* [PULL 00/27] target-arm queue
@ 2020-08-24 9:47 Peter Maydell
2020-08-24 9:47 ` [PULL 01/27] hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores Peter Maydell
` (27 more replies)
0 siblings, 28 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
First arm pullreq for 5.2: Eric's SMMU stuff, and a bunch of
cleanup/refactoring from me.
thanks
-- PMM
The following changes since commit 8367a77c4d3f6e1e60890f5510304feb2c621611:
Merge remote-tracking branch 'remotes/vivier2/tags/linux-user-for-5.2-pull-request' into staging (2020-08-23 16:34:43 +0100)
are available in the Git repository at:
https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20200824
for you to fetch changes up to b34aa5129e9c3aff890b4f4bcc84962e94185629:
target/arm: Use correct FPST for VCMLA, VCADD on fp16 (2020-08-24 10:15:12 +0100)
----------------------------------------------------------------
target-arm queue:
* hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores
* hw/arm/smmuv3: Implement SMMUv3.2 range-invalidation
* docs/system/arm: Document the Xilinx Versal Virt board
* target/arm: Make M-profile NOCP take precedence over UNDEF
* target/arm: Use correct FPST for VCMLA, VCADD on fp16
* target/arm: Various cleanups preparing for fp16 support
----------------------------------------------------------------
Edgar E. Iglesias (1):
docs/system/arm: Document the Xilinx Versal Virt board
Eric Auger (11):
hw/arm/smmu-common: Factorize some code in smmu_ptw_64()
hw/arm/smmu-common: Add IOTLB helpers
hw/arm/smmu: Introduce smmu_get_iotlb_key()
hw/arm/smmu: Introduce SMMUTLBEntry for PTW and IOTLB value
hw/arm/smmu-common: Manage IOTLB block entries
hw/arm/smmuv3: Introduce smmuv3_s1_range_inval() helper
hw/arm/smmuv3: Get prepared for range invalidation
hw/arm/smmuv3: Fix IIDR offset
hw/arm/smmuv3: Let AIDR advertise SMMUv3.0 support
hw/arm/smmuv3: Support HAD and advertise SMMUv3.1 support
hw/arm/smmuv3: Advertise SMMUv3.2 range invalidation
Peter Maydell (14):
target/arm: Pull handling of XScale insns out of disas_coproc_insn()
target/arm: Separate decode from handling of coproc insns
target/arm: Convert A32 coprocessor insns to decodetree
target/arm: Tidy up disas_arm_insn()
target/arm: Do M-profile NOCP checks early and via decodetree
target/arm: Convert T32 coprocessor insns to decodetree
target/arm: Remove ARCH macro
target/arm: Delete unused VFP_DREG macros
target/arm/translate.c: Delete/amend incorrect comments
target/arm: Delete unused ARM_FEATURE_CRC
target/arm: Replace A64 get_fpstatus_ptr() with generic fpstatus_ptr()
target/arm: Make A32/T32 use new fpstatus_ptr() API
target/arm: Implement FPST_STD_F16 fpstatus
target/arm: Use correct FPST for VCMLA, VCADD on fp16
Philippe Mathieu-Daudé (1):
hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores
docs/system/arm/xlnx-versal-virt.rst | 176 +++++++++++++++++++++++
docs/system/target-arm.rst | 1 +
hw/arm/smmu-internal.h | 8 ++
hw/arm/smmuv3-internal.h | 10 +-
include/hw/arm/smmu-common.h | 19 ++-
include/hw/arm/smmuv3.h | 1 +
target/arm/cpu.h | 10 +-
target/arm/translate-a64.h | 1 -
target/arm/translate.h | 52 +++++++
target/arm/a32.decode | 19 +++
target/arm/m-nocp.decode | 42 ++++++
target/arm/t32.decode | 19 +++
target/arm/vfp.decode | 2 -
hw/arm/smmu-common.c | 214 ++++++++++++++++++---------
hw/arm/smmuv3.c | 142 +++++++++---------
hw/cpu/a9mpcore.c | 12 +-
target/arm/cpu.c | 3 +
target/arm/helper.c | 29 ++++
target/arm/translate-a64.c | 89 +++++-------
target/arm/translate-sve.c | 34 ++---
target/arm/translate.c | 272 +++++++++++++++++------------------
target/arm/vfp_helper.c | 5 +
MAINTAINERS | 3 +-
hw/arm/trace-events | 12 +-
target/arm/meson.build | 1 +
target/arm/translate-neon.c.inc | 28 ++--
target/arm/translate-vfp.c.inc | 96 ++++++++-----
27 files changed, 885 insertions(+), 415 deletions(-)
create mode 100644 docs/system/arm/xlnx-versal-virt.rst
create mode 100644 target/arm/m-nocp.decode
^ permalink raw reply [flat|nested] 29+ messages in thread
* [PULL 01/27] hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 02/27] hw/arm/smmu-common: Factorize some code in smmu_ptw_64() Peter Maydell
` (26 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Philippe Mathieu-Daudé <f4bug@amsat.org>
The 'Cortex-A9MPCore internal peripheral' block can only be
used with Cortex A5 and A9 cores. As we don't model the A5
yet, simply check the machine cpu core is a Cortex A9. If
not return an error.
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-id: 20200709152337.15533-1-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/cpu/a9mpcore.c | 12 +++++++++++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/hw/cpu/a9mpcore.c b/hw/cpu/a9mpcore.c
index 351295e5187..ec186d49abd 100644
--- a/hw/cpu/a9mpcore.c
+++ b/hw/cpu/a9mpcore.c
@@ -15,6 +15,7 @@
#include "hw/irq.h"
#include "hw/qdev-properties.h"
#include "hw/core/cpu.h"
+#include "cpu.h"
#define A9_GIC_NUM_PRIORITY_BITS 5
@@ -52,8 +53,18 @@ static void a9mp_priv_realize(DeviceState *dev, Error **errp)
*wdtbusdev;
int i;
bool has_el3;
+ CPUState *cpu0;
Object *cpuobj;
+ cpu0 = qemu_get_cpu(0);
+ cpuobj = OBJECT(cpu0);
+ if (strcmp(object_get_typename(cpuobj), ARM_CPU_TYPE_NAME("cortex-a9"))) {
+ /* We might allow Cortex-A5 once we model it */
+ error_setg(errp,
+ "Cortex-A9MPCore peripheral can only use Cortex-A9 CPU");
+ return;
+ }
+
scudev = DEVICE(&s->scu);
qdev_prop_set_uint32(scudev, "num-cpu", s->num_cpu);
if (!sysbus_realize(SYS_BUS_DEVICE(&s->scu), errp)) {
@@ -70,7 +81,6 @@ static void a9mp_priv_realize(DeviceState *dev, Error **errp)
/* Make the GIC's TZ support match the CPUs. We assume that
* either all the CPUs have TZ, or none do.
*/
- cpuobj = OBJECT(qemu_get_cpu(0));
has_el3 = object_property_find(cpuobj, "has_el3", NULL) &&
object_property_get_bool(cpuobj, "has_el3", &error_abort);
qdev_prop_set_bit(gicdev, "has-security-extensions", has_el3);
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 02/27] hw/arm/smmu-common: Factorize some code in smmu_ptw_64()
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
2020-08-24 9:47 ` [PULL 01/27] hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 03/27] hw/arm/smmu-common: Add IOTLB helpers Peter Maydell
` (25 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Page and block PTE decoding can share some code. Let's
first handle table PTE and factorize some code shared by
page and block PTEs.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-2-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmu-common.c | 48 ++++++++++++++++----------------------------
1 file changed, 17 insertions(+), 31 deletions(-)
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index e13a5f4a7cb..a3f9e473985 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -186,7 +186,7 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
uint64_t subpage_size = 1ULL << level_shift(level, granule_sz);
uint64_t mask = subpage_size - 1;
uint32_t offset = iova_level_offset(iova, inputsize, level, granule_sz);
- uint64_t pte;
+ uint64_t pte, gpa;
dma_addr_t pte_addr = baseaddr + offset * sizeof(pte);
uint8_t ap;
@@ -199,56 +199,42 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
if (is_invalid_pte(pte) || is_reserved_pte(pte, level)) {
trace_smmu_ptw_invalid_pte(stage, level, baseaddr,
pte_addr, offset, pte);
- info->type = SMMU_PTW_ERR_TRANSLATION;
- goto error;
+ break;
}
- if (is_page_pte(pte, level)) {
- uint64_t gpa = get_page_pte_address(pte, granule_sz);
+ if (is_table_pte(pte, level)) {
+ ap = PTE_APTABLE(pte);
- ap = PTE_AP(pte);
if (is_permission_fault(ap, perm)) {
info->type = SMMU_PTW_ERR_PERMISSION;
goto error;
}
-
- tlbe->translated_addr = gpa + (iova & mask);
- tlbe->perm = PTE_AP_TO_PERM(ap);
+ baseaddr = get_table_pte_address(pte, granule_sz);
+ level++;
+ continue;
+ } else if (is_page_pte(pte, level)) {
+ gpa = get_page_pte_address(pte, granule_sz);
trace_smmu_ptw_page_pte(stage, level, iova,
baseaddr, pte_addr, pte, gpa);
- return 0;
- }
- if (is_block_pte(pte, level)) {
+ } else {
uint64_t block_size;
- hwaddr gpa = get_block_pte_address(pte, level, granule_sz,
- &block_size);
-
- ap = PTE_AP(pte);
- if (is_permission_fault(ap, perm)) {
- info->type = SMMU_PTW_ERR_PERMISSION;
- goto error;
- }
+ gpa = get_block_pte_address(pte, level, granule_sz,
+ &block_size);
trace_smmu_ptw_block_pte(stage, level, baseaddr,
pte_addr, pte, iova, gpa,
block_size >> 20);
-
- tlbe->translated_addr = gpa + (iova & mask);
- tlbe->perm = PTE_AP_TO_PERM(ap);
- return 0;
}
-
- /* table pte */
- ap = PTE_APTABLE(pte);
-
+ ap = PTE_AP(pte);
if (is_permission_fault(ap, perm)) {
info->type = SMMU_PTW_ERR_PERMISSION;
goto error;
}
- baseaddr = get_table_pte_address(pte, granule_sz);
- level++;
- }
+ tlbe->translated_addr = gpa + (iova & mask);
+ tlbe->perm = PTE_AP_TO_PERM(ap);
+ return 0;
+ }
info->type = SMMU_PTW_ERR_TRANSLATION;
error:
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 03/27] hw/arm/smmu-common: Add IOTLB helpers
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
2020-08-24 9:47 ` [PULL 01/27] hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores Peter Maydell
2020-08-24 9:47 ` [PULL 02/27] hw/arm/smmu-common: Factorize some code in smmu_ptw_64() Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 04/27] hw/arm/smmu: Introduce smmu_get_iotlb_key() Peter Maydell
` (24 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Add two helpers: one to lookup for a given IOTLB entry and
one to insert a new entry. We also move the tracing there.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-3-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
include/hw/arm/smmu-common.h | 2 ++
hw/arm/smmu-common.c | 36 ++++++++++++++++++++++++++++++++++++
hw/arm/smmuv3.c | 26 ++------------------------
hw/arm/trace-events | 5 +++--
4 files changed, 43 insertions(+), 26 deletions(-)
diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
index ca4a4b1ad1e..1dceec5cb1f 100644
--- a/include/hw/arm/smmu-common.h
+++ b/include/hw/arm/smmu-common.h
@@ -153,6 +153,8 @@ IOMMUMemoryRegion *smmu_iommu_mr(SMMUState *s, uint32_t sid);
#define SMMU_IOTLB_MAX_SIZE 256
+IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg, hwaddr iova);
+void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry);
void smmu_iotlb_inv_all(SMMUState *s);
void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid);
void smmu_iotlb_inv_iova(SMMUState *s, uint16_t asid, dma_addr_t iova);
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index a3f9e473985..f3aa581f807 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -32,6 +32,42 @@
/* IOTLB Management */
+IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
+ hwaddr iova)
+{
+ SMMUIOTLBKey key = {.asid = cfg->asid, .iova = iova};
+ IOMMUTLBEntry *entry = g_hash_table_lookup(bs->iotlb, &key);
+
+ if (entry) {
+ cfg->iotlb_hits++;
+ trace_smmu_iotlb_lookup_hit(cfg->asid, iova,
+ cfg->iotlb_hits, cfg->iotlb_misses,
+ 100 * cfg->iotlb_hits /
+ (cfg->iotlb_hits + cfg->iotlb_misses));
+ } else {
+ cfg->iotlb_misses++;
+ trace_smmu_iotlb_lookup_miss(cfg->asid, iova,
+ cfg->iotlb_hits, cfg->iotlb_misses,
+ 100 * cfg->iotlb_hits /
+ (cfg->iotlb_hits + cfg->iotlb_misses));
+ }
+ return entry;
+}
+
+void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry)
+{
+ SMMUIOTLBKey *key = g_new0(SMMUIOTLBKey, 1);
+
+ if (g_hash_table_size(bs->iotlb) >= SMMU_IOTLB_MAX_SIZE) {
+ smmu_iotlb_inv_all(bs);
+ }
+
+ key->asid = cfg->asid;
+ key->iova = entry->iova;
+ trace_smmu_iotlb_insert(cfg->asid, entry->iova);
+ g_hash_table_insert(bs->iotlb, key, entry);
+}
+
inline void smmu_iotlb_inv_all(SMMUState *s)
{
trace_smmu_iotlb_inv_all();
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index 57a79df55b5..cd2a2e7e148 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -636,7 +636,6 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
.addr_mask = ~(hwaddr)0,
.perm = IOMMU_NONE,
};
- SMMUIOTLBKey key, *new_key;
qemu_mutex_lock(&s->mutex);
@@ -675,16 +674,8 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
page_mask = (1ULL << (tt->granule_sz)) - 1;
aligned_addr = addr & ~page_mask;
- key.asid = cfg->asid;
- key.iova = aligned_addr;
-
- cached_entry = g_hash_table_lookup(bs->iotlb, &key);
+ cached_entry = smmu_iotlb_lookup(bs, cfg, aligned_addr);
if (cached_entry) {
- cfg->iotlb_hits++;
- trace_smmu_iotlb_cache_hit(cfg->asid, aligned_addr,
- cfg->iotlb_hits, cfg->iotlb_misses,
- 100 * cfg->iotlb_hits /
- (cfg->iotlb_hits + cfg->iotlb_misses));
if ((flag & IOMMU_WO) && !(cached_entry->perm & IOMMU_WO)) {
status = SMMU_TRANS_ERROR;
if (event.record_trans_faults) {
@@ -698,16 +689,6 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
goto epilogue;
}
- cfg->iotlb_misses++;
- trace_smmu_iotlb_cache_miss(cfg->asid, addr & ~page_mask,
- cfg->iotlb_hits, cfg->iotlb_misses,
- 100 * cfg->iotlb_hits /
- (cfg->iotlb_hits + cfg->iotlb_misses));
-
- if (g_hash_table_size(bs->iotlb) >= SMMU_IOTLB_MAX_SIZE) {
- smmu_iotlb_inv_all(bs);
- }
-
cached_entry = g_new0(IOMMUTLBEntry, 1);
if (smmu_ptw(cfg, aligned_addr, flag, cached_entry, &ptw_info)) {
@@ -753,10 +734,7 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
}
status = SMMU_TRANS_ERROR;
} else {
- new_key = g_new0(SMMUIOTLBKey, 1);
- new_key->asid = cfg->asid;
- new_key->iova = aligned_addr;
- g_hash_table_insert(bs->iotlb, new_key, cached_entry);
+ smmu_iotlb_insert(bs, cfg, cached_entry);
status = SMMU_TRANS_SUCCESS;
}
diff --git a/hw/arm/trace-events b/hw/arm/trace-events
index 0acedcedc6f..b808a1bfc19 100644
--- a/hw/arm/trace-events
+++ b/hw/arm/trace-events
@@ -14,6 +14,9 @@ smmu_iotlb_inv_all(void) "IOTLB invalidate all"
smmu_iotlb_inv_asid(uint16_t asid) "IOTLB invalidate asid=%d"
smmu_iotlb_inv_iova(uint16_t asid, uint64_t addr) "IOTLB invalidate asid=%d addr=0x%"PRIx64
smmu_inv_notifiers_mr(const char *name) "iommu mr=%s"
+smmu_iotlb_lookup_hit(uint16_t asid, uint64_t addr, uint32_t hit, uint32_t miss, uint32_t p) "IOTLB cache HIT asid=%d addr=0x%"PRIx64" hit=%d miss=%d hit rate=%d"
+smmu_iotlb_lookup_miss(uint16_t asid, uint64_t addr, uint32_t hit, uint32_t miss, uint32_t p) "IOTLB cache MISS asid=%d addr=0x%"PRIx64" hit=%d miss=%d hit rate=%d"
+smmu_iotlb_insert(uint16_t asid, uint64_t addr) "IOTLB ++ asid=%d addr=0x%"PRIx64
# smmuv3.c
smmuv3_read_mmio(uint64_t addr, uint64_t val, unsigned size, uint32_t r) "addr: 0x%"PRIx64" val:0x%"PRIx64" size: 0x%x(%d)"
@@ -46,8 +49,6 @@ smmuv3_cmdq_tlbi_nh_va(int vmid, int asid, uint64_t addr, bool leaf) "vmid =%d a
smmuv3_cmdq_tlbi_nh_vaa(int vmid, uint64_t addr) "vmid =%d addr=0x%"PRIx64
smmuv3_cmdq_tlbi_nh(void) ""
smmuv3_cmdq_tlbi_nh_asid(uint16_t asid) "asid=%d"
-smmu_iotlb_cache_hit(uint16_t asid, uint64_t addr, uint32_t hit, uint32_t miss, uint32_t p) "IOTLB cache HIT asid=%d addr=0x%"PRIx64" hit=%d miss=%d hit rate=%d"
-smmu_iotlb_cache_miss(uint16_t asid, uint64_t addr, uint32_t hit, uint32_t miss, uint32_t p) "IOTLB cache MISS asid=%d addr=0x%"PRIx64" hit=%d miss=%d hit rate=%d"
smmuv3_config_cache_inv(uint32_t sid) "Config cache INV for sid %d"
smmuv3_notify_flag_add(const char *iommu) "ADD SMMUNotifier node for iommu mr=%s"
smmuv3_notify_flag_del(const char *iommu) "DEL SMMUNotifier node for iommu mr=%s"
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 04/27] hw/arm/smmu: Introduce smmu_get_iotlb_key()
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (2 preceding siblings ...)
2020-08-24 9:47 ` [PULL 03/27] hw/arm/smmu-common: Add IOTLB helpers Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 05/27] hw/arm/smmu: Introduce SMMUTLBEntry for PTW and IOTLB value Peter Maydell
` (23 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Introduce the smmu_get_iotlb_key() helper and the
SMMU_IOTLB_ASID() macro. Also move smmu_get_iotlb_key and
smmu_iotlb_key_hash in the IOTLB related code section.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-4-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmu-internal.h | 1 +
include/hw/arm/smmu-common.h | 1 +
hw/arm/smmu-common.c | 66 ++++++++++++++++++++----------------
3 files changed, 38 insertions(+), 30 deletions(-)
diff --git a/hw/arm/smmu-internal.h b/hw/arm/smmu-internal.h
index 7794d6d3947..3104f768cd2 100644
--- a/hw/arm/smmu-internal.h
+++ b/hw/arm/smmu-internal.h
@@ -96,4 +96,5 @@ uint64_t iova_level_offset(uint64_t iova, int inputsize,
MAKE_64BIT_MASK(0, gsz - 3);
}
+#define SMMU_IOTLB_ASID(key) ((key).asid)
#endif
diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
index 1dceec5cb1f..5f9f3535d2a 100644
--- a/include/hw/arm/smmu-common.h
+++ b/include/hw/arm/smmu-common.h
@@ -155,6 +155,7 @@ IOMMUMemoryRegion *smmu_iommu_mr(SMMUState *s, uint32_t sid);
IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg, hwaddr iova);
void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry);
+SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova);
void smmu_iotlb_inv_all(SMMUState *s);
void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid);
void smmu_iotlb_inv_iova(SMMUState *s, uint16_t asid, dma_addr_t iova);
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index f3aa581f807..7dc8541e8b4 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -32,10 +32,42 @@
/* IOTLB Management */
+static guint smmu_iotlb_key_hash(gconstpointer v)
+{
+ SMMUIOTLBKey *key = (SMMUIOTLBKey *)v;
+ uint32_t a, b, c;
+
+ /* Jenkins hash */
+ a = b = c = JHASH_INITVAL + sizeof(*key);
+ a += key->asid;
+ b += extract64(key->iova, 0, 32);
+ c += extract64(key->iova, 32, 32);
+
+ __jhash_mix(a, b, c);
+ __jhash_final(a, b, c);
+
+ return c;
+}
+
+static gboolean smmu_iotlb_key_equal(gconstpointer v1, gconstpointer v2)
+{
+ const SMMUIOTLBKey *k1 = v1;
+ const SMMUIOTLBKey *k2 = v2;
+
+ return (k1->asid == k2->asid) && (k1->iova == k2->iova);
+}
+
+SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova)
+{
+ SMMUIOTLBKey key = {.asid = asid, .iova = iova};
+
+ return key;
+}
+
IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
hwaddr iova)
{
- SMMUIOTLBKey key = {.asid = cfg->asid, .iova = iova};
+ SMMUIOTLBKey key = smmu_get_iotlb_key(cfg->asid, iova);
IOMMUTLBEntry *entry = g_hash_table_lookup(bs->iotlb, &key);
if (entry) {
@@ -62,8 +94,7 @@ void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry)
smmu_iotlb_inv_all(bs);
}
- key->asid = cfg->asid;
- key->iova = entry->iova;
+ *key = smmu_get_iotlb_key(cfg->asid, entry->iova);
trace_smmu_iotlb_insert(cfg->asid, entry->iova);
g_hash_table_insert(bs->iotlb, key, entry);
}
@@ -80,12 +111,12 @@ static gboolean smmu_hash_remove_by_asid(gpointer key, gpointer value,
uint16_t asid = *(uint16_t *)user_data;
SMMUIOTLBKey *iotlb_key = (SMMUIOTLBKey *)key;
- return iotlb_key->asid == asid;
+ return SMMU_IOTLB_ASID(*iotlb_key) == asid;
}
inline void smmu_iotlb_inv_iova(SMMUState *s, uint16_t asid, dma_addr_t iova)
{
- SMMUIOTLBKey key = {.asid = asid, .iova = iova};
+ SMMUIOTLBKey key = smmu_get_iotlb_key(asid, iova);
trace_smmu_iotlb_inv_iova(asid, iova);
g_hash_table_remove(s->iotlb, &key);
@@ -383,31 +414,6 @@ IOMMUMemoryRegion *smmu_iommu_mr(SMMUState *s, uint32_t sid)
return NULL;
}
-static guint smmu_iotlb_key_hash(gconstpointer v)
-{
- SMMUIOTLBKey *key = (SMMUIOTLBKey *)v;
- uint32_t a, b, c;
-
- /* Jenkins hash */
- a = b = c = JHASH_INITVAL + sizeof(*key);
- a += key->asid;
- b += extract64(key->iova, 0, 32);
- c += extract64(key->iova, 32, 32);
-
- __jhash_mix(a, b, c);
- __jhash_final(a, b, c);
-
- return c;
-}
-
-static gboolean smmu_iotlb_key_equal(gconstpointer v1, gconstpointer v2)
-{
- const SMMUIOTLBKey *k1 = v1;
- const SMMUIOTLBKey *k2 = v2;
-
- return (k1->asid == k2->asid) && (k1->iova == k2->iova);
-}
-
/* Unmap the whole notifier's range */
static void smmu_unmap_notifier_range(IOMMUNotifier *n)
{
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 05/27] hw/arm/smmu: Introduce SMMUTLBEntry for PTW and IOTLB value
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (3 preceding siblings ...)
