[PATCH v2 19/37] target/riscv: Partial-SIMD Miscellaneous Instructions

From: LIU Zhiwei <zhiwei_liu@c-sky.com>
To: qemu-devel@nongnu.org, qemu-riscv@nongnu.org
Cc: palmer@dabbelt.com, richard.henderson@linaro.org,
	bin.meng@windriver.com,
	Alistair Francis <alistair.francis@wdc.com>,
	LIU Zhiwei <zhiwei_liu@c-sky.com>
Subject: [PATCH v2 19/37] target/riscv: Partial-SIMD Miscellaneous Instructions
Date: Thu, 10 Jun 2021 15:58:50 +0800	[thread overview]
Message-ID: <20210610075908.3305506-20-zhiwei_liu@c-sky.com> (raw)
In-Reply-To: <20210610075908.3305506-1-zhiwei_liu@c-sky.com>

32-bit signed or unsigned clip value. 32-bit leading
redundant sign, leading zero, leading one count. Parallel
byte sum of absolute difference or parallel byte sum of
absolute difference accumulation.

Signed-off-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
---
 target/riscv/helper.h                   |  8 +++
 target/riscv/insn32.decode              |  8 +++
 target/riscv/insn_trans/trans_rvp.c.inc |  9 +++
 target/riscv/packed_helper.c            | 75 +++++++++++++++++++++++++
 4 files changed, 100 insertions(+)

diff --git a/target/riscv/helper.h b/target/riscv/helper.h
index a37b023c53..35c8c61b00 100644
--- a/target/riscv/helper.h
+++ b/target/riscv/helper.h
@@ -1318,3 +1318,11 @@ DEF_HELPER_4(kmsda, tl, env, tl, tl, tl)
 DEF_HELPER_4(kmsxda, tl, env, tl, tl, tl)
 
 DEF_HELPER_3(smal, i64, env, i64, tl)
+
+DEF_HELPER_3(sclip32, tl, env, tl, tl)
+DEF_HELPER_3(uclip32, tl, env, tl, tl)
+DEF_HELPER_2(clrs32, tl, env, tl)
+DEF_HELPER_2(clz32, tl, env, tl)
+DEF_HELPER_2(clo32, tl, env, tl)
+DEF_HELPER_3(pbsad, tl, env, tl, tl)
+DEF_HELPER_4(pbsada, tl, env, tl, tl, tl)
diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 233df941b4..ce8bdee34b 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -921,3 +921,11 @@ kmsda      0100110  ..... ..... 001 ..... 1110111 @r
 kmsxda     0100111  ..... ..... 001 ..... 1110111 @r
 
 smal       0101111  ..... ..... 001 ..... 1110111 @r
+
+sclip32    1110010  ..... ..... 000 ..... 1110111 @sh5
+uclip32    1111010  ..... ..... 000 ..... 1110111 @sh5
+clrs32     1010111  11000 ..... 000 ..... 1110111 @r2
+clz32      1010111  11001 ..... 000 ..... 1110111 @r2
+clo32      1010111  11011 ..... 000 ..... 1110111 @r2
+pbsad      1111110  ..... ..... 000 ..... 1110111 @r
+pbsada     1111111  ..... ..... 000 ..... 1110111 @r
diff --git a/target/riscv/insn_trans/trans_rvp.c.inc b/target/riscv/insn_trans/trans_rvp.c.inc
index 8b0728fc5a..43e7e5a75d 100644
--- a/target/riscv/insn_trans/trans_rvp.c.inc
+++ b/target/riscv/insn_trans/trans_rvp.c.inc
@@ -502,3 +502,12 @@ static bool trans_##NAME(DisasContext *s, arg_r *a)    \
 }
 
 GEN_RVP_R_D64_S64_OOL(smal);
+
+/* Partial-SIMD Miscellaneous Instructions */
+GEN_RVP_SHIFTI(sclip32, NULL, gen_helper_sclip32);
+GEN_RVP_SHIFTI(uclip32, NULL, gen_helper_uclip32);
+GEN_RVP_R2_OOL(clrs32);
+GEN_RVP_R2_OOL(clz32);
+GEN_RVP_R2_OOL(clo32);
+GEN_RVP_R_OOL(pbsad);
+GEN_RVP_R_ACC_OOL(pbsada);
diff --git a/target/riscv/packed_helper.c b/target/riscv/packed_helper.c
index 1f9a5d620f..1f2b90c394 100644
--- a/target/riscv/packed_helper.c
+++ b/target/riscv/packed_helper.c
@@ -1969,3 +1969,78 @@ uint64_t helper_smal(CPURISCVState *env, uint64_t a, target_ulong b)
     }
     return result;
 }
+
+/* Partial-SIMD Miscellaneous Instructions */
+static inline void do_sclip32(CPURISCVState *env, void *vd, void *va,
+                              void *vb, uint8_t i)
+{
+    int32_t *d = vd, *a = va;
+    uint8_t shift = *(uint8_t *)vb & 0x1f;
+
+    d[i] = sat64(env, a[i], shift);
+}
+
+RVPR(sclip32, 1, 4);
+
+static inline void do_uclip32(CPURISCVState *env, void *vd, void *va,
+                              void *vb, uint8_t i)
+{
+    int32_t *d = vd, *a = va;
+    uint8_t shift = *(uint8_t *)vb & 0x1f;
+
+    if (a[i] < 0) {
+        d[i] = 0;
+        env->vxsat = 0x1;
+    } else {
+        d[i] = satu64(env, a[i], shift);
+    }
+}
+
+RVPR(uclip32, 1, 4);
+
+static inline void do_clrs32(CPURISCVState *env, void *vd, void *va, uint8_t i)
+{
+    int32_t *d = vd, *a = va;
+    d[i] = clrsb32(a[i]);
+}
+
+RVPR2(clrs32, 1, 4);
+
+static inline void do_clz32(CPURISCVState *env, void *vd, void *va, uint8_t i)
+{
+    int32_t *d = vd, *a = va;
+    d[i] = clz32(a[i]);
+}
+
+RVPR2(clz32, 1, 4);
+
+static inline void do_clo32(CPURISCVState *env, void *vd, void *va, uint8_t i)
+{
+    int32_t *d = vd, *a = va;
+    d[i] = clo32(a[i]);
+}
+
+RVPR2(clo32, 1, 4);
+
+static inline void do_pbsad(CPURISCVState *env, void *vd, void *va,
+                            void *vb, uint8_t i)
+{
+    target_ulong *d = vd;
+    uint8_t *a = va, *b = vb;
+    *d += abs(a[i] - b[i]);
+}
+
+RVPR(pbsad, 1, 1);
+
+static inline void do_pbsada(CPURISCVState *env, void *vd, void *va,
+                             void *vb, void *vc, uint8_t i)
+{
+    target_ulong *d = vd, *c = vc;
+    uint8_t *a = va, *b = vb;
+    if (i == 0) {
+        *d += *c;
+    }
+    *d += abs(a[i] - b[i]);
+}
+
+RVPR_ACC(pbsada, 1, 1);
-- 
2.25.1


