Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
>>> On 07.11.17 at 10:34,wrote: > My understanding is i need to implement x86 emulator for legacy and VEX > CPU features(GFNI,VAES and VPCLMULQDQ), right? Yes. > As for this patch, whether it is suitable for merge into Xen upstream > this time? At this time no in any event - the tree is frozen for 4.10. And once the tree re-opens, the requested emulator additions are then a prereq for the one here to go in. Jan ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
On Tue, Nov 07, 2017 at 01:11:02AM -0700, Jan Beulich wrote: > >>> On 07.11.17 at 07:28,wrote: > > For those new instructions, you mean i also need to support those > > three instructions(GFNI,VAES and VPCLMULQDQ) in x86_emulate() in PV? > > Why three instructions? And why PV? I'm afraid I'm confused, and > hence I'm afraid simply saying "yes" to your question might not be > enough. > > Jan Hello Jan, Sorry for my unclear answer! My understanding is i need to implement x86 emulator for legacy and VEX CPU features(GFNI,VAES and VPCLMULQDQ), right? As for this patch, whether it is suitable for merge into Xen upstream this time? x86 emulator patches for some intel's CPU features are in our plan and we will send those related patches in future. Regards, Yang ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
>>> On 07.11.17 at 07:28,wrote: > For those new instructions, you mean i also need to support those > three instructions(GFNI,VAES and VPCLMULQDQ) in x86_emulate() in PV? Why three instructions? And why PV? I'm afraid I'm confused, and hence I'm afraid simply saying "yes" to your question might not be enough. Jan ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
On Mon, Nov 06, 2017 at 03:39:45AM -0700, Jan Beulich wrote: > >>> On 27.10.17 at 16:18,wrote: > > Intel IceLake cpu has added new cpu features: AVX512VBMI2/GFNI/ > > VAES/AVX512VNNI/AVX512BITALG/VPCLMULQDQ. Those new cpu features > > need expose to guest.wq > > First of all, please don't forget to Cc relevant maintainers. > Thanks Jan's remind. > > The bit definition: > > CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512VBMI2 > > CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI > > CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES > > CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ > > These last three have VEX (and for GFNI even legacy) encodings. > While it wouldn't be reasonable yet to request EVEX encoded insn > support to be added to the emulator while enabling new ISA > extensions, I think legacy and VEX encoded ones should be taken > care of with the state the emulator is currently in. > > Jan Hello Jan, Thanks for reviewing my patch! For those new instructions, you mean i also need to support those three instructions(GFNI,VAES and VPCLMULQDQ) in x86_emulate() in PV? Many thanks! Regards, Yang ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
>>> On 27.10.17 at 16:18,wrote: > Intel IceLake cpu has added new cpu features: AVX512VBMI2/GFNI/ > VAES/AVX512VNNI/AVX512BITALG/VPCLMULQDQ. Those new cpu features > need expose to guest.wq First of all, please don't forget to Cc relevant maintainers. > The bit definition: > CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512VBMI2 > CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI > CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES > CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ These last three have VEX (and for GFNI even legacy) encodings. While it wouldn't be reasonable yet to request EVEX encoded insn support to be added to the emulator while enabling new ISA extensions, I think legacy and VEX encoded ones should be taken care of with the state the emulator is currently in. Jan ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