2020-08-24 9:47 ` [PULL 04/27] hw/arm/smmu: Introduce smmu_get_iotlb_key() Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 06/27] hw/arm/smmu-common: Manage IOTLB block entries Peter Maydell
` (22 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Introduce a specialized SMMUTLBEntry to store the result of
the PTW and cache in the IOTLB. This structure extends the
generic IOMMUTLBEntry struct with the level of the entry and
the granule size.
Those latter will be useful when implementing range invalidation.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-5-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
include/hw/arm/smmu-common.h | 12 +++++++++---
hw/arm/smmu-common.c | 32 +++++++++++++++++---------------
hw/arm/smmuv3.c | 10 +++++-----
3 files changed, 31 insertions(+), 23 deletions(-)
diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
index 5f9f3535d2a..79c2c6486ad 100644
--- a/include/hw/arm/smmu-common.h
+++ b/include/hw/arm/smmu-common.h
@@ -52,6 +52,12 @@ typedef struct SMMUTransTableInfo {
uint8_t granule_sz; /* granule page shift */
} SMMUTransTableInfo;
+typedef struct SMMUTLBEntry {
+ IOMMUTLBEntry entry;
+ uint8_t level;
+ uint8_t granule;
+} SMMUTLBEntry;
+
/*
* Generic structure populated by derived SMMU devices
* after decoding the configuration information and used as
@@ -140,7 +146,7 @@ static inline uint16_t smmu_get_sid(SMMUDevice *sdev)
* pair, according to @cfg translation config
*/
int smmu_ptw(SMMUTransCfg *cfg, dma_addr_t iova, IOMMUAccessFlags perm,
- IOMMUTLBEntry *tlbe, SMMUPTWEventInfo *info);
+ SMMUTLBEntry *tlbe, SMMUPTWEventInfo *info);
/**
* select_tt - compute which translation table shall be used according to
@@ -153,8 +159,8 @@ IOMMUMemoryRegion *smmu_iommu_mr(SMMUState *s, uint32_t sid);
#define SMMU_IOTLB_MAX_SIZE 256
-IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg, hwaddr iova);
-void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry);
+SMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg, hwaddr iova);
+void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, SMMUTLBEntry *entry);
SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova);
void smmu_iotlb_inv_all(SMMUState *s);
void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid);
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index 7dc8541e8b4..398e958bb44 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -64,11 +64,11 @@ SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova)
return key;
}
-IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
- hwaddr iova)
+SMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
+ hwaddr iova)
{
SMMUIOTLBKey key = smmu_get_iotlb_key(cfg->asid, iova);
- IOMMUTLBEntry *entry = g_hash_table_lookup(bs->iotlb, &key);
+ SMMUTLBEntry *entry = g_hash_table_lookup(bs->iotlb, &key);
if (entry) {
cfg->iotlb_hits++;
@@ -86,7 +86,7 @@ IOMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
return entry;
}
-void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry)
+void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, SMMUTLBEntry *new)
{
SMMUIOTLBKey *key = g_new0(SMMUIOTLBKey, 1);
@@ -94,9 +94,9 @@ void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, IOMMUTLBEntry *entry)
smmu_iotlb_inv_all(bs);
}
- *key = smmu_get_iotlb_key(cfg->asid, entry->iova);
- trace_smmu_iotlb_insert(cfg->asid, entry->iova);
- g_hash_table_insert(bs->iotlb, key, entry);
+ *key = smmu_get_iotlb_key(cfg->asid, new->entry.iova);
+ trace_smmu_iotlb_insert(cfg->asid, new->entry.iova);
+ g_hash_table_insert(bs->iotlb, key, new);
}
inline void smmu_iotlb_inv_all(SMMUState *s)
@@ -216,7 +216,7 @@ SMMUTransTableInfo *select_tt(SMMUTransCfg *cfg, dma_addr_t iova)
* @cfg: translation config
* @iova: iova to translate
* @perm: access type
- * @tlbe: IOMMUTLBEntry (out)
+ * @tlbe: SMMUTLBEntry (out)
* @info: handle to an error info
*
* Return 0 on success, < 0 on error. In case of error, @info is filled
@@ -226,7 +226,7 @@ SMMUTransTableInfo *select_tt(SMMUTransCfg *cfg, dma_addr_t iova)
*/
static int smmu_ptw_64(SMMUTransCfg *cfg,
dma_addr_t iova, IOMMUAccessFlags perm,
- IOMMUTLBEntry *tlbe, SMMUPTWEventInfo *info)
+ SMMUTLBEntry *tlbe, SMMUPTWEventInfo *info)
{
dma_addr_t baseaddr, indexmask;
int stage = cfg->stage;
@@ -246,8 +246,8 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
baseaddr = extract64(tt->ttb, 0, 48);
baseaddr &= ~indexmask;
- tlbe->iova = iova;
- tlbe->addr_mask = (1 << granule_sz) - 1;
+ tlbe->entry.iova = iova;
+ tlbe->entry.addr_mask = (1 << granule_sz) - 1;
while (level <= 3) {
uint64_t subpage_size = 1ULL << level_shift(level, granule_sz);
@@ -298,14 +298,16 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
goto error;
}
- tlbe->translated_addr = gpa + (iova & mask);
- tlbe->perm = PTE_AP_TO_PERM(ap);
+ tlbe->entry.translated_addr = gpa + (iova & mask);
+ tlbe->entry.perm = PTE_AP_TO_PERM(ap);
+ tlbe->level = level;
+ tlbe->granule = granule_sz;
return 0;
}
info->type = SMMU_PTW_ERR_TRANSLATION;
error:
- tlbe->perm = IOMMU_NONE;
+ tlbe->entry.perm = IOMMU_NONE;
return -EINVAL;
}
@@ -321,7 +323,7 @@ error:
* return 0 on success
*/
inline int smmu_ptw(SMMUTransCfg *cfg, dma_addr_t iova, IOMMUAccessFlags perm,
- IOMMUTLBEntry *tlbe, SMMUPTWEventInfo *info)
+ SMMUTLBEntry *tlbe, SMMUPTWEventInfo *info)
{
if (!cfg->aa64) {
/*
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index cd2a2e7e148..db74d27add5 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -626,7 +626,7 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
SMMUTranslationStatus status;
SMMUState *bs = ARM_SMMU(s);
uint64_t page_mask, aligned_addr;
- IOMMUTLBEntry *cached_entry = NULL;
+ SMMUTLBEntry *cached_entry = NULL;
SMMUTransTableInfo *tt;
SMMUTransCfg *cfg = NULL;
IOMMUTLBEntry entry = {
@@ -676,7 +676,7 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
cached_entry = smmu_iotlb_lookup(bs, cfg, aligned_addr);
if (cached_entry) {
- if ((flag & IOMMU_WO) && !(cached_entry->perm & IOMMU_WO)) {
+ if ((flag & IOMMU_WO) && !(cached_entry->entry.perm & IOMMU_WO)) {
status = SMMU_TRANS_ERROR;
if (event.record_trans_faults) {
event.type = SMMU_EVT_F_PERMISSION;
@@ -689,7 +689,7 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
goto epilogue;
}
- cached_entry = g_new0(IOMMUTLBEntry, 1);
+ cached_entry = g_new0(SMMUTLBEntry, 1);
if (smmu_ptw(cfg, aligned_addr, flag, cached_entry, &ptw_info)) {
g_free(cached_entry);
@@ -743,9 +743,9 @@ epilogue:
switch (status) {
case SMMU_TRANS_SUCCESS:
entry.perm = flag;
- entry.translated_addr = cached_entry->translated_addr +
+ entry.translated_addr = cached_entry->entry.translated_addr +
(addr & page_mask);
- entry.addr_mask = cached_entry->addr_mask;
+ entry.addr_mask = cached_entry->entry.addr_mask;
trace_smmuv3_translate_success(mr->parent_obj.name, sid, addr,
entry.translated_addr, entry.perm);
break;
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 06/27] hw/arm/smmu-common: Manage IOTLB block entries
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (4 preceding siblings ...)
2020-08-24 9:47 ` [PULL 05/27] hw/arm/smmu: Introduce SMMUTLBEntry for PTW and IOTLB value Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 07/27] hw/arm/smmuv3: Introduce smmuv3_s1_range_inval() helper Peter Maydell
` (21 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
At the moment each entry in the IOTLB corresponds to a page sized
mapping (4K, 16K or 64K), even if the page belongs to a mapped
block. In case of block mapping this unefficiently consumes IOTLB
entries.
Change the value of the entry so that it reflects the actual
mapping it belongs to (block or page start address and size).
Also the level/tg of the entry is encoded in the key. In subsequent
patches we will enable range invalidation. This latter is able
to provide the level/tg of the entry.
Encoding the level/tg directly in the key will allow to invalidate
using g_hash_table_remove() when num_pages equals to 1.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-6-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmu-internal.h | 7 ++++
include/hw/arm/smmu-common.h | 10 ++++--
hw/arm/smmu-common.c | 67 ++++++++++++++++++++++++++----------
hw/arm/smmuv3.c | 6 ++--
hw/arm/trace-events | 2 +-
5 files changed, 67 insertions(+), 25 deletions(-)
diff --git a/hw/arm/smmu-internal.h b/hw/arm/smmu-internal.h
index 3104f768cd2..55147f29be4 100644
--- a/hw/arm/smmu-internal.h
+++ b/hw/arm/smmu-internal.h
@@ -97,4 +97,11 @@ uint64_t iova_level_offset(uint64_t iova, int inputsize,
}
#define SMMU_IOTLB_ASID(key) ((key).asid)
+
+typedef struct SMMUIOTLBPageInvInfo {
+ int asid;
+ uint64_t iova;
+ uint64_t mask;
+} SMMUIOTLBPageInvInfo;
+
#endif
diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
index 79c2c6486ad..ece62c36523 100644
--- a/include/hw/arm/smmu-common.h
+++ b/include/hw/arm/smmu-common.h
@@ -97,6 +97,8 @@ typedef struct SMMUPciBus {
typedef struct SMMUIOTLBKey {
uint64_t iova;
uint16_t asid;
+ uint8_t tg;
+ uint8_t level;
} SMMUIOTLBKey;
typedef struct SMMUState {
@@ -159,12 +161,14 @@ IOMMUMemoryRegion *smmu_iommu_mr(SMMUState *s, uint32_t sid);
#define SMMU_IOTLB_MAX_SIZE 256
-SMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg, hwaddr iova);
+SMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
+ SMMUTransTableInfo *tt, hwaddr iova);
void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, SMMUTLBEntry *entry);
-SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova);
+SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova,
+ uint8_t tg, uint8_t level);
void smmu_iotlb_inv_all(SMMUState *s);
void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid);
-void smmu_iotlb_inv_iova(SMMUState *s, uint16_t asid, dma_addr_t iova);
+void smmu_iotlb_inv_iova(SMMUState *s, int asid, dma_addr_t iova);
/* Unmap the range of all the notifiers registered to any IOMMU mr */
void smmu_inv_notifiers_all(SMMUState *s);
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index 398e958bb44..2922deec6fc 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -39,7 +39,7 @@ static guint smmu_iotlb_key_hash(gconstpointer v)
/* Jenkins hash */
a = b = c = JHASH_INITVAL + sizeof(*key);
- a += key->asid;
+ a += key->asid + key->level + key->tg;
b += extract64(key->iova, 0, 32);
c += extract64(key->iova, 32, 32);
@@ -51,24 +51,41 @@ static guint smmu_iotlb_key_hash(gconstpointer v)
static gboolean smmu_iotlb_key_equal(gconstpointer v1, gconstpointer v2)
{
- const SMMUIOTLBKey *k1 = v1;
- const SMMUIOTLBKey *k2 = v2;
+ SMMUIOTLBKey *k1 = (SMMUIOTLBKey *)v1, *k2 = (SMMUIOTLBKey *)v2;
- return (k1->asid == k2->asid) && (k1->iova == k2->iova);
+ return (k1->asid == k2->asid) && (k1->iova == k2->iova) &&
+ (k1->level == k2->level) && (k1->tg == k2->tg);
}
-SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova)
+SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova,
+ uint8_t tg, uint8_t level)
{
- SMMUIOTLBKey key = {.asid = asid, .iova = iova};
+ SMMUIOTLBKey key = {.asid = asid, .iova = iova, .tg = tg, .level = level};
return key;
}
SMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
- hwaddr iova)
+ SMMUTransTableInfo *tt, hwaddr iova)
{
- SMMUIOTLBKey key = smmu_get_iotlb_key(cfg->asid, iova);
- SMMUTLBEntry *entry = g_hash_table_lookup(bs->iotlb, &key);
+ uint8_t tg = (tt->granule_sz - 10) / 2;
+ uint8_t inputsize = 64 - tt->tsz;
+ uint8_t stride = tt->granule_sz - 3;
+ uint8_t level = 4 - (inputsize - 4) / stride;
+ SMMUTLBEntry *entry = NULL;
+
+ while (level <= 3) {
+ uint64_t subpage_size = 1ULL << level_shift(level, tt->granule_sz);
+ uint64_t mask = subpage_size - 1;
+ SMMUIOTLBKey key;
+
+ key = smmu_get_iotlb_key(cfg->asid, iova & ~mask, tg, level);
+ entry = g_hash_table_lookup(bs->iotlb, &key);
+ if (entry) {
+ break;
+ }
+ level++;
+ }
if (entry) {
cfg->iotlb_hits++;
@@ -89,13 +106,14 @@ SMMUTLBEntry *smmu_iotlb_lookup(SMMUState *bs, SMMUTransCfg *cfg,
void smmu_iotlb_insert(SMMUState *bs, SMMUTransCfg *cfg, SMMUTLBEntry *new)
{
SMMUIOTLBKey *key = g_new0(SMMUIOTLBKey, 1);
+ uint8_t tg = (new->granule - 10) / 2;
if (g_hash_table_size(bs->iotlb) >= SMMU_IOTLB_MAX_SIZE) {
smmu_iotlb_inv_all(bs);
}
- *key = smmu_get_iotlb_key(cfg->asid, new->entry.iova);
- trace_smmu_iotlb_insert(cfg->asid, new->entry.iova);
+ *key = smmu_get_iotlb_key(cfg->asid, new->entry.iova, tg, new->level);
+ trace_smmu_iotlb_insert(cfg->asid, new->entry.iova, tg, new->level);
g_hash_table_insert(bs->iotlb, key, new);
}
@@ -114,12 +132,26 @@ static gboolean smmu_hash_remove_by_asid(gpointer key, gpointer value,
return SMMU_IOTLB_ASID(*iotlb_key) == asid;
}
-inline void smmu_iotlb_inv_iova(SMMUState *s, uint16_t asid, dma_addr_t iova)
+static gboolean smmu_hash_remove_by_asid_iova(gpointer key, gpointer value,
+ gpointer user_data)
{
- SMMUIOTLBKey key = smmu_get_iotlb_key(asid, iova);
+ SMMUTLBEntry *iter = (SMMUTLBEntry *)value;
+ IOMMUTLBEntry *entry = &iter->entry;
+ SMMUIOTLBPageInvInfo *info = (SMMUIOTLBPageInvInfo *)user_data;
+ SMMUIOTLBKey iotlb_key = *(SMMUIOTLBKey *)key;
+
+ if (info->asid >= 0 && info->asid != SMMU_IOTLB_ASID(iotlb_key)) {
+ return false;
+ }
+ return (info->iova & ~entry->addr_mask) == entry->iova;
+}
+
+inline void smmu_iotlb_inv_iova(SMMUState *s, int asid, dma_addr_t iova)
+{
+ SMMUIOTLBPageInvInfo info = {.asid = asid, .iova = iova};
trace_smmu_iotlb_inv_iova(asid, iova);
- g_hash_table_remove(s->iotlb, &key);
+ g_hash_table_foreach_remove(s->iotlb, smmu_hash_remove_by_asid_iova, &info);
}
inline void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid)
@@ -246,9 +278,6 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
baseaddr = extract64(tt->ttb, 0, 48);
baseaddr &= ~indexmask;
- tlbe->entry.iova = iova;
- tlbe->entry.addr_mask = (1 << granule_sz) - 1;
-
while (level <= 3) {
uint64_t subpage_size = 1ULL << level_shift(level, granule_sz);
uint64_t mask = subpage_size - 1;
@@ -298,7 +327,9 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
goto error;
}
- tlbe->entry.translated_addr = gpa + (iova & mask);
+ tlbe->entry.translated_addr = gpa;
+ tlbe->entry.iova = iova & ~mask;
+ tlbe->entry.addr_mask = mask;
tlbe->entry.perm = PTE_AP_TO_PERM(ap);
tlbe->level = level;
tlbe->granule = granule_sz;
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index db74d27add5..b717bde8320 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -674,7 +674,7 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
page_mask = (1ULL << (tt->granule_sz)) - 1;
aligned_addr = addr & ~page_mask;
- cached_entry = smmu_iotlb_lookup(bs, cfg, aligned_addr);
+ cached_entry = smmu_iotlb_lookup(bs, cfg, tt, aligned_addr);
if (cached_entry) {
if ((flag & IOMMU_WO) && !(cached_entry->entry.perm & IOMMU_WO)) {
status = SMMU_TRANS_ERROR;
@@ -744,7 +744,7 @@ epilogue:
case SMMU_TRANS_SUCCESS:
entry.perm = flag;
entry.translated_addr = cached_entry->entry.translated_addr +
- (addr & page_mask);
+ (addr & cached_entry->entry.addr_mask);
entry.addr_mask = cached_entry->entry.addr_mask;
trace_smmuv3_translate_success(mr->parent_obj.name, sid, addr,
entry.translated_addr, entry.perm);
@@ -972,7 +972,7 @@ static int smmuv3_cmdq_consume(SMMUv3State *s)
trace_smmuv3_cmdq_tlbi_nh_vaa(vmid, addr);
smmuv3_inv_notifiers_iova(bs, -1, addr);
- smmu_iotlb_inv_all(bs);
+ smmu_iotlb_inv_iova(bs, -1, addr);
break;
}
case SMMU_CMD_TLBI_NH_VA:
diff --git a/hw/arm/trace-events b/hw/arm/trace-events
index b808a1bfc19..f74d3e920f1 100644
--- a/hw/arm/trace-events
+++ b/hw/arm/trace-events
@@ -16,7 +16,7 @@ smmu_iotlb_inv_iova(uint16_t asid, uint64_t addr) "IOTLB invalidate asid=%d addr
smmu_inv_notifiers_mr(const char *name) "iommu mr=%s"
smmu_iotlb_lookup_hit(uint16_t asid, uint64_t addr, uint32_t hit, uint32_t miss, uint32_t p) "IOTLB cache HIT asid=%d addr=0x%"PRIx64" hit=%d miss=%d hit rate=%d"
smmu_iotlb_lookup_miss(uint16_t asid, uint64_t addr, uint32_t hit, uint32_t miss, uint32_t p) "IOTLB cache MISS asid=%d addr=0x%"PRIx64" hit=%d miss=%d hit rate=%d"
-smmu_iotlb_insert(uint16_t asid, uint64_t addr) "IOTLB ++ asid=%d addr=0x%"PRIx64
+smmu_iotlb_insert(uint16_t asid, uint64_t addr, uint8_t tg, uint8_t level) "IOTLB ++ asid=%d addr=0x%"PRIx64" tg=%d level=%d"
# smmuv3.c
smmuv3_read_mmio(uint64_t addr, uint64_t val, unsigned size, uint32_t r) "addr: 0x%"PRIx64" val:0x%"PRIx64" size: 0x%x(%d)"
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 07/27] hw/arm/smmuv3: Introduce smmuv3_s1_range_inval() helper
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (5 preceding siblings ...)
2020-08-24 9:47 ` [PULL 06/27] hw/arm/smmu-common: Manage IOTLB block entries Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 08/27] hw/arm/smmuv3: Get prepared for range invalidation Peter Maydell
` (20 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Let's introduce an helper for S1 IOVA range invalidation.
This will be used for NH_VA and NH_VAA commands. It decodes
the same fields, trace, calls the UNMAP notifiers and
invalidate the corresponding IOTLB entries.
At the moment, we do not support 3.2 range invalidation yet.
So it reduces to a single IOVA invalidation.
Note the leaf bit now is also decoded for the CMD_TLBI_NH_VAA
command. At the moment it is only used for tracing.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-7-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmuv3.c | 36 +++++++++++++++++-------------------
hw/arm/trace-events | 3 +--
2 files changed, 18 insertions(+), 21 deletions(-)
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index b717bde8320..e4a2cea7adc 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -836,6 +836,22 @@ static void smmuv3_inv_notifiers_iova(SMMUState *s, int asid, dma_addr_t iova)
}
}
+static void smmuv3_s1_range_inval(SMMUState *s, Cmd *cmd)
+{
+ dma_addr_t addr = CMD_ADDR(cmd);
+ uint8_t type = CMD_TYPE(cmd);
+ uint16_t vmid = CMD_VMID(cmd);
+ bool leaf = CMD_LEAF(cmd);
+ int asid = -1;
+
+ if (type == SMMU_CMD_TLBI_NH_VA) {
+ asid = CMD_ASID(cmd);
+ }
+ trace_smmuv3_s1_range_inval(vmid, asid, addr, leaf);
+ smmuv3_inv_notifiers_iova(s, asid, addr);
+ smmu_iotlb_inv_iova(s, asid, addr);
+}
+
static int smmuv3_cmdq_consume(SMMUv3State *s)
{
SMMUState *bs = ARM_SMMU(s);
@@ -966,27 +982,9 @@ static int smmuv3_cmdq_consume(SMMUv3State *s)
smmu_iotlb_inv_all(bs);
break;
case SMMU_CMD_TLBI_NH_VAA:
- {
- dma_addr_t addr = CMD_ADDR(&cmd);
- uint16_t vmid = CMD_VMID(&cmd);
-
- trace_smmuv3_cmdq_tlbi_nh_vaa(vmid, addr);
- smmuv3_inv_notifiers_iova(bs, -1, addr);
- smmu_iotlb_inv_iova(bs, -1, addr);
- break;
- }
case SMMU_CMD_TLBI_NH_VA:
- {
- uint16_t asid = CMD_ASID(&cmd);
- uint16_t vmid = CMD_VMID(&cmd);
- dma_addr_t addr = CMD_ADDR(&cmd);
- bool leaf = CMD_LEAF(&cmd);
-
- trace_smmuv3_cmdq_tlbi_nh_va(vmid, asid, addr, leaf);
- smmuv3_inv_notifiers_iova(bs, asid, addr);
- smmu_iotlb_inv_iova(bs, asid, addr);
+ smmuv3_s1_range_inval(bs, &cmd);
break;
- }
case SMMU_CMD_TLBI_EL3_ALL:
case SMMU_CMD_TLBI_EL3_VA:
case SMMU_CMD_TLBI_EL2_ALL:
diff --git a/hw/arm/trace-events b/hw/arm/trace-events
index f74d3e920f1..c219fe9e828 100644
--- a/hw/arm/trace-events
+++ b/hw/arm/trace-events
@@ -45,8 +45,7 @@ smmuv3_cmdq_cfgi_ste_range(int start, int end) "start=0x%d - end=0x%d"
smmuv3_cmdq_cfgi_cd(uint32_t sid) "streamid = %d"
smmuv3_config_cache_hit(uint32_t sid, uint32_t hits, uint32_t misses, uint32_t perc) "Config cache HIT for sid %d (hits=%d, misses=%d, hit rate=%d)"
smmuv3_config_cache_miss(uint32_t sid, uint32_t hits, uint32_t misses, uint32_t perc) "Config cache MISS for sid %d (hits=%d, misses=%d, hit rate=%d)"
-smmuv3_cmdq_tlbi_nh_va(int vmid, int asid, uint64_t addr, bool leaf) "vmid =%d asid =%d addr=0x%"PRIx64" leaf=%d"
-smmuv3_cmdq_tlbi_nh_vaa(int vmid, uint64_t addr) "vmid =%d addr=0x%"PRIx64
+smmuv3_s1_range_inval(int vmid, int asid, uint64_t addr, bool leaf) "vmid =%d asid =%d addr=0x%"PRIx64" leaf=%d"
smmuv3_cmdq_tlbi_nh(void) ""
smmuv3_cmdq_tlbi_nh_asid(uint16_t asid) "asid=%d"
smmuv3_config_cache_inv(uint32_t sid) "Config cache INV for sid %d"
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 08/27] hw/arm/smmuv3: Get prepared for range invalidation
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (6 preceding siblings ...)
2020-08-24 9:47 ` [PULL 07/27] hw/arm/smmuv3: Introduce smmuv3_s1_range_inval() helper Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 09/27] hw/arm/smmuv3: Fix IIDR offset Peter Maydell
` (19 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Enhance the smmu_iotlb_inv_iova() helper with range invalidation.
This uses the new fields passed in the NH_VA and NH_VAA commands:
the size of the range, the level and the granule.
As NH_VA and NH_VAA both use those fields, their decoding and
handling is factorized in a new smmuv3_s1_range_inval() helper.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-8-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmuv3-internal.h | 4 +++
include/hw/arm/smmu-common.h | 3 +-
hw/arm/smmu-common.c | 25 +++++++++++---
hw/arm/smmuv3.c | 64 +++++++++++++++++++++++-------------
hw/arm/trace-events | 4 +--
5 files changed, 69 insertions(+), 31 deletions(-)
diff --git a/hw/arm/smmuv3-internal.h b/hw/arm/smmuv3-internal.h
index 4112394129e..5babf72f7d5 100644
--- a/hw/arm/smmuv3-internal.h
+++ b/hw/arm/smmuv3-internal.h
@@ -298,6 +298,8 @@ enum { /* Command completion notification */
};
#define CMD_TYPE(x) extract32((x)->word[0], 0 , 8)
+#define CMD_NUM(x) extract32((x)->word[0], 12 , 5)
+#define CMD_SCALE(x) extract32((x)->word[0], 20 , 5)
#define CMD_SSEC(x) extract32((x)->word[0], 10, 1)
#define CMD_SSV(x) extract32((x)->word[0], 11, 1)
#define CMD_RESUME_AC(x) extract32((x)->word[0], 12, 1)
@@ -310,6 +312,8 @@ enum { /* Command completion notification */
#define CMD_RESUME_STAG(x) extract32((x)->word[2], 0 , 16)
#define CMD_RESP(x) extract32((x)->word[2], 11, 2)
#define CMD_LEAF(x) extract32((x)->word[2], 0 , 1)
+#define CMD_TTL(x) extract32((x)->word[2], 8 , 2)
+#define CMD_TG(x) extract32((x)->word[2], 10, 2)
#define CMD_STE_RANGE(x) extract32((x)->word[2], 0 , 5)
#define CMD_ADDR(x) ({ \
uint64_t high = (uint64_t)(x)->word[3]; \
diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
index ece62c36523..4f6acf634cf 100644
--- a/include/hw/arm/smmu-common.h
+++ b/include/hw/arm/smmu-common.h
@@ -168,7 +168,8 @@ SMMUIOTLBKey smmu_get_iotlb_key(uint16_t asid, uint64_t iova,
uint8_t tg, uint8_t level);
void smmu_iotlb_inv_all(SMMUState *s);
void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid);
-void smmu_iotlb_inv_iova(SMMUState *s, int asid, dma_addr_t iova);
+void smmu_iotlb_inv_iova(SMMUState *s, int asid, dma_addr_t iova,
+ uint8_t tg, uint64_t num_pages, uint8_t ttl);
/* Unmap the range of all the notifiers registered to any IOMMU mr */
void smmu_inv_notifiers_all(SMMUState *s);
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index 2922deec6fc..8d89a86699a 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -143,15 +143,30 @@ static gboolean smmu_hash_remove_by_asid_iova(gpointer key, gpointer value,
if (info->asid >= 0 && info->asid != SMMU_IOTLB_ASID(iotlb_key)) {
return false;
}
- return (info->iova & ~entry->addr_mask) == entry->iova;
+ return ((info->iova & ~entry->addr_mask) == entry->iova) ||
+ ((entry->iova & ~info->mask) == info->iova);
}
-inline void smmu_iotlb_inv_iova(SMMUState *s, int asid, dma_addr_t iova)
+inline void
+smmu_iotlb_inv_iova(SMMUState *s, int asid, dma_addr_t iova,
+ uint8_t tg, uint64_t num_pages, uint8_t ttl)
{
- SMMUIOTLBPageInvInfo info = {.asid = asid, .iova = iova};
+ if (ttl && (num_pages == 1)) {
+ SMMUIOTLBKey key = smmu_get_iotlb_key(asid, iova, tg, ttl);
- trace_smmu_iotlb_inv_iova(asid, iova);
- g_hash_table_foreach_remove(s->iotlb, smmu_hash_remove_by_asid_iova, &info);
+ g_hash_table_remove(s->iotlb, &key);
+ } else {
+ /* if tg is not set we use 4KB range invalidation */
+ uint8_t granule = tg ? tg * 2 + 10 : 12;
+
+ SMMUIOTLBPageInvInfo info = {
+ .asid = asid, .iova = iova,
+ .mask = (num_pages * 1 << granule) - 1};
+
+ g_hash_table_foreach_remove(s->iotlb,
+ smmu_hash_remove_by_asid_iova,
+ &info);
+ }
}
inline void smmu_iotlb_inv_asid(SMMUState *s, uint16_t asid)
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index e4a2cea7adc..89ab11fc36a 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -785,42 +785,49 @@ epilogue:
* @n: notifier to be called
* @asid: address space ID or negative value if we don't care
* @iova: iova
+ * @tg: translation granule (if communicated through range invalidation)
+ * @num_pages: number of @granule sized pages (if tg != 0), otherwise 1
*/
static void smmuv3_notify_iova(IOMMUMemoryRegion *mr,
IOMMUNotifier *n,
- int asid,
- dma_addr_t iova)
+ int asid, dma_addr_t iova,
+ uint8_t tg, uint64_t num_pages)
{
SMMUDevice *sdev = container_of(mr, SMMUDevice, iommu);
- SMMUEventInfo event = {.inval_ste_allowed = true};
- SMMUTransTableInfo *tt;
- SMMUTransCfg *cfg;
IOMMUTLBEntry entry;
+ uint8_t granule = tg;
- cfg = smmuv3_get_config(sdev, &event);
- if (!cfg) {
- return;
- }
+ if (!tg) {
+ SMMUEventInfo event = {.inval_ste_allowed = true};
+ SMMUTransCfg *cfg = smmuv3_get_config(sdev, &event);
+ SMMUTransTableInfo *tt;
- if (asid >= 0 && cfg->asid != asid) {
- return;
- }
+ if (!cfg) {
+ return;
+ }
- tt = select_tt(cfg, iova);
- if (!tt) {
- return;
+ if (asid >= 0 && cfg->asid != asid) {
+ return;
+ }
+
+ tt = select_tt(cfg, iova);
+ if (!tt) {
+ return;
+ }
+ granule = tt->granule_sz;
}
entry.target_as = &address_space_memory;
entry.iova = iova;
- entry.addr_mask = (1 << tt->granule_sz) - 1;
+ entry.addr_mask = num_pages * (1 << granule) - 1;
entry.perm = IOMMU_NONE;
memory_region_notify_one(n, &entry);
}
-/* invalidate an asid/iova tuple in all mr's */
-static void smmuv3_inv_notifiers_iova(SMMUState *s, int asid, dma_addr_t iova)
+/* invalidate an asid/iova range tuple in all mr's */
+static void smmuv3_inv_notifiers_iova(SMMUState *s, int asid, dma_addr_t iova,
+ uint8_t tg, uint64_t num_pages)
{
SMMUDevice *sdev;
@@ -828,28 +835,39 @@ static void smmuv3_inv_notifiers_iova(SMMUState *s, int asid, dma_addr_t iova)
IOMMUMemoryRegion *mr = &sdev->iommu;
IOMMUNotifier *n;
- trace_smmuv3_inv_notifiers_iova(mr->parent_obj.name, asid, iova);
+ trace_smmuv3_inv_notifiers_iova(mr->parent_obj.name, asid, iova,
+ tg, num_pages);
IOMMU_NOTIFIER_FOREACH(n, mr) {
- smmuv3_notify_iova(mr, n, asid, iova);
+ smmuv3_notify_iova(mr, n, asid, iova, tg, num_pages);
}
}
}
static void smmuv3_s1_range_inval(SMMUState *s, Cmd *cmd)
{
+ uint8_t scale = 0, num = 0, ttl = 0;
dma_addr_t addr = CMD_ADDR(cmd);
uint8_t type = CMD_TYPE(cmd);
uint16_t vmid = CMD_VMID(cmd);
bool leaf = CMD_LEAF(cmd);
+ uint8_t tg = CMD_TG(cmd);
+ hwaddr num_pages = 1;
int asid = -1;
+ if (tg) {
+ scale = CMD_SCALE(cmd);
+ num = CMD_NUM(cmd);
+ ttl = CMD_TTL(cmd);
+ num_pages = (num + 1) * (1 << (scale));
+ }
+
if (type == SMMU_CMD_TLBI_NH_VA) {
asid = CMD_ASID(cmd);
}
- trace_smmuv3_s1_range_inval(vmid, asid, addr, leaf);
- smmuv3_inv_notifiers_iova(s, asid, addr);
- smmu_iotlb_inv_iova(s, asid, addr);
+ trace_smmuv3_s1_range_inval(vmid, asid, addr, tg, num_pages, ttl, leaf);
+ smmuv3_inv_notifiers_iova(s, asid, addr, tg, num_pages);
+ smmu_iotlb_inv_iova(s, asid, addr, tg, num_pages, ttl);
}
static int smmuv3_cmdq_consume(SMMUv3State *s)
diff --git a/hw/arm/trace-events b/hw/arm/trace-events
index c219fe9e828..3d905e0f7d0 100644
--- a/hw/arm/trace-events
+++ b/hw/arm/trace-events
@@ -45,11 +45,11 @@ smmuv3_cmdq_cfgi_ste_range(int start, int end) "start=0x%d - end=0x%d"
smmuv3_cmdq_cfgi_cd(uint32_t sid) "streamid = %d"
smmuv3_config_cache_hit(uint32_t sid, uint32_t hits, uint32_t misses, uint32_t perc) "Config cache HIT for sid %d (hits=%d, misses=%d, hit rate=%d)"
smmuv3_config_cache_miss(uint32_t sid, uint32_t hits, uint32_t misses, uint32_t perc) "Config cache MISS for sid %d (hits=%d, misses=%d, hit rate=%d)"
-smmuv3_s1_range_inval(int vmid, int asid, uint64_t addr, bool leaf) "vmid =%d asid =%d addr=0x%"PRIx64" leaf=%d"
+smmuv3_s1_range_inval(int vmid, int asid, uint64_t addr, uint8_t tg, uint64_t num_pages, uint8_t ttl, bool leaf) "vmid =%d asid =%d addr=0x%"PRIx64" tg=%d num_pages=0x%"PRIx64" ttl=%d leaf=%d"
smmuv3_cmdq_tlbi_nh(void) ""
smmuv3_cmdq_tlbi_nh_asid(uint16_t asid) "asid=%d"
smmuv3_config_cache_inv(uint32_t sid) "Config cache INV for sid %d"
smmuv3_notify_flag_add(const char *iommu) "ADD SMMUNotifier node for iommu mr=%s"
smmuv3_notify_flag_del(const char *iommu) "DEL SMMUNotifier node for iommu mr=%s"
-smmuv3_inv_notifiers_iova(const char *name, uint16_t asid, uint64_t iova) "iommu mr=%s asid=%d iova=0x%"PRIx64
+smmuv3_inv_notifiers_iova(const char *name, uint16_t asid, uint64_t iova, uint8_t tg, uint64_t num_pages) "iommu mr=%s asid=%d iova=0x%"PRIx64" tg=%d num_pages=0x%"PRIx64
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 09/27] hw/arm/smmuv3: Fix IIDR offset
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (7 preceding siblings ...)
2020-08-24 9:47 ` [PULL 08/27] hw/arm/smmuv3: Get prepared for range invalidation Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 10/27] hw/arm/smmuv3: Let AIDR advertise SMMUv3.0 support Peter Maydell
` (18 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
The SMMU IIDR register is at 0x018 offset.
Fixes: 10a83cb9887 ("hw/arm/smmuv3: Skeleton")
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-9-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmuv3-internal.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/arm/smmuv3-internal.h b/hw/arm/smmuv3-internal.h
index 5babf72f7d5..ef093eaff50 100644
--- a/hw/arm/smmuv3-internal.h
+++ b/hw/arm/smmuv3-internal.h
@@ -63,7 +63,7 @@ REG32(IDR5, 0x14)
#define SMMU_IDR5_OAS 4
-REG32(IIDR, 0x1c)
+REG32(IIDR, 0x18)
REG32(CR0, 0x20)
FIELD(CR0, SMMU_ENABLE, 0, 1)
FIELD(CR0, EVENTQEN, 2, 1)
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 10/27] hw/arm/smmuv3: Let AIDR advertise SMMUv3.0 support
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (8 preceding siblings ...)
2020-08-24 9:47 ` [PULL 09/27] hw/arm/smmuv3: Fix IIDR offset Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 11/27] hw/arm/smmuv3: Support HAD and advertise SMMUv3.1 support Peter Maydell
` (17 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Add the support for AIDR register. It currently advertises
SMMU V3.0 spec.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-10-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmuv3-internal.h | 1 +
include/hw/arm/smmuv3.h | 1 +
hw/arm/smmuv3.c | 3 +++
3 files changed, 5 insertions(+)
diff --git a/hw/arm/smmuv3-internal.h b/hw/arm/smmuv3-internal.h
index ef093eaff50..bd34a4f3300 100644
--- a/hw/arm/smmuv3-internal.h
+++ b/hw/arm/smmuv3-internal.h
@@ -64,6 +64,7 @@ REG32(IDR5, 0x14)
#define SMMU_IDR5_OAS 4
REG32(IIDR, 0x18)
+REG32(AIDR, 0x1c)
REG32(CR0, 0x20)
FIELD(CR0, SMMU_ENABLE, 0, 1)
FIELD(CR0, EVENTQEN, 2, 1)
diff --git a/include/hw/arm/smmuv3.h b/include/hw/arm/smmuv3.h
index 36b2f452539..68d7a963e0f 100644
--- a/include/hw/arm/smmuv3.h
+++ b/include/hw/arm/smmuv3.h
@@ -41,6 +41,7 @@ typedef struct SMMUv3State {
uint32_t idr[6];
uint32_t iidr;
+ uint32_t aidr;
uint32_t cr[3];
uint32_t cr0ack;
uint32_t statusr;
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index 89ab11fc36a..718f28462ea 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -1251,6 +1251,9 @@ static MemTxResult smmu_readl(SMMUv3State *s, hwaddr offset,
case A_IIDR:
*data = s->iidr;
return MEMTX_OK;
+ case A_AIDR:
+ *data = s->aidr;
+ return MEMTX_OK;
case A_CR0:
*data = s->cr[0];
return MEMTX_OK;
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 11/27] hw/arm/smmuv3: Support HAD and advertise SMMUv3.1 support
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (9 preceding siblings ...)
2020-08-24 9:47 ` [PULL 10/27] hw/arm/smmuv3: Let AIDR advertise SMMUv3.0 support Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 12/27] hw/arm/smmuv3: Advertise SMMUv3.2 range invalidation Peter Maydell
` (16 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
HAD is a mandatory features with SMMUv3.1 if S1P is set, which is
our case. Other 3.1 mandatory features come with S2P which we don't
have.
So let's support HAD and advertise SMMUv3.1 support in AIDR.
