|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-changelog] [xen staging] x86emul: support AVX512F fused-multiply-add insns
commit d49dbd3266c0fda9bada669d72526cc33c48b761
Author: Jan Beulich <jbeulich@xxxxxxxx>
AuthorDate: Tue Nov 20 15:10:30 2018 +0100
Commit: Jan Beulich <jbeulich@xxxxxxxx>
CommitDate: Tue Nov 20 15:10:30 2018 +0100
x86emul: support AVX512F fused-multiply-add insns
Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
Acked-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
---
tools/tests/x86_emulator/evex-disp8.c | 30 +++++++++++
xen/arch/x86/x86_emulate/x86_emulate.c | 91 +++++++++++++++++++++++++---------
2 files changed, 97 insertions(+), 24 deletions(-)
diff --git a/tools/tests/x86_emulator/evex-disp8.c
b/tools/tests/x86_emulator/evex-disp8.c
index 9263bcd575..b92851b8c4 100644
--- a/tools/tests/x86_emulator/evex-disp8.c
+++ b/tools/tests/x86_emulator/evex-disp8.c
@@ -107,6 +107,36 @@ static const struct test avx512f_all[] = {
INSN_FP(add, 0f, 58),
INSN_FP(cmp, 0f, c2),
INSN_FP(div, 0f, 5e),
+ INSN(fmadd132, 66, 0f38, 98, vl, sd, vl),
+ INSN(fmadd132, 66, 0f38, 99, el, sd, el),
+ INSN(fmadd213, 66, 0f38, a8, vl, sd, vl),
+ INSN(fmadd213, 66, 0f38, a9, el, sd, el),
+ INSN(fmadd231, 66, 0f38, b8, vl, sd, vl),
+ INSN(fmadd231, 66, 0f38, b9, el, sd, el),
+ INSN(fmaddsub132, 66, 0f38, 96, vl, sd, vl),
+ INSN(fmaddsub213, 66, 0f38, a6, vl, sd, vl),
+ INSN(fmaddsub231, 66, 0f38, b6, vl, sd, vl),
+ INSN(fmsub132, 66, 0f38, 9a, vl, sd, vl),
+ INSN(fmsub132, 66, 0f38, 9b, el, sd, el),
+ INSN(fmsub213, 66, 0f38, aa, vl, sd, vl),
+ INSN(fmsub213, 66, 0f38, ab, el, sd, el),
+ INSN(fmsub231, 66, 0f38, ba, vl, sd, vl),
+ INSN(fmsub231, 66, 0f38, bb, el, sd, el),
+ INSN(fmsubadd132, 66, 0f38, 97, vl, sd, vl),
+ INSN(fmsubadd213, 66, 0f38, a7, vl, sd, vl),
+ INSN(fmsubadd231, 66, 0f38, b7, vl, sd, vl),
+ INSN(fnmadd132, 66, 0f38, 9c, vl, sd, vl),
+ INSN(fnmadd132, 66, 0f38, 9d, el, sd, el),
+ INSN(fnmadd213, 66, 0f38, ac, vl, sd, vl),
+ INSN(fnmadd213, 66, 0f38, ad, el, sd, el),
+ INSN(fnmadd231, 66, 0f38, bc, vl, sd, vl),
+ INSN(fnmadd231, 66, 0f38, bd, el, sd, el),
+ INSN(fnmsub132, 66, 0f38, 9e, vl, sd, vl),
+ INSN(fnmsub132, 66, 0f38, 9f, el, sd, el),
+ INSN(fnmsub213, 66, 0f38, ae, vl, sd, vl),
+ INSN(fnmsub213, 66, 0f38, af, el, sd, el),
+ INSN(fnmsub231, 66, 0f38, be, vl, sd, vl),
+ INSN(fnmsub231, 66, 0f38, bf, el, sd, el),
INSN_FP(max, 0f, 5f),
INSN_FP(min, 0f, 5d),
INSN_SFP(mov, 0f, 10),
diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c
b/xen/arch/x86/x86_emulate/x86_emulate.c
index 7e8a8e23bf..773b971369 100644
--- a/xen/arch/x86/x86_emulate/x86_emulate.c
+++ b/xen/arch/x86/x86_emulate/x86_emulate.c
@@ -452,30 +452,30 @@ static const struct ext0f38_table {
[0x8c] = { .simd_size = simd_packed_int },
[0x8e] = { .simd_size = simd_packed_int, .to_mem = 1 },
[0x90 ... 0x93] = { .simd_size = simd_other, .vsib = 1 },
- [0x96 ... 0x98] = { .simd_size = simd_packed_fp },
- [0x99] = { .simd_size = simd_scalar_vexw },
- [0x9a] = { .simd_size = simd_packed_fp },
- [0x9b] = { .simd_size = simd_scalar_vexw },
- [0x9c] = { .simd_size = simd_packed_fp },
- [0x9d] = { .simd_size = simd_scalar_vexw },
- [0x9e] = { .simd_size = simd_packed_fp },
- [0x9f] = { .simd_size = simd_scalar_vexw },
- [0xa6 ... 0xa8] = { .simd_size = simd_packed_fp },
- [0xa9] = { .simd_size = simd_scalar_vexw },
- [0xaa] = { .simd_size = simd_packed_fp },
- [0xab] = { .simd_size = simd_scalar_vexw },
- [0xac] = { .simd_size = simd_packed_fp },
- [0xad] = { .simd_size = simd_scalar_vexw },
- [0xae] = { .simd_size = simd_packed_fp },
- [0xaf] = { .simd_size = simd_scalar_vexw },
- [0xb6 ... 0xb8] = { .simd_size = simd_packed_fp },
- [0xb9] = { .simd_size = simd_scalar_vexw },
- [0xba] = { .simd_size = simd_packed_fp },
- [0xbb] = { .simd_size = simd_scalar_vexw },
- [0xbc] = { .simd_size = simd_packed_fp },
- [0xbd] = { .simd_size = simd_scalar_vexw },
- [0xbe] = { .simd_size = simd_packed_fp },
- [0xbf] = { .simd_size = simd_scalar_vexw },
+ [0x96 ... 0x98] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0x99] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0x9a] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0x9b] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0x9c] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0x9d] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0x9e] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0x9f] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xa6 ... 0xa8] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xa9] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xaa] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xab] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xac] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xad] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xae] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xaf] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xb6 ... 0xb8] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xb9] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xba] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xbb] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xbc] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xbd] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
+ [0xbe] = { .simd_size = simd_packed_fp, .d8s = d8s_vl },
+ [0xbf] = { .simd_size = simd_scalar_vexw, .d8s = d8s_dq },
[0xc8 ... 0xcd] = { .simd_size = simd_other },
[0xdb] = { .simd_size = simd_packed_int, .two_op = 1 },
[0xdc ... 0xdf] = { .simd_size = simd_packed_int },
@@ -8301,6 +8301,49 @@ x86_emulate(
host_and_vcpu_must_have(fma);
goto simd_0f_ymm;
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x96): /* vfmaddsub132p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x97): /* vfmsubadd132p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x98): /* vfmadd132p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x9a): /* vfmsub132p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x9c): /* vfnmadd132p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x9e): /* vfnmsub132p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xa6): /* vfmaddsub213p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xa7): /* vfmsubadd213p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xa8): /* vfmadd213p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xaa): /* vfmsub213p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xac): /* vfnmadd213p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xae): /* vfnmsub213p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xb6): /* vfmaddsub231p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xb7): /* vfmsubadd231p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xb8): /* vfmadd231p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xba): /* vfmsub231p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xbc): /* vfnmadd231p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xbe): /* vfnmsub231p{s,d}
[xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+ host_and_vcpu_must_have(avx512f);
+ if ( ea.type == OP_MEM || !evex.br )
+ avx512_vlen_check(false);
+ goto simd_zmm;
+
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x99): /* vfmadd132s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x9b): /* vfmsub132s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x9d): /* vfnmadd132s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0x9f): /* vfnmsub132s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xa9): /* vfmadd213s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xab): /* vfmsub213s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xad): /* vfnmadd213s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xaf): /* vfnmsub213s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xb9): /* vfmadd231s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xbb): /* vfmsub231s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xbd): /* vfnmadd231s{s,d}
xmm/mem,xmm,xmm{k} */
+ case X86EMUL_OPC_EVEX_66(0x0f38, 0xbf): /* vfnmsub231s{s,d}
xmm/mem,xmm,xmm{k} */
+ host_and_vcpu_must_have(avx512f);
+ if ( ea.type == OP_MEM )
+ {
+ generate_exception_if(evex.br, EXC_UD);
+ avx512_vlen_check(true);
+ }
+ goto simd_zmm;
+
case X86EMUL_OPC(0x0f38, 0xc8): /* sha1nexte xmm/m128,xmm */
case X86EMUL_OPC(0x0f38, 0xc9): /* sha1msg1 xmm/m128,xmm */
case X86EMUL_OPC(0x0f38, 0xca): /* sha1msg2 xmm/m128,xmm */
--
generated by git-patchbot for /home/xen/git/xen.git#staging
_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/xen-changelog
|
 |
Lists.xenproject.org is hosted with RackSpace, monitoring our |