aboutsummaryrefslogtreecommitdiff
path: root/target/i386/tcg
diff options
context:
space:
mode:
Diffstat (limited to 'target/i386/tcg')
-rw-r--r--target/i386/tcg/decode-new.c.inc40
-rw-r--r--target/i386/tcg/decode-new.h1
-rw-r--r--target/i386/tcg/emit.c.inc51
-rw-r--r--target/i386/tcg/translate.c1
4 files changed, 93 insertions, 0 deletions
diff --git a/target/i386/tcg/decode-new.c.inc b/target/i386/tcg/decode-new.c.inc
index 8baee90..e4878b9 100644
--- a/target/i386/tcg/decode-new.c.inc
+++ b/target/i386/tcg/decode-new.c.inc
@@ -376,6 +376,16 @@ static const X86OpEntry opcodes_0F38_00toEF[240] = {
[0x92] = X86_OP_ENTRY3(VPGATHERD, V,x, H,x, M,d, vex12 cpuid(AVX2) p_66), /* vgatherdps/d */
[0x93] = X86_OP_ENTRY3(VPGATHERQ, V,x, H,x, M,q, vex12 cpuid(AVX2) p_66), /* vgatherqps/d */
+ /* Should be exception type 2 but they do not have legacy SSE equivalents? */
+ [0x96] = X86_OP_ENTRY3(VFMADDSUB132Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x97] = X86_OP_ENTRY3(VFMSUBADD132Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+
+ [0xa6] = X86_OP_ENTRY3(VFMADDSUB213Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xa7] = X86_OP_ENTRY3(VFMSUBADD213Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+
+ [0xb6] = X86_OP_ENTRY3(VFMADDSUB231Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xb7] = X86_OP_ENTRY3(VFMSUBADD231Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+
[0x08] = X86_OP_ENTRY3(PSIGNB, V,x, H,x, W,x, vex4 cpuid(SSSE3) mmx avx2_256 p_00_66),
[0x09] = X86_OP_ENTRY3(PSIGNW, V,x, H,x, W,x, vex4 cpuid(SSSE3) mmx avx2_256 p_00_66),
[0x0a] = X86_OP_ENTRY3(PSIGND, V,x, H,x, W,x, vex4 cpuid(SSSE3) mmx avx2_256 p_00_66),
@@ -421,6 +431,34 @@ static const X86OpEntry opcodes_0F38_00toEF[240] = {
[0x8c] = X86_OP_ENTRY3(VPMASKMOV, V,x, H,x, WM,x, vex6 cpuid(AVX2) p_66),
[0x8e] = X86_OP_ENTRY3(VPMASKMOV_st, M,x, V,x, H,x, vex6 cpuid(AVX2) p_66),
+ /* Should be exception type 2 or 3 but they do not have legacy SSE equivalents? */
+ [0x98] = X86_OP_ENTRY3(VFMADD132Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x99] = X86_OP_ENTRY3(VFMADD132Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x9a] = X86_OP_ENTRY3(VFMSUB132Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x9b] = X86_OP_ENTRY3(VFMSUB132Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x9c] = X86_OP_ENTRY3(VFNMADD132Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x9d] = X86_OP_ENTRY3(VFNMADD132Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x9e] = X86_OP_ENTRY3(VFNMSUB132Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0x9f] = X86_OP_ENTRY3(VFNMSUB132Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+
+ [0xa8] = X86_OP_ENTRY3(VFMADD213Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xa9] = X86_OP_ENTRY3(VFMADD213Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xaa] = X86_OP_ENTRY3(VFMSUB213Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xab] = X86_OP_ENTRY3(VFMSUB213Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xac] = X86_OP_ENTRY3(VFNMADD213Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xad] = X86_OP_ENTRY3(VFNMADD213Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xae] = X86_OP_ENTRY3(VFNMSUB213Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xaf] = X86_OP_ENTRY3(VFNMSUB213Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+
+ [0xb8] = X86_OP_ENTRY3(VFMADD231Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xb9] = X86_OP_ENTRY3(VFMADD231Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xba] = X86_OP_ENTRY3(VFMSUB231Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xbb] = X86_OP_ENTRY3(VFMSUB231Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xbc] = X86_OP_ENTRY3(VFNMADD231Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xbd] = X86_OP_ENTRY3(VFNMADD231Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xbe] = X86_OP_ENTRY3(VFNMSUB231Px, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+ [0xbf] = X86_OP_ENTRY3(VFNMSUB231Sx, V,x, H,x, W,x, vex6 cpuid(FMA) p_66),
+
[0xdb] = X86_OP_ENTRY3(VAESIMC, V,dq, None,None, W,dq, vex4 cpuid(AES) p_66),
[0xdc] = X86_OP_ENTRY3(VAESENC, V,x, H,x, W,x, vex4 cpuid(AES) p_66),
[0xdd] = X86_OP_ENTRY3(VAESENCLAST, V,x, H,x, W,x, vex4 cpuid(AES) p_66),
@@ -1350,6 +1388,8 @@ static bool has_cpuid_feature(DisasContext *s, X86CPUIDFeature cpuid)
return true;
case X86_FEAT_F16C:
return (s->cpuid_ext_features & CPUID_EXT_F16C);
+ case X86_FEAT_FMA:
+ return (s->cpuid_ext_features & CPUID_EXT_FMA);
case X86_FEAT_MOVBE:
return (s->cpuid_ext_features & CPUID_EXT_MOVBE);
case X86_FEAT_PCLMULQDQ:
diff --git a/target/i386/tcg/decode-new.h b/target/i386/tcg/decode-new.h
index 0ef5462..cb6b8bc 100644
--- a/target/i386/tcg/decode-new.h
+++ b/target/i386/tcg/decode-new.h
@@ -105,6 +105,7 @@ typedef enum X86CPUIDFeature {
X86_FEAT_BMI1,
X86_FEAT_BMI2,
X86_FEAT_F16C,
+ X86_FEAT_FMA,
X86_FEAT_MOVBE,
X86_FEAT_PCLMULQDQ,
X86_FEAT_SSE,
diff --git a/target/i386/tcg/emit.c.inc b/target/i386/tcg/emit.c.inc
index 9334f09..7037ff9 100644
--- a/target/i386/tcg/emit.c.inc
+++ b/target/i386/tcg/emit.c.inc
@@ -39,6 +39,11 @@ typedef void (*SSEFunc_0_eppt)(TCGv_ptr env, TCGv_ptr reg_a, TCGv_ptr reg_b,
TCGv val);
typedef void (*SSEFunc_0_epppti)(TCGv_ptr env, TCGv_ptr reg_a, TCGv_ptr reg_b,
TCGv_ptr reg_c, TCGv a0, TCGv_i32 scale);
+typedef void (*SSEFunc_0_eppppi)(TCGv_ptr env, TCGv_ptr reg_a, TCGv_ptr reg_b,
+ TCGv_ptr reg_c, TCGv_ptr reg_d, TCGv_i32 flags);
+typedef void (*SSEFunc_0_eppppii)(TCGv_ptr env, TCGv_ptr reg_a, TCGv_ptr reg_b,
+ TCGv_ptr reg_c, TCGv_ptr reg_d, TCGv_i32 even,
+ TCGv_i32 odd);
static inline TCGv_i32 tcg_constant8u_i32(uint8_t val)
{
@@ -491,6 +496,52 @@ FP_SSE(VMIN, min)
FP_SSE(VDIV, div)
FP_SSE(VMAX, max)
+#define FMA_SSE_PACKED(uname, ptr0, ptr1, ptr2, even, odd) \
+static void gen_##uname##Px(DisasContext *s, CPUX86State *env, X86DecodedInsn *decode) \
+{ \
+ SSEFunc_0_eppppii xmm = s->vex_w ? gen_helper_fma4pd_xmm : gen_helper_fma4ps_xmm; \
+ SSEFunc_0_eppppii ymm = s->vex_w ? gen_helper_fma4pd_ymm : gen_helper_fma4ps_ymm; \
+ SSEFunc_0_eppppii fn = s->vex_l ? ymm : xmm; \
+ \
+ fn(cpu_env, OP_PTR0, ptr0, ptr1, ptr2, \
+ tcg_constant_i32(even), \
+ tcg_constant_i32((even) ^ (odd))); \
+}
+
+#define FMA_SSE(uname, ptr0, ptr1, ptr2, flags) \
+FMA_SSE_PACKED(uname, ptr0, ptr1, ptr2, flags, flags) \
+static void gen_##uname##Sx(DisasContext *s, CPUX86State *env, X86DecodedInsn *decode) \
+{ \
+ SSEFunc_0_eppppi fn = s->vex_w ? gen_helper_fma4sd : gen_helper_fma4ss; \
+ \
+ fn(cpu_env, OP_PTR0, ptr0, ptr1, ptr2, \
+ tcg_constant_i32(flags)); \
+} \
+
+FMA_SSE(VFMADD231, OP_PTR1, OP_PTR2, OP_PTR0, 0)
+FMA_SSE(VFMADD213, OP_PTR1, OP_PTR0, OP_PTR2, 0)
+FMA_SSE(VFMADD132, OP_PTR0, OP_PTR2, OP_PTR1, 0)
+
+FMA_SSE(VFNMADD231, OP_PTR1, OP_PTR2, OP_PTR0, float_muladd_negate_product)
+FMA_SSE(VFNMADD213, OP_PTR1, OP_PTR0, OP_PTR2, float_muladd_negate_product)
+FMA_SSE(VFNMADD132, OP_PTR0, OP_PTR2, OP_PTR1, float_muladd_negate_product)
+
+FMA_SSE(VFMSUB231, OP_PTR1, OP_PTR2, OP_PTR0, float_muladd_negate_c)
+FMA_SSE(VFMSUB213, OP_PTR1, OP_PTR0, OP_PTR2, float_muladd_negate_c)
+FMA_SSE(VFMSUB132, OP_PTR0, OP_PTR2, OP_PTR1, float_muladd_negate_c)
+
+FMA_SSE(VFNMSUB231, OP_PTR1, OP_PTR2, OP_PTR0, float_muladd_negate_c|float_muladd_negate_product)
+FMA_SSE(VFNMSUB213, OP_PTR1, OP_PTR0, OP_PTR2, float_muladd_negate_c|float_muladd_negate_product)
+FMA_SSE(VFNMSUB132, OP_PTR0, OP_PTR2, OP_PTR1, float_muladd_negate_c|float_muladd_negate_product)
+
+FMA_SSE_PACKED(VFMADDSUB231, OP_PTR1, OP_PTR2, OP_PTR0, float_muladd_negate_c, 0)
+FMA_SSE_PACKED(VFMADDSUB213, OP_PTR1, OP_PTR0, OP_PTR2, float_muladd_negate_c, 0)
+FMA_SSE_PACKED(VFMADDSUB132, OP_PTR0, OP_PTR2, OP_PTR1, float_muladd_negate_c, 0)
+
+FMA_SSE_PACKED(VFMSUBADD231, OP_PTR1, OP_PTR2, OP_PTR0, 0, float_muladd_negate_c)
+FMA_SSE_PACKED(VFMSUBADD213, OP_PTR1, OP_PTR0, OP_PTR2, 0, float_muladd_negate_c)
+FMA_SSE_PACKED(VFMSUBADD132, OP_PTR0, OP_PTR2, OP_PTR1, 0, float_muladd_negate_c)
+
#define FP_UNPACK_SSE(uname, lname) \
static void gen_##uname(DisasContext *s, CPUX86State *env, X86DecodedInsn *decode) \
{ \
diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c
index e19d5c1..85be2e5 100644
--- a/target/i386/tcg/translate.c
+++ b/target/i386/tcg/translate.c
@@ -26,6 +26,7 @@
#include "tcg/tcg-op-gvec.h"
#include "exec/cpu_ldst.h"
#include "exec/translator.h"
+#include "fpu/softfloat.h"
#include "exec/helper-proto.h"
#include "exec/helper-gen.h"
generated by cgit v1.1 at 2025-01-11 12:25:25 +0000