HAD support allows the CD to disable hierarchical attributes, ie.
if the HAD0/1 bit is set, the APTable field of table descriptors
walked through TTB0/1 is ignored.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-11-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmuv3-internal.h | 2 ++
include/hw/arm/smmu-common.h | 1 +
hw/arm/smmu-common.c | 2 +-
hw/arm/smmuv3.c | 6 +++++-
hw/arm/trace-events | 2 +-
5 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/hw/arm/smmuv3-internal.h b/hw/arm/smmuv3-internal.h
index bd34a4f3300..9ae7d97fafd 100644
--- a/hw/arm/smmuv3-internal.h
+++ b/hw/arm/smmuv3-internal.h
@@ -54,6 +54,7 @@ REG32(IDR1, 0x4)
REG32(IDR2, 0x8)
REG32(IDR3, 0xc)
+ FIELD(IDR3, HAD, 2, 1);
REG32(IDR4, 0x10)
REG32(IDR5, 0x14)
FIELD(IDR5, OAS, 0, 3);
@@ -578,6 +579,7 @@ static inline int pa_range(STE *ste)
lo = (x)->word[(sel) * 2 + 2] & ~0xfULL; \
hi | lo; \
})
+#define CD_HAD(x, sel) extract32((x)->word[(sel) * 2 + 2], 1, 1)
#define CD_TSZ(x, sel) extract32((x)->word[0], (16 * (sel)) + 0, 6)
#define CD_TG(x, sel) extract32((x)->word[0], (16 * (sel)) + 6, 2)
diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
index 4f6acf634cf..880dccd7c04 100644
--- a/include/hw/arm/smmu-common.h
+++ b/include/hw/arm/smmu-common.h
@@ -50,6 +50,7 @@ typedef struct SMMUTransTableInfo {
uint64_t ttb; /* TT base address */
uint8_t tsz; /* input range, ie. 2^(64 -tsz)*/
uint8_t granule_sz; /* granule page shift */
+ bool had; /* hierarchical attribute disable */
} SMMUTransTableInfo;
typedef struct SMMUTLBEntry {
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index 8d89a86699a..3838db13952 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -316,7 +316,7 @@ static int smmu_ptw_64(SMMUTransCfg *cfg,
if (is_table_pte(pte, level)) {
ap = PTE_APTABLE(pte);
- if (is_permission_fault(ap, perm)) {
+ if (is_permission_fault(ap, perm) && !tt->had) {
info->type = SMMU_PTW_ERR_PERMISSION;
goto error;
}
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index 718f28462ea..b262f0e4a74 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -254,6 +254,8 @@ static void smmuv3_init_regs(SMMUv3State *s)
s->idr[1] = FIELD_DP32(s->idr[1], IDR1, EVENTQS, SMMU_EVENTQS);
s->idr[1] = FIELD_DP32(s->idr[1], IDR1, CMDQS, SMMU_CMDQS);
+ s->idr[3] = FIELD_DP32(s->idr[3], IDR3, HAD, 1);
+
/* 4K and 64K granule support */
s->idr[5] = FIELD_DP32(s->idr[5], IDR5, GRAN4K, 1);
s->idr[5] = FIELD_DP32(s->idr[5], IDR5, GRAN64K, 1);
@@ -270,6 +272,7 @@ static void smmuv3_init_regs(SMMUv3State *s)
s->features = 0;
s->sid_split = 0;
+ s->aidr = 0x1;
}
static int smmu_get_ste(SMMUv3State *s, dma_addr_t addr, STE *buf,
@@ -506,7 +509,8 @@ static int decode_cd(SMMUTransCfg *cfg, CD *cd, SMMUEventInfo *event)
if (tt->ttb & ~(MAKE_64BIT_MASK(0, cfg->oas))) {
goto bad_cd;
}
- trace_smmuv3_decode_cd_tt(i, tt->tsz, tt->ttb, tt->granule_sz);
+ tt->had = CD_HAD(cd, i);
+ trace_smmuv3_decode_cd_tt(i, tt->tsz, tt->ttb, tt->granule_sz, tt->had);
}
event->record_trans_faults = CD_R(cd);
diff --git a/hw/arm/trace-events b/hw/arm/trace-events
index 3d905e0f7d0..c8a4d80f6bd 100644
--- a/hw/arm/trace-events
+++ b/hw/arm/trace-events
@@ -39,7 +39,7 @@ smmuv3_translate_abort(const char *n, uint16_t sid, uint64_t addr, bool is_write
smmuv3_translate_success(const char *n, uint16_t sid, uint64_t iova, uint64_t translated, int perm) "%s sid=%d iova=0x%"PRIx64" translated=0x%"PRIx64" perm=0x%x"
smmuv3_get_cd(uint64_t addr) "CD addr: 0x%"PRIx64
smmuv3_decode_cd(uint32_t oas) "oas=%d"
-smmuv3_decode_cd_tt(int i, uint32_t tsz, uint64_t ttb, uint32_t granule_sz) "TT[%d]:tsz:%d ttb:0x%"PRIx64" granule_sz:%d"
+smmuv3_decode_cd_tt(int i, uint32_t tsz, uint64_t ttb, uint32_t granule_sz, bool had) "TT[%d]:tsz:%d ttb:0x%"PRIx64" granule_sz:%d had:%d"
smmuv3_cmdq_cfgi_ste(int streamid) "streamid =%d"
smmuv3_cmdq_cfgi_ste_range(int start, int end) "start=0x%d - end=0x%d"
smmuv3_cmdq_cfgi_cd(uint32_t sid) "streamid = %d"
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 12/27] hw/arm/smmuv3: Advertise SMMUv3.2 range invalidation
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (10 preceding siblings ...)
2020-08-24 9:47 ` [PULL 11/27] hw/arm/smmuv3: Support HAD and advertise SMMUv3.1 support Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 13/27] docs/system/arm: Document the Xilinx Versal Virt board Peter Maydell
` (15 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: Eric Auger <eric.auger@redhat.com>
Expose the RIL bit so that the guest driver uses range
invalidation. Although RIL is a 3.2 features, We let
the AIDR advertise SMMUv3.1 support as v3.x implementation
is allowed to implement features from v3.(x+1).
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200728150815.11446-12-eric.auger@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
hw/arm/smmuv3-internal.h | 1 +
hw/arm/smmuv3.c | 1 +
2 files changed, 2 insertions(+)
diff --git a/hw/arm/smmuv3-internal.h b/hw/arm/smmuv3-internal.h
index 9ae7d97fafd..fa3c088972e 100644
--- a/hw/arm/smmuv3-internal.h
+++ b/hw/arm/smmuv3-internal.h
@@ -55,6 +55,7 @@ REG32(IDR1, 0x4)
REG32(IDR2, 0x8)
REG32(IDR3, 0xc)
FIELD(IDR3, HAD, 2, 1);
+ FIELD(IDR3, RIL, 10, 1);
REG32(IDR4, 0x10)
REG32(IDR5, 0x14)
FIELD(IDR5, OAS, 0, 3);
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index b262f0e4a74..0122700e725 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -254,6 +254,7 @@ static void smmuv3_init_regs(SMMUv3State *s)
s->idr[1] = FIELD_DP32(s->idr[1], IDR1, EVENTQS, SMMU_EVENTQS);
s->idr[1] = FIELD_DP32(s->idr[1], IDR1, CMDQS, SMMU_CMDQS);
+ s->idr[3] = FIELD_DP32(s->idr[3], IDR3, RIL, 1);
s->idr[3] = FIELD_DP32(s->idr[3], IDR3, HAD, 1);
/* 4K and 64K granule support */
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 13/27] docs/system/arm: Document the Xilinx Versal Virt board
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (11 preceding siblings ...)
2020-08-24 9:47 ` [PULL 12/27] hw/arm/smmuv3: Advertise SMMUv3.2 range invalidation Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 14/27] target/arm: Pull handling of XScale insns out of disas_coproc_insn() Peter Maydell
` (14 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com>
Document the Xilinx Versal Virt board.
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>
Message-id: 20200803164749.301971-2-edgar.iglesias@gmail.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
docs/system/arm/xlnx-versal-virt.rst | 176 +++++++++++++++++++++++++++
docs/system/target-arm.rst | 1 +
MAINTAINERS | 3 +-
3 files changed, 179 insertions(+), 1 deletion(-)
create mode 100644 docs/system/arm/xlnx-versal-virt.rst
diff --git a/docs/system/arm/xlnx-versal-virt.rst b/docs/system/arm/xlnx-versal-virt.rst
new file mode 100644
index 00000000000..2602d0f9953
--- /dev/null
+++ b/docs/system/arm/xlnx-versal-virt.rst
@@ -0,0 +1,176 @@
+Xilinx Versal Virt (``xlnx-versal-virt``)
+=========================================
+
+Xilinx Versal is a family of heterogeneous multi-core SoCs
+(System on Chip) that combine traditional hardened CPUs and I/O
+peripherals in a Processing System (PS) with runtime programmable
+FPGA logic (PL) and an Artificial Intelligence Engine (AIE).
+
+More details here:
+https://www.xilinx.com/products/silicon-devices/acap/versal.html
+
+The family of Versal SoCs share a single architecture but come in
+different parts with different speed grades, amounts of PL and
+other differences.
+
+The Xilinx Versal Virt board in QEMU is a model of a virtual board
+(does not exist in reality) with a virtual Versal SoC without I/O
+limitations. Currently, we support the following cores and devices:
+
+Implemented CPU cores:
+
+- 2 ACPUs (ARM Cortex-A72)
+
+Implemented devices:
+
+- Interrupt controller (ARM GICv3)
+- 2 UARTs (ARM PL011)
+- An RTC (Versal built-in)
+- 2 GEMs (Cadence MACB Ethernet MACs)
+- 8 ADMA (Xilinx zDMA) channels
+- 2 SD Controllers
+- OCM (256KB of On Chip Memory)
+- DDR memory
+
+QEMU does not yet model any other devices, including the PL and the AI Engine.
+
+Other differences between the hardware and the QEMU model:
+
+- QEMU allows the amount of DDR memory provided to be specified with the
+ ``-m`` argument. If a DTB is provided on the command line then QEMU will
+ edit it to include suitable entries describing the Versal DDR memory ranges.
+
+- QEMU provides 8 virtio-mmio virtio transports; these start at
+ address ``0xa0000000`` and have IRQs from 111 and upwards.
+
+Running
+"""""""
+If the user provides an Operating System to be loaded, we expect users
+to use the ``-kernel`` command line option.
+
+Users can load firmware or boot-loaders with the ``-device loader`` options.
+
+When loading an OS, QEMU generates a DTB and selects an appropriate address
+where it gets loaded. This DTB will be passed to the kernel in register x0.
+
+If there's no ``-kernel`` option, we generate a DTB and place it at 0x1000
+for boot-loaders or firmware to pick it up.
+
+If users want to provide their own DTB, they can use the ``-dtb`` option.
+These DTBs will have their memory nodes modified to match QEMU's
+selected ram_size option before they get passed to the kernel or FW.
+
+When loading an OS, we turn on QEMU's PSCI implementation with SMC
+as the PSCI conduit. When there's no ``-kernel`` option, we assume the user
+provides EL3 firmware to handle PSCI.
+
+A few examples:
+
+Direct Linux boot of a generic ARM64 upstream Linux kernel:
+
+.. code-block:: bash
+
+ $ qemu-system-aarch64 -M xlnx-versal-virt -m 2G \
+ -serial mon:stdio -display none \
+ -kernel arch/arm64/boot/Image \
+ -nic user -nic user \
+ -device virtio-rng-device,bus=virtio-mmio-bus.0 \
+ -drive if=none,index=0,file=hd0.qcow2,id=hd0,snapshot \
+ -drive file=qemu_sd.qcow2,if=sd,index=0,snapshot \
+ -device virtio-blk-device,drive=hd0 -append root=/dev/vda
+
+Direct Linux boot of PetaLinux 2019.2:
+
+.. code-block:: bash
+
+ $ qemu-system-aarch64 -M xlnx-versal-virt -m 2G \
+ -serial mon:stdio -display none \
+ -kernel petalinux-v2019.2/Image \
+ -append "rdinit=/sbin/init console=ttyAMA0,115200n8 earlycon=pl011,mmio,0xFF000000,115200n8" \
+ -net nic,model=cadence_gem,netdev=net0 -netdev user,id=net0 \
+ -device virtio-rng-device,bus=virtio-mmio-bus.0,rng=rng0 \
+ -object rng-random,filename=/dev/urandom,id=rng0
+
+Boot PetaLinux 2019.2 via ARM Trusted Firmware (2018.3 because the 2019.2
+version of ATF tries to configure the CCI which we don't model) and U-boot:
+
+.. code-block:: bash
+
+ $ qemu-system-aarch64 -M xlnx-versal-virt -m 2G \
+ -serial stdio -display none \
+ -device loader,file=petalinux-v2018.3/bl31.elf,cpu-num=0 \
+ -device loader,file=petalinux-v2019.2/u-boot.elf \
+ -device loader,addr=0x20000000,file=petalinux-v2019.2/Image \
+ -nic user -nic user \
+ -device virtio-rng-device,bus=virtio-mmio-bus.0,rng=rng0 \
+ -object rng-random,filename=/dev/urandom,id=rng0
+
+Run the following at the U-Boot prompt:
+
+.. code-block:: bash
+
+ Versal>
+ fdt addr $fdtcontroladdr
+ fdt move $fdtcontroladdr 0x40000000
+ fdt set /timer clock-frequency <0x3dfd240>
+ setenv bootargs "rdinit=/sbin/init maxcpus=1 console=ttyAMA0,115200n8 earlycon=pl011,mmio,0xFF000000,115200n8"
+ booti 20000000 - 40000000
+ fdt addr $fdtcontroladdr
+
+Boot Linux as DOM0 on Xen via U-Boot:
+
+.. code-block:: bash
+
+ $ qemu-system-aarch64 -M xlnx-versal-virt -m 4G \
+ -serial stdio -display none \
+ -device loader,file=petalinux-v2019.2/u-boot.elf,cpu-num=0 \
+ -device loader,addr=0x30000000,file=linux/2018-04-24/xen \
+ -device loader,addr=0x40000000,file=petalinux-v2019.2/Image \
+ -nic user -nic user \
+ -device virtio-rng-device,bus=virtio-mmio-bus.0,rng=rng0 \
+ -object rng-random,filename=/dev/urandom,id=rng0
+
+Run the following at the U-Boot prompt:
+
+.. code-block:: bash
+
+ Versal>
+ fdt addr $fdtcontroladdr
+ fdt move $fdtcontroladdr 0x20000000
+ fdt set /timer clock-frequency <0x3dfd240>
+ fdt set /chosen xen,xen-bootargs "console=dtuart dtuart=/uart@ff000000 dom0_mem=640M bootscrub=0 maxcpus=1 timer_slop=0"
+ fdt set /chosen xen,dom0-bootargs "rdinit=/sbin/init clk_ignore_unused console=hvc0 maxcpus=1"
+ fdt mknode /chosen dom0
+ fdt set /chosen/dom0 compatible "xen,multiboot-module"
+ fdt set /chosen/dom0 reg <0x00000000 0x40000000 0x0 0x03100000>
+ booti 30000000 - 20000000
+
+Boot Linux as Dom0 on Xen via ARM Trusted Firmware and U-Boot:
+
+.. code-block:: bash
+
+ $ qemu-system-aarch64 -M xlnx-versal-virt -m 4G \
+ -serial stdio -display none \
+ -device loader,file=petalinux-v2018.3/bl31.elf,cpu-num=0 \
+ -device loader,file=petalinux-v2019.2/u-boot.elf \
+ -device loader,addr=0x30000000,file=linux/2018-04-24/xen \
+ -device loader,addr=0x40000000,file=petalinux-v2019.2/Image \
+ -nic user -nic user \
+ -device virtio-rng-device,bus=virtio-mmio-bus.0,rng=rng0 \
+ -object rng-random,filename=/dev/urandom,id=rng0
+
+Run the following at the U-Boot prompt:
+
+.. code-block:: bash
+
+ Versal>
+ fdt addr $fdtcontroladdr
+ fdt move $fdtcontroladdr 0x20000000
+ fdt set /timer clock-frequency <0x3dfd240>
+ fdt set /chosen xen,xen-bootargs "console=dtuart dtuart=/uart@ff000000 dom0_mem=640M bootscrub=0 maxcpus=1 timer_slop=0"
+ fdt set /chosen xen,dom0-bootargs "rdinit=/sbin/init clk_ignore_unused console=hvc0 maxcpus=1"
+ fdt mknode /chosen dom0
+ fdt set /chosen/dom0 compatible "xen,multiboot-module"
+ fdt set /chosen/dom0 reg <0x00000000 0x40000000 0x0 0x03100000>
+ booti 30000000 - 20000000
+
diff --git a/docs/system/target-arm.rst b/docs/system/target-arm.rst
index 4c5b0e4aab8..afdb37e7384 100644
--- a/docs/system/target-arm.rst
+++ b/docs/system/target-arm.rst
@@ -93,6 +93,7 @@ undocumented; you can get a complete list by running
arm/sx1
arm/stellaris
arm/virt
+ arm/xlnx-versal-virt
Arm CPU features
================
diff --git a/MAINTAINERS b/MAINTAINERS
index 0886eb3d2b5..5a22c8be429 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -894,7 +894,7 @@ F: hw/misc/zynq*
F: include/hw/misc/zynq*
X: hw/ssi/xilinx_*
-Xilinx ZynqMP
+Xilinx ZynqMP and Versal
M: Alistair Francis <alistair@alistair23.me>
M: Edgar E. Iglesias <edgar.iglesias@gmail.com>
M: Peter Maydell <peter.maydell@linaro.org>
@@ -905,6 +905,7 @@ F: include/hw/*/xlnx*.h
F: include/hw/ssi/xilinx_spips.h
F: hw/display/dpcd.c
F: include/hw/display/dpcd.h
+F: docs/system/arm/xlnx-versal-virt.rst
ARM ACPI Subsystem
M: Shannon Zhao <shannon.zhaosl@gmail.com>
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 14/27] target/arm: Pull handling of XScale insns out of disas_coproc_insn()
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (12 preceding siblings ...)
2020-08-24 9:47 ` [PULL 13/27] docs/system/arm: Document the Xilinx Versal Virt board Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:47 ` [PULL 15/27] target/arm: Separate decode from handling of coproc insns Peter Maydell
` (13 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
At the moment we check for XScale/iwMMXt insns inside
disas_coproc_insn(): for CPUs with ARM_FEATURE_XSCALE all copro insns
with cp 0 or 1 are handled specially. This works, but is an odd
place for this check, because disas_coproc_insn() is called from both
the Arm and Thumb decoders but the XScale case never applies for
Thumb (all the XScale CPUs were ARMv5, which has only Thumb1, not
Thumb2 with the 32-bit coprocessor insn encodings). It also makes it
awkward to convert the real copro access insns to decodetree.
Move the identification of XScale out to its own function
which is only called from disas_arm_insn().
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-2-peter.maydell@linaro.org
---
target/arm/translate.c | 44 ++++++++++++++++++++++++++++--------------
1 file changed, 29 insertions(+), 15 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 556588d92fe..99663236aa9 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -4551,20 +4551,6 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
cpnum = (insn >> 8) & 0xf;
- /* First check for coprocessor space used for XScale/iwMMXt insns */
- if (arm_dc_feature(s, ARM_FEATURE_XSCALE) && (cpnum < 2)) {
- if (extract32(s->c15_cpar, cpnum, 1) == 0) {
- return 1;
- }
- if (arm_dc_feature(s, ARM_FEATURE_IWMMXT)) {
- return disas_iwmmxt_insn(s, insn);
- } else if (arm_dc_feature(s, ARM_FEATURE_XSCALE)) {
- return disas_dsp_insn(s, insn);
- }
- return 1;
- }
-
- /* Otherwise treat as a generic register access */
is64 = (insn & (1 << 25)) == 0;
if (!is64 && ((insn & (1 << 4)) == 0)) {
/* cdp */
@@ -4823,6 +4809,23 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
return 1;
}
+/* Decode XScale DSP or iWMMXt insn (in the copro space, cp=0 or 1) */
+static void disas_xscale_insn(DisasContext *s, uint32_t insn)
+{
+ int cpnum = (insn >> 8) & 0xf;
+
+ if (extract32(s->c15_cpar, cpnum, 1) == 0) {
+ unallocated_encoding(s);
+ } else if (arm_dc_feature(s, ARM_FEATURE_IWMMXT)) {
+ if (disas_iwmmxt_insn(s, insn)) {
+ unallocated_encoding(s);
+ }
+ } else if (arm_dc_feature(s, ARM_FEATURE_XSCALE)) {
+ if (disas_dsp_insn(s, insn)) {
+ unallocated_encoding(s);
+ }
+ }
+}
/* Store a 64-bit value to a register pair. Clobbers val. */
static void gen_storeq_reg(DisasContext *s, int rlow, int rhigh, TCGv_i64 val)
@@ -8270,15 +8273,26 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
case 0xc:
case 0xd:
case 0xe:
- if (((insn >> 8) & 0xe) == 10) {
+ {
+ /* First check for coprocessor space used for XScale/iwMMXt insns */
+ int cpnum = (insn >> 8) & 0xf;
+
+ if (arm_dc_feature(s, ARM_FEATURE_XSCALE) && (cpnum < 2)) {
+ disas_xscale_insn(s, insn);
+ break;
+ }
+
+ if ((cpnum & 0xe) == 10) {
/* VFP, but failed disas_vfp. */
goto illegal_op;
}
+
if (disas_coproc_insn(s, insn)) {
/* Coprocessor. */
goto illegal_op;
}
break;
+ }
default:
illegal_op:
unallocated_encoding(s);
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 15/27] target/arm: Separate decode from handling of coproc insns
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (13 preceding siblings ...)