On Thu, Nov 02, 2017 at 08:59:20AM +0800, Zhong Yang wrote: > On Wed, Nov 01, 2017 at 03:29:16PM -0400, Konrad Rzeszutek Wilk wrote: > > On Fri, Oct 27, 2017 at 10:18:04PM +0800, Yang Zhong wrote: > > > Intel IceLake cpu has added new cpu features: AVX512VBMI2/GFNI/ > > > VAES/AVX512VNNI/AVX512BITALG/VPCLMULQDQ. Those new cpu features > > > need expose to guest.wq > > > > s/wq// > > Hello Konrad, > > Thanks for reviewing my patch, i will remove this .wq in next > version,thanks! Heh. I also did an review and it looked OK to me. You can also add Reviewed-by: Konrad Rzeszutek Wilk> > Regards, > > Yang > > > > > > > The bit definition: > > > CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512VBMI2 > > > CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI > > > CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES > > > CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ > > > CPUID.(EAX=7,ECX=0):ECX[bit 11] AVX512VNNI > > > CPUID.(EAX=7,ECX=0):ECX[bit 12] AVX512_BITALG > > > > > > The release document ref below link: > > > https://software.intel.com/sites/default/files/managed/c5/15/\ > > > architecture-instruction-set-extensions-programming-reference.pdf > > > > Ah! Thank you! > > > > > > Signed-off-by: Yang Zhong > > > --- > > > docs/man/xl.cfg.pod.5.in| 3 ++- > > > tools/libxl/libxl_cpuid.c | 6 ++ > > > tools/misc/xen-cpuid.c | 13 +++-- > > > xen/include/public/arch-x86/cpufeatureset.h | 6 ++ > > > xen/tools/gen-cpuid.py | 3 ++- > > > 5 files changed, 23 insertions(+), 8 deletions(-) > > > > > > diff --git a/docs/man/xl.cfg.pod.5.in b/docs/man/xl.cfg.pod.5.in > > > index b7b91d8..d056768 100644 > > > --- a/docs/man/xl.cfg.pod.5.in > > > +++ b/docs/man/xl.cfg.pod.5.in > > > @@ -1731,7 +1731,8 @@ perfctr_core perfctr_nb pge pku popcnt pse pse36 > > > psn rdrand rdseed rdtscp rtm > > > sha skinit smap smep smx ss sse sse2 sse3 sse4.1 sse4.2 sse4_1 sse4_2 > > > sse4a > > > ssse3 svm svm_decode svm_lbrv svm_npt svm_nrips svm_pausefilt svm_tscrate > > > svm_vmcbclean syscall sysenter tbm tm tm2 topoext tsc tsc-deadline > > > tsc_adjust > > > -umip vme vmx wdt x2apic xop xsave xtpr > > > +umip vme vmx wdt x2apic xop xsave xtpr avx512_vbmi2 gfni vaes vpclmulqdq > > > +avx512_vnni avx512_bitalg > > > > > > > > > The xend syntax is a list of values in the form of > > > diff --git a/tools/libxl/libxl_cpuid.c b/tools/libxl/libxl_cpuid.c > > > index e692b61..614991f 100644 > > > --- a/tools/libxl/libxl_cpuid.c > > > +++ b/tools/libxl/libxl_cpuid.c > > > @@ -199,6 +199,12 @@ int libxl_cpuid_parse_config(libxl_cpuid_policy_list > > > *cpuid, const char* str) > > > {"umip", 0x0007, 0, CPUID_REG_ECX, 2, 1}, > > > {"pku", 0x0007, 0, CPUID_REG_ECX, 3, 1}, > > > {"ospke",0x0007, 0, CPUID_REG_ECX, 4, 1}, > > > +{"avx512_vbmi2", 0x0007, 0, CPUID_REG_ECX, 6, 1}, > > > +{"gfni", 0x0007, 0, CPUID_REG_ECX, 8, 1}, > > > +{"vaes", 0x0007, 0, CPUID_REG_ECX, 9, 1}, > > > +{"vpclmulqdq", 0x0007, 0, CPUID_REG_ECX, 10, 1}, > > > +{"avx512_vnni", 0x0007, 0, CPUID_REG_ECX, 11, 1}, > > > +{"avx512_bitalg",0x0007, 0, CPUID_REG_ECX, 12, 1}, > > > > > > {"avx512-4vnniw",0x0007, 0, CPUID_REG_EDX, 2, 1}, > > > {"avx512-4fmaps",0x0007, 0, CPUID_REG_EDX, 3, 1}, > > > diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c > > > index 106be0f..985deea 100644 > > > --- a/tools/misc/xen-cpuid.c > > > +++ b/tools/misc/xen-cpuid.c > > > @@ -120,12 +120,13 @@ static const char *str_Da1[32] = > > > > > > static const char *str_7c0[32] = > > > { > > > -[ 0] = "prechwt1", [ 1] = "avx512vbmi", > > > -[ 2] = "REZ", [ 3] = "pku", > > > -[ 4] = "ospke", > > > - > > > -[5 ... 13] