2)
-GEN_VEXT_VX(vwsub_wx_h, 2, 4)
-GEN_VEXT_VX(vwsub_wx_w, 4, 8)
+GEN_VEXT_VX(vwaddu_vx_b)
+GEN_VEXT_VX(vwaddu_vx_h)
+GEN_VEXT_VX(vwaddu_vx_w)
+GEN_VEXT_VX(vwsubu_vx_b)
+GEN_VEXT_VX(vwsubu_vx_h)
+GEN_VEXT_VX(vwsubu_vx_w)
+GEN_VEXT_VX(vwadd_vx_b)
+GEN_VEXT_VX(vwadd_vx_h)
+GEN_VEXT_VX(vwadd_vx_w)
+GEN_VEXT_VX(vwsub_vx_b)
+GEN_VEXT_VX(vwsub_vx_h)
+GEN_VEXT_VX(vwsub_vx_w)
+GEN_VEXT_VX(vwaddu_wx_b)
+GEN_VEXT_VX(vwaddu_wx_h)
+GEN_VEXT_VX(vwaddu_wx_w)
+GEN_VEXT_VX(vwsubu_wx_b)
+GEN_VEXT_VX(vwsubu_wx_h)
+GEN_VEXT_VX(vwsubu_wx_w)
+GEN_VEXT_VX(vwadd_wx_b)
+GEN_VEXT_VX(vwadd_wx_h)
+GEN_VEXT_VX(vwadd_wx_w)
+GEN_VEXT_VX(vwsub_wx_b)
+GEN_VEXT_VX(vwsub_wx_h)
+GEN_VEXT_VX(vwsub_wx_w)
 
 /* Vector Integer Add-with-Carry / Subtract-with-Borrow Instructions */
 #define DO_VADC(N, M, C) (N + M + C)
@@ -1091,18 +1089,18 @@ RVVCALL(OPIVV2, vxor_vv_b, OP_SSS_B, H1, H1, H1, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_h, OP_SSS_H, H2, H2, H2, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_w, OP_SSS_W, H4, H4, H4, DO_XOR)
 RVVCALL(OPIVV2, vxor_vv_d, OP_SSS_D, H8, H8, H8, DO_XOR)
-GEN_VEXT_VV(vand_vv_b, 1, 1)
-GEN_VEXT_VV(vand_vv_h, 2, 2)
-GEN_VEXT_VV(vand_vv_w, 4, 4)
-GEN_VEXT_VV(vand_vv_d, 8, 8)
-GEN_VEXT_VV(vor_vv_b, 1, 1)
-GEN_VEXT_VV(vor_vv_h, 2, 2)
-GEN_VEXT_VV(vor_vv_w, 4, 4)
-GEN_VEXT_VV(vor_vv_d, 8, 8)
-GEN_VEXT_VV(vxor_vv_b, 1, 1)
-GEN_VEXT_VV(vxor_vv_h, 2, 2)
-GEN_VEXT_VV(vxor_vv_w, 4, 4)
-GEN_VEXT_VV(vxor_vv_d, 8, 8)
+GEN_VEXT_VV(vand_vv_b)
+GEN_VEXT_VV(vand_vv_h)
+GEN_VEXT_VV(vand_vv_w)
+GEN_VEXT_VV(vand_vv_d)
+GEN_VEXT_VV(vor_vv_b)
+GEN_VEXT_VV(vor_vv_h)
+GEN_VEXT_VV(vor_vv_w)
+GEN_VEXT_VV(vor_vv_d)
+GEN_VEXT_VV(vxor_vv_b)
+GEN_VEXT_VV(vxor_vv_h)
+GEN_VEXT_VV(vxor_vv_w)
+GEN_VEXT_VV(vxor_vv_d)
 
 RVVCALL(OPIVX2, vand_vx_b, OP_SSS_B, H1, H1, DO_AND)
 RVVCALL(OPIVX2, vand_vx_h, OP_SSS_H, H2, H2, DO_AND)
@@ -1116,18 +1114,18 @@ RVVCALL(OPIVX2, vxor_vx_b, OP_SSS_B, H1, H1, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_h, OP_SSS_H, H2, H2, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_w, OP_SSS_W, H4, H4, DO_XOR)
 RVVCALL(OPIVX2, vxor_vx_d, OP_SSS_D, H8, H8, DO_XOR)
-GEN_VEXT_VX(vand_vx_b, 1, 1)
-GEN_VEXT_VX(vand_vx_h, 2, 2)
-GEN_VEXT_VX(vand_vx_w, 4, 4)
-GEN_VEXT_VX(vand_vx_d, 8, 8)
-GEN_VEXT_VX(vor_vx_b, 1, 1)
-GEN_VEXT_VX(vor_vx_h, 2, 2)
-GEN_VEXT_VX(vor_vx_w, 4, 4)
-GEN_VEXT_VX(vor_vx_d, 8, 8)
-GEN_VEXT_VX(vxor_vx_b, 1, 1)
-GEN_VEXT_VX(vxor_vx_h, 2, 2)
-GEN_VEXT_VX(vxor_vx_w, 4, 4)
-GEN_VEXT_VX(vxor_vx_d, 8, 8)
+GEN_VEXT_VX(vand_vx_b)
+GEN_VEXT_VX(vand_vx_h)
+GEN_VEXT_VX(vand_vx_w)
+GEN_VEXT_VX(vand_vx_d)
+GEN_VEXT_VX(vor_vx_b)
+GEN_VEXT_VX(vor_vx_h)
+GEN_VEXT_VX(vor_vx_w)
+GEN_VEXT_VX(vor_vx_d)
+GEN_VEXT_VX(vxor_vx_b)
+GEN_VEXT_VX(vxor_vx_h)
+GEN_VEXT_VX(vxor_vx_w)
+GEN_VEXT_VX(vxor_vx_d)
 
 /* Vector Single-Width Bit Shift Instructions */
 #define DO_SLL(N, M)  (N << (M))
@@ -1348,22 +1346,22 @@ RVVCALL(OPIVV2, vmax_vv_b, OP_SSS_B, H1, H1, H1, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_h, OP_SSS_H, H2, H2, H2, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_w, OP_SSS_W, H4, H4, H4, DO_MAX)
 RVVCALL(OPIVV2, vmax_vv_d, OP_SSS_D, H8, H8, H8, DO_MAX)
-GEN_VEXT_VV(vminu_vv_b, 1, 1)
-GEN_VEXT_VV(vminu_vv_h, 2, 2)
-GEN_VEXT_VV(vminu_vv_w, 4, 4)
-GEN_VEXT_VV(vminu_vv_d, 8, 8)
-GEN_VEXT_VV(vmin_vv_b, 1, 1)
-GEN_VEXT_VV(vmin_vv_h, 2, 2)
-GEN_VEXT_VV(vmin_vv_w, 4, 4)
-GEN_VEXT_VV(vmin_vv_d, 8, 8)
-GEN_VEXT_VV(vmaxu_vv_b, 1, 1)
-GEN_VEXT_VV(vmaxu_vv_h, 2, 2)
-GEN_VEXT_VV(vmaxu_vv_w, 4, 4)
-GEN_VEXT_VV(vmaxu_vv_d, 8, 8)
-GEN_VEXT_VV(vmax_vv_b, 1, 1)
-GEN_VEXT_VV(vmax_vv_h, 2, 2)
-GEN_VEXT_VV(vmax_vv_w, 4, 4)
-GEN_VEXT_VV(vmax_vv_d, 8, 8)
+GEN_VEXT_VV(vminu_vv_b)
+GEN_VEXT_VV(vminu_vv_h)
+GEN_VEXT_VV(vminu_vv_w)
+GEN_VEXT_VV(vminu_vv_d)
+GEN_VEXT_VV(vmin_vv_b)
+GEN_VEXT_VV(vmin_vv_h)
+GEN_VEXT_VV(vmin_vv_w)
+GEN_VEXT_VV(vmin_vv_d)
+GEN_VEXT_VV(vmaxu_vv_b)
+GEN_VEXT_VV(vmaxu_vv_h)
+GEN_VEXT_VV(vmaxu_vv_w)
+GEN_VEXT_VV(vmaxu_vv_d)
+GEN_VEXT_VV(vmax_vv_b)
+GEN_VEXT_VV(vmax_vv_h)
+GEN_VEXT_VV(vmax_vv_w)
+GEN_VEXT_VV(vmax_vv_d)
 
 RVVCALL(OPIVX2, vminu_vx_b, OP_UUU_B, H1, H1, DO_MIN)
 RVVCALL(OPIVX2, vminu_vx_h, OP_UUU_H, H2, H2, DO_MIN)
@@ -1381,22 +1379,22 @@ RVVCALL(OPIVX2, vmax_vx_b, OP_SSS_B, H1, H1, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_h, OP_SSS_H, H2, H2, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_w, OP_SSS_W, H4, H4, DO_MAX)
 RVVCALL(OPIVX2, vmax_vx_d, OP_SSS_D, H8, H8, DO_MAX)
-GEN_VEXT_VX(vminu_vx_b, 1, 1)
-GEN_VEXT_VX(vminu_vx_h, 2, 2)
-GEN_VEXT_VX(vminu_vx_w, 4, 4)
-GEN_VEXT_VX(vminu_vx_d, 8, 8)
-GEN_VEXT_VX(vmin_vx_b, 1, 1)
-GEN_VEXT_VX(vmin_vx_h, 2, 2)
-GEN_VEXT_VX(vmin_vx_w, 4, 4)
-GEN_VEXT_VX(vmin_vx_d, 8, 8)
-GEN_VEXT_VX(vmaxu_vx_b, 1, 1)
-GEN_VEXT_VX(vmaxu_vx_h, 2, 2)
-GEN_VEXT_VX(vmaxu_vx_w, 4, 4)
-GEN_VEXT_VX(vmaxu_vx_d, 8, 8)
-GEN_VEXT_VX(vmax_vx_b, 1, 1)
-GEN_VEXT_VX(vmax_vx_h, 2, 2)
-GEN_VEXT_VX(vmax_vx_w, 4, 4)
-GEN_VEXT_VX(vmax_vx_d, 8, 8)
+GEN_VEXT_VX(vminu_vx_b)
+GEN_VEXT_VX(vminu_vx_h)
+GEN_VEXT_VX(vminu_vx_w)
+GEN_VEXT_VX(vminu_vx_d)
+GEN_VEXT_VX(vmin_vx_b)
+GEN_VEXT_VX(vmin_vx_h)
+GEN_VEXT_VX(vmin_vx_w)
+GEN_VEXT_VX(vmin_vx_d)
+GEN_VEXT_VX(vmaxu_vx_b)
+GEN_VEXT_VX(vmaxu_vx_h)
+GEN_VEXT_VX(vmaxu_vx_w)
+GEN_VEXT_VX(vmaxu_vx_d)
+GEN_VEXT_VX(vmax_vx_b)
+GEN_VEXT_VX(vmax_vx_h)
+GEN_VEXT_VX(vmax_vx_w)
+GEN_VEXT_VX(vmax_vx_d)
 
 /* Vector Single-Width Integer Multiply Instructions */
 #define DO_MUL(N, M) (N * M)
@@ -1404,10 +1402,10 @@ RVVCALL(OPIVV2, vmul_vv_b, OP_SSS_B, H1, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_h, OP_SSS_H, H2, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_w, OP_SSS_W, H4, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vmul_vv_d, OP_SSS_D, H8, H8, H8, DO_MUL)
-GEN_VEXT_VV(vmul_vv_b, 1, 1)
-GEN_VEXT_VV(vmul_vv_h, 2, 2)
-GEN_VEXT_VV(vmul_vv_w, 4, 4)
-GEN_VEXT_VV(vmul_vv_d, 8, 8)
+GEN_VEXT_VV(vmul_vv_b)
+GEN_VEXT_VV(vmul_vv_h)
+GEN_VEXT_VV(vmul_vv_w)
+GEN_VEXT_VV(vmul_vv_d)
 
 static int8_t do_mulh_b(int8_t s2, int8_t s1)
 {
@@ -1511,18 +1509,18 @@ RVVCALL(OPIVV2, vmulhsu_vv_b, OP_SUS_B, H1, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVV2, vmulhsu_vv_h, OP_SUS_H, H2, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVV2, vmulhsu_vv_w, OP_SUS_W, H4, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVV2, vmulhsu_vv_d, OP_SUS_D, H8, H8, H8, do_mulhsu_d)
-GEN_VEXT_VV(vmulh_vv_b, 1, 1)
-GEN_VEXT_VV(vmulh_vv_h, 2, 2)
-GEN_VEXT_VV(vmulh_vv_w, 4, 4)
-GEN_VEXT_VV(vmulh_vv_d, 8, 8)
-GEN_VEXT_VV(vmulhu_vv_b, 1, 1)
-GEN_VEXT_VV(vmulhu_vv_h, 2, 2)
-GEN_VEXT_VV(vmulhu_vv_w, 4, 4)
-GEN_VEXT_VV(vmulhu_vv_d, 8, 8)
-GEN_VEXT_VV(vmulhsu_vv_b, 1, 1)
-GEN_VEXT_VV(vmulhsu_vv_h, 2, 2)
-GEN_VEXT_VV(vmulhsu_vv_w, 4, 4)
-GEN_VEXT_VV(vmulhsu_vv_d, 8, 8)
+GEN_VEXT_VV(vmulh_vv_b)
+GEN_VEXT_VV(vmulh_vv_h)
+GEN_VEXT_VV(vmulh_vv_w)
+GEN_VEXT_VV(vmulh_vv_d)
+GEN_VEXT_VV(vmulhu_vv_b)
+GEN_VEXT_VV(vmulhu_vv_h)
+GEN_VEXT_VV(vmulhu_vv_w)
+GEN_VEXT_VV(vmulhu_vv_d)
+GEN_VEXT_VV(vmulhsu_vv_b)
+GEN_VEXT_VV(vmulhsu_vv_h)
+GEN_VEXT_VV(vmulhsu_vv_w)
+GEN_VEXT_VV(vmulhsu_vv_d)
 
 RVVCALL(OPIVX2, vmul_vx_b, OP_SSS_B, H1, H1, DO_MUL)
 RVVCALL(OPIVX2, vmul_vx_h, OP_SSS_H, H2, H2, DO_MUL)
@@ -1540,22 +1538,22 @@ RVVCALL(OPIVX2, vmulhsu_vx_b, OP_SUS_B, H1, H1, do_mulhsu_b)
 RVVCALL(OPIVX2, vmulhsu_vx_h, OP_SUS_H, H2, H2, do_mulhsu_h)
 RVVCALL(OPIVX2, vmulhsu_vx_w, OP_SUS_W, H4, H4, do_mulhsu_w)
 RVVCALL(OPIVX2, vmulhsu_vx_d, OP_SUS_D, H8, H8, do_mulhsu_d)
-GEN_VEXT_VX(vmul_vx_b, 1, 1)
-GEN_VEXT_VX(vmul_vx_h, 2, 2)
-GEN_VEXT_VX(vmul_vx_w, 4, 4)
-GEN_VEXT_VX(vmul_vx_d, 8, 8)
-GEN_VEXT_VX(vmulh_vx_b, 1, 1)
-GEN_VEXT_VX(vmulh_vx_h, 2, 2)
-GEN_VEXT_VX(vmulh_vx_w, 4, 4)
-GEN_VEXT_VX(vmulh_vx_d, 8, 8)
-GEN_VEXT_VX(vmulhu_vx_b, 1, 1)
-GEN_VEXT_VX(vmulhu_vx_h, 2, 2)
-GEN_VEXT_VX(vmulhu_vx_w, 4, 4)
-GEN_VEXT_VX(vmulhu_vx_d, 8, 8)
-GEN_VEXT_VX(vmulhsu_vx_b, 1, 1)
-GEN_VEXT_VX(vmulhsu_vx_h, 2, 2)
-GEN_VEXT_VX(vmulhsu_vx_w, 4, 4)
-GEN_VEXT_VX(vmulhsu_vx_d, 8, 8)
+GEN_VEXT_VX(vmul_vx_b)
+GEN_VEXT_VX(vmul_vx_h)
+GEN_VEXT_VX(vmul_vx_w)
+GEN_VEXT_VX(vmul_vx_d)
+GEN_VEXT_VX(vmulh_vx_b)
+GEN_VEXT_VX(vmulh_vx_h)
+GEN_VEXT_VX(vmulh_vx_w)
+GEN_VEXT_VX(vmulh_vx_d)
+GEN_VEXT_VX(vmulhu_vx_b)
+GEN_VEXT_VX(vmulhu_vx_h)
+GEN_VEXT_VX(vmulhu_vx_w)
+GEN_VEXT_VX(vmulhu_vx_d)
+GEN_VEXT_VX(vmulhsu_vx_b)
+GEN_VEXT_VX(vmulhsu_vx_h)
+GEN_VEXT_VX(vmulhsu_vx_w)
+GEN_VEXT_VX(vmulhsu_vx_d)
 
 /* Vector Integer Divide Instructions */
 #define DO_DIVU(N, M) (unlikely(M == 0) ? (__typeof(N))(-1) : N / M)
@@ -1581,22 +1579,22 @@ RVVCALL(OPIVV2, vrem_vv_b, OP_SSS_B, H1, H1, H1, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_h, OP_SSS_H, H2, H2, H2, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_w, OP_SSS_W, H4, H4, H4, DO_REM)
 RVVCALL(OPIVV2, vrem_vv_d, OP_SSS_D, H8, H8, H8, DO_REM)
-GEN_VEXT_VV(vdivu_vv_b, 1, 1)
-GEN_VEXT_VV(vdivu_vv_h, 2, 2)
-GEN_VEXT_VV(vdivu_vv_w, 4, 4)
-GEN_VEXT_VV(vdivu_vv_d, 8, 8)
-GEN_VEXT_VV(vdiv_vv_b, 1, 1)
-GEN_VEXT_VV(vdiv_vv_h, 2, 2)
-GEN_VEXT_VV(vdiv_vv_w, 4, 4)
-GEN_VEXT_VV(vdiv_vv_d, 8, 8)
-GEN_VEXT_VV(vremu_vv_b, 1, 1)
-GEN_VEXT_VV(vremu_vv_h, 2, 2)
-GEN_VEXT_VV(vremu_vv_w, 4, 4)
-GEN_VEXT_VV(vremu_vv_d, 8, 8)
-GEN_VEXT_VV(vrem_vv_b, 1, 1)
-GEN_VEXT_VV(vrem_vv_h, 2, 2)
-GEN_VEXT_VV(vrem_vv_w, 4, 4)
-GEN_VEXT_VV(vrem_vv_d, 8, 8)
+GEN_VEXT_VV(vdivu_vv_b)
+GEN_VEXT_VV(vdivu_vv_h)
+GEN_VEXT_VV(vdivu_vv_w)
+GEN_VEXT_VV(vdivu_vv_d)
+GEN_VEXT_VV(vdiv_vv_b)
+GEN_VEXT_VV(vdiv_vv_h)
+GEN_VEXT_VV(vdiv_vv_w)
+GEN_VEXT_VV(vdiv_vv_d)
+GEN_VEXT_VV(vremu_vv_b)
+GEN_VEXT_VV(vremu_vv_h)
+GEN_VEXT_VV(vremu_vv_w)
+GEN_VEXT_VV(vremu_vv_d)
+GEN_VEXT_VV(vrem_vv_b)
+GEN_VEXT_VV(vrem_vv_h)
+GEN_VEXT_VV(vrem_vv_w)
+GEN_VEXT_VV(vrem_vv_d)
 
 RVVCALL(OPIVX2, vdivu_vx_b, OP_UUU_B, H1, H1, DO_DIVU)
 RVVCALL(OPIVX2, vdivu_vx_h, OP_UUU_H, H2, H2, DO_DIVU)
@@ -1614,22 +1612,22 @@ RVVCALL(OPIVX2, vrem_vx_b, OP_SSS_B, H1, H1, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_h, OP_SSS_H, H2, H2, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_w, OP_SSS_W, H4, H4, DO_REM)
 RVVCALL(OPIVX2, vrem_vx_d, OP_SSS_D, H8, H8, DO_REM)
-GEN_VEXT_VX(vdivu_vx_b, 1, 1)
-GEN_VEXT_VX(vdivu_vx_h, 2, 2)
-GEN_VEXT_VX(vdivu_vx_w, 4, 4)
-GEN_VEXT_VX(vdivu_vx_d, 8, 8)
-GEN_VEXT_VX(vdiv_vx_b, 1, 1)
-GEN_VEXT_VX(vdiv_vx_h, 2, 2)
-GEN_VEXT_VX(vdiv_vx_w, 4, 4)
-GEN_VEXT_VX(vdiv_vx_d, 8, 8)
-GEN_VEXT_VX(vremu_vx_b, 1, 1)
-GEN_VEXT_VX(vremu_vx_h, 2, 2)
-GEN_VEXT_VX(vremu_vx_w, 4, 4)
-GEN_VEXT_VX(vremu_vx_d, 8, 8)
-GEN_VEXT_VX(vrem_vx_b, 1, 1)
-GEN_VEXT_VX(vrem_vx_h, 2, 2)
-GEN_VEXT_VX(vrem_vx_w, 4, 4)
-GEN_VEXT_VX(vrem_vx_d, 8, 8)
+GEN_VEXT_VX(vdivu_vx_b)
+GEN_VEXT_VX(vdivu_vx_h)
+GEN_VEXT_VX(vdivu_vx_w)
+GEN_VEXT_VX(vdivu_vx_d)
+GEN_VEXT_VX(vdiv_vx_b)
+GEN_VEXT_VX(vdiv_vx_h)
+GEN_VEXT_VX(vdiv_vx_w)
+GEN_VEXT_VX(vdiv_vx_d)
+GEN_VEXT_VX(vremu_vx_b)
+GEN_VEXT_VX(vremu_vx_h)
+GEN_VEXT_VX(vremu_vx_w)
+GEN_VEXT_VX(vremu_vx_d)
+GEN_VEXT_VX(vrem_vx_b)
+GEN_VEXT_VX(vrem_vx_h)
+GEN_VEXT_VX(vrem_vx_w)
+GEN_VEXT_VX(vrem_vx_d)
 
 /* Vector Widening Integer Multiply Instructions */
 RVVCALL(OPIVV2, vwmul_vv_b, WOP_SSS_B, H2, H1, H1, DO_MUL)
@@ -1641,15 +1639,15 @@ RVVCALL(OPIVV2, vwmulu_vv_w, WOP_UUU_W, H8, H4, H4, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_b, WOP_SUS_B, H2, H1, H1, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_h, WOP_SUS_H, H4, H2, H2, DO_MUL)
 RVVCALL(OPIVV2, vwmulsu_vv_w, WOP_SUS_W, H8, H4, H4, DO_MUL)
-GEN_VEXT_VV(vwmul_vv_b, 1, 2)
-GEN_VEXT_VV(vwmul_vv_h, 2, 4)
-GEN_VEXT_VV(vwmul_vv_w, 4, 8)
-GEN_VEXT_VV(vwmulu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmulu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmulu_vv_w, 4, 8)
-GEN_VEXT_VV(vwmulsu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmulsu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmulsu_vv_w, 4, 8)
+GEN_VEXT_VV(vwmul_vv_b)
+GEN_VEXT_VV(vwmul_vv_h)
+GEN_VEXT_VV(vwmul_vv_w)
+GEN_VEXT_VV(vwmulu_vv_b)
+GEN_VEXT_VV(vwmulu_vv_h)
+GEN_VEXT_VV(vwmulu_vv_w)
+GEN_VEXT_VV(vwmulsu_vv_b)
+GEN_VEXT_VV(vwmulsu_vv_h)
+GEN_VEXT_VV(vwmulsu_vv_w)
 
 RVVCALL(OPIVX2, vwmul_vx_b, WOP_SSS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmul_vx_h, WOP_SSS_H, H4, H2, DO_MUL)
@@ -1660,15 +1658,15 @@ RVVCALL(OPIVX2, vwmulu_vx_w, WOP_UUU_W, H8, H4, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_b, WOP_SUS_B, H2, H1, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_h, WOP_SUS_H, H4, H2, DO_MUL)
 RVVCALL(OPIVX2, vwmulsu_vx_w, WOP_SUS_W, H8, H4, DO_MUL)
-GEN_VEXT_VX(vwmul_vx_b, 1, 2)
-GEN_VEXT_VX(vwmul_vx_h, 2, 4)
-GEN_VEXT_VX(vwmul_vx_w, 4, 8)
-GEN_VEXT_VX(vwmulu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmulu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmulu_vx_w, 4, 8)
-GEN_VEXT_VX(vwmulsu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmulsu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmulsu_vx_w, 4, 8)
+GEN_VEXT_VX(vwmul_vx_b)
+GEN_VEXT_VX(vwmul_vx_h)
+GEN_VEXT_VX(vwmul_vx_w)
+GEN_VEXT_VX(vwmulu_vx_b)
+GEN_VEXT_VX(vwmulu_vx_h)
+GEN_VEXT_VX(vwmulu_vx_w)
+GEN_VEXT_VX(vwmulsu_vx_b)
+GEN_VEXT_VX(vwmulsu_vx_h)
+GEN_VEXT_VX(vwmulsu_vx_w)
 
 /* Vector Single-Width Integer Multiply-Add Instructions */
 #define OPIVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)   \
@@ -1700,22 +1698,22 @@ RVVCALL(OPIVV3, vnmsub_vv_b, OP_SSS_B, H1, H1, H1, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_h, OP_SSS_H, H2, H2, H2, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_w, OP_SSS_W, H4, H4, H4, DO_NMSUB)
 RVVCALL(OPIVV3, vnmsub_vv_d, OP_SSS_D, H8, H8, H8, DO_NMSUB)
-GEN_VEXT_VV(vmacc_vv_b, 1, 1)
-GEN_VEXT_VV(vmacc_vv_h, 2, 2)
-GEN_VEXT_VV(vmacc_vv_w, 4, 4)
-GEN_VEXT_VV(vmacc_vv_d, 8, 8)
-GEN_VEXT_VV(vnmsac_vv_b, 1, 1)
-GEN_VEXT_VV(vnmsac_vv_h, 2, 2)
-GEN_VEXT_VV(vnmsac_vv_w, 4, 4)
-GEN_VEXT_VV(vnmsac_vv_d, 8, 8)
-GEN_VEXT_VV(vmadd_vv_b, 1, 1)
-GEN_VEXT_VV(vmadd_vv_h, 2, 2)
-GEN_VEXT_VV(vmadd_vv_w, 4, 4)
-GEN_VEXT_VV(vmadd_vv_d, 8, 8)
-GEN_VEXT_VV(vnmsub_vv_b, 1, 1)
-GEN_VEXT_VV(vnmsub_vv_h, 2, 2)
-GEN_VEXT_VV(vnmsub_vv_w, 4, 4)
-GEN_VEXT_VV(vnmsub_vv_d, 8, 8)
+GEN_VEXT_VV(vmacc_vv_b)
+GEN_VEXT_VV(vmacc_vv_h)
+GEN_VEXT_VV(vmacc_vv_w)
+GEN_VEXT_VV(vmacc_vv_d)
+GEN_VEXT_VV(vnmsac_vv_b)
+GEN_VEXT_VV(vnmsac_vv_h)
+GEN_VEXT_VV(vnmsac_vv_w)
+GEN_VEXT_VV(vnmsac_vv_d)
+GEN_VEXT_VV(vmadd_vv_b)
+GEN_VEXT_VV(vmadd_vv_h)
+GEN_VEXT_VV(vmadd_vv_w)
+GEN_VEXT_VV(vmadd_vv_d)
+GEN_VEXT_VV(vnmsub_vv_b)
+GEN_VEXT_VV(vnmsub_vv_h)
+GEN_VEXT_VV(vnmsub_vv_w)
+GEN_VEXT_VV(vnmsub_vv_d)
 
 #define OPIVX3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)             \
 static void do_##NAME(void *vd, target_long s1, void *vs2, int i)   \
@@ -1741,22 +1739,22 @@ RVVCALL(OPIVX3, vnmsub_vx_b, OP_SSS_B, H1, H1, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_h, OP_SSS_H, H2, H2, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_w, OP_SSS_W, H4, H4, DO_NMSUB)
 RVVCALL(OPIVX3, vnmsub_vx_d, OP_SSS_D, H8, H8, DO_NMSUB)
-GEN_VEXT_VX(vmacc_vx_b, 1, 1)
-GEN_VEXT_VX(vmacc_vx_h, 2, 2)
-GEN_VEXT_VX(vmacc_vx_w, 4, 4)
-GEN_VEXT_VX(vmacc_vx_d, 8, 8)
-GEN_VEXT_VX(vnmsac_vx_b, 1, 1)
-GEN_VEXT_VX(vnmsac_vx_h, 2, 2)
-GEN_VEXT_VX(vnmsac_vx_w, 4, 4)
-GEN_VEXT_VX(vnmsac_vx_d, 8, 8)
-GEN_VEXT_VX(vmadd_vx_b, 1, 1)
-GEN_VEXT_VX(vmadd_vx_h, 2, 2)
-GEN_VEXT_VX(vmadd_vx_w, 4, 4)
-GEN_VEXT_VX(vmadd_vx_d, 8, 8)
-GEN_VEXT_VX(vnmsub_vx_b, 1, 1)
-GEN_VEXT_VX(vnmsub_vx_h, 2, 2)
-GEN_VEXT_VX(vnmsub_vx_w, 4, 4)
-GEN_VEXT_VX(vnmsub_vx_d, 8, 8)
+GEN_VEXT_VX(vmacc_vx_b)
+GEN_VEXT_VX(vmacc_vx_h)
+GEN_VEXT_VX(vmacc_vx_w)
+GEN_VEXT_VX(vmacc_vx_d)
+GEN_VEXT_VX(vnmsac_vx_b)
+GEN_VEXT_VX(vnmsac_vx_h)
+GEN_VEXT_VX(vnmsac_vx_w)
+GEN_VEXT_VX(vnmsac_vx_d)
+GEN_VEXT_VX(vmadd_vx_b)
+GEN_VEXT_VX(vmadd_vx_h)
+GEN_VEXT_VX(vmadd_vx_w)
+GEN_VEXT_VX(vmadd_vx_d)
+GEN_VEXT_VX(vnmsub_vx_b)
+GEN_VEXT_VX(vnmsub_vx_h)
+GEN_VEXT_VX(vnmsub_vx_w)
+GEN_VEXT_VX(vnmsub_vx_d)
 
 /* Vector Widening Integer Multiply-Add Instructions */
 RVVCALL(OPIVV3, vwmaccu_vv_b, WOP_UUU_B, H2, H1, H1, DO_MACC)
@@ -1768,15 +1766,15 @@ RVVCALL(OPIVV3, vwmacc_vv_w, WOP_SSS_W, H8, H4, H4, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_b, WOP_SSU_B, H2, H1, H1, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_h, WOP_SSU_H, H4, H2, H2, DO_MACC)
 RVVCALL(OPIVV3, vwmaccsu_vv_w, WOP_SSU_W, H8, H4, H4, DO_MACC)
-GEN_VEXT_VV(vwmaccu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmaccu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmaccu_vv_w, 4, 8)
-GEN_VEXT_VV(vwmacc_vv_b, 1, 2)
-GEN_VEXT_VV(vwmacc_vv_h, 2, 4)
-GEN_VEXT_VV(vwmacc_vv_w, 4, 8)
-GEN_VEXT_VV(vwmaccsu_vv_b, 1, 2)
-GEN_VEXT_VV(vwmaccsu_vv_h, 2, 4)
-GEN_VEXT_VV(vwmaccsu_vv_w, 4, 8)
+GEN_VEXT_VV(vwmaccu_vv_b)
+GEN_VEXT_VV(vwmaccu_vv_h)
+GEN_VEXT_VV(vwmaccu_vv_w)
+GEN_VEXT_VV(vwmacc_vv_b)
+GEN_VEXT_VV(vwmacc_vv_h)
+GEN_VEXT_VV(vwmacc_vv_w)
+GEN_VEXT_VV(vwmaccsu_vv_b)
+GEN_VEXT_VV(vwmaccsu_vv_h)
+GEN_VEXT_VV(vwmaccsu_vv_w)
 
 RVVCALL(OPIVX3, vwmaccu_vx_b, WOP_UUU_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccu_vx_h, WOP_UUU_H, H4, H2, DO_MACC)
@@ -1790,18 +1788,18 @@ RVVCALL(OPIVX3, vwmaccsu_vx_w, WOP_SSU_W, H8, H4, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_b, WOP_SUS_B, H2, H1, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_h, WOP_SUS_H, H4, H2, DO_MACC)
 RVVCALL(OPIVX3, vwmaccus_vx_w, WOP_SUS_W, H8, H4, DO_MACC)
-GEN_VEXT_VX(vwmaccu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmaccu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmaccu_vx_w, 4, 8)
-GEN_VEXT_VX(vwmacc_vx_b, 1, 2)
-GEN_VEXT_VX(vwmacc_vx_h, 2, 4)
-GEN_VEXT_VX(vwmacc_vx_w, 4, 8)
-GEN_VEXT_VX(vwmaccsu_vx_b, 1, 2)
-GEN_VEXT_VX(vwmaccsu_vx_h, 2, 4)
-GEN_VEXT_VX(vwmaccsu_vx_w, 4, 8)
-GEN_VEXT_VX(vwmaccus_vx_b, 1, 2)
-GEN_VEXT_VX(vwmaccus_vx_h, 2, 4)
-GEN_VEXT_VX(vwmaccus_vx_w, 4, 8)
+GEN_VEXT_VX(vwmaccu_vx_b)
+GEN_VEXT_VX(vwmaccu_vx_h)
+GEN_VEXT_VX(vwmaccu_vx_w)
+GEN_VEXT_VX(vwmacc_vx_b)
+GEN_VEXT_VX(vwmacc_vx_h)
+GEN_VEXT_VX(vwmacc_vx_w)
+GEN_VEXT_VX(vwmaccsu_vx_b)
+GEN_VEXT_VX(vwmaccsu_vx_h)
+GEN_VEXT_VX(vwmaccsu_vx_w)
+GEN_VEXT_VX(vwmaccus_vx_b)
+GEN_VEXT_VX(vwmaccus_vx_h)
+GEN_VEXT_VX(vwmaccus_vx_w)
 
 /* Vector Integer Merge and Move Instructions */
 #define GEN_VEXT_VMV_VV(NAME, ETYPE, H)                              \
@@ -1922,7 +1920,7 @@ vext_vv_rm_1(void *vd, void *v0, void *vs1, void *vs2,
 static inline void
 vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
              CPURISCVState *env,
-             uint32_t desc, uint32_t esz, uint32_t dsz,
+             uint32_t desc,
              opivv2_rm_fn *fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -1949,11 +1947,11 @@ vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
 }
 
 /* generate helpers for fixed point instructions with OPIVV format */
-#define GEN_VEXT_VV_RM(NAME, ESZ, DSZ)                          \
+#define GEN_VEXT_VV_RM(NAME)                                    \
 void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,     \
                   CPURISCVState *env, uint32_t desc)            \
 {                                                               \
-    vext_vv_rm_2(vd, v0, vs1, vs2, env, desc, ESZ, DSZ,         \
+    vext_vv_rm_2(vd, v0, vs1, vs2, env, desc,                   \
                  do_##NAME);                                    \
 }
 
@@ -2004,10 +2002,10 @@ RVVCALL(OPIVV2_RM, vsaddu_vv_b, OP_UUU_B, H1, H1, H1, saddu8)
 RVVCALL(OPIVV2_RM, vsaddu_vv_h, OP_UUU_H, H2, H2, H2, saddu16)
 RVVCALL(OPIVV2_RM, vsaddu_vv_w, OP_UUU_W, H4, H4, H4, saddu32)
 RVVCALL(OPIVV2_RM, vsaddu_vv_d, OP_UUU_D, H8, H8, H8, saddu64)
-GEN_VEXT_VV_RM(vsaddu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vsaddu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vsaddu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vsaddu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vsaddu_vv_b)
+GEN_VEXT_VV_RM(vsaddu_vv_h)
+GEN_VEXT_VV_RM(vsaddu_vv_w)
+GEN_VEXT_VV_RM(vsaddu_vv_d)
 
 typedef void opivx2_rm_fn(void *vd, target_long s1, void *vs2, int i,
                           CPURISCVState *env, int vxrm);
@@ -2039,7 +2037,7 @@ vext_vx_rm_1(void *vd, void *v0, target_long s1, void *vs2,
 static inline void
 vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
              CPURISCVState *env,
-             uint32_t desc, uint32_t esz, uint32_t dsz,
+             uint32_t desc,
              opivx2_rm_fn *fn)
 {
     uint32_t vm = vext_vm(desc);
@@ -2066,11 +2064,11 @@ vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
 }
 
 /* generate helpers for fixed point instructions with OPIVX format */
-#define GEN_VEXT_VX_RM(NAME, ESZ, DSZ)                    \
+#define GEN_VEXT_VX_RM(NAME)                              \
 void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
         void *vs2, CPURISCVState *env, uint32_t desc)     \
 {                                                         \
-    vext_vx_rm_2(vd, v0, s1, vs2, env, desc, ESZ, DSZ,    \
+    vext_vx_rm_2(vd, v0, s1, vs2, env, desc,              \
                  do_##NAME);                              \
 }
 
@@ -2078,10 +2076,10 @@ RVVCALL(OPIVX2_RM, vsaddu_vx_b, OP_UUU_B, H1, H1, saddu8)
 RVVCALL(OPIVX2_RM, vsaddu_vx_h, OP_UUU_H, H2, H2, saddu16)
 RVVCALL(OPIVX2_RM, vsaddu_vx_w, OP_UUU_W, H4, H4, saddu32)
 RVVCALL(OPIVX2_RM, vsaddu_vx_d, OP_UUU_D, H8, H8, saddu64)
-GEN_VEXT_VX_RM(vsaddu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vsaddu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vsaddu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vsaddu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vsaddu_vx_b)
+GEN_VEXT_VX_RM(vsaddu_vx_h)
+GEN_VEXT_VX_RM(vsaddu_vx_w)
+GEN_VEXT_VX_RM(vsaddu_vx_d)
 
 static inline int8_t sadd8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2127,19 +2125,19 @@ RVVCALL(OPIVV2_RM, vsadd_vv_b, OP_SSS_B, H1, H1, H1, sadd8)
 RVVCALL(OPIVV2_RM, vsadd_vv_h, OP_SSS_H, H2, H2, H2, sadd16)
 RVVCALL(OPIVV2_RM, vsadd_vv_w, OP_SSS_W, H4, H4, H4, sadd32)
 RVVCALL(OPIVV2_RM, vsadd_vv_d, OP_SSS_D, H8, H8, H8, sadd64)
-GEN_VEXT_VV_RM(vsadd_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vsadd_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vsadd_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vsadd_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vsadd_vv_b)
+GEN_VEXT_VV_RM(vsadd_vv_h)
+GEN_VEXT_VV_RM(vsadd_vv_w)
+GEN_VEXT_VV_RM(vsadd_vv_d)
 
 RVVCALL(OPIVX2_RM, vsadd_vx_b, OP_SSS_B, H1, H1, sadd8)
 RVVCALL(OPIVX2_RM, vsadd_vx_h, OP_SSS_H, H2, H2, sadd16)
 RVVCALL(OPIVX2_RM, vsadd_vx_w, OP_SSS_W, H4, H4, sadd32)
 RVVCALL(OPIVX2_RM, vsadd_vx_d, OP_SSS_D, H8, H8, sadd64)
-GEN_VEXT_VX_RM(vsadd_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vsadd_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vsadd_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vsadd_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vsadd_vx_b)
+GEN_VEXT_VX_RM(vsadd_vx_h)
+GEN_VEXT_VX_RM(vsadd_vx_w)
+GEN_VEXT_VX_RM(vsadd_vx_d)
 
 static inline uint8_t ssubu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
 {
@@ -2188,19 +2186,19 @@ RVVCALL(OPIVV2_RM, vssubu_vv_b, OP_UUU_B, H1, H1, H1, ssubu8)
 RVVCALL(OPIVV2_RM, vssubu_vv_h, OP_UUU_H, H2, H2, H2, ssubu16)
 RVVCALL(OPIVV2_RM, vssubu_vv_w, OP_UUU_W, H4, H4, H4, ssubu32)
 RVVCALL(OPIVV2_RM, vssubu_vv_d, OP_UUU_D, H8, H8, H8, ssubu64)
-GEN_VEXT_VV_RM(vssubu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssubu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssubu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssubu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssubu_vv_b)
+GEN_VEXT_VV_RM(vssubu_vv_h)
+GEN_VEXT_VV_RM(vssubu_vv_w)
+GEN_VEXT_VV_RM(vssubu_vv_d)
 
 RVVCALL(OPIVX2_RM, vssubu_vx_b, OP_UUU_B, H1, H1, ssubu8)
 RVVCALL(OPIVX2_RM, vssubu_vx_h, OP_UUU_H, H2, H2, ssubu16)
 RVVCALL(OPIVX2_RM, vssubu_vx_w, OP_UUU_W, H4, H4, ssubu32)
 RVVCALL(OPIVX2_RM, vssubu_vx_d, OP_UUU_D, H8, H8, ssubu64)
-GEN_VEXT_VX_RM(vssubu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssubu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssubu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssubu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssubu_vx_b)
+GEN_VEXT_VX_RM(vssubu_vx_h)
+GEN_VEXT_VX_RM(vssubu_vx_w)
+GEN_VEXT_VX_RM(vssubu_vx_d)
 
 static inline int8_t ssub8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
 {
@@ -2246,19 +2244,19 @@ RVVCALL(OPIVV2_RM, vssub_vv_b, OP_SSS_B, H1, H1, H1, ssub8)
 RVVCALL(OPIVV2_RM, vssub_vv_h, OP_SSS_H, H2, H2, H2, ssub16)
 RVVCALL(OPIVV2_RM, vssub_vv_w, OP_SSS_W, H4, H4, H4, ssub32)
 RVVCALL(OPIVV2_RM, vssub_vv_d, OP_SSS_D, H8, H8, H8, ssub64)
-GEN_VEXT_VV_RM(vssub_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssub_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssub_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssub_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssub_vv_b)
+GEN_VEXT_VV_RM(vssub_vv_h)
+GEN_VEXT_VV_RM(vssub_vv_w)
+GEN_VEXT_VV_RM(vssub_vv_d)
 
 RVVCALL(OPIVX2_RM, vssub_vx_b, OP_SSS_B, H1, H1, ssub8)
 RVVCALL(OPIVX2_RM, vssub_vx_h, OP_SSS_H, H2, H2, ssub16)
 RVVCALL(OPIVX2_RM, vssub_vx_w, OP_SSS_W, H4, H4, ssub32)
 RVVCALL(OPIVX2_RM, vssub_vx_d, OP_SSS_D, H8, H8, ssub64)
-GEN_VEXT_VX_RM(vssub_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssub_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssub_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssub_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssub_vx_b)
+GEN_VEXT_VX_RM(vssub_vx_h)
+GEN_VEXT_VX_RM(vssub_vx_w)
+GEN_VEXT_VX_RM(vssub_vx_d)
 
 /* Vector Single-Width Averaging Add and Subtract */
 static inline uint8_t get_round(int vxrm, uint64_t v, uint8_t shift)
@@ -2310,19 +2308,19 @@ RVVCALL(OPIVV2_RM, vaadd_vv_b, OP_SSS_B, H1, H1, H1, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_h, OP_SSS_H, H2, H2, H2, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_w, OP_SSS_W, H4, H4, H4, aadd32)
 RVVCALL(OPIVV2_RM, vaadd_vv_d, OP_SSS_D, H8, H8, H8, aadd64)
-GEN_VEXT_VV_RM(vaadd_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vaadd_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vaadd_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vaadd_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vaadd_vv_b)
+GEN_VEXT_VV_RM(vaadd_vv_h)
+GEN_VEXT_VV_RM(vaadd_vv_w)
+GEN_VEXT_VV_RM(vaadd_vv_d)
 
 RVVCALL(OPIVX2_RM, vaadd_vx_b, OP_SSS_B, H1, H1, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_h, OP_SSS_H, H2, H2, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_w, OP_SSS_W, H4, H4, aadd32)
 RVVCALL(OPIVX2_RM, vaadd_vx_d, OP_SSS_D, H8, H8, aadd64)
-GEN_VEXT_VX_RM(vaadd_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vaadd_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vaadd_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vaadd_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vaadd_vx_b)
+GEN_VEXT_VX_RM(vaadd_vx_h)
+GEN_VEXT_VX_RM(vaadd_vx_w)
+GEN_VEXT_VX_RM(vaadd_vx_d)
 
 static inline uint32_t aaddu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2347,19 +2345,19 @@ RVVCALL(OPIVV2_RM, vaaddu_vv_b, OP_UUU_B, H1, H1, H1, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_h, OP_UUU_H, H2, H2, H2, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_w, OP_UUU_W, H4, H4, H4, aaddu32)
 RVVCALL(OPIVV2_RM, vaaddu_vv_d, OP_UUU_D, H8, H8, H8, aaddu64)
-GEN_VEXT_VV_RM(vaaddu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vaaddu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vaaddu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vaaddu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vaaddu_vv_b)
+GEN_VEXT_VV_RM(vaaddu_vv_h)
+GEN_VEXT_VV_RM(vaaddu_vv_w)
+GEN_VEXT_VV_RM(vaaddu_vv_d)
 
 RVVCALL(OPIVX2_RM, vaaddu_vx_b, OP_UUU_B, H1, H1, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_h, OP_UUU_H, H2, H2, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_w, OP_UUU_W, H4, H4, aaddu32)
 RVVCALL(OPIVX2_RM, vaaddu_vx_d, OP_UUU_D, H8, H8, aaddu64)
-GEN_VEXT_VX_RM(vaaddu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vaaddu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vaaddu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vaaddu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vaaddu_vx_b)
+GEN_VEXT_VX_RM(vaaddu_vx_h)
+GEN_VEXT_VX_RM(vaaddu_vx_w)
+GEN_VEXT_VX_RM(vaaddu_vx_d)
 
 static inline int32_t asub32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
 {
@@ -2383,19 +2381,19 @@ RVVCALL(OPIVV2_RM, vasub_vv_b, OP_SSS_B, H1, H1, H1, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_h, OP_SSS_H, H2, H2, H2, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_w, OP_SSS_W, H4, H4, H4, asub32)
 RVVCALL(OPIVV2_RM, vasub_vv_d, OP_SSS_D, H8, H8, H8, asub64)
-GEN_VEXT_VV_RM(vasub_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vasub_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vasub_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vasub_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vasub_vv_b)
+GEN_VEXT_VV_RM(vasub_vv_h)
+GEN_VEXT_VV_RM(vasub_vv_w)
+GEN_VEXT_VV_RM(vasub_vv_d)
 
 RVVCALL(OPIVX2_RM, vasub_vx_b, OP_SSS_B, H1, H1, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_h, OP_SSS_H, H2, H2, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_w, OP_SSS_W, H4, H4, asub32)
 RVVCALL(OPIVX2_RM, vasub_vx_d, OP_SSS_D, H8, H8, asub64)
-GEN_VEXT_VX_RM(vasub_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vasub_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vasub_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vasub_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vasub_vx_b)
+GEN_VEXT_VX_RM(vasub_vx_h)
+GEN_VEXT_VX_RM(vasub_vx_w)
+GEN_VEXT_VX_RM(vasub_vx_d)
 
 static inline uint32_t asubu32(CPURISCVState *env, int vxrm,
                                uint32_t a, uint32_t b)
@@ -2420,19 +2418,19 @@ RVVCALL(OPIVV2_RM, vasubu_vv_b, OP_UUU_B, H1, H1, H1, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_h, OP_UUU_H, H2, H2, H2, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_w, OP_UUU_W, H4, H4, H4, asubu32)
 RVVCALL(OPIVV2_RM, vasubu_vv_d, OP_UUU_D, H8, H8, H8, asubu64)
-GEN_VEXT_VV_RM(vasubu_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vasubu_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vasubu_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vasubu_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vasubu_vv_b)
+GEN_VEXT_VV_RM(vasubu_vv_h)
+GEN_VEXT_VV_RM(vasubu_vv_w)
+GEN_VEXT_VV_RM(vasubu_vv_d)
 
 RVVCALL(OPIVX2_RM, vasubu_vx_b, OP_UUU_B, H1, H1, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_h, OP_UUU_H, H2, H2, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_w, OP_UUU_W, H4, H4, asubu32)
 RVVCALL(OPIVX2_RM, vasubu_vx_d, OP_UUU_D, H8, H8, asubu64)
-GEN_VEXT_VX_RM(vasubu_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vasubu_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vasubu_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vasubu_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vasubu_vx_b)
+GEN_VEXT_VX_RM(vasubu_vx_h)
+GEN_VEXT_VX_RM(vasubu_vx_w)
+GEN_VEXT_VX_RM(vasubu_vx_d)
 
 /* Vector Single-Width Fractional Multiply with Rounding and Saturation */
 static inline int8_t vsmul8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2527,19 +2525,19 @@ RVVCALL(OPIVV2_RM, vsmul_vv_b, OP_SSS_B, H1, H1, H1, vsmul8)
 RVVCALL(OPIVV2_RM, vsmul_vv_h, OP_SSS_H, H2, H2, H2, vsmul16)
 RVVCALL(OPIVV2_RM, vsmul_vv_w, OP_SSS_W, H4, H4, H4, vsmul32)
 RVVCALL(OPIVV2_RM, vsmul_vv_d, OP_SSS_D, H8, H8, H8, vsmul64)
-GEN_VEXT_VV_RM(vsmul_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vsmul_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vsmul_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vsmul_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vsmul_vv_b)
+GEN_VEXT_VV_RM(vsmul_vv_h)
+GEN_VEXT_VV_RM(vsmul_vv_w)
+GEN_VEXT_VV_RM(vsmul_vv_d)
 
 RVVCALL(OPIVX2_RM, vsmul_vx_b, OP_SSS_B, H1, H1, vsmul8)
 RVVCALL(OPIVX2_RM, vsmul_vx_h, OP_SSS_H, H2, H2, vsmul16)
 RVVCALL(OPIVX2_RM, vsmul_vx_w, OP_SSS_W, H4, H4, vsmul32)
 RVVCALL(OPIVX2_RM, vsmul_vx_d, OP_SSS_D, H8, H8, vsmul64)
-GEN_VEXT_VX_RM(vsmul_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vsmul_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vsmul_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vsmul_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vsmul_vx_b)
+GEN_VEXT_VX_RM(vsmul_vx_h)
+GEN_VEXT_VX_RM(vsmul_vx_w)
+GEN_VEXT_VX_RM(vsmul_vx_d)
 
 /* Vector Single-Width Scaling Shift Instructions */
 static inline uint8_t
@@ -2586,19 +2584,19 @@ RVVCALL(OPIVV2_RM, vssrl_vv_b, OP_UUU_B, H1, H1, H1, vssrl8)
 RVVCALL(OPIVV2_RM, vssrl_vv_h, OP_UUU_H, H2, H2, H2, vssrl16)
 RVVCALL(OPIVV2_RM, vssrl_vv_w, OP_UUU_W, H4, H4, H4, vssrl32)
 RVVCALL(OPIVV2_RM, vssrl_vv_d, OP_UUU_D, H8, H8, H8, vssrl64)
-GEN_VEXT_VV_RM(vssrl_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssrl_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssrl_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssrl_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssrl_vv_b)
+GEN_VEXT_VV_RM(vssrl_vv_h)
+GEN_VEXT_VV_RM(vssrl_vv_w)
+GEN_VEXT_VV_RM(vssrl_vv_d)
 
 RVVCALL(OPIVX2_RM, vssrl_vx_b, OP_UUU_B, H1, H1, vssrl8)
 RVVCALL(OPIVX2_RM, vssrl_vx_h, OP_UUU_H, H2, H2, vssrl16)
 RVVCALL(OPIVX2_RM, vssrl_vx_w, OP_UUU_W, H4, H4, vssrl32)
 RVVCALL(OPIVX2_RM, vssrl_vx_d, OP_UUU_D, H8, H8, vssrl64)
-GEN_VEXT_VX_RM(vssrl_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssrl_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssrl_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssrl_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssrl_vx_b)
+GEN_VEXT_VX_RM(vssrl_vx_h)
+GEN_VEXT_VX_RM(vssrl_vx_w)
+GEN_VEXT_VX_RM(vssrl_vx_d)
 
 static inline int8_t
 vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
@@ -2645,19 +2643,19 @@ RVVCALL(OPIVV2_RM, vssra_vv_b, OP_SSS_B, H1, H1, H1, vssra8)
 RVVCALL(OPIVV2_RM, vssra_vv_h, OP_SSS_H, H2, H2, H2, vssra16)
 RVVCALL(OPIVV2_RM, vssra_vv_w, OP_SSS_W, H4, H4, H4, vssra32)
 RVVCALL(OPIVV2_RM, vssra_vv_d, OP_SSS_D, H8, H8, H8, vssra64)
-GEN_VEXT_VV_RM(vssra_vv_b, 1, 1)
-GEN_VEXT_VV_RM(vssra_vv_h, 2, 2)
-GEN_VEXT_VV_RM(vssra_vv_w, 4, 4)
-GEN_VEXT_VV_RM(vssra_vv_d, 8, 8)
+GEN_VEXT_VV_RM(vssra_vv_b)
+GEN_VEXT_VV_RM(vssra_vv_h)
+GEN_VEXT_VV_RM(vssra_vv_w)
+GEN_VEXT_VV_RM(vssra_vv_d)
 
 RVVCALL(OPIVX2_RM, vssra_vx_b, OP_SSS_B, H1, H1, vssra8)
 RVVCALL(OPIVX2_RM, vssra_vx_h, OP_SSS_H, H2, H2, vssra16)
 RVVCALL(OPIVX2_RM, vssra_vx_w, OP_SSS_W, H4, H4, vssra32)
 RVVCALL(OPIVX2_RM, vssra_vx_d, OP_SSS_D, H8, H8, vssra64)
-GEN_VEXT_VX_RM(vssra_vx_b, 1, 1)
-GEN_VEXT_VX_RM(vssra_vx_h, 2, 2)
-GEN_VEXT_VX_RM(vssra_vx_w, 4, 4)
-GEN_VEXT_VX_RM(vssra_vx_d, 8, 8)
+GEN_VEXT_VX_RM(vssra_vx_b)
+GEN_VEXT_VX_RM(vssra_vx_h)
+GEN_VEXT_VX_RM(vssra_vx_w)
+GEN_VEXT_VX_RM(vssra_vx_d)
 
 /* Vector Narrowing Fixed-Point Clip Instructions */
 static inline int8_t
@@ -2720,16 +2718,16 @@ vnclip32(CPURISCVState *env, int vxrm, int64_t a, int32_t b)
 RVVCALL(OPIVV2_RM, vnclip_wv_b, NOP_SSS_B, H1, H2, H1, vnclip8)
 RVVCALL(OPIVV2_RM, vnclip_wv_h, NOP_SSS_H, H2, H4, H2, vnclip16)
 RVVCALL(OPIVV2_RM, vnclip_wv_w, NOP_SSS_W, H4, H8, H4, vnclip32)
-GEN_VEXT_VV_RM(vnclip_wv_b, 1, 1)
-GEN_VEXT_VV_RM(vnclip_wv_h, 2, 2)
-GEN_VEXT_VV_RM(vnclip_wv_w, 4, 4)
+GEN_VEXT_VV_RM(vnclip_wv_b)
+GEN_VEXT_VV_RM(vnclip_wv_h)
+GEN_VEXT_VV_RM(vnclip_wv_w)
 
 RVVCALL(OPIVX2_RM, vnclip_wx_b, NOP_SSS_B, H1, H2, vnclip8)
 RVVCALL(OPIVX2_RM, vnclip_wx_h, NOP_SSS_H, H2, H4, vnclip16)
 RVVCALL(OPIVX2_RM, vnclip_wx_w, NOP_SSS_W, H4, H8, vnclip32)
-GEN_VEXT_VX_RM(vnclip_wx_b, 1, 1)
-GEN_VEXT_VX_RM(vnclip_wx_h, 2, 2)
-GEN_VEXT_VX_RM(vnclip_wx_w, 4, 4)
+GEN_VEXT_VX_RM(vnclip_wx_b)
+GEN_VEXT_VX_RM(vnclip_wx_h)
+GEN_VEXT_VX_RM(vnclip_wx_w)
 
 static inline uint8_t
 vnclipu8(CPURISCVState *env, int vxrm, uint16_t a, uint8_t b)
@@ -2782,16 +2780,16 @@ vnclipu32(CPURISCVState *env, int vxrm, uint64_t a, uint32_t b)
 RVVCALL(OPIVV2_RM, vnclipu_wv_b, NOP_UUU_B, H1, H2, H1, vnclipu8)
 RVVCALL(OPIVV2_RM, vnclipu_wv_h, NOP_UUU_H, H2, H4, H2, vnclipu16)
 RVVCALL(OPIVV2_RM, vnclipu_wv_w, NOP_UUU_W, H4, H8, H4, vnclipu32)
-GEN_VEXT_VV_RM(vnclipu_wv_b, 1, 1)
-GEN_VEXT_VV_RM(vnclipu_wv_h, 2, 2)
-GEN_VEXT_VV_RM(vnclipu_wv_w, 4, 4)
+GEN_VEXT_VV_RM(vnclipu_wv_b)
+GEN_VEXT_VV_RM(vnclipu_wv_h)
+GEN_VEXT_VV_RM(vnclipu_wv_w)
 
 RVVCALL(OPIVX2_RM, vnclipu_wx_b, NOP_UUU_B, H1, H2, vnclipu8)
 RVVCALL(OPIVX2_RM, vnclipu_wx_h, NOP_UUU_H, H2, H4, vnclipu16)
 RVVCALL(OPIVX2_RM, vnclipu_wx_w, NOP_UUU_W, H4, H8, vnclipu32)
-GEN_VEXT_VX_RM(vnclipu_wx_b, 1, 1)
-GEN_VEXT_VX_RM(vnclipu_wx_h, 2, 2)
-GEN_VEXT_VX_RM(vnclipu_wx_w, 4, 4)
+GEN_VEXT_VX_RM(vnclipu_wx_b)
+GEN_VEXT_VX_RM(vnclipu_wx_h)
+GEN_VEXT_VX_RM(vnclipu_wx_w)
 
 /*
  *** Vector Float Point Arithmetic Instructions
@@ -2806,7 +2804,7 @@ static void do_##NAME(void *vd, void *vs1, void *vs2, int i,   \
     *((TD *)vd + HD(i)) = OP(s2, s1, &env->fp_status);         \
 }
 
-#define GEN_VEXT_VV_ENV(NAME, ESZ, DSZ)                   \
+#define GEN_VEXT_VV_ENV(NAME)                             \
 void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
@@ -2827,9 +2825,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
 RVVCALL(OPFVV2, vfadd_vv_h, OP_UUU_H, H2, H2, H2, float16_add)
 RVVCALL(OPFVV2, vfadd_vv_w, OP_UUU_W, H4, H4, H4, float32_add)
 RVVCALL(OPFVV2, vfadd_vv_d, OP_UUU_D, H8, H8, H8, float64_add)
-GEN_VEXT_VV_ENV(vfadd_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfadd_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfadd_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfadd_vv_h)
+GEN_VEXT_VV_ENV(vfadd_vv_w)
+GEN_VEXT_VV_ENV(vfadd_vv_d)
 
 #define OPFVF2(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)        \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
@@ -2839,7 +2837,7 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i, \
     *((TD *)vd + HD(i)) = OP(s2, (TX1)(T1)s1, &env->fp_status);\
 }
 
-#define GEN_VEXT_VF(NAME, ESZ, DSZ)                       \
+#define GEN_VEXT_VF(NAME)                                 \
 void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
                   void *vs2, CPURISCVState *env,          \
                   uint32_t desc)                          \
@@ -2860,22 +2858,22 @@ void HELPER(NAME)(void *vd, void *v0, uint64_t s1,        \
 RVVCALL(OPFVF2, vfadd_vf_h, OP_UUU_H, H2, H2, float16_add)
 RVVCALL(OPFVF2, vfadd_vf_w, OP_UUU_W, H4, H4, float32_add)
 RVVCALL(OPFVF2, vfadd_vf_d, OP_UUU_D, H8, H8, float64_add)
-GEN_VEXT_VF(vfadd_vf_h, 2, 2)
-GEN_VEXT_VF(vfadd_vf_w, 4, 4)
-GEN_VEXT_VF(vfadd_vf_d, 8, 8)
+GEN_VEXT_VF(vfadd_vf_h)
+GEN_VEXT_VF(vfadd_vf_w)
+GEN_VEXT_VF(vfadd_vf_d)
 
 RVVCALL(OPFVV2, vfsub_vv_h, OP_UUU_H, H2, H2, H2, float16_sub)
 RVVCALL(OPFVV2, vfsub_vv_w, OP_UUU_W, H4, H4, H4, float32_sub)
 RVVCALL(OPFVV2, vfsub_vv_d, OP_UUU_D, H8, H8, H8, float64_sub)
-GEN_VEXT_VV_ENV(vfsub_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsub_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsub_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsub_vv_h)
+GEN_VEXT_VV_ENV(vfsub_vv_w)
+GEN_VEXT_VV_ENV(vfsub_vv_d)
 RVVCALL(OPFVF2, vfsub_vf_h, OP_UUU_H, H2, H2, float16_sub)
 RVVCALL(OPFVF2, vfsub_vf_w, OP_UUU_W, H4, H4, float32_sub)
 RVVCALL(OPFVF2, vfsub_vf_d, OP_UUU_D, H8, H8, float64_sub)
-GEN_VEXT_VF(vfsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfsub_vf_h)
+GEN_VEXT_VF(vfsub_vf_w)
+GEN_VEXT_VF(vfsub_vf_d)
 
 static uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s)
 {
@@ -2895,9 +2893,9 @@ static uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrsub_vf_h, OP_UUU_H, H2, H2, float16_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_w, OP_UUU_W, H4, H4, float32_rsub)
 RVVCALL(OPFVF2, vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub)
-GEN_VEXT_VF(vfrsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfrsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfrsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfrsub_vf_h)
+GEN_VEXT_VF(vfrsub_vf_w)
+GEN_VEXT_VF(vfrsub_vf_d)
 
 /* Vector Widening Floating-Point Add/Subtract Instructions */
 static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s)
@@ -2915,12 +2913,12 @@ static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16)
 RVVCALL(OPFVV2, vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32)
-GEN_VEXT_VV_ENV(vfwadd_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwadd_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwadd_vv_h)
+GEN_VEXT_VV_ENV(vfwadd_vv_w)
 RVVCALL(OPFVF2, vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16)
 RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32)
-GEN_VEXT_VF(vfwadd_vf_h, 2, 4)
-GEN_VEXT_VF(vfwadd_vf_w, 4, 8)
+GEN_VEXT_VF(vfwadd_vf_h)
+GEN_VEXT_VF(vfwadd_vf_w)
 
 static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -2937,12 +2935,12 @@ static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16)
 RVVCALL(OPFVV2, vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32)
-GEN_VEXT_VV_ENV(vfwsub_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwsub_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwsub_vv_h)
+GEN_VEXT_VV_ENV(vfwsub_vv_w)
 RVVCALL(OPFVF2, vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16)
 RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32)
-GEN_VEXT_VF(vfwsub_vf_h, 2, 4)
-GEN_VEXT_VF(vfwsub_vf_w, 4, 8)
+GEN_VEXT_VF(vfwsub_vf_h)
+GEN_VEXT_VF(vfwsub_vf_w)
 
 static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -2956,12 +2954,12 @@ static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16)
 RVVCALL(OPFVV2, vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32)
-GEN_VEXT_VV_ENV(vfwadd_wv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwadd_wv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwadd_wv_h)
+GEN_VEXT_VV_ENV(vfwadd_wv_w)
 RVVCALL(OPFVF2, vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16)
 RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32)
-GEN_VEXT_VF(vfwadd_wf_h, 2, 4)
-GEN_VEXT_VF(vfwadd_wf_w, 4, 8)
+GEN_VEXT_VF(vfwadd_wf_h)
+GEN_VEXT_VF(vfwadd_wf_w)
 
 static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s)
 {
@@ -2975,39 +2973,39 @@ static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s)
 
 RVVCALL(OPFVV2, vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16)
 RVVCALL(OPFVV2, vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32)
-GEN_VEXT_VV_ENV(vfwsub_wv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwsub_wv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwsub_wv_h)
+GEN_VEXT_VV_ENV(vfwsub_wv_w)
 RVVCALL(OPFVF2, vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16)
 RVVCALL(OPFVF2, vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32)
-GEN_VEXT_VF(vfwsub_wf_h, 2, 4)
-GEN_VEXT_VF(vfwsub_wf_w, 4, 8)
+GEN_VEXT_VF(vfwsub_wf_h)
+GEN_VEXT_VF(vfwsub_wf_w)
 
 /* Vector Single-Width Floating-Point Multiply/Divide Instructions */
 RVVCALL(OPFVV2, vfmul_vv_h, OP_UUU_H, H2, H2, H2, float16_mul)
 RVVCALL(OPFVV2, vfmul_vv_w, OP_UUU_W, H4, H4, H4, float32_mul)
 RVVCALL(OPFVV2, vfmul_vv_d, OP_UUU_D, H8, H8, H8, float64_mul)
-GEN_VEXT_VV_ENV(vfmul_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmul_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmul_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmul_vv_h)
+GEN_VEXT_VV_ENV(vfmul_vv_w)
+GEN_VEXT_VV_ENV(vfmul_vv_d)
 RVVCALL(OPFVF2, vfmul_vf_h, OP_UUU_H, H2, H2, float16_mul)
 RVVCALL(OPFVF2, vfmul_vf_w, OP_UUU_W, H4, H4, float32_mul)
 RVVCALL(OPFVF2, vfmul_vf_d, OP_UUU_D, H8, H8, float64_mul)
-GEN_VEXT_VF(vfmul_vf_h, 2, 2)
-GEN_VEXT_VF(vfmul_vf_w, 4, 4)
-GEN_VEXT_VF(vfmul_vf_d, 8, 8)
+GEN_VEXT_VF(vfmul_vf_h)
+GEN_VEXT_VF(vfmul_vf_w)
+GEN_VEXT_VF(vfmul_vf_d)
 
 RVVCALL(OPFVV2, vfdiv_vv_h, OP_UUU_H, H2, H2, H2, float16_div)
 RVVCALL(OPFVV2, vfdiv_vv_w, OP_UUU_W, H4, H4, H4, float32_div)
 RVVCALL(OPFVV2, vfdiv_vv_d, OP_UUU_D, H8, H8, H8, float64_div)
-GEN_VEXT_VV_ENV(vfdiv_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfdiv_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfdiv_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfdiv_vv_h)
+GEN_VEXT_VV_ENV(vfdiv_vv_w)
+GEN_VEXT_VV_ENV(vfdiv_vv_d)
 RVVCALL(OPFVF2, vfdiv_vf_h, OP_UUU_H, H2, H2, float16_div)
 RVVCALL(OPFVF2, vfdiv_vf_w, OP_UUU_W, H4, H4, float32_div)
 RVVCALL(OPFVF2, vfdiv_vf_d, OP_UUU_D, H8, H8, float64_div)
-GEN_VEXT_VF(vfdiv_vf_h, 2, 2)
-GEN_VEXT_VF(vfdiv_vf_w, 4, 4)
-GEN_VEXT_VF(vfdiv_vf_d, 8, 8)
+GEN_VEXT_VF(vfdiv_vf_h)
+GEN_VEXT_VF(vfdiv_vf_w)
+GEN_VEXT_VF(vfdiv_vf_d)
 
 static uint16_t float16_rdiv(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3027,9 +3025,9 @@ static uint64_t float64_rdiv(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVF2, vfrdiv_vf_h, OP_UUU_H, H2, H2, float16_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_w, OP_UUU_W, H4, H4, float32_rdiv)
 RVVCALL(OPFVF2, vfrdiv_vf_d, OP_UUU_D, H8, H8, float64_rdiv)
-GEN_VEXT_VF(vfrdiv_vf_h, 2, 2)
-GEN_VEXT_VF(vfrdiv_vf_w, 4, 4)
-GEN_VEXT_VF(vfrdiv_vf_d, 8, 8)
+GEN_VEXT_VF(vfrdiv_vf_h)
+GEN_VEXT_VF(vfrdiv_vf_w)
+GEN_VEXT_VF(vfrdiv_vf_d)
 
 /* Vector Widening Floating-Point Multiply */
 static uint32_t vfwmul16(uint16_t a, uint16_t b, float_status *s)
@@ -3046,12 +3044,12 @@ static uint64_t vfwmul32(uint32_t a, uint32_t b, float_status *s)
 }
 RVVCALL(OPFVV2, vfwmul_vv_h, WOP_UUU_H, H4, H2, H2, vfwmul16)
 RVVCALL(OPFVV2, vfwmul_vv_w, WOP_UUU_W, H8, H4, H4, vfwmul32)
-GEN_VEXT_VV_ENV(vfwmul_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwmul_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwmul_vv_h)
+GEN_VEXT_VV_ENV(vfwmul_vv_w)
 RVVCALL(OPFVF2, vfwmul_vf_h, WOP_UUU_H, H4, H2, vfwmul16)
 RVVCALL(OPFVF2, vfwmul_vf_w, WOP_UUU_W, H8, H4, vfwmul32)
-GEN_VEXT_VF(vfwmul_vf_h, 2, 4)
-GEN_VEXT_VF(vfwmul_vf_w, 4, 8)
+GEN_VEXT_VF(vfwmul_vf_h)
+GEN_VEXT_VF(vfwmul_vf_w)
 
 /* Vector Single-Width Floating-Point Fused Multiply-Add Instructions */
 #define OPFVV3(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)       \
@@ -3082,9 +3080,9 @@ static uint64_t fmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmacc_vv_h, OP_UUU_H, H2, H2, H2, fmacc16)
 RVVCALL(OPFVV3, vfmacc_vv_w, OP_UUU_W, H4, H4, H4, fmacc32)
 RVVCALL(OPFVV3, vfmacc_vv_d, OP_UUU_D, H8, H8, H8, fmacc64)
-GEN_VEXT_VV_ENV(vfmacc_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmacc_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmacc_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmacc_vv_h)
+GEN_VEXT_VV_ENV(vfmacc_vv_w)
+GEN_VEXT_VV_ENV(vfmacc_vv_d)
 
 #define OPFVF3(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)           \
 static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
@@ -3098,9 +3096,9 @@ static void do_##NAME(void *vd, uint64_t s1, void *vs2, int i,    \
 RVVCALL(OPFVF3, vfmacc_vf_h, OP_UUU_H, H2, H2, fmacc16)
 RVVCALL(OPFVF3, vfmacc_vf_w, OP_UUU_W, H4, H4, fmacc32)
 RVVCALL(OPFVF3, vfmacc_vf_d, OP_UUU_D, H8, H8, fmacc64)
-GEN_VEXT_VF(vfmacc_vf_h, 2, 2)
-GEN_VEXT_VF(vfmacc_vf_w, 4, 4)
-GEN_VEXT_VF(vfmacc_vf_d, 8, 8)
+GEN_VEXT_VF(vfmacc_vf_h)
+GEN_VEXT_VF(vfmacc_vf_w)
+GEN_VEXT_VF(vfmacc_vf_d)
 
 static uint16_t fnmacc16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3123,15 +3121,15 @@ static uint64_t fnmacc64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmacc_vv_h, OP_UUU_H, H2, H2, H2, fnmacc16)
 RVVCALL(OPFVV3, vfnmacc_vv_w, OP_UUU_W, H4, H4, H4, fnmacc32)
 RVVCALL(OPFVV3, vfnmacc_vv_d, OP_UUU_D, H8, H8, H8, fnmacc64)
-GEN_VEXT_VV_ENV(vfnmacc_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmacc_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmacc_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmacc_vv_h)
+GEN_VEXT_VV_ENV(vfnmacc_vv_w)
+GEN_VEXT_VV_ENV(vfnmacc_vv_d)
 RVVCALL(OPFVF3, vfnmacc_vf_h, OP_UUU_H, H2, H2, fnmacc16)
 RVVCALL(OPFVF3, vfnmacc_vf_w, OP_UUU_W, H4, H4, fnmacc32)
 RVVCALL(OPFVF3, vfnmacc_vf_d, OP_UUU_D, H8, H8, fnmacc64)
-GEN_VEXT_VF(vfnmacc_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmacc_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmacc_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmacc_vf_h)
+GEN_VEXT_VF(vfnmacc_vf_w)
+GEN_VEXT_VF(vfnmacc_vf_d)
 
 static uint16_t fmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3151,15 +3149,15 @@ static uint64_t fmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsac_vv_h, OP_UUU_H, H2, H2, H2, fmsac16)
 RVVCALL(OPFVV3, vfmsac_vv_w, OP_UUU_W, H4, H4, H4, fmsac32)
 RVVCALL(OPFVV3, vfmsac_vv_d, OP_UUU_D, H8, H8, H8, fmsac64)
-GEN_VEXT_VV_ENV(vfmsac_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmsac_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmsac_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmsac_vv_h)
+GEN_VEXT_VV_ENV(vfmsac_vv_w)
+GEN_VEXT_VV_ENV(vfmsac_vv_d)
 RVVCALL(OPFVF3, vfmsac_vf_h, OP_UUU_H, H2, H2, fmsac16)
 RVVCALL(OPFVF3, vfmsac_vf_w, OP_UUU_W, H4, H4, fmsac32)
 RVVCALL(OPFVF3, vfmsac_vf_d, OP_UUU_D, H8, H8, fmsac64)
-GEN_VEXT_VF(vfmsac_vf_h, 2, 2)
-GEN_VEXT_VF(vfmsac_vf_w, 4, 4)
-GEN_VEXT_VF(vfmsac_vf_d, 8, 8)
+GEN_VEXT_VF(vfmsac_vf_h)
+GEN_VEXT_VF(vfmsac_vf_w)
+GEN_VEXT_VF(vfmsac_vf_d)
 
 static uint16_t fnmsac16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3179,15 +3177,15 @@ static uint64_t fnmsac64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsac_vv_h, OP_UUU_H, H2, H2, H2, fnmsac16)
 RVVCALL(OPFVV3, vfnmsac_vv_w, OP_UUU_W, H4, H4, H4, fnmsac32)
 RVVCALL(OPFVV3, vfnmsac_vv_d, OP_UUU_D, H8, H8, H8, fnmsac64)
-GEN_VEXT_VV_ENV(vfnmsac_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmsac_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmsac_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmsac_vv_h)
+GEN_VEXT_VV_ENV(vfnmsac_vv_w)
+GEN_VEXT_VV_ENV(vfnmsac_vv_d)
 RVVCALL(OPFVF3, vfnmsac_vf_h, OP_UUU_H, H2, H2, fnmsac16)
 RVVCALL(OPFVF3, vfnmsac_vf_w, OP_UUU_W, H4, H4, fnmsac32)
 RVVCALL(OPFVF3, vfnmsac_vf_d, OP_UUU_D, H8, H8, fnmsac64)
-GEN_VEXT_VF(vfnmsac_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmsac_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmsac_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmsac_vf_h)
+GEN_VEXT_VF(vfnmsac_vf_w)
+GEN_VEXT_VF(vfnmsac_vf_d)
 
 static uint16_t fmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3207,15 +3205,15 @@ static uint64_t fmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmadd_vv_h, OP_UUU_H, H2, H2, H2, fmadd16)
 RVVCALL(OPFVV3, vfmadd_vv_w, OP_UUU_W, H4, H4, H4, fmadd32)
 RVVCALL(OPFVV3, vfmadd_vv_d, OP_UUU_D, H8, H8, H8, fmadd64)
-GEN_VEXT_VV_ENV(vfmadd_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmadd_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmadd_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmadd_vv_h)
+GEN_VEXT_VV_ENV(vfmadd_vv_w)
+GEN_VEXT_VV_ENV(vfmadd_vv_d)
 RVVCALL(OPFVF3, vfmadd_vf_h, OP_UUU_H, H2, H2, fmadd16)
 RVVCALL(OPFVF3, vfmadd_vf_w, OP_UUU_W, H4, H4, fmadd32)
 RVVCALL(OPFVF3, vfmadd_vf_d, OP_UUU_D, H8, H8, fmadd64)
-GEN_VEXT_VF(vfmadd_vf_h, 2, 2)
-GEN_VEXT_VF(vfmadd_vf_w, 4, 4)
-GEN_VEXT_VF(vfmadd_vf_d, 8, 8)
+GEN_VEXT_VF(vfmadd_vf_h)
+GEN_VEXT_VF(vfmadd_vf_w)
+GEN_VEXT_VF(vfmadd_vf_d)
 
 static uint16_t fnmadd16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3238,15 +3236,15 @@ static uint64_t fnmadd64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmadd_vv_h, OP_UUU_H, H2, H2, H2, fnmadd16)
 RVVCALL(OPFVV3, vfnmadd_vv_w, OP_UUU_W, H4, H4, H4, fnmadd32)
 RVVCALL(OPFVV3, vfnmadd_vv_d, OP_UUU_D, H8, H8, H8, fnmadd64)
-GEN_VEXT_VV_ENV(vfnmadd_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmadd_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmadd_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmadd_vv_h)
+GEN_VEXT_VV_ENV(vfnmadd_vv_w)
+GEN_VEXT_VV_ENV(vfnmadd_vv_d)
 RVVCALL(OPFVF3, vfnmadd_vf_h, OP_UUU_H, H2, H2, fnmadd16)
 RVVCALL(OPFVF3, vfnmadd_vf_w, OP_UUU_W, H4, H4, fnmadd32)
 RVVCALL(OPFVF3, vfnmadd_vf_d, OP_UUU_D, H8, H8, fnmadd64)
-GEN_VEXT_VF(vfnmadd_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmadd_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmadd_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmadd_vf_h)
+GEN_VEXT_VF(vfnmadd_vf_w)
+GEN_VEXT_VF(vfnmadd_vf_d)
 
 static uint16_t fmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3266,15 +3264,15 @@ static uint64_t fmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfmsub_vv_h, OP_UUU_H, H2, H2, H2, fmsub16)
 RVVCALL(OPFVV3, vfmsub_vv_w, OP_UUU_W, H4, H4, H4, fmsub32)
 RVVCALL(OPFVV3, vfmsub_vv_d, OP_UUU_D, H8, H8, H8, fmsub64)
-GEN_VEXT_VV_ENV(vfmsub_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmsub_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmsub_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmsub_vv_h)
+GEN_VEXT_VV_ENV(vfmsub_vv_w)
+GEN_VEXT_VV_ENV(vfmsub_vv_d)
 RVVCALL(OPFVF3, vfmsub_vf_h, OP_UUU_H, H2, H2, fmsub16)
 RVVCALL(OPFVF3, vfmsub_vf_w, OP_UUU_W, H4, H4, fmsub32)
 RVVCALL(OPFVF3, vfmsub_vf_d, OP_UUU_D, H8, H8, fmsub64)
-GEN_VEXT_VF(vfmsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfmsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfmsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfmsub_vf_h)
+GEN_VEXT_VF(vfmsub_vf_w)
+GEN_VEXT_VF(vfmsub_vf_d)
 
 static uint16_t fnmsub16(uint16_t a, uint16_t b, uint16_t d, float_status *s)
 {
@@ -3294,15 +3292,15 @@ static uint64_t fnmsub64(uint64_t a, uint64_t b, uint64_t d, float_status *s)
 RVVCALL(OPFVV3, vfnmsub_vv_h, OP_UUU_H, H2, H2, H2, fnmsub16)
 RVVCALL(OPFVV3, vfnmsub_vv_w, OP_UUU_W, H4, H4, H4, fnmsub32)
 RVVCALL(OPFVV3, vfnmsub_vv_d, OP_UUU_D, H8, H8, H8, fnmsub64)
-GEN_VEXT_VV_ENV(vfnmsub_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfnmsub_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfnmsub_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfnmsub_vv_h)
+GEN_VEXT_VV_ENV(vfnmsub_vv_w)
+GEN_VEXT_VV_ENV(vfnmsub_vv_d)
 RVVCALL(OPFVF3, vfnmsub_vf_h, OP_UUU_H, H2, H2, fnmsub16)
 RVVCALL(OPFVF3, vfnmsub_vf_w, OP_UUU_W, H4, H4, fnmsub32)
 RVVCALL(OPFVF3, vfnmsub_vf_d, OP_UUU_D, H8, H8, fnmsub64)
-GEN_VEXT_VF(vfnmsub_vf_h, 2, 2)
-GEN_VEXT_VF(vfnmsub_vf_w, 4, 4)
-GEN_VEXT_VF(vfnmsub_vf_d, 8, 8)
+GEN_VEXT_VF(vfnmsub_vf_h)
+GEN_VEXT_VF(vfnmsub_vf_w)
+GEN_VEXT_VF(vfnmsub_vf_d)
 
 /* Vector Widening Floating-Point Fused Multiply-Add Instructions */
 static uint32_t fwmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
@@ -3319,12 +3317,12 @@ static uint64_t fwmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwmacc16)
 RVVCALL(OPFVV3, vfwmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwmacc32)
-GEN_VEXT_VV_ENV(vfwmacc_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwmacc_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwmacc_vv_h)
+GEN_VEXT_VV_ENV(vfwmacc_vv_w)
 RVVCALL(OPFVF3, vfwmacc_vf_h, WOP_UUU_H, H4, H2, fwmacc16)
 RVVCALL(OPFVF3, vfwmacc_vf_w, WOP_UUU_W, H8, H4, fwmacc32)
-GEN_VEXT_VF(vfwmacc_vf_h, 2, 4)
-GEN_VEXT_VF(vfwmacc_vf_w, 4, 8)
+GEN_VEXT_VF(vfwmacc_vf_h)
+GEN_VEXT_VF(vfwmacc_vf_w)
 
 static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3342,12 +3340,12 @@ static uint64_t fwnmacc32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmacc_vv_h, WOP_UUU_H, H4, H2, H2, fwnmacc16)
 RVVCALL(OPFVV3, vfwnmacc_vv_w, WOP_UUU_W, H8, H4, H4, fwnmacc32)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwnmacc_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_h)
+GEN_VEXT_VV_ENV(vfwnmacc_vv_w)
 RVVCALL(OPFVF3, vfwnmacc_vf_h, WOP_UUU_H, H4, H2, fwnmacc16)
 RVVCALL(OPFVF3, vfwnmacc_vf_w, WOP_UUU_W, H8, H4, fwnmacc32)
-GEN_VEXT_VF(vfwnmacc_vf_h, 2, 4)
-GEN_VEXT_VF(vfwnmacc_vf_w, 4, 8)
+GEN_VEXT_VF(vfwnmacc_vf_h)
+GEN_VEXT_VF(vfwnmacc_vf_w)
 
 static uint32_t fwmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3365,12 +3363,12 @@ static uint64_t fwmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwmsac16)
 RVVCALL(OPFVV3, vfwmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwmsac32)
-GEN_VEXT_VV_ENV(vfwmsac_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwmsac_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwmsac_vv_h)
+GEN_VEXT_VV_ENV(vfwmsac_vv_w)
 RVVCALL(OPFVF3, vfwmsac_vf_h, WOP_UUU_H, H4, H2, fwmsac16)
 RVVCALL(OPFVF3, vfwmsac_vf_w, WOP_UUU_W, H8, H4, fwmsac32)
-GEN_VEXT_VF(vfwmsac_vf_h, 2, 4)
-GEN_VEXT_VF(vfwmsac_vf_w, 4, 8)
+GEN_VEXT_VF(vfwmsac_vf_h)
+GEN_VEXT_VF(vfwmsac_vf_w)
 
 static uint32_t fwnmsac16(uint16_t a, uint16_t b, uint32_t d, float_status *s)
 {
@@ -3388,12 +3386,12 @@ static uint64_t fwnmsac32(uint32_t a, uint32_t b, uint64_t d, float_status *s)
 
 RVVCALL(OPFVV3, vfwnmsac_vv_h, WOP_UUU_H, H4, H2, H2, fwnmsac16)
 RVVCALL(OPFVV3, vfwnmsac_vv_w, WOP_UUU_W, H8, H4, H4, fwnmsac32)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_h, 2, 4)
-GEN_VEXT_VV_ENV(vfwnmsac_vv_w, 4, 8)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_h)
+GEN_VEXT_VV_ENV(vfwnmsac_vv_w)
 RVVCALL(OPFVF3, vfwnmsac_vf_h, WOP_UUU_H, H4, H2, fwnmsac16)
 RVVCALL(OPFVF3, vfwnmsac_vf_w, WOP_UUU_W, H8, H4, fwnmsac32)
-GEN_VEXT_VF(vfwnmsac_vf_h, 2, 4)
-GEN_VEXT_VF(vfwnmsac_vf_w, 4, 8)
+GEN_VEXT_VF(vfwnmsac_vf_h)
+GEN_VEXT_VF(vfwnmsac_vf_w)
 
 /* Vector Floating-Point Square-Root Instruction */
 /* (TD, T2, TX2) */
@@ -3409,7 +3407,7 @@ static void do_##NAME(void *vd, void *vs2, int i,      \
     *((TD *)vd + HD(i)) = OP(s2, &env->fp_status);     \
 }
 
-#define GEN_VEXT_V_ENV(NAME, ESZ, DSZ)                 \
+#define GEN_VEXT_V_ENV(NAME)                           \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
         CPURISCVState *env, uint32_t desc)             \
 {                                                      \
@@ -3432,9 +3430,9 @@ void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
 RVVCALL(OPFVV1, vfsqrt_v_h, OP_UU_H, H2, H2, float16_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_w, OP_UU_W, H4, H4, float32_sqrt)
 RVVCALL(OPFVV1, vfsqrt_v_d, OP_UU_D, H8, H8, float64_sqrt)
-GEN_VEXT_V_ENV(vfsqrt_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfsqrt_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfsqrt_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfsqrt_v_h)
+GEN_VEXT_V_ENV(vfsqrt_v_w)
+GEN_VEXT_V_ENV(vfsqrt_v_d)
 
 /*
  * Vector Floating-Point Reciprocal Square-Root Estimate Instruction
@@ -3614,9 +3612,9 @@ static float64 frsqrt7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrsqrt7_v_h, OP_UU_H, H2, H2, frsqrt7_h)
 RVVCALL(OPFVV1, vfrsqrt7_v_w, OP_UU_W, H4, H4, frsqrt7_s)
 RVVCALL(OPFVV1, vfrsqrt7_v_d, OP_UU_D, H8, H8, frsqrt7_d)
-GEN_VEXT_V_ENV(vfrsqrt7_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfrsqrt7_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfrsqrt7_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfrsqrt7_v_h)
+GEN_VEXT_V_ENV(vfrsqrt7_v_w)
+GEN_VEXT_V_ENV(vfrsqrt7_v_d)
 
 /*
  * Vector Floating-Point Reciprocal Estimate Instruction
@@ -3805,36 +3803,36 @@ static float64 frec7_d(float64 f, float_status *s)
 RVVCALL(OPFVV1, vfrec7_v_h, OP_UU_H, H2, H2, frec7_h)
 RVVCALL(OPFVV1, vfrec7_v_w, OP_UU_W, H4, H4, frec7_s)
 RVVCALL(OPFVV1, vfrec7_v_d, OP_UU_D, H8, H8, frec7_d)
-GEN_VEXT_V_ENV(vfrec7_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfrec7_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfrec7_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfrec7_v_h)
+GEN_VEXT_V_ENV(vfrec7_v_w)
+GEN_VEXT_V_ENV(vfrec7_v_d)
 
 /* Vector Floating-Point MIN/MAX Instructions */
 RVVCALL(OPFVV2, vfmin_vv_h, OP_UUU_H, H2, H2, H2, float16_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_w, OP_UUU_W, H4, H4, H4, float32_minimum_number)
 RVVCALL(OPFVV2, vfmin_vv_d, OP_UUU_D, H8, H8, H8, float64_minimum_number)
-GEN_VEXT_VV_ENV(vfmin_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmin_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmin_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmin_vv_h)
+GEN_VEXT_VV_ENV(vfmin_vv_w)
+GEN_VEXT_VV_ENV(vfmin_vv_d)
 RVVCALL(OPFVF2, vfmin_vf_h, OP_UUU_H, H2, H2, float16_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_w, OP_UUU_W, H4, H4, float32_minimum_number)
 RVVCALL(OPFVF2, vfmin_vf_d, OP_UUU_D, H8, H8, float64_minimum_number)
-GEN_VEXT_VF(vfmin_vf_h, 2, 2)
-GEN_VEXT_VF(vfmin_vf_w, 4, 4)
-GEN_VEXT_VF(vfmin_vf_d, 8, 8)
+GEN_VEXT_VF(vfmin_vf_h)
+GEN_VEXT_VF(vfmin_vf_w)
+GEN_VEXT_VF(vfmin_vf_d)
 
 RVVCALL(OPFVV2, vfmax_vv_h, OP_UUU_H, H2, H2, H2, float16_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_w, OP_UUU_W, H4, H4, H4, float32_maximum_number)
 RVVCALL(OPFVV2, vfmax_vv_d, OP_UUU_D, H8, H8, H8, float64_maximum_number)
-GEN_VEXT_VV_ENV(vfmax_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfmax_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfmax_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfmax_vv_h)
+GEN_VEXT_VV_ENV(vfmax_vv_w)
+GEN_VEXT_VV_ENV(vfmax_vv_d)
 RVVCALL(OPFVF2, vfmax_vf_h, OP_UUU_H, H2, H2, float16_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_w, OP_UUU_W, H4, H4, float32_maximum_number)
 RVVCALL(OPFVF2, vfmax_vf_d, OP_UUU_D, H8, H8, float64_maximum_number)
-GEN_VEXT_VF(vfmax_vf_h, 2, 2)
-GEN_VEXT_VF(vfmax_vf_w, 4, 4)
-GEN_VEXT_VF(vfmax_vf_d, 8, 8)
+GEN_VEXT_VF(vfmax_vf_h)
+GEN_VEXT_VF(vfmax_vf_w)
+GEN_VEXT_VF(vfmax_vf_d)
 
 /* Vector Floating-Point Sign-Injection Instructions */
 static uint16_t fsgnj16(uint16_t a, uint16_t b, float_status *s)
@@ -3855,15 +3853,15 @@ static uint64_t fsgnj64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnj_vv_h, OP_UUU_H, H2, H2, H2, fsgnj16)
 RVVCALL(OPFVV2, vfsgnj_vv_w, OP_UUU_W, H4, H4, H4, fsgnj32)
 RVVCALL(OPFVV2, vfsgnj_vv_d, OP_UUU_D, H8, H8, H8, fsgnj64)
-GEN_VEXT_VV_ENV(vfsgnj_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsgnj_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsgnj_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsgnj_vv_h)
+GEN_VEXT_VV_ENV(vfsgnj_vv_w)
+GEN_VEXT_VV_ENV(vfsgnj_vv_d)
 RVVCALL(OPFVF2, vfsgnj_vf_h, OP_UUU_H, H2, H2, fsgnj16)
 RVVCALL(OPFVF2, vfsgnj_vf_w, OP_UUU_W, H4, H4, fsgnj32)
 RVVCALL(OPFVF2, vfsgnj_vf_d, OP_UUU_D, H8, H8, fsgnj64)
-GEN_VEXT_VF(vfsgnj_vf_h, 2, 2)
-GEN_VEXT_VF(vfsgnj_vf_w, 4, 4)
-GEN_VEXT_VF(vfsgnj_vf_d, 8, 8)
+GEN_VEXT_VF(vfsgnj_vf_h)
+GEN_VEXT_VF(vfsgnj_vf_w)
+GEN_VEXT_VF(vfsgnj_vf_d)
 
 static uint16_t fsgnjn16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3883,15 +3881,15 @@ static uint64_t fsgnjn64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjn_vv_h, OP_UUU_H, H2, H2, H2, fsgnjn16)
 RVVCALL(OPFVV2, vfsgnjn_vv_w, OP_UUU_W, H4, H4, H4, fsgnjn32)
 RVVCALL(OPFVV2, vfsgnjn_vv_d, OP_UUU_D, H8, H8, H8, fsgnjn64)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsgnjn_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_h)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_w)
+GEN_VEXT_VV_ENV(vfsgnjn_vv_d)
 RVVCALL(OPFVF2, vfsgnjn_vf_h, OP_UUU_H, H2, H2, fsgnjn16)
 RVVCALL(OPFVF2, vfsgnjn_vf_w, OP_UUU_W, H4, H4, fsgnjn32)
 RVVCALL(OPFVF2, vfsgnjn_vf_d, OP_UUU_D, H8, H8, fsgnjn64)
-GEN_VEXT_VF(vfsgnjn_vf_h, 2, 2)
-GEN_VEXT_VF(vfsgnjn_vf_w, 4, 4)
-GEN_VEXT_VF(vfsgnjn_vf_d, 8, 8)
+GEN_VEXT_VF(vfsgnjn_vf_h)
+GEN_VEXT_VF(vfsgnjn_vf_w)
+GEN_VEXT_VF(vfsgnjn_vf_d)
 
 static uint16_t fsgnjx16(uint16_t a, uint16_t b, float_status *s)
 {
@@ -3911,15 +3909,15 @@ static uint64_t fsgnjx64(uint64_t a, uint64_t b, float_status *s)
 RVVCALL(OPFVV2, vfsgnjx_vv_h, OP_UUU_H, H2, H2, H2, fsgnjx16)
 RVVCALL(OPFVV2, vfsgnjx_vv_w, OP_UUU_W, H4, H4, H4, fsgnjx32)
 RVVCALL(OPFVV2, vfsgnjx_vv_d, OP_UUU_D, H8, H8, H8, fsgnjx64)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_h, 2, 2)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_w, 4, 4)
-GEN_VEXT_VV_ENV(vfsgnjx_vv_d, 8, 8)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_h)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_w)
+GEN_VEXT_VV_ENV(vfsgnjx_vv_d)
 RVVCALL(OPFVF2, vfsgnjx_vf_h, OP_UUU_H, H2, H2, fsgnjx16)
 RVVCALL(OPFVF2, vfsgnjx_vf_w, OP_UUU_W, H4, H4, fsgnjx32)
 RVVCALL(OPFVF2, vfsgnjx_vf_d, OP_UUU_D, H8, H8, fsgnjx64)
-GEN_VEXT_VF(vfsgnjx_vf_h, 2, 2)
-GEN_VEXT_VF(vfsgnjx_vf_w, 4, 4)
-GEN_VEXT_VF(vfsgnjx_vf_d, 8, 8)
+GEN_VEXT_VF(vfsgnjx_vf_h)
+GEN_VEXT_VF(vfsgnjx_vf_w)
+GEN_VEXT_VF(vfsgnjx_vf_d)
 
 /* Vector Floating-Point Compare Instructions */
 #define GEN_VEXT_CMP_VV_ENV(NAME, ETYPE, H, DO_OP)            \
@@ -4063,7 +4061,7 @@ static void do_##NAME(void *vd, void *vs2, int i)      \
     *((TD *)vd + HD(i)) = OP(s2);                      \
 }
 
-#define GEN_VEXT_V(NAME, ESZ, DSZ)                     \
+#define GEN_VEXT_V(NAME)                               \
 void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
                   CPURISCVState *env, uint32_t desc)   \
 {                                                      \
@@ -4140,9 +4138,9 @@ target_ulong fclass_d(uint64_t frs1)
 RVVCALL(OPIVV1, vfclass_v_h, OP_UU_H, H2, H2, fclass_h)
 RVVCALL(OPIVV1, vfclass_v_w, OP_UU_W, H4, H4, fclass_s)
 RVVCALL(OPIVV1, vfclass_v_d, OP_UU_D, H8, H8, fclass_d)
-GEN_VEXT_V(vfclass_v_h, 2, 2)
-GEN_VEXT_V(vfclass_v_w, 4, 4)
-GEN_VEXT_V(vfclass_v_d, 8, 8)
+GEN_VEXT_V(vfclass_v_h)
+GEN_VEXT_V(vfclass_v_w)
+GEN_VEXT_V(vfclass_v_d)
 
 /* Vector Floating-Point Merge Instruction */
 #define GEN_VFMERGE_VF(NAME, ETYPE, H)                        \
@@ -4170,33 +4168,33 @@ GEN_VFMERGE_VF(vfmerge_vfm_d, int64_t, H8)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_h, OP_UU_H, H2, H2, float16_to_uint16)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_w, OP_UU_W, H4, H4, float32_to_uint32)
 RVVCALL(OPFVV1, vfcvt_xu_f_v_d, OP_UU_D, H8, H8, float64_to_uint64)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_xu_f_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_h)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_w)
+GEN_VEXT_V_ENV(vfcvt_xu_f_v_d)
 
 /* vfcvt.x.f.v vd, vs2, vm # Convert float to signed integer. */
 RVVCALL(OPFVV1, vfcvt_x_f_v_h, OP_UU_H, H2, H2, float16_to_int16)
 RVVCALL(OPFVV1, vfcvt_x_f_v_w, OP_UU_W, H4, H4, float32_to_int32)
 RVVCALL(OPFVV1, vfcvt_x_f_v_d, OP_UU_D, H8, H8, float64_to_int64)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_x_f_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_h)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_w)
+GEN_VEXT_V_ENV(vfcvt_x_f_v_d)
 
 /* vfcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_xu_v_h, OP_UU_H, H2, H2, uint16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_w, OP_UU_W, H4, H4, uint32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_xu_v_d, OP_UU_D, H8, H8, uint64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_f_xu_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_h)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_w)
+GEN_VEXT_V_ENV(vfcvt_f_xu_v_d)
 
 /* vfcvt.f.x.v vd, vs2, vm # Convert integer to float. */
 RVVCALL(OPFVV1, vfcvt_f_x_v_h, OP_UU_H, H2, H2, int16_to_float16)
 RVVCALL(OPFVV1, vfcvt_f_x_v_w, OP_UU_W, H4, H4, int32_to_float32)
 RVVCALL(OPFVV1, vfcvt_f_x_v_d, OP_UU_D, H8, H8, int64_to_float64)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_h, 2, 2)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_w, 4, 4)
-GEN_VEXT_V_ENV(vfcvt_f_x_v_d, 8, 8)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_h)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_w)
+GEN_VEXT_V_ENV(vfcvt_f_x_v_d)
 
 /* Widening Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4206,30 +4204,30 @@ GEN_VEXT_V_ENV(vfcvt_f_x_v_d, 8, 8)
 /* vfwcvt.xu.f.v vd, vs2, vm # Convert float to double-width unsigned integer.*/
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_h, WOP_UU_H, H4, H2, float16_to_uint32)
 RVVCALL(OPFVV1, vfwcvt_xu_f_v_w, WOP_UU_W, H8, H4, float32_to_uint64)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_h)
+GEN_VEXT_V_ENV(vfwcvt_xu_f_v_w)
 
 /* vfwcvt.x.f.v vd, vs2, vm # Convert float to double-width signed integer. */
 RVVCALL(OPFVV1, vfwcvt_x_f_v_h, WOP_UU_H, H4, H2, float16_to_int32)
 RVVCALL(OPFVV1, vfwcvt_x_f_v_w, WOP_UU_W, H8, H4, float32_to_int64)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_x_f_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_h)
+GEN_VEXT_V_ENV(vfwcvt_x_f_v_w)
 
 /* vfwcvt.f.xu.v vd, vs2, vm # Convert unsigned integer to double-width float */
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_b, WOP_UU_B, H2, H1, uint8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_h, WOP_UU_H, H4, H2, uint16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_xu_v_w, WOP_UU_W, H8, H4, uint32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b, 1, 2)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_b)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_h)
+GEN_VEXT_V_ENV(vfwcvt_f_xu_v_w)
 
 /* vfwcvt.f.x.v vd, vs2, vm # Convert integer to double-width float. */
 RVVCALL(OPFVV1, vfwcvt_f_x_v_b, WOP_UU_B, H2, H1, int8_to_float16)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_h, WOP_UU_H, H4, H2, int16_to_float32)
 RVVCALL(OPFVV1, vfwcvt_f_x_v_w, WOP_UU_W, H8, H4, int32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_b, 1, 2)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_f_x_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_b)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_h)
+GEN_VEXT_V_ENV(vfwcvt_f_x_v_w)
 
 /*
  * vfwcvt.f.f.v vd, vs2, vm
@@ -4242,8 +4240,8 @@ static uint32_t vfwcvtffv16(uint16_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfwcvt_f_f_v_h, WOP_UU_H, H4, H2, vfwcvtffv16)
 RVVCALL(OPFVV1, vfwcvt_f_f_v_w, WOP_UU_W, H8, H4, float32_to_float64)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_h, 2, 4)
-GEN_VEXT_V_ENV(vfwcvt_f_f_v_w, 4, 8)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_h)
+GEN_VEXT_V_ENV(vfwcvt_f_f_v_w)
 
 /* Narrowing Floating-Point/Integer Type-Convert Instructions */
 /* (TD, T2, TX2) */
@@ -4254,29 +4252,29 @@ GEN_VEXT_V_ENV(vfwcvt_f_f_v_w, 4, 8)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_b, NOP_UU_B, H1, H2, float16_to_uint8)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_h, NOP_UU_H, H2, H4, float32_to_uint16)
 RVVCALL(OPFVV1, vfncvt_xu_f_w_w, NOP_UU_W, H4, H8, float64_to_uint32)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_b, 1, 1)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_xu_f_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_b)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_h)
+GEN_VEXT_V_ENV(vfncvt_xu_f_w_w)
 
 /* vfncvt.x.f.v vd, vs2, vm # Convert double-width float to signed integer. */
 RVVCALL(OPFVV1, vfncvt_x_f_w_b, NOP_UU_B, H1, H2, float16_to_int8)
 RVVCALL(OPFVV1, vfncvt_x_f_w_h, NOP_UU_H, H2, H4, float32_to_int16)
 RVVCALL(OPFVV1, vfncvt_x_f_w_w, NOP_UU_W, H4, H8, float64_to_int32)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_b, 1, 1)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_x_f_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_b)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_h)
+GEN_VEXT_V_ENV(vfncvt_x_f_w_w)
 
 /* vfncvt.f.xu.v vd, vs2, vm # Convert double-width unsigned integer to float */
 RVVCALL(OPFVV1, vfncvt_f_xu_w_h, NOP_UU_H, H2, H4, uint32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_xu_w_w, NOP_UU_W, H4, H8, uint64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_f_xu_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_h)
+GEN_VEXT_V_ENV(vfncvt_f_xu_w_w)
 
 /* vfncvt.f.x.v vd, vs2, vm # Convert double-width integer to float. */
 RVVCALL(OPFVV1, vfncvt_f_x_w_h, NOP_UU_H, H2, H4, int32_to_float16)
 RVVCALL(OPFVV1, vfncvt_f_x_w_w, NOP_UU_W, H4, H8, int64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_f_x_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_h)
+GEN_VEXT_V_ENV(vfncvt_f_x_w_w)
 
 /* vfncvt.f.f.v vd, vs2, vm # Convert double float to single-width float. */
 static uint16_t vfncvtffv16(uint32_t a, float_status *s)
@@ -4286,8 +4284,8 @@ static uint16_t vfncvtffv16(uint32_t a, float_status *s)
 
 RVVCALL(OPFVV1, vfncvt_f_f_w_h, NOP_UU_H, H2, H4, vfncvtffv16)
 RVVCALL(OPFVV1, vfncvt_f_f_w_w, NOP_UU_W, H4, H8, float64_to_float32)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_h, 2, 2)
-GEN_VEXT_V_ENV(vfncvt_f_f_w_w, 4, 4)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_h)
+GEN_VEXT_V_ENV(vfncvt_f_f_w_w)
 
 /*
  *** Vector Reduction Operations
-- 
2.34.1



^ permalink raw reply related	[flat|nested] 42+ messages in thread

end of thread, other threads:[~2022-03-25 10:52 UTC | newest]

Thread overview: 42+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-03-24 19:03 [PATCH qemu v4 00/14] Add tail agnostic behavior for rvv instructions ~eopxd
2022-03-24 19:03 ` ~eopxd
2022-03-01  9:07 ` [PATCH qemu v4 04/14] target/riscv: rvv: Add tail agnostic for vv instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-25  9:10   ` Weiwei Li
2022-03-25  9:10     ` Weiwei Li
2022-03-07  7:10 ` [PATCH qemu v4 05/14] target/riscv: rvv: Add tail agnostic for vector load / store instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-07  7:32 ` [PATCH qemu v4 06/14] target/riscv: rvv: Add tail agnostic for vx, vvm, vxm instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-25 10:32   ` Weiwei Li
2022-03-25 10:32     ` Weiwei Li
2022-03-07  9:38 ` [PATCH qemu v4 07/14] target/riscv: rvv: Add tail agnostic for vector integer shift instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-07  9:43 ` [PATCH qemu v4 08/14] target/riscv: rvv: Add tail agnostic for vector integer comparison instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-07  9:53 ` [PATCH qemu v4 09/14] target/riscv: rvv: Add tail agnostic for vector integer merge and move instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-25 10:38   ` Weiwei Li
2022-03-25 10:38     ` Weiwei Li
2022-03-07 10:04 ` [PATCH qemu v4 10/14] target/riscv: rvv: Add tail agnostic for vector fix-point arithmetic instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-07 10:05 ` [PATCH qemu v4 11/14] target/riscv: rvv: Add tail agnostic for vector floating-point instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-25 10:42   ` Weiwei Li
2022-03-25 10:42     ` Weiwei Li
2022-03-07 12:21 ` [PATCH qemu v4 12/14] target/riscv: rvv: Add tail agnostic for vector reduction instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-25 10:47   ` Weiwei Li
2022-03-25 10:47     ` Weiwei Li
2022-03-07 15:26 ` [PATCH qemu v4 13/14] target/riscv: rvv: Add tail agnostic for vector mask instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-07 15:59 ` [PATCH qemu v4 14/14] target/riscv: rvv: Add tail agnostic for vector permutation instructions ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-25 10:51   ` Weiwei Li
2022-03-25 10:51     ` Weiwei Li
2022-03-09  8:34 ` [PATCH qemu v4 02/14] target/riscv: rvv: Rename ambiguous esz ~eopxd
2022-03-24 19:03   ` ~eopxd
2022-03-12  6:28 ` [PATCH qemu v4 03/14] target/riscv: rvv: Early exit when vstart >= vl ~eopxd
2022-03-12  6:28   ` ~eopxd
2022-03-14  7:38 ` [PATCH qemu v4 01/14] target/riscv: rvv: Prune redundant ESZ, DSZ parameter passed ~eopxd
2022-03-14  7:38   ` ~eopxd

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.