2020-08-24 9:47 ` [PULL 14/27] target/arm: Pull handling of XScale insns out of disas_coproc_insn() Peter Maydell
@ 2020-08-24 9:47 ` Peter Maydell
2020-08-24 9:48 ` [PULL 16/27] target/arm: Convert A32 coprocessor insns to decodetree Peter Maydell
` (12 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:47 UTC (permalink / raw)
To: qemu-devel
As a prelude to making coproc insns use decodetree, split out the
part of disas_coproc_insn() which does instruction decoding from the
part which does the actual work, and make do_coproc_insn() handle the
UNDEF-on-bad-permissions and similar cases itself rather than
returning 1 to eventually percolate up to a callsite that calls
unallocated_encoding() for it.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-3-peter.maydell@linaro.org
---
target/arm/translate.c | 76 ++++++++++++++++++++++++------------------
1 file changed, 44 insertions(+), 32 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 99663236aa9..01591f0e325 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -4544,34 +4544,12 @@ void gen_gvec_uaba(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]);
}
-static int disas_coproc_insn(DisasContext *s, uint32_t insn)
+static void do_coproc_insn(DisasContext *s, int cpnum, int is64,
+ int opc1, int crn, int crm, int opc2,
+ bool isread, int rt, int rt2)
{
- int cpnum, is64, crn, crm, opc1, opc2, isread, rt, rt2;
const ARMCPRegInfo *ri;
- cpnum = (insn >> 8) & 0xf;
-
- is64 = (insn & (1 << 25)) == 0;
- if (!is64 && ((insn & (1 << 4)) == 0)) {
- /* cdp */
- return 1;
- }
-
- crm = insn & 0xf;
- if (is64) {
- crn = 0;
- opc1 = (insn >> 4) & 0xf;
- opc2 = 0;
- rt2 = (insn >> 16) & 0xf;
- } else {
- crn = (insn >> 16) & 0xf;
- opc1 = (insn >> 21) & 7;
- opc2 = (insn >> 5) & 7;
- rt2 = 0;
- }
- isread = (insn >> 20) & 1;
- rt = (insn >> 12) & 0xf;
-
ri = get_arm_cp_reginfo(s->cp_regs,
ENCODE_CP_REG(cpnum, is64, s->ns, crn, crm, opc1, opc2));
if (ri) {
@@ -4579,7 +4557,8 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
/* Check access permissions */
if (!cp_access_ok(s->current_el, ri, isread)) {
- return 1;
+ unallocated_encoding(s);
+ return;
}
if (s->hstr_active || ri->accessfn ||
@@ -4653,14 +4632,15 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
/* Handle special cases first */
switch (ri->type & ~(ARM_CP_FLAG_MASK & ~ARM_CP_SPECIAL)) {
case ARM_CP_NOP:
- return 0;
+ return;
case ARM_CP_WFI:
if (isread) {
- return 1;
+ unallocated_encoding(s);
+ return;
}
gen_set_pc_im(s, s->base.pc_next);
s->base.is_jmp = DISAS_WFI;
- return 0;
+ return;
default:
break;
}
@@ -4720,7 +4700,7 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
/* Write */
if (ri->type & ARM_CP_CONST) {
/* If not forbidden by access permissions, treat as WI */
- return 0;
+ return;
}
if (is64) {
@@ -4786,7 +4766,7 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
gen_lookup_tb(s);
}
- return 0;
+ return;
}
/* Unknown register; this might be a guest error or a QEMU
@@ -4806,7 +4786,39 @@ static int disas_coproc_insn(DisasContext *s, uint32_t insn)
s->ns ? "non-secure" : "secure");
}
- return 1;
+ unallocated_encoding(s);
+ return;
+}
+
+static int disas_coproc_insn(DisasContext *s, uint32_t insn)
+{
+ int cpnum, is64, crn, crm, opc1, opc2, isread, rt, rt2;
+
+ cpnum = (insn >> 8) & 0xf;
+
+ is64 = (insn & (1 << 25)) == 0;
+ if (!is64 && ((insn & (1 << 4)) == 0)) {
+ /* cdp */
+ return 1;
+ }
+
+ crm = insn & 0xf;
+ if (is64) {
+ crn = 0;
+ opc1 = (insn >> 4) & 0xf;
+ opc2 = 0;
+ rt2 = (insn >> 16) & 0xf;
+ } else {
+ crn = (insn >> 16) & 0xf;
+ opc1 = (insn >> 21) & 7;
+ opc2 = (insn >> 5) & 7;
+ rt2 = 0;
+ }
+ isread = (insn >> 20) & 1;
+ rt = (insn >> 12) & 0xf;
+
+ do_coproc_insn(s, cpnum, is64, opc1, crn, crm, opc2, isread, rt, rt2);
+ return 0;
}
/* Decode XScale DSP or iWMMXt insn (in the copro space, cp=0 or 1) */
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 16/27] target/arm: Convert A32 coprocessor insns to decodetree
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (14 preceding siblings ...)
2020-08-24 9:47 ` [PULL 15/27] target/arm: Separate decode from handling of coproc insns Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 17/27] target/arm: Tidy up disas_arm_insn() Peter Maydell
` (11 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
Convert the A32 coprocessor instructions to decodetree.
Note that this corrects an underdecoding: for the 64-bit access case
(MRRC/MCRR) we did not check that bits [24:21] were 0b0010, so we
would incorrectly treat LDC/STC as MRRC/MCRR rather than UNDEFing
them.
The decodetree versions of these insns assume the coprocessor
is in the range 0..7 or 14..15. This is architecturally sensible
(as per the comments) and OK in practice for QEMU because the only
uses of the ARMCPRegInfo infrastructure we have that aren't
for coprocessors 14 or 15 are the pxa2xx use of coprocessor 6.
We add an assertion to the define_one_arm_cp_reg_with_opaque()
function to catch any accidental future attempts to use it to
define coprocessor registers for invalid coprocessors.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-4-peter.maydell@linaro.org
---
target/arm/a32.decode | 19 +++++++++++
target/arm/helper.c | 29 +++++++++++++++++
target/arm/translate.c | 74 +++++++++++++++++++++++++++++++++++-------
3 files changed, 111 insertions(+), 11 deletions(-)
diff --git a/target/arm/a32.decode b/target/arm/a32.decode
index 0bd952c0692..4dfd9139bf3 100644
--- a/target/arm/a32.decode
+++ b/target/arm/a32.decode
@@ -47,6 +47,8 @@
&bfi rd rn lsb msb
&sat rd rn satimm imm sh
&pkh rd rn rm imm tb
+&mcr cp opc1 crn crm opc2 rt
+&mcrr cp opc1 crm rt rt2
# Data-processing (register)
@@ -529,6 +531,23 @@ LDM_a32 ---- 100 b:1 i:1 u:1 w:1 1 rn:4 list:16 &ldst_block
B .... 1010 ........................ @branch
BL .... 1011 ........................ @branch
+# Coprocessor instructions
+
+# We decode MCR, MCR, MRRC and MCRR only, because for QEMU the
+# other coprocessor instructions always UNDEF.
+# The trans_ functions for these will ignore cp values 8..13 for v7 or
+# earlier, and 0..13 for v8 and later, because those areas of the
+# encoding space may be used for other things, such as VFP or Neon.
+
+@mcr ---- .... opc1:3 . crn:4 rt:4 cp:4 opc2:3 . crm:4 &mcr
+@mcrr ---- .... .... rt2:4 rt:4 cp:4 opc1:4 crm:4 &mcrr
+
+MCRR .... 1100 0100 .... .... .... .... .... @mcrr
+MRRC .... 1100 0101 .... .... .... .... .... @mcrr
+
+MCR .... 1110 ... 0 .... .... .... ... 1 .... @mcr
+MRC .... 1110 ... 1 .... .... .... ... 1 .... @mcr
+
# Supervisor call
SVC ---- 1111 imm:24 &i
diff --git a/target/arm/helper.c b/target/arm/helper.c
index 455c92b8915..6b4f0eb5334 100644
--- a/target/arm/helper.c
+++ b/target/arm/helper.c
@@ -8462,6 +8462,35 @@ void define_one_arm_cp_reg_with_opaque(ARMCPU *cpu,
assert((r->state != ARM_CP_STATE_AA32) || (r->opc0 == 0));
/* AArch64 regs are all 64 bit so ARM_CP_64BIT is meaningless */
assert((r->state != ARM_CP_STATE_AA64) || !(r->type & ARM_CP_64BIT));
+ /*
+ * This API is only for Arm's system coprocessors (14 and 15) or
+ * (M-profile or v7A-and-earlier only) for implementation defined
+ * coprocessors in the range 0..7. Our decode assumes this, since
+ * 8..13 can be used for other insns including VFP and Neon. See
+ * valid_cp() in translate.c. Assert here that we haven't tried
+ * to use an invalid coprocessor number.
+ */
+ switch (r->state) {
+ case ARM_CP_STATE_BOTH:
+ /* 0 has a special meaning, but otherwise the same rules as AA32. */
+ if (r->cp == 0) {
+ break;
+ }
+ /* fall through */
+ case ARM_CP_STATE_AA32:
+ if (arm_feature(&cpu->env, ARM_FEATURE_V8) &&
+ !arm_feature(&cpu->env, ARM_FEATURE_M)) {
+ assert(r->cp >= 14 && r->cp <= 15);
+ } else {
+ assert(r->cp < 8 || (r->cp >= 14 && r->cp <= 15));
+ }
+ break;
+ case ARM_CP_STATE_AA64:
+ assert(r->cp == 0 || r->cp == CP_REG_ARM64_SYSREG_CP);
+ break;
+ default:
+ g_assert_not_reached();
+ }
/* The AArch64 pseudocode CheckSystemAccess() specifies that op1
* encodes a minimum access level for the register. We roll this
* runtime check into our general permission check code, so check
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 01591f0e325..6ee920eec53 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -5237,6 +5237,68 @@ static int t16_pop_list(DisasContext *s, int x)
#include "decode-t32.c.inc"
#include "decode-t16.c.inc"
+static bool valid_cp(DisasContext *s, int cp)
+{
+ /*
+ * Return true if this coprocessor field indicates something
+ * that's really a possible coprocessor.
+ * For v7 and earlier, coprocessors 8..15 were reserved for Arm use,
+ * and of those only cp14 and cp15 were used for registers.
+ * cp10 and cp11 were used for VFP and Neon, whose decode is
+ * dealt with elsewhere. With the advent of fp16, cp9 is also
+ * now part of VFP.
+ * For v8A and later, the encoding has been tightened so that
+ * only cp14 and cp15 are valid, and other values aren't considered
+ * to be in the coprocessor-instruction space at all. v8M still
+ * permits coprocessors 0..7.
+ */
+ if (arm_dc_feature(s, ARM_FEATURE_V8) &&
+ !arm_dc_feature(s, ARM_FEATURE_M)) {
+ return cp >= 14;
+ }
+ return cp < 8 || cp >= 14;
+}
+
+static bool trans_MCR(DisasContext *s, arg_MCR *a)
+{
+ if (!valid_cp(s, a->cp)) {
+ return false;
+ }
+ do_coproc_insn(s, a->cp, false, a->opc1, a->crn, a->crm, a->opc2,
+ false, a->rt, 0);
+ return true;
+}
+
+static bool trans_MRC(DisasContext *s, arg_MRC *a)
+{
+ if (!valid_cp(s, a->cp)) {
+ return false;
+ }
+ do_coproc_insn(s, a->cp, false, a->opc1, a->crn, a->crm, a->opc2,
+ true, a->rt, 0);
+ return true;
+}
+
+static bool trans_MCRR(DisasContext *s, arg_MCRR *a)
+{
+ if (!valid_cp(s, a->cp)) {
+ return false;
+ }
+ do_coproc_insn(s, a->cp, true, a->opc1, 0, a->crm, 0,
+ false, a->rt, a->rt2);
+ return true;
+}
+
+static bool trans_MRRC(DisasContext *s, arg_MRRC *a)
+{
+ if (!valid_cp(s, a->cp)) {
+ return false;
+ }
+ do_coproc_insn(s, a->cp, true, a->opc1, 0, a->crm, 0,
+ true, a->rt, a->rt2);
+ return true;
+}
+
/* Helpers to swap operands for reverse-subtract. */
static void gen_rsb(TCGv_i32 dst, TCGv_i32 a, TCGv_i32 b)
{
@@ -8293,17 +8355,7 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
disas_xscale_insn(s, insn);
break;
}
-
- if ((cpnum & 0xe) == 10) {
- /* VFP, but failed disas_vfp. */
- goto illegal_op;
- }
-
- if (disas_coproc_insn(s, insn)) {
- /* Coprocessor. */
- goto illegal_op;
- }
- break;
+ /* fall through */
}
default:
illegal_op:
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 17/27] target/arm: Tidy up disas_arm_insn()
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (15 preceding siblings ...)
2020-08-24 9:48 ` [PULL 16/27] target/arm: Convert A32 coprocessor insns to decodetree Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 18/27] target/arm: Do M-profile NOCP checks early and via decodetree Peter Maydell
` (10 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
The only thing left in the "legacy decoder" is the handling
of disas_xscale_insn(), and we can simplify the code.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-5-peter.maydell@linaro.org
---
target/arm/translate.c | 26 +++++++++-----------------
1 file changed, 9 insertions(+), 17 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 6ee920eec53..362d1cc50fb 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -8342,26 +8342,18 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
return;
}
/* fall back to legacy decoder */
-
- switch ((insn >> 24) & 0xf) {
- case 0xc:
- case 0xd:
- case 0xe:
- {
- /* First check for coprocessor space used for XScale/iwMMXt insns */
- int cpnum = (insn >> 8) & 0xf;
-
- if (arm_dc_feature(s, ARM_FEATURE_XSCALE) && (cpnum < 2)) {
+ /* TODO: convert xscale/iwmmxt decoder to decodetree ?? */
+ if (arm_dc_feature(s, ARM_FEATURE_XSCALE)) {
+ if (((insn & 0x0c000e00) == 0x0c000000)
+ && ((insn & 0x03000000) != 0x03000000)) {
+ /* Coprocessor insn, coprocessor 0 or 1 */
disas_xscale_insn(s, insn);
- break;
+ return;
}
- /* fall through */
- }
- default:
- illegal_op:
- unallocated_encoding(s);
- break;
}
+
+illegal_op:
+ unallocated_encoding(s);
}
static bool thumb_insn_is_16bit(DisasContext *s, uint32_t pc, uint32_t insn)
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 18/27] target/arm: Do M-profile NOCP checks early and via decodetree
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (16 preceding siblings ...)
2020-08-24 9:48 ` [PULL 17/27] target/arm: Tidy up disas_arm_insn() Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 19/27] target/arm: Convert T32 coprocessor insns to decodetree Peter Maydell
` (9 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
For M-profile CPUs, the architecture specifies that the NOCP
exception when a coprocessor is not present or disabled should cover
the entire wide range of coprocessor-space encodings, and should take
precedence over UNDEF exceptions. (This is the opposite of
A-profile, where checking for a disabled FPU has to happen last.)
Implement this with decodetree patterns that cover the specified
ranges of the encoding space. There are a few instructions (VLLDM,
VLSTM, and in v8.1 also VSCCLRM) which are in copro-space but must
not be NOCP'd: these must be handled also in the new m-nocp.decode so
they take precedence.
This is a minor behaviour change: for unallocated insn patterns in
the VFP area (cp=10,11) we will now NOCP rather than UNDEF when the
FPU is disabled.
As well as giving us the correct architectural behaviour for v8.1M
and the recommended behaviour for v8.0M, this refactoring also
removes the old NOCP handling from the remains of the 'legacy
decoder' in disas_thumb2_insn(), paving the way for cleaning that up.
Since we don't currently have a v8.1M feature bit or any v8.1M CPUs,
the minor changes to this logic that we'll need for v8.1M are marked
up with TODO comments.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-6-peter.maydell@linaro.org
---
target/arm/m-nocp.decode | 42 +++++++++++++++++++++++++++
target/arm/vfp.decode | 2 --
target/arm/translate.c | 30 ++++++++++----------
target/arm/meson.build | 1 +
target/arm/translate-vfp.c.inc | 52 +++++++++++++++++++++++++++-------
5 files changed, 100 insertions(+), 27 deletions(-)
create mode 100644 target/arm/m-nocp.decode
diff --git a/target/arm/m-nocp.decode b/target/arm/m-nocp.decode
new file mode 100644
index 00000000000..7182d7d1217
--- /dev/null
+++ b/target/arm/m-nocp.decode
@@ -0,0 +1,42 @@
+# M-profile UserFault.NOCP exception handling
+#
+# Copyright (c) 2020 Linaro, Ltd
+#
+# This library is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public
+# License as published by the Free Software Foundation; either
+# version 2.1 of the License, or (at your option) any later version.
+#
+# This library is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+# Lesser General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public
+# License along with this library; if not, see <http://www.gnu.org/licenses/>.
+
+#
+# This file is processed by scripts/decodetree.py
+#
+# For M-profile, the architecture specifies that NOCP UsageFaults
+# should take precedence over UNDEF faults over the whole wide
+# range of coprocessor-space encodings, with the exception of
+# VLLDM and VLSTM. (Compare v8.1M IsCPInstruction() pseudocode and
+# v8M Arm ARM rule R_QLGM.) This isn't mandatory for v8.0M but we choose
+# to behave the same as v8.1M.
+# This decode is handled before any others (and in particular before
+# decoding FP instructions which are in the coprocessor space).
+# If the coprocessor is not present or disabled then we will generate
+# the NOCP exception; otherwise we let the insn through to the main decode.
+
+{
+ # Special cases which do not take an early NOCP: VLLDM and VLSTM
+ VLLDM_VLSTM 1110 1100 001 l:1 rn:4 0000 1010 0000 0000
+ # TODO: VSCCLRM (new in v8.1M) is similar:
+ #VSCCLRM 1110 1100 1-01 1111 ---- 1011 ---- ---0
+
+ NOCP 111- 1110 ---- ---- ---- cp:4 ---- ----
+ NOCP 111- 110- ---- ---- ---- cp:4 ---- ----
+ # TODO: From v8.1M onwards we will also want this range to NOCP
+ #NOCP_8_1 111- 1111 ---- ---- ---- ---- ---- ---- cp=10
+}
diff --git a/target/arm/vfp.decode b/target/arm/vfp.decode
index 5fd70f975ae..2c793e3e87f 100644
--- a/target/arm/vfp.decode
+++ b/target/arm/vfp.decode
@@ -213,5 +213,3 @@ VCVT_sp_int ---- 1110 1.11 110 s:1 .... 1010 rz:1 1.0 .... \
vd=%vd_sp vm=%vm_sp
VCVT_dp_int ---- 1110 1.11 110 s:1 .... 1011 rz:1 1.0 .... \
vd=%vd_sp vm=%vm_dp
-
-VLLDM_VLSTM 1110 1100 001 l:1 rn:4 0000 1010 0000 0000
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 362d1cc50fb..958e9b6699f 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -1176,6 +1176,7 @@ static TCGv_ptr vfp_reg_ptr(bool dp, int reg)
#define ARM_CP_RW_BIT (1 << 20)
/* Include the VFP and Neon decoders */
+#include "decode-m-nocp.c.inc"
#include "translate-vfp.c.inc"
#include "translate-neon.c.inc"
@@ -8433,6 +8434,19 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
ARCH(6T2);
}
+ if (arm_dc_feature(s, ARM_FEATURE_M)) {
+ /*
+ * NOCP takes precedence over any UNDEF for (almost) the
+ * entire wide range of coprocessor-space encodings, so check
+ * for it first before proceeding to actually decode eg VFP
+ * insns. This decode also handles the few insns which are
+ * in copro space but do not have NOCP checks (eg VLLDM, VLSTM).
+ */
+ if (disas_m_nocp(s, insn)) {
+ return;
+ }
+ }
+
if ((insn & 0xef000000) == 0xef000000) {
/*
* T32 encodings 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq
@@ -8481,21 +8495,7 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
/* Coprocessor. */
if (arm_dc_feature(s, ARM_FEATURE_M)) {
/* 0b111x_11xx_xxxx_xxxx_xxxx_xxxx_xxxx_xxxx */
- if (extract32(insn, 24, 2) == 3) {
- goto illegal_op; /* op0 = 0b11 : unallocated */
- }
-
- if (((insn >> 8) & 0xe) == 10 &&
- dc_isar_feature(aa32_fpsp_v2, s)) {
- /* FP, and the CPU supports it */
- goto illegal_op;
- } else {
- /* All other insns: NOCP */
- gen_exception_insn(s, s->pc_curr, EXCP_NOCP,
- syn_uncategorized(),
- default_exception_el(s));
- }
- break;
+ goto illegal_op;
}
if (((insn >> 24) & 3) == 3) {
/* Neon DP, but failed disas_neon_dp() */
diff --git a/target/arm/meson.build b/target/arm/meson.build
index bd46cdb5239..89900907125 100644
--- a/target/arm/meson.build
+++ b/target/arm/meson.build
@@ -5,6 +5,7 @@ gen = [
decodetree.process('neon-ls.decode', extra_args: '--static-decode=disas_neon_ls'),
decodetree.process('vfp.decode', extra_args: '--static-decode=disas_vfp'),
decodetree.process('vfp-uncond.decode', extra_args: '--static-decode=disas_vfp_uncond'),
+ decodetree.process('m-nocp.decode', extra_args: '--static-decode=disas_m_nocp'),
decodetree.process('a32.decode', extra_args: '--static-decode=disas_a32'),
decodetree.process('a32-uncond.decode', extra_args: '--static-decode=disas_a32_uncond'),
decodetree.process('t32.decode', extra_args: '--static-decode=disas_t32'),
diff --git a/target/arm/translate-vfp.c.inc b/target/arm/translate-vfp.c.inc
index 2d63fa0d399..d376bd1c1ad 100644
--- a/target/arm/translate-vfp.c.inc
+++ b/target/arm/translate-vfp.c.inc
@@ -95,14 +95,11 @@ static inline long vfp_f16_offset(unsigned reg, bool top)
static bool full_vfp_access_check(DisasContext *s, bool ignore_vfp_enabled)
{
if (s->fp_excp_el) {
- if (arm_dc_feature(s, ARM_FEATURE_M)) {
- gen_exception_insn(s, s->pc_curr, EXCP_NOCP, syn_uncategorized(),
- s->fp_excp_el);
- } else {
- gen_exception_insn(s, s->pc_curr, EXCP_UDEF,
- syn_fp_access_trap(1, 0xe, false),
- s->fp_excp_el);
- }
+ /* M-profile handled this earlier, in disas_m_nocp() */
+ assert (!arm_dc_feature(s, ARM_FEATURE_M));
+ gen_exception_insn(s, s->pc_curr, EXCP_UDEF,
+ syn_fp_access_trap(1, 0xe, false),
+ s->fp_excp_el);
return false;
}
@@ -2842,9 +2839,14 @@ static bool trans_VLLDM_VLSTM(DisasContext *s, arg_VLLDM_VLSTM *a)
!arm_dc_feature(s, ARM_FEATURE_V8)) {
return false;
}
- /* If not secure, UNDEF. */
+ /*
+ * If not secure, UNDEF. We must emit code for this
+ * rather than returning false so that this takes
+ * precedence over the m-nocp.decode NOCP fallback.
+ */
if (!s->v8m_secure) {
- return false;
+ unallocated_encoding(s);
+ return true;
}
/* If no fpu, NOP. */
if (!dc_isar_feature(aa32_vfp, s)) {
@@ -2863,3 +2865,33 @@ static bool trans_VLLDM_VLSTM(DisasContext *s, arg_VLLDM_VLSTM *a)
s->base.is_jmp = DISAS_UPDATE_EXIT;
return true;
}
+
+static bool trans_NOCP(DisasContext *s, arg_NOCP *a)
+{
+ /*
+ * Handle M-profile early check for disabled coprocessor:
+ * all we need to do here is emit the NOCP exception if
+ * the coprocessor is disabled. Otherwise we return false
+ * and the real VFP/etc decode will handle the insn.
+ */
+ assert(arm_dc_feature(s, ARM_FEATURE_M));
+
+ if (a->cp == 11) {
+ a->cp = 10;
+ }
+ /* TODO: in v8.1M cp 8, 9, 14, 15 also are governed by the cp10 enable */
+
+ if (a->cp != 10) {
+ gen_exception_insn(s, s->pc_curr, EXCP_NOCP,
+ syn_uncategorized(), default_exception_el(s));
+ return true;
+ }
+
+ if (s->fp_excp_el != 0) {
+ gen_exception_insn(s, s->pc_curr, EXCP_NOCP,
+ syn_uncategorized(), s->fp_excp_el);
+ return true;
+ }
+
+ return false;
+}
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 19/27] target/arm: Convert T32 coprocessor insns to decodetree
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (17 preceding siblings ...)
2020-08-24 9:48 ` [PULL 18/27] target/arm: Do M-profile NOCP checks early and via decodetree Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 20/27] target/arm: Remove ARCH macro Peter Maydell
` (8 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
Convert the T32 coprocessor instructions to decodetree.
As with the A32 conversion, this corrects an underdecoding
where we did not check that MRRC/MCRR [24:21] were 0b0010
and so treated some kinds of LDC/STC and MRRC/MCRR rather
than UNDEFing them.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-7-peter.maydell@linaro.org
---
target/arm/t32.decode | 19 +++++++++++++
target/arm/translate.c | 64 ++----------------------------------------
2 files changed, 21 insertions(+), 62 deletions(-)
diff --git a/target/arm/t32.decode b/target/arm/t32.decode
index c21a988f971..7069d821fde 100644
--- a/target/arm/t32.decode
+++ b/target/arm/t32.decode
@@ -45,6 +45,8 @@
&sat !extern rd rn satimm imm sh
&pkh !extern rd rn rm imm tb
&cps !extern mode imod M A I F
+&mcr !extern cp opc1 crn crm opc2 rt
+&mcrr !extern cp opc1 crm rt rt2
# Data-processing (register)
@@ -621,6 +623,23 @@ RFE 1110 1001 10.1 .... 1100000000000000 @rfe pu=1
SRS 1110 1000 00.0 1101 1100 0000 000. .... @srs pu=2
SRS 1110 1001 10.0 1101 1100 0000 000. .... @srs pu=1
+# Coprocessor instructions
+
+# We decode MCR, MCR, MRRC and MCRR only, because for QEMU the
+# other coprocessor instructions always UNDEF.
+# The trans_ functions for these will ignore cp values 8..13 for v7 or
+# earlier, and 0..13 for v8 and later, because those areas of the
+# encoding space may be used for other things, such as VFP or Neon.
+
+@mcr .... .... opc1:3 . crn:4 rt:4 cp:4 opc2:3 . crm:4
+@mcrr .... .... .... rt2:4 rt:4 cp:4 opc1:4 crm:4
+
+MCRR 1110 1100 0100 .... .... .... .... .... @mcrr
+MRRC 1110 1100 0101 .... .... .... .... .... @mcrr
+
+MCR 1110 1110 ... 0 .... .... .... ... 1 .... @mcr
+MRC 1110 1110 ... 1 .... .... .... ... 1 .... @mcr
+
# Branches
%imm24 26:s1 13:1 11:1 16:10 0:11 !function=t32_branch24
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 958e9b6699f..e2ae4f79445 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -4791,37 +4791,6 @@ static void do_coproc_insn(DisasContext *s, int cpnum, int is64,
return;
}
-static int disas_coproc_insn(DisasContext *s, uint32_t insn)
-{
- int cpnum, is64, crn, crm, opc1, opc2, isread, rt, rt2;
-
- cpnum = (insn >> 8) & 0xf;
-
- is64 = (insn & (1 << 25)) == 0;
- if (!is64 && ((insn & (1 << 4)) == 0)) {
- /* cdp */
- return 1;
- }
-
- crm = insn & 0xf;
- if (is64) {
- crn = 0;
- opc1 = (insn >> 4) & 0xf;
- opc2 = 0;
- rt2 = (insn >> 16) & 0xf;
- } else {
- crn = (insn >> 16) & 0xf;
- opc1 = (insn >> 21) & 7;
- opc2 = (insn >> 5) & 7;
- rt2 = 0;
- }
- isread = (insn >> 20) & 1;
- rt = (insn >> 12) & 0xf;
-
- do_coproc_insn(s, cpnum, is64, opc1, crn, crm, opc2, isread, rt, rt2);
- return 0;
-}
-
/* Decode XScale DSP or iWMMXt insn (in the copro space, cp=0 or 1) */
static void disas_xscale_insn(DisasContext *s, uint32_t insn)
{
@@ -8485,38 +8454,9 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
((insn >> 28) == 0xe && disas_vfp(s, insn))) {
return;
}
- /* fall back to legacy decoder */
- switch ((insn >> 25) & 0xf) {
- case 0: case 1: case 2: case 3:
- /* 16-bit instructions. Should never happen. */
- abort();
- case 6: case 7: case 14: case 15:
- /* Coprocessor. */
- if (arm_dc_feature(s, ARM_FEATURE_M)) {
- /* 0b111x_11xx_xxxx_xxxx_xxxx_xxxx_xxxx_xxxx */
- goto illegal_op;
- }
- if (((insn >> 24) & 3) == 3) {
- /* Neon DP, but failed disas_neon_dp() */
- goto illegal_op;
- } else if (((insn >> 8) & 0xe) == 10) {
- /* VFP, but failed disas_vfp. */
- goto illegal_op;
- } else {
- if (insn & (1 << 28))
- goto illegal_op;
- if (disas_coproc_insn(s, insn)) {
- goto illegal_op;
- }
- }
- break;
- case 12:
- goto illegal_op;
- default:
- illegal_op:
- unallocated_encoding(s);
- }
+illegal_op:
+ unallocated_encoding(s);
}
static void disas_thumb_insn(DisasContext *s, uint32_t insn)
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 20/27] target/arm: Remove ARCH macro
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (18 preceding siblings ...)
2020-08-24 9:48 ` [PULL 19/27] target/arm: Convert T32 coprocessor insns to decodetree Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 21/27] target/arm: Delete unused VFP_DREG macros Peter Maydell
` (7 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
The ARCH() macro was used a lot in the legacy decoder, but
there are now just two uses of it left. Since a macro which
expands out to a goto is liable to be confusing when reading
code, replace the last two uses with a simple open-coded
qeuivalent.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803111849.13368-8-peter.maydell@linaro.org
---
target/arm/translate.c | 14 +++++++++-----
1 file changed, 9 insertions(+), 5 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index e2ae4f79445..b74a4350e39 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -49,8 +49,6 @@
#define ENABLE_ARCH_7 arm_dc_feature(s, ARM_FEATURE_V7)
#define ENABLE_ARCH_8 arm_dc_feature(s, ARM_FEATURE_V8)
-#define ARCH(x) do { if (!ENABLE_ARCH_##x) goto illegal_op; } while(0)
-
#include "translate.h"
#if defined(CONFIG_USER_ONLY)
@@ -7909,7 +7907,7 @@ static bool trans_BLX_i(DisasContext *s, arg_BLX_i *a)
{
TCGv_i32 tmp;
- /* For A32, ARCH(5) is checked near the start of the uncond block. */
+ /* For A32, ARM_FEATURE_V5 is checked near the start of the uncond block. */
if (s->thumb && (a->imm & 2)) {
return false;
}
@@ -8275,7 +8273,10 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
* choose to UNDEF. In ARMv5 and above the space is used
* for miscellaneous unconditional instructions.
*/
- ARCH(5);
+ if (!arm_dc_feature(s, ARM_FEATURE_V5)) {
+ unallocated_encoding(s);
+ return;
+ }
/* Unconditional instructions. */
/* TODO: Perhaps merge these into one decodetree output file. */
@@ -8400,7 +8401,10 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
goto illegal_op;
}
} else if ((insn & 0xf800e800) != 0xf000e800) {
- ARCH(6T2);
+ if (!arm_dc_feature(s, ARM_FEATURE_THUMB2)) {
+ unallocated_encoding(s);
+ return;
+ }
}
if (arm_dc_feature(s, ARM_FEATURE_M)) {
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 21/27] target/arm: Delete unused VFP_DREG macros
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (19 preceding siblings ...)
2020-08-24 9:48 ` [PULL 20/27] target/arm: Remove ARCH macro Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 22/27] target/arm/translate.c: Delete/amend incorrect comments Peter Maydell
` (6 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
As part of the Neon decodetree conversion we removed all
the uses of the VFP_DREG macros, but forgot to remove the
macro definitions. Do so now.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200803124848.18295-1-peter.maydell@linaro.org
---
target/arm/translate.c | 15 ---------------
1 file changed, 15 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index b74a4350e39..4ffd8b1fbe5 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -2470,21 +2470,6 @@ static int disas_dsp_insn(DisasContext *s, uint32_t insn)
return 1;
}
-#define VFP_REG_SHR(x, n) (((n) > 0) ? (x) >> (n) : (x) << -(n))
-#define VFP_DREG(reg, insn, bigbit, smallbit) do { \
- if (dc_isar_feature(aa32_simd_r32, s)) { \
- reg = (((insn) >> (bigbit)) & 0x0f) \
- | (((insn) >> ((smallbit) - 4)) & 0x10); \
- } else { \
- if (insn & (1 << (smallbit))) \
- return 1; \
- reg = ((insn) >> (bigbit)) & 0x0f; \
- }} while (0)
-
-#define VFP_DREG_D(reg, insn) VFP_DREG(reg, insn, 12, 22)
-#define VFP_DREG_N(reg, insn) VFP_DREG(reg, insn, 16, 7)
-#define VFP_DREG_M(reg, insn) VFP_DREG(reg, insn, 0, 5)
-
static inline bool use_goto_tb(DisasContext *s, target_ulong dest)
{
#ifndef CONFIG_USER_ONLY
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 22/27] target/arm/translate.c: Delete/amend incorrect comments
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (20 preceding siblings ...)
2020-08-24 9:48 ` [PULL 21/27] target/arm: Delete unused VFP_DREG macros Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 23/27] target/arm: Delete unused ARM_FEATURE_CRC Peter Maydell
` (5 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
In arm_tr_init_disas_context() we have a FIXME comment that suggests
"cpu_M0 can probably be the same as cpu_V0". This isn't in fact
possible: cpu_V0 is used as a temporary inside gen_iwmmxt_shift(),
and that function is called in various places where cpu_M0 contains a
live value (i.e. between gen_op_iwmmxt_movq_M0_wRn() and
gen_op_iwmmxt_movq_wRn_M0() calls). Remove the comment.
We also have a comment on the declarations of cpu_V0/V1/M0 which
claims they're "for efficiency". This isn't true with modern TCG, so
replace this comment with one which notes that they're only used with
the iwmmxt decode.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200803132815.3861-1-peter.maydell@linaro.org
---
target/arm/translate.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index 4ffd8b1fbe5..dd25adcf402 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -57,8 +57,9 @@
#define IS_USER(s) (s->user)
#endif
-/* We reuse the same 64-bit temporaries for efficiency. */
+/* These are TCG temporaries used only by the legacy iwMMXt decoder */
static TCGv_i64 cpu_V0, cpu_V1, cpu_M0;
+/* These are TCG globals which alias CPUARMState fields */
static TCGv_i32 cpu_R[16];
TCGv_i32 cpu_CF, cpu_NF, cpu_VF, cpu_ZF;
TCGv_i64 cpu_exclusive_addr;
@@ -8566,7 +8567,6 @@ static void arm_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
cpu_V0 = tcg_temp_new_i64();
cpu_V1 = tcg_temp_new_i64();
- /* FIXME: cpu_M0 can probably be the same as cpu_V0. */
cpu_M0 = tcg_temp_new_i64();
}
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 23/27] target/arm: Delete unused ARM_FEATURE_CRC
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (21 preceding siblings ...)
2020-08-24 9:48 ` [PULL 22/27] target/arm/translate.c: Delete/amend incorrect comments Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 24/27] target/arm: Replace A64 get_fpstatus_ptr() with generic fpstatus_ptr() Peter Maydell
` (4 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
In commit 962fcbf2efe57231a9f5df we converted the uses of the
ARM_FEATURE_CRC bit to use the aa32_crc32 isar_feature test
instead. However we forgot to remove the now-unused definition
of the feature name in the enum. Delete it now.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20200805210848.6688-1-peter.maydell@linaro.org
---
target/arm/cpu.h | 1 -
1 file changed, 1 deletion(-)
diff --git a/target/arm/cpu.h b/target/arm/cpu.h
index 9e8ed423ea1..9d2845c1797 100644
--- a/target/arm/cpu.h
+++ b/target/arm/cpu.h
@@ -1950,7 +1950,6 @@ enum arm_features {
ARM_FEATURE_V8,
ARM_FEATURE_AARCH64, /* supports 64 bit mode */
ARM_FEATURE_CBAR, /* has cp15 CBAR */
- ARM_FEATURE_CRC, /* ARMv8 CRC instructions */
ARM_FEATURE_CBAR_RO, /* has cp15 CBAR and it is read-only */
ARM_FEATURE_EL2, /* has EL2 Virtualization support */
ARM_FEATURE_EL3, /* has EL3 Secure monitor support */
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 24/27] target/arm: Replace A64 get_fpstatus_ptr() with generic fpstatus_ptr()
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (22 preceding siblings ...)
2020-08-24 9:48 ` [PULL 23/27] target/arm: Delete unused ARM_FEATURE_CRC Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 25/27] target/arm: Make A32/T32 use new fpstatus_ptr() API Peter Maydell
` (3 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
We currently have two versions of get_fpstatus_ptr(), which both take
an effectively boolean argument:
* the one for A64 takes "bool is_f16" to distinguish fp16 from other ops
* the one for A32/T32 takes "int neon" to distinguish Neon from other ops
This is confusing, and to implement ARMv8.2-FP16 the A32/T32 one will
need to make a four-way distinction between "non-Neon, FP16",
"non-Neon, single/double", "Neon, FP16" and "Neon, single/double".
The A64 version will then be a strict subset of the A32/T32 version.
To clean this all up, we want to go to a single implementation which
takes an enum argument with values FPST_FPCR, FPST_STD,
FPST_FPCR_F16, and FPST_STD_F16. We rename the function to
fpstatus_ptr() so that unconverted code gets a compilation error
rather than silently passing the wrong thing to the new function.
This commit implements that new API, and converts A64 to use it:
get_fpstatus_ptr(false) -> fpstatus_ptr(FPST_FPCR)
get_fpstatus_ptr(true) -> fpstatus_ptr(FPST_FPCR_F16)
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200806104453.30393-2-peter.maydell@linaro.org
---
target/arm/translate-a64.h | 1 -
target/arm/translate.h | 51 ++++++++++++++++++++++
target/arm/translate-a64.c | 89 +++++++++++++++-----------------------
target/arm/translate-sve.c | 34 +++++++--------
4 files changed, 103 insertions(+), 72 deletions(-)
diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h
index 647f0c74f62..2e0d16da259 100644
--- a/target/arm/translate-a64.h
+++ b/target/arm/translate-a64.h
@@ -37,7 +37,6 @@ TCGv_i64 cpu_reg_sp(DisasContext *s, int reg);
TCGv_i64 read_cpu_reg(DisasContext *s, int reg, int sf);
TCGv_i64 read_cpu_reg_sp(DisasContext *s, int reg, int sf);
void write_fp_dreg(DisasContext *s, int reg, TCGv_i64 v);
-TCGv_ptr get_fpstatus_ptr(bool);
bool logic_imm_decode_wmask(uint64_t *result, unsigned int immn,
unsigned int imms, unsigned int immr);
bool sve_access_check(DisasContext *s);
diff --git a/target/arm/translate.h b/target/arm/translate.h
index 16f2699ad72..e3680e65479 100644
--- a/target/arm/translate.h
+++ b/target/arm/translate.h
@@ -393,4 +393,55 @@ typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32);
typedef void CryptoThreeOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr);
typedef void AtomicThreeOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGArg, MemOp);
+/*
+ * Enum for argument to fpstatus_ptr().
+ */
+typedef enum ARMFPStatusFlavour {
+ FPST_FPCR,
+ FPST_FPCR_F16,
+ FPST_STD,
+ FPST_STD_F16,
+} ARMFPStatusFlavour;
+
+/**
+ * fpstatus_ptr: return TCGv_ptr to the specified fp_status field
+ *
+ * We have multiple softfloat float_status fields in the Arm CPU state struct
+ * (see the comment in cpu.h for details). Return a TCGv_ptr which has
+ * been set up to point to the requested field in the CPU state struct.
+ * The options are:
+ *
+ * FPST_FPCR
+ * for non-FP16 operations controlled by the FPCR
+ * FPST_FPCR_F16
+ * for operations controlled by the FPCR where FPCR.FZ16 is to be used
+ * FPST_STD
+ * for A32/T32 Neon operations using the "standard FPSCR value"
+ * FPST_STD_F16
+ * as FPST_STD, but where FPCR.FZ16 is to be used
+ */
+static inline TCGv_ptr fpstatus_ptr(ARMFPStatusFlavour flavour)
+{
+ TCGv_ptr statusptr = tcg_temp_new_ptr();
+ int offset;
+
+ switch (flavour) {
+ case FPST_FPCR:
+ offset = offsetof(CPUARMState, vfp.fp_status);
+ break;
+ case FPST_FPCR_F16:
+ offset = offsetof(CPUARMState, vfp.fp_status_f16);
+ break;
+ case FPST_STD:
+ offset = offsetof(CPUARMState, vfp.standard_fp_status);
+ break;
+ case FPST_STD_F16:
+ /* Not yet used or implemented: fall through to assert */
+ default:
+ g_assert_not_reached();
+ }
+ tcg_gen_addi_ptr(statusptr, cpu_env, offset);
+ return statusptr;
+}
+
#endif /* TARGET_ARM_TRANSLATE_H */
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index 534c3ff5f37..0fc5e12fab4 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -609,25 +609,6 @@ static void write_fp_sreg(DisasContext *s, int reg, TCGv_i32 v)
tcg_temp_free_i64(tmp);
}
-TCGv_ptr get_fpstatus_ptr(bool is_f16)
-{
- TCGv_ptr statusptr = tcg_temp_new_ptr();
- int offset;
-
- /* In A64 all instructions (both FP and Neon) use the FPCR; there
- * is no equivalent of the A32 Neon "standard FPSCR value".
- * However half-precision operations operate under a different
- * FZ16 flag and use vfp.fp_status_f16 instead of vfp.fp_status.
- */
- if (is_f16) {
- offset = offsetof(CPUARMState, vfp.fp_status_f16);
- } else {
- offset = offsetof(CPUARMState, vfp.fp_status);
- }
- tcg_gen_addi_ptr(statusptr, cpu_env, offset);
- return statusptr;
-}
-
/* Expand a 2-operand AdvSIMD vector operation using an expander function. */
static void gen_gvec_fn2(DisasContext *s, bool is_q, int rd, int rn,
GVecGen2Fn *gvec_fn, int vece)
@@ -689,7 +670,7 @@ static void gen_gvec_op3_fpst(DisasContext *s, bool is_q, int rd, int rn,
int rm, bool is_fp16, int data,
gen_helper_gvec_3_ptr *fn)
{
- TCGv_ptr fpst = get_fpstatus_ptr(is_fp16);
+ TCGv_ptr fpst = fpstatus_ptr(is_fp16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(vec_full_reg_offset(s, rd),
vec_full_reg_offset(s, rn),
vec_full_reg_offset(s, rm), fpst,
@@ -5898,7 +5879,7 @@ static void handle_fp_compare(DisasContext *s, int size,
bool cmp_with_zero, bool signal_all_nans)
{
TCGv_i64 tcg_flags = tcg_temp_new_i64();
- TCGv_ptr fpst = get_fpstatus_ptr(size == MO_16);
+ TCGv_ptr fpst = fpstatus_ptr(size == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
if (size == MO_64) {
TCGv_i64 tcg_vn, tcg_vm;
@@ -6157,7 +6138,7 @@ static void handle_fp_1src_half(DisasContext *s, int opcode, int rd, int rn)
tcg_gen_xori_i32(tcg_res, tcg_op, 0x8000);
break;
case 0x3: /* FSQRT */
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
gen_helper_sqrt_f16(tcg_res, tcg_op, fpst);
break;
case 0x8: /* FRINTN */
@@ -6167,7 +6148,7 @@ static void handle_fp_1src_half(DisasContext *s, int opcode, int rd, int rn)
case 0xc: /* FRINTA */
{
TCGv_i32 tcg_rmode = tcg_const_i32(arm_rmode_to_sf(opcode & 7));
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, fpst);
gen_helper_advsimd_rinth(tcg_res, tcg_op, fpst);
@@ -6177,11 +6158,11 @@ static void handle_fp_1src_half(DisasContext *s, int opcode, int rd, int rn)
break;
}
case 0xe: /* FRINTX */
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
gen_helper_advsimd_rinth_exact(tcg_res, tcg_op, fpst);
break;
case 0xf: /* FRINTI */
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
gen_helper_advsimd_rinth(tcg_res, tcg_op, fpst);
break;
default:
@@ -6253,7 +6234,7 @@ static void handle_fp_1src_single(DisasContext *s, int opcode, int rd, int rn)
g_assert_not_reached();
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
if (rmode >= 0) {
TCGv_i32 tcg_rmode = tcg_const_i32(rmode);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, fpst);
@@ -6330,7 +6311,7 @@ static void handle_fp_1src_double(DisasContext *s, int opcode, int rd, int rn)
g_assert_not_reached();
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
if (rmode >= 0) {
TCGv_i32 tcg_rmode = tcg_const_i32(rmode);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, fpst);
@@ -6365,7 +6346,7 @@ static void handle_fp_fcvt(DisasContext *s, int opcode,
/* Single to half */
TCGv_i32 tcg_rd = tcg_temp_new_i32();
TCGv_i32 ahp = get_ahp_flag();
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_vfp_fcvt_f32_to_f16(tcg_rd, tcg_rn, fpst, ahp);
/* write_fp_sreg is OK here because top half of tcg_rd is zero */
@@ -6385,7 +6366,7 @@ static void handle_fp_fcvt(DisasContext *s, int opcode,
/* Double to single */
gen_helper_vfp_fcvtsd(tcg_rd, tcg_rn, cpu_env);
} else {
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
TCGv_i32 ahp = get_ahp_flag();
/* Double to half */
gen_helper_vfp_fcvt_f64_to_f16(tcg_rd, tcg_rn, fpst, ahp);
@@ -6401,7 +6382,7 @@ static void handle_fp_fcvt(DisasContext *s, int opcode,
case 0x3:
{
TCGv_i32 tcg_rn = read_fp_sreg(s, rn);
- TCGv_ptr tcg_fpst = get_fpstatus_ptr(false);
+ TCGv_ptr tcg_fpst = fpstatus_ptr(FPST_FPCR);
TCGv_i32 tcg_ahp = get_ahp_flag();
tcg_gen_ext16u_i32(tcg_rn, tcg_rn);
if (dtype == 0) {
@@ -6518,7 +6499,7 @@ static void handle_fp_2src_single(DisasContext *s, int opcode,
TCGv_ptr fpst;
tcg_res = tcg_temp_new_i32();
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
tcg_op1 = read_fp_sreg(s, rn);
tcg_op2 = read_fp_sreg(s, rm);
@@ -6571,7 +6552,7 @@ static void handle_fp_2src_double(DisasContext *s, int opcode,
TCGv_ptr fpst;
tcg_res = tcg_temp_new_i64();
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
tcg_op1 = read_fp_dreg(s, rn);
tcg_op2 = read_fp_dreg(s, rm);
@@ -6624,7 +6605,7 @@ static void handle_fp_2src_half(DisasContext *s, int opcode,
TCGv_ptr fpst;
tcg_res = tcg_temp_new_i32();
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
tcg_op1 = read_fp_hreg(s, rn);
tcg_op2 = read_fp_hreg(s, rm);
@@ -6723,7 +6704,7 @@ static void handle_fp_3src_single(DisasContext *s, bool o0, bool o1,
{
TCGv_i32 tcg_op1, tcg_op2, tcg_op3;
TCGv_i32 tcg_res = tcg_temp_new_i32();
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
tcg_op1 = read_fp_sreg(s, rn);
tcg_op2 = read_fp_sreg(s, rm);
@@ -6761,7 +6742,7 @@ static void handle_fp_3src_double(DisasContext *s, bool o0, bool o1,
{
TCGv_i64 tcg_op1, tcg_op2, tcg_op3;
TCGv_i64 tcg_res = tcg_temp_new_i64();
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
tcg_op1 = read_fp_dreg(s, rn);
tcg_op2 = read_fp_dreg(s, rm);
@@ -6799,7 +6780,7 @@ static void handle_fp_3src_half(DisasContext *s, bool o0, bool o1,
{
TCGv_i32 tcg_op1, tcg_op2, tcg_op3;
TCGv_i32 tcg_res = tcg_temp_new_i32();
- TCGv_ptr fpst = get_fpstatus_ptr(true);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR_F16);
tcg_op1 = read_fp_hreg(s, rn);
tcg_op2 = read_fp_hreg(s, rm);
@@ -6945,7 +6926,7 @@ static void handle_fpfpcvt(DisasContext *s, int rd, int rn, int opcode,
TCGv_i32 tcg_shift, tcg_single;
TCGv_i64 tcg_double;
- tcg_fpstatus = get_fpstatus_ptr(type == 3);
+ tcg_fpstatus = fpstatus_ptr(type == 3 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_shift = tcg_const_i32(64 - scale);
@@ -7233,7 +7214,7 @@ static void handle_fmov(DisasContext *s, int rd, int rn, int type, bool itof)
static void handle_fjcvtzs(DisasContext *s, int rd, int rn)
{
TCGv_i64 t = read_fp_dreg(s, rn);
- TCGv_ptr fpstatus = get_fpstatus_ptr(false);
+ TCGv_ptr fpstatus = fpstatus_ptr(FPST_FPCR);
gen_helper_fjcvtzs(t, t, fpstatus);
@@ -7847,7 +7828,7 @@ static void disas_simd_across_lanes(DisasContext *s, uint32_t insn)
* Note that correct NaN propagation requires that we do these
* operations in exactly the order specified by the pseudocode.
*/
- TCGv_ptr fpst = get_fpstatus_ptr(size == MO_16);
+ TCGv_ptr fpst = fpstatus_ptr(size == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
int fpopcode = opcode | is_min << 4 | is_u << 5;
int vmap = (1 << elements) - 1;
TCGv_i32 tcg_res32 = do_reduction_op(s, fpopcode, rn, esize,
@@ -8359,7 +8340,7 @@ static void disas_simd_scalar_pairwise(DisasContext *s, uint32_t insn)
return;
}
- fpst = get_fpstatus_ptr(size == MO_16);
+ fpst = fpstatus_ptr(size == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
break;
default:
unallocated_encoding(s);
@@ -8872,7 +8853,7 @@ static void handle_simd_intfp_conv(DisasContext *s, int rd, int rn,
int elements, int is_signed,
int fracbits, int size)
{
- TCGv_ptr tcg_fpst = get_fpstatus_ptr(size == MO_16);
+ TCGv_ptr tcg_fpst = fpstatus_ptr(size == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
TCGv_i32 tcg_shift = NULL;
MemOp mop = size | (is_signed ? MO_SIGN : 0);
@@ -9053,7 +9034,7 @@ static void handle_simd_shift_fpint_conv(DisasContext *s, bool is_scalar,
assert(!(is_scalar && is_q));
tcg_rmode = tcg_const_i32(arm_rmode_to_sf(FPROUNDING_ZERO));
- tcg_fpstatus = get_fpstatus_ptr(size == MO_16);
+ tcg_fpstatus = fpstatus_ptr(size == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, tcg_fpstatus);
fracbits = (16 << size) - immhb;
tcg_shift = tcg_const_i32(fracbits);
@@ -9392,7 +9373,7 @@ static void handle_3same_float(DisasContext *s, int size, int elements,
int fpopcode, int rd, int rn, int rm)
{
int pass;
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
for (pass = 0; pass < elements; pass++) {
if (size) {
@@ -9785,7 +9766,7 @@ static void disas_simd_scalar_three_reg_same_fp16(DisasContext *s,
return;
}
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
tcg_op1 = read_fp_hreg(s, rn);
tcg_op2 = read_fp_hreg(s, rm);
@@ -10038,7 +10019,7 @@ static void handle_2misc_fcmp_zero(DisasContext *s, int opcode,
return;
}
- fpst = get_fpstatus_ptr(size == MO_16);
+ fpst = fpstatus_ptr(size == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
if (is_double) {
TCGv_i64 tcg_op = tcg_temp_new_i64();
@@ -10168,7 +10149,7 @@ static void handle_2misc_reciprocal(DisasContext *s, int opcode,
int size, int rn, int rd)
{
bool is_double = (size == 3);
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
if (is_double) {
TCGv_i64 tcg_op = tcg_temp_new_i64();
@@ -10309,7 +10290,7 @@ static void handle_2misc_narrow(DisasContext *s, bool scalar,
} else {
TCGv_i32 tcg_lo = tcg_temp_new_i32();
TCGv_i32 tcg_hi = tcg_temp_new_i32();
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
TCGv_i32 ahp = get_ahp_flag();
tcg_gen_extr_i64_i32(tcg_lo, tcg_hi, tcg_op);
@@ -10571,7 +10552,7 @@ static void disas_simd_scalar_two_reg_misc(DisasContext *s, uint32_t insn)
if (is_fcvt) {
tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode));
- tcg_fpstatus = get_fpstatus_ptr(false);
+ tcg_fpstatus = fpstatus_ptr(FPST_FPCR);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, tcg_fpstatus);
} else {
tcg_rmode = NULL;
@@ -11396,7 +11377,7 @@ static void handle_simd_3same_pair(DisasContext *s, int is_q, int u, int opcode,
/* Floating point operations need fpst */
if (opcode >= 0x58) {
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
} else {
fpst = NULL;
}
@@ -11994,7 +11975,7 @@ static void disas_simd_three_reg_same_fp16(DisasContext *s, uint32_t insn)
break;
}
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_FPCR_F16);
if (pairwise) {
int maxpass = is_q ? 8 : 4;
@@ -12287,7 +12268,7 @@ static void handle_2misc_widening(DisasContext *s, int opcode, bool is_q,
/* 16 -> 32 bit fp conversion */
int srcelt = is_q ? 4 : 0;
TCGv_i32 tcg_res[4];
- TCGv_ptr fpst = get_fpstatus_ptr(false);
+ TCGv_ptr fpst = fpstatus_ptr(FPST_FPCR);
TCGv_i32 ahp = get_ahp_flag();
for (pass = 0; pass < 4; pass++) {
@@ -12759,7 +12740,7 @@ static void disas_simd_two_reg_misc(DisasContext *s, uint32_t insn)
}
if (need_fpstatus || need_rmode) {
- tcg_fpstatus = get_fpstatus_ptr(false);
+ tcg_fpstatus = fpstatus_ptr(FPST_FPCR);
} else {
tcg_fpstatus = NULL;
}
@@ -13149,7 +13130,7 @@ static void disas_simd_two_reg_misc_fp16(DisasContext *s, uint32_t insn)
}
if (need_rmode || need_fpst) {
- tcg_fpstatus = get_fpstatus_ptr(true);
+ tcg_fpstatus = fpstatus_ptr(FPST_FPCR_F16);
}
if (need_rmode) {
@@ -13458,7 +13439,7 @@ static void disas_simd_indexed(DisasContext *s, uint32_t insn)
}
if (is_fp) {
- fpst = get_fpstatus_ptr(is_fp16);
+ fpst = fpstatus_ptr(is_fp16 ? FPST_FPCR_F16 : FPST_FPCR);
} else {
fpst = NULL;
}
diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c
index 8c7fbbd5033..d97cb37d83f 100644
--- a/target/arm/translate-sve.c
+++ b/target/arm/translate-sve.c
@@ -3470,7 +3470,7 @@ static bool trans_FMLA_zzxz(DisasContext *s, arg_FMLA_zzxz *a)
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_4_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3496,7 +3496,7 @@ static bool trans_FMUL_zzx(DisasContext *s, arg_FMUL_zzx *a)
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3528,7 +3528,7 @@ static void do_reduce(DisasContext *s, arg_rpr_esz *a,
tcg_gen_addi_ptr(t_zn, cpu_env, vec_full_reg_offset(s, a->rn));
tcg_gen_addi_ptr(t_pg, cpu_env, pred_full_reg_offset(s, a->pg));
- status = get_fpstatus_ptr(a->esz == MO_16);
+ status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
fn(temp, t_zn, t_pg, status, t_desc);
tcg_temp_free_ptr(t_zn);
@@ -3570,7 +3570,7 @@ DO_VPZ(FMAXV, fmaxv)
static void do_zz_fp(DisasContext *s, arg_rr_esz *a, gen_helper_gvec_2_ptr *fn)
{
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_2_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
@@ -3618,7 +3618,7 @@ static void do_ppz_fp(DisasContext *s, arg_rpr_esz *a,
gen_helper_gvec_3_ptr *fn)
{
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(pred_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
@@ -3670,7 +3670,7 @@ static bool trans_FTMAD(DisasContext *s, arg_FTMAD *a)
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3710,7 +3710,7 @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a)
t_pg = tcg_temp_new_ptr();
tcg_gen_addi_ptr(t_rm, cpu_env, vec_full_reg_offset(s, a->rm));
tcg_gen_addi_ptr(t_pg, cpu_env, pred_full_reg_offset(s, a->pg));
- t_fpst = get_fpstatus_ptr(a->esz == MO_16);
+ t_fpst = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
t_desc = tcg_const_i32(simd_desc(vsz, vsz, 0));
fns[a->esz - 1](t_val, t_val, t_rm, t_pg, t_fpst, t_desc);
@@ -3737,7 +3737,7 @@ static bool do_zzz_fp(DisasContext *s, arg_rrr_esz *a,
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3779,7 +3779,7 @@ static bool do_zpzz_fp(DisasContext *s, arg_rprr_esz *a,
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_4_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3831,7 +3831,7 @@ static void do_fp_scalar(DisasContext *s, int zd, int zn, int pg, bool is_fp16,
tcg_gen_addi_ptr(t_zn, cpu_env, vec_full_reg_offset(s, zn));
tcg_gen_addi_ptr(t_pg, cpu_env, pred_full_reg_offset(s, pg));
- status = get_fpstatus_ptr(is_fp16);
+ status = fpstatus_ptr(is_fp16 ? FPST_FPCR_F16 : FPST_FPCR);
desc = tcg_const_i32(simd_desc(vsz, vsz, 0));
fn(t_zd, t_zn, t_pg, scalar, status, desc);
@@ -3895,7 +3895,7 @@ static bool do_fp_cmp(DisasContext *s, arg_rprr_esz *a,
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_4_ptr(pred_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3939,7 +3939,7 @@ static bool trans_FCADD(DisasContext *s, arg_FCADD *a)
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_4_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -3958,7 +3958,7 @@ static bool do_fmla(DisasContext *s, arg_rprrr_esz *a,
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_5_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -4001,7 +4001,7 @@ static bool trans_FCMLA_zpzzz(DisasContext *s, arg_FCMLA_zpzzz *a)
}
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_5_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -4024,7 +4024,7 @@ static bool trans_FCMLA_zzxz(DisasContext *s, arg_FCMLA_zzxz *a)
tcg_debug_assert(a->rd == a->ra);
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(vec_full_reg_offset(s, a->rd),
vec_full_reg_offset(s, a->rn),
vec_full_reg_offset(s, a->rm),
@@ -4045,7 +4045,7 @@ static bool do_zpz_ptr(DisasContext *s, int rd, int rn, int pg,
{
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
- TCGv_ptr status = get_fpstatus_ptr(is_fp16);
+ TCGv_ptr status = fpstatus_ptr(is_fp16 ? FPST_FPCR_F16 : FPST_FPCR);
tcg_gen_gvec_3_ptr(vec_full_reg_offset(s, rd),
vec_full_reg_offset(s, rn),
pred_full_reg_offset(s, pg),
@@ -4191,7 +4191,7 @@ static bool do_frint_mode(DisasContext *s, arg_rpr_esz *a, int mode)
if (sve_access_check(s)) {
unsigned vsz = vec_full_reg_size(s);
TCGv_i32 tmode = tcg_const_i32(mode);
- TCGv_ptr status = get_fpstatus_ptr(a->esz == MO_16);
+ TCGv_ptr status = fpstatus_ptr(a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR);
gen_helper_set_rmode(tmode, tmode, status);
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 25/27] target/arm: Make A32/T32 use new fpstatus_ptr() API
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (23 preceding siblings ...)
2020-08-24 9:48 ` [PULL 24/27] target/arm: Replace A64 get_fpstatus_ptr() with generic fpstatus_ptr() Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 26/27] target/arm: Implement FPST_STD_F16 fpstatus Peter Maydell
` (2 subsequent siblings)
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
Make A32/T32 code use the new fpstatus_ptr() API:
get_fpstatus_ptr(0) -> fpstatus_ptr(FPST_FPCR)
get_fpstatus_ptr(1) -> fpstatus_ptr(FPST_STD)
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200806104453.30393-3-peter.maydell@linaro.org
---
target/arm/translate.c | 13 ----------
target/arm/translate-neon.c.inc | 28 ++++++++++-----------
target/arm/translate-vfp.c.inc | 44 ++++++++++++++++-----------------
3 files changed, 36 insertions(+), 49 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index dd25adcf402..d34c1d351a6 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -1094,19 +1094,6 @@ static inline void gen_hlt(DisasContext *s, int imm)
unallocated_encoding(s);
}
-static TCGv_ptr get_fpstatus_ptr(int neon)
-{
- TCGv_ptr statusptr = tcg_temp_new_ptr();
- int offset;
- if (neon) {
- offset = offsetof(CPUARMState, vfp.standard_fp_status);
- } else {
- offset = offsetof(CPUARMState, vfp.fp_status);
- }
- tcg_gen_addi_ptr(statusptr, cpu_env, offset);
- return statusptr;
-}
-
static inline long vfp_reg_offset(bool dp, unsigned reg)
{
if (dp) {
diff --git a/target/arm/translate-neon.c.inc b/target/arm/translate-neon.c.inc
index 8fbe8cef9f6..f8f2cb1869f 100644
--- a/target/arm/translate-neon.c.inc
+++ b/target/arm/translate-neon.c.inc
@@ -181,7 +181,7 @@ static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a)
}
opr_sz = (1 + a->q) * 8;
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
fn_gvec_ptr = a->size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah;
tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
@@ -218,7 +218,7 @@ static bool trans_VCADD(DisasContext *s, arg_VCADD *a)
}
opr_sz = (1 + a->q) * 8;
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
fn_gvec_ptr = a->size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh;
tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
@@ -322,7 +322,7 @@ static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a)
fn_gvec_ptr = (a->size ? gen_helper_gvec_fcmlas_idx
: gen_helper_gvec_fcmlah_idx);
opr_sz = (1 + a->q) * 8;
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
vfp_reg_offset(1, a->vm),
@@ -358,7 +358,7 @@ static bool trans_VDOT_scalar(DisasContext *s, arg_VDOT_scalar *a)
fn_gvec = a->u ? gen_helper_gvec_udot_idx_b : gen_helper_gvec_sdot_idx_b;
opr_sz = (1 + a->q) * 8;
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
tcg_gen_gvec_3_ool(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
vfp_reg_offset(1, a->rm),
@@ -1063,7 +1063,7 @@ static bool do_3same_fp(DisasContext *s, arg_3same *a, VFPGen3OpSPFn *fn,
return true;
}
- TCGv_ptr fpstatus = get_fpstatus_ptr(1);
+ TCGv_ptr fpstatus = fpstatus_ptr(FPST_STD);
for (pass = 0; pass < (a->q ? 4 : 2); pass++) {
tmp = neon_load_reg(a->vn, pass);
tmp2 = neon_load_reg(a->vm, pass);
@@ -1091,7 +1091,7 @@ static bool do_3same_fp(DisasContext *s, arg_3same *a, VFPGen3OpSPFn *fn,
uint32_t rn_ofs, uint32_t rm_ofs, \
uint32_t oprsz, uint32_t maxsz) \
{ \
- TCGv_ptr fpst = get_fpstatus_ptr(1); \
+ TCGv_ptr fpst = fpstatus_ptr(FPST_STD); \
tcg_gen_gvec_3_ptr(rd_ofs, rn_ofs, rm_ofs, fpst, \
oprsz, maxsz, 0, FUNC); \
tcg_temp_free_ptr(fpst); \
@@ -1287,7 +1287,7 @@ static bool do_3same_fp_pair(DisasContext *s, arg_3same *a, VFPGen3OpSPFn *fn)
* early. Since Q is 0 there are always just two passes, so instead
* of a complicated loop over each pass we just unroll.
*/
- fpstatus = get_fpstatus_ptr(1);
+ fpstatus = fpstatus_ptr(FPST_STD);
tmp = neon_load_reg(a->vn, 0);
tmp2 = neon_load_reg(a->vn, 1);
fn(tmp, tmp, tmp2, fpstatus);
@@ -1790,7 +1790,7 @@ static bool do_fp_2sh(DisasContext *s, arg_2reg_shift *a,
return true;
}
- fpstatus = get_fpstatus_ptr(1);
+ fpstatus = fpstatus_ptr(FPST_STD);
shiftv = tcg_const_i32(a->shift);
for (pass = 0; pass < (a->q ? 4 : 2); pass++) {
tmp = neon_load_reg(a->vm, pass);
@@ -2591,7 +2591,7 @@ static bool trans_VMLS_2sc(DisasContext *s, arg_2scalar *a)
#define WRAP_FP_FN(WRAPNAME, FUNC) \
static void WRAPNAME(TCGv_i32 rd, TCGv_i32 rn, TCGv_i32 rm) \
{ \
- TCGv_ptr fpstatus = get_fpstatus_ptr(1); \
+ TCGv_ptr fpstatus = fpstatus_ptr(FPST_STD); \
FUNC(rd, rn, rm, fpstatus); \
tcg_temp_free_ptr(fpstatus); \
}
@@ -3480,7 +3480,7 @@ static bool trans_VCVT_F16_F32(DisasContext *s, arg_2misc *a)
return true;
}
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_STD);
ahp = get_ahp_flag();
tmp = neon_load_reg(a->vm, 0);
gen_helper_vfp_fcvt_f32_to_f16(tmp, tmp, fpst, ahp);
@@ -3528,7 +3528,7 @@ static bool trans_VCVT_F32_F16(DisasContext *s, arg_2misc *a)
return true;
}
- fpst = get_fpstatus_ptr(true);
+ fpst = fpstatus_ptr(FPST_STD);
ahp = get_ahp_flag();
tmp3 = tcg_temp_new_i32();
tmp = neon_load_reg(a->vm, 0);
@@ -3838,7 +3838,7 @@ static bool do_2misc_fp(DisasContext *s, arg_2misc *a,
return true;
}
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
for (pass = 0; pass < (a->q ? 4 : 2); pass++) {
TCGv_i32 tmp = neon_load_reg(a->vm, pass);
fn(tmp, tmp, fpst);
@@ -3932,7 +3932,7 @@ static bool do_vrint(DisasContext *s, arg_2misc *a, int rmode)
return true;
}
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode));
gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, cpu_env);
for (pass = 0; pass < (a->q ? 4 : 2); pass++) {
@@ -3993,7 +3993,7 @@ static bool do_vcvt(DisasContext *s, arg_2misc *a, int rmode, bool is_signed)
return true;
}
- fpst = get_fpstatus_ptr(1);
+ fpst = fpstatus_ptr(FPST_STD);
tcg_shift = tcg_const_i32(0);
tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode));
gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, cpu_env);
diff --git a/target/arm/translate-vfp.c.inc b/target/arm/translate-vfp.c.inc
index d376bd1c1ad..4eeafb494ad 100644
--- a/target/arm/translate-vfp.c.inc
+++ b/target/arm/translate-vfp.c.inc
@@ -359,7 +359,7 @@ static bool trans_VRINT(DisasContext *s, arg_VRINT *a)
return true;
}
- fpst = get_fpstatus_ptr(0);
+ fpst = fpstatus_ptr(FPST_FPCR);
tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rounding));
gen_helper_set_rmode(tcg_rmode, tcg_rmode, fpst);
@@ -422,7 +422,7 @@ static bool trans_VCVT(DisasContext *s, arg_VCVT *a)
return true;
}
- fpst = get_fpstatus_ptr(0);
+ fpst = fpstatus_ptr(FPST_FPCR);
tcg_shift = tcg_const_i32(0);
@@ -1231,7 +1231,7 @@ static bool do_vfp_3op_sp(DisasContext *s, VFPGen3OpSPFn *fn,
f0 = tcg_temp_new_i32();
f1 = tcg_temp_new_i32();
fd = tcg_temp_new_i32();
- fpst = get_fpstatus_ptr(0);
+ fpst = fpstatus_ptr(FPST_FPCR);
neon_load_reg32(f0, vn);
neon_load_reg32(f1, vm);
@@ -1314,7 +1314,7 @@ static bool do_vfp_3op_dp(DisasContext *s, VFPGen3OpDPFn *fn,
f0 = tcg_temp_new_i64();
f1 = tcg_temp_new_i64();
fd = tcg_temp_new_i64();
- fpst = get_fpstatus_ptr(0);
+ fpst = fpstatus_ptr(FPST_FPCR);
neon_load_reg64(f0, vn);
neon_load_reg64(f1, vm);
@@ -1796,7 +1796,7 @@ static bool do_vfm_sp(DisasContext *s, arg_VFMA_sp *a, bool neg_n, bool neg_d)
/* VFNMA, VFNMS */
gen_helper_vfp_negs(vd, vd);
}
- fpst = get_fpstatus_ptr(0);
+ fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_vfp_muladds(vd, vn, vm, vd, fpst);
neon_store_reg32(vd, a->vd);
@@ -1887,7 +1887,7 @@ static bool do_vfm_dp(DisasContext *s, arg_VFMA_dp *a, bool neg_n, bool neg_d)
/* VFNMA, VFNMS */
gen_helper_vfp_negd(vd, vd);
}
- fpst = get_fpstatus_ptr(0);
+ fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_vfp_muladdd(vd, vn, vm, vd, fpst);
neon_store_reg64(vd, a->vd);
@@ -2171,7 +2171,7 @@ static bool trans_VCVT_f32_f16(DisasContext *s, arg_VCVT_f32_f16 *a)
return true;
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
ahp_mode = get_ahp_flag();
tmp = tcg_temp_new_i32();
/* The T bit tells us if we want the low or high 16 bits of Vm */
@@ -2208,7 +2208,7 @@ static bool trans_VCVT_f64_f16(DisasContext *s, arg_VCVT_f64_f16 *a)
return true;
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
ahp_mode = get_ahp_flag();
tmp = tcg_temp_new_i32();
/* The T bit tells us if we want the low or high 16 bits of Vm */
@@ -2237,7 +2237,7 @@ static bool trans_VCVT_f16_f32(DisasContext *s, arg_VCVT_f16_f32 *a)
return true;
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
ahp_mode = get_ahp_flag();
tmp = tcg_temp_new_i32();
@@ -2274,7 +2274,7 @@ static bool trans_VCVT_f16_f64(DisasContext *s, arg_VCVT_f16_f64 *a)
return true;
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
ahp_mode = get_ahp_flag();
tmp = tcg_temp_new_i32();
vm = tcg_temp_new_i64();
@@ -2304,7 +2304,7 @@ static bool trans_VRINTR_sp(DisasContext *s, arg_VRINTR_sp *a)
tmp = tcg_temp_new_i32();
neon_load_reg32(tmp, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_rints(tmp, tmp, fpst);
neon_store_reg32(tmp, a->vd);
tcg_temp_free_ptr(fpst);
@@ -2336,7 +2336,7 @@ static bool trans_VRINTR_dp(DisasContext *s, arg_VRINTR_dp *a)
tmp = tcg_temp_new_i64();
neon_load_reg64(tmp, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_rintd(tmp, tmp, fpst);
neon_store_reg64(tmp, a->vd);
tcg_temp_free_ptr(fpst);
@@ -2360,7 +2360,7 @@ static bool trans_VRINTZ_sp(DisasContext *s, arg_VRINTZ_sp *a)
tmp = tcg_temp_new_i32();
neon_load_reg32(tmp, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
tcg_rmode = tcg_const_i32(float_round_to_zero);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, fpst);
gen_helper_rints(tmp, tmp, fpst);
@@ -2397,7 +2397,7 @@ static bool trans_VRINTZ_dp(DisasContext *s, arg_VRINTZ_dp *a)
tmp = tcg_temp_new_i64();
neon_load_reg64(tmp, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
tcg_rmode = tcg_const_i32(float_round_to_zero);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, fpst);
gen_helper_rintd(tmp, tmp, fpst);
@@ -2424,7 +2424,7 @@ static bool trans_VRINTX_sp(DisasContext *s, arg_VRINTX_sp *a)
tmp = tcg_temp_new_i32();
neon_load_reg32(tmp, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_rints_exact(tmp, tmp, fpst);
neon_store_reg32(tmp, a->vd);
tcg_temp_free_ptr(fpst);
@@ -2456,7 +2456,7 @@ static bool trans_VRINTX_dp(DisasContext *s, arg_VRINTX_dp *a)
tmp = tcg_temp_new_i64();
neon_load_reg64(tmp, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
gen_helper_rintd_exact(tmp, tmp, fpst);
neon_store_reg64(tmp, a->vd);
tcg_temp_free_ptr(fpst);
@@ -2535,7 +2535,7 @@ static bool trans_VCVT_int_sp(DisasContext *s, arg_VCVT_int_sp *a)
vm = tcg_temp_new_i32();
neon_load_reg32(vm, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
if (a->s) {
/* i32 -> f32 */
gen_helper_vfp_sitos(vm, vm, fpst);
@@ -2571,7 +2571,7 @@ static bool trans_VCVT_int_dp(DisasContext *s, arg_VCVT_int_dp *a)
vm = tcg_temp_new_i32();
vd = tcg_temp_new_i64();
neon_load_reg32(vm, a->vm);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
if (a->s) {
/* i32 -> f64 */
gen_helper_vfp_sitod(vd, vm, fpst);
@@ -2637,7 +2637,7 @@ static bool trans_VCVT_fix_sp(DisasContext *s, arg_VCVT_fix_sp *a)
vd = tcg_temp_new_i32();
neon_load_reg32(vd, a->vd);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
shift = tcg_const_i32(frac_bits);
/* Switch on op:U:sx bits */
@@ -2702,7 +2702,7 @@ static bool trans_VCVT_fix_dp(DisasContext *s, arg_VCVT_fix_dp *a)
vd = tcg_temp_new_i64();
neon_load_reg64(vd, a->vd);
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
shift = tcg_const_i32(frac_bits);
/* Switch on op:U:sx bits */
@@ -2755,7 +2755,7 @@ static bool trans_VCVT_sp_int(DisasContext *s, arg_VCVT_sp_int *a)
return true;
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
vm = tcg_temp_new_i32();
neon_load_reg32(vm, a->vm);
@@ -2797,7 +2797,7 @@ static bool trans_VCVT_dp_int(DisasContext *s, arg_VCVT_dp_int *a)
return true;
}
- fpst = get_fpstatus_ptr(false);
+ fpst = fpstatus_ptr(FPST_FPCR);
vm = tcg_temp_new_i64();
vd = tcg_temp_new_i32();
neon_load_reg64(vm, a->vm);
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 26/27] target/arm: Implement FPST_STD_F16 fpstatus
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (24 preceding siblings ...)
2020-08-24 9:48 ` [PULL 25/27] target/arm: Make A32/T32 use new fpstatus_ptr() API Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 9:48 ` [PULL 27/27] target/arm: Use correct FPST for VCMLA, VCADD on fp16 Peter Maydell
2020-08-24 13:53 ` [PULL 00/27] target-arm queue Peter Maydell
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
Architecturally, Neon FP16 operations use the "standard FPSCR" like
all other Neon operations. However, this is defined in the Arm ARM
pseudocode as "a fixed value, except that FZ16 (and AHP) follow the
FPSCR bits". In QEMU, the softfloat float_status doesn't include
separate flush-to-zero for FP16 operations, so we must keep separate
fp_status for "Neon non-FP16" and "Neon fp16" operations, in the
same way we do already for the non-Neon "fp_status" vs "fp_status_f16".
Add the extra float_status field to the CPU state structure,
ensure it is correctly initialized and updated on FPSCR writes,
and make fpstatus_ptr(FPST_STD_F16) return a pointer to it.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200806104453.30393-4-peter.maydell@linaro.org
---
target/arm/cpu.h | 9 ++++++++-
target/arm/translate.h | 3 ++-
target/arm/cpu.c | 3 +++
target/arm/vfp_helper.c | 5 +++++
4 files changed, 18 insertions(+), 2 deletions(-)
diff --git a/target/arm/cpu.h b/target/arm/cpu.h
index 9d2845c1797..ac857bdc2c1 100644
--- a/target/arm/cpu.h
+++ b/target/arm/cpu.h
@@ -609,6 +609,8 @@ typedef struct CPUARMState {
* fp_status: is the "normal" fp status.
* fp_status_fp16: used for half-precision calculations
* standard_fp_status : the ARM "Standard FPSCR Value"
+ * standard_fp_status_fp16 : used for half-precision
+ * calculations with the ARM "Standard FPSCR Value"
*
* Half-precision operations are governed by a separate
* flush-to-zero control bit in FPSCR:FZ16. We pass a separate
@@ -619,15 +621,20 @@ typedef struct CPUARMState {
* Neon) which the architecture defines as controlled by the
* standard FPSCR value rather than the FPSCR.
*
+ * The "standard FPSCR but for fp16 ops" is needed because
+ * the "standard FPSCR" tracks the FPSCR.FZ16 bit rather than
+ * using a fixed value for it.
+ *
* To avoid having to transfer exception bits around, we simply
* say that the FPSCR cumulative exception flags are the logical
- * OR of the flags in the three fp statuses. This relies on the
+ * OR of the flags in the four fp statuses. This relies on the
* only thing which needs to read the exception flags being
* an explicit FPSCR read.
*/
float_status fp_status;
float_status fp_status_f16;
float_status standard_fp_status;
+ float_status standard_fp_status_f16;
/* ZCR_EL[1-3] */
uint64_t zcr_el[4];
diff --git a/target/arm/translate.h b/target/arm/translate.h
index e3680e65479..6d6d4c0f425 100644
--- a/target/arm/translate.h
+++ b/target/arm/translate.h
@@ -436,7 +436,8 @@ static inline TCGv_ptr fpstatus_ptr(ARMFPStatusFlavour flavour)
offset = offsetof(CPUARMState, vfp.standard_fp_status);
break;
case FPST_STD_F16:
- /* Not yet used or implemented: fall through to assert */
+ offset = offsetof(CPUARMState, vfp.standard_fp_status_f16);
+ break;
default:
g_assert_not_reached();
}
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 111579554fb..6b382fcd60e 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -391,12 +391,15 @@ static void arm_cpu_reset(DeviceState *dev)
set_flush_to_zero(1, &env->vfp.standard_fp_status);
set_flush_inputs_to_zero(1, &env->vfp.standard_fp_status);
set_default_nan_mode(1, &env->vfp.standard_fp_status);
+ set_default_nan_mode(1, &env->vfp.standard_fp_status_f16);
set_float_detect_tininess(float_tininess_before_rounding,
&env->vfp.fp_status);
set_float_detect_tininess(float_tininess_before_rounding,
&env->vfp.standard_fp_status);
set_float_detect_tininess(float_tininess_before_rounding,
&env->vfp.fp_status_f16);
+ set_float_detect_tininess(float_tininess_before_rounding,
+ &env->vfp.standard_fp_status_f16);
#ifndef CONFIG_USER_ONLY
if (kvm_enabled()) {
kvm_arm_reset_vcpu(cpu);
diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c
index 60dcd4bf145..64266ece620 100644
--- a/target/arm/vfp_helper.c
+++ b/target/arm/vfp_helper.c
@@ -93,6 +93,8 @@ static uint32_t vfp_get_fpscr_from_host(CPUARMState *env)
/* FZ16 does not generate an input denormal exception. */
i |= (get_float_exception_flags(&env->vfp.fp_status_f16)
& ~float_flag_input_denormal);
+ i |= (get_float_exception_flags(&env->vfp.standard_fp_status_f16)
+ & ~float_flag_input_denormal);
return vfp_exceptbits_from_host(i);
}
@@ -124,7 +126,9 @@ static void vfp_set_fpscr_to_host(CPUARMState *env, uint32_t val)
if (changed & FPCR_FZ16) {
bool ftz_enabled = val & FPCR_FZ16;
set_flush_to_zero(ftz_enabled, &env->vfp.fp_status_f16);
+ set_flush_to_zero(ftz_enabled, &env->vfp.standard_fp_status_f16);
set_flush_inputs_to_zero(ftz_enabled, &env->vfp.fp_status_f16);
+ set_flush_inputs_to_zero(ftz_enabled, &env->vfp.standard_fp_status_f16);
}
if (changed & FPCR_FZ) {
bool ftz_enabled = val & FPCR_FZ;
@@ -146,6 +150,7 @@ static void vfp_set_fpscr_to_host(CPUARMState *env, uint32_t val)
set_float_exception_flags(i, &env->vfp.fp_status);
set_float_exception_flags(0, &env->vfp.fp_status_f16);
set_float_exception_flags(0, &env->vfp.standard_fp_status);
+ set_float_exception_flags(0, &env->vfp.standard_fp_status_f16);
}
#else
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* [PULL 27/27] target/arm: Use correct FPST for VCMLA, VCADD on fp16
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (25 preceding siblings ...)
2020-08-24 9:48 ` [PULL 26/27] target/arm: Implement FPST_STD_F16 fpstatus Peter Maydell
@ 2020-08-24 9:48 ` Peter Maydell
2020-08-24 13:53 ` [PULL 00/27] target-arm queue Peter Maydell
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 9:48 UTC (permalink / raw)
To: qemu-devel
When we implemented the VCMLA and VCADD insns we put in the
code to handle fp16, but left it using the standard fp status
flags. Correct them to use FPST_STD_F16 for fp16 operations.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200806104453.30393-5-peter.maydell@linaro.org
---
target/arm/translate-neon.c.inc | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/target/arm/translate-neon.c.inc b/target/arm/translate-neon.c.inc
index f8f2cb1869f..9879731a521 100644
--- a/target/arm/translate-neon.c.inc
+++ b/target/arm/translate-neon.c.inc
@@ -181,7 +181,7 @@ static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a)
}
opr_sz = (1 + a->q) * 8;
- fpst = fpstatus_ptr(FPST_STD);
+ fpst = fpstatus_ptr(a->size == 0 ? FPST_STD_F16 : FPST_STD);
fn_gvec_ptr = a->size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah;
tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
@@ -218,7 +218,7 @@ static bool trans_VCADD(DisasContext *s, arg_VCADD *a)
}
opr_sz = (1 + a->q) * 8;
- fpst = fpstatus_ptr(FPST_STD);
+ fpst = fpstatus_ptr(a->size == 0 ? FPST_STD_F16 : FPST_STD);
fn_gvec_ptr = a->size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh;
tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
@@ -322,7 +322,7 @@ static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a)
fn_gvec_ptr = (a->size ? gen_helper_gvec_fcmlas_idx
: gen_helper_gvec_fcmlah_idx);
opr_sz = (1 + a->q) * 8;
- fpst = fpstatus_ptr(FPST_STD);
+ fpst = fpstatus_ptr(a->size == 0 ? FPST_STD_F16 : FPST_STD);
tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn),
vfp_reg_offset(1, a->vm),
--
2.20.1
^ permalink raw reply related [flat|nested] 29+ messages in thread
* Re: [PULL 00/27] target-arm queue
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
` (26 preceding siblings ...)
2020-08-24 9:48 ` [PULL 27/27] target/arm: Use correct FPST for VCMLA, VCADD on fp16 Peter Maydell
@ 2020-08-24 13:53 ` Peter Maydell
27 siblings, 0 replies; 29+ messages in thread
From: Peter Maydell @ 2020-08-24 13:53 UTC (permalink / raw)
To: QEMU Developers
On Mon, 24 Aug 2020 at 10:48, Peter Maydell <peter.maydell@linaro.org> wrote:
>
> First arm pullreq for 5.2: Eric's SMMU stuff, and a bunch of
> cleanup/refactoring from me.
>
> thanks
> -- PMM
>
> The following changes since commit 8367a77c4d3f6e1e60890f5510304feb2c621611:
>
> Merge remote-tracking branch 'remotes/vivier2/tags/linux-user-for-5.2-pull-request' into staging (2020-08-23 16:34:43 +0100)
>
> are available in the Git repository at:
>
> https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20200824
>
> for you to fetch changes up to b34aa5129e9c3aff890b4f4bcc84962e94185629:
>
> target/arm: Use correct FPST for VCMLA, VCADD on fp16 (2020-08-24 10:15:12 +0100)
>
> ----------------------------------------------------------------
> target-arm queue:
> * hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores
> * hw/arm/smmuv3: Implement SMMUv3.2 range-invalidation
> * docs/system/arm: Document the Xilinx Versal Virt board
> * target/arm: Make M-profile NOCP take precedence over UNDEF
> * target/arm: Use correct FPST for VCMLA, VCADD on fp16
> * target/arm: Various cleanups preparing for fp16 support
Applied, thanks.
Please update the changelog at https://wiki.qemu.org/ChangeLog/5.2
for any user-visible changes.
-- PMM
^ permalink raw reply [flat|nested] 29+ messages in thread
end of thread, other threads:[~2020-08-24 13:53 UTC | newest]
Thread overview: 29+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-08-24 9:47 [PULL 00/27] target-arm queue Peter Maydell
2020-08-24 9:47 ` [PULL 01/27] hw/cpu/a9mpcore: Verify the machine use Cortex-A9 cores Peter Maydell
2020-08-24 9:47 ` [PULL 02/27] hw/arm/smmu-common: Factorize some code in smmu_ptw_64() Peter Maydell
2020-08-24 9:47 ` [PULL 03/27] hw/arm/smmu-common: Add IOTLB helpers Peter Maydell
2020-08-24 9:47 ` [PULL 04/27] hw/arm/smmu: Introduce smmu_get_iotlb_key() Peter Maydell
2020-08-24 9:47 ` [PULL 05/27] hw/arm/smmu: Introduce SMMUTLBEntry for PTW and IOTLB value Peter Maydell
2020-08-24 9:47 ` [PULL 06/27] hw/arm/smmu-common: Manage IOTLB block entries Peter Maydell
2020-08-24 9:47 ` [PULL 07/27] hw/arm/smmuv3: Introduce smmuv3_s1_range_inval() helper Peter Maydell
2020-08-24 9:47 ` [PULL 08/27] hw/arm/smmuv3: Get prepared for range invalidation Peter Maydell
2020-08-24 9:47 ` [PULL 09/27] hw/arm/smmuv3: Fix IIDR offset Peter Maydell
2020-08-24 9:47 ` [PULL 10/27] hw/arm/smmuv3: Let AIDR advertise SMMUv3.0 support Peter Maydell
2020-08-24 9:47 ` [PULL 11/27] hw/arm/smmuv3: Support HAD and advertise SMMUv3.1 support Peter Maydell
2020-08-24 9:47 ` [PULL 12/27] hw/arm/smmuv3: Advertise SMMUv3.2 range invalidation Peter Maydell
2020-08-24 9:47 ` [PULL 13/27] docs/system/arm: Document the Xilinx Versal Virt board Peter Maydell
2020-08-24 9:47 ` [PULL 14/27] target/arm: Pull handling of XScale insns out of disas_coproc_insn() Peter Maydell
2020-08-24 9:47 ` [PULL 15/27] target/arm: Separate decode from handling of coproc insns Peter Maydell
2020-08-24 9:48 ` [PULL 16/27] target/arm: Convert A32 coprocessor insns to decodetree Peter Maydell
2020-08-24 9:48 ` [PULL 17/27] target/arm: Tidy up disas_arm_insn() Peter Maydell
2020-08-24 9:48 ` [PULL 18/27] target/arm: Do M-profile NOCP checks early and via decodetree Peter Maydell
2020-08-24 9:48 ` [PULL 19/27] target/arm: Convert T32 coprocessor insns to decodetree Peter Maydell
2020-08-24 9:48 ` [PULL 20/27] target/arm: Remove ARCH macro Peter Maydell
2020-08-24 9:48 ` [PULL 21/27] target/arm: Delete unused VFP_DREG macros Peter Maydell
2020-08-24 9:48 ` [PULL 22/27] target/arm/translate.c: Delete/amend incorrect comments Peter Maydell
2020-08-24 9:48 ` [PULL 23/27] target/arm: Delete unused ARM_FEATURE_CRC Peter Maydell
2020-08-24 9:48 ` [PULL 24/27] target/arm: Replace A64 get_fpstatus_ptr() with generic fpstatus_ptr() Peter Maydell
2020-08-24 9:48 ` [PULL 25/27] target/arm: Make A32/T32 use new fpstatus_ptr() API Peter Maydell
2020-08-24 9:48 ` [PULL 26/27] target/arm: Implement FPST_STD_F16 fpstatus Peter Maydell
2020-08-24 9:48 ` [PULL 27/27] target/arm: Use correct FPST for VCMLA, VCADD on fp16 Peter Maydell
2020-08-24 13:53 ` [PULL 00/27] target-arm queue Peter Maydell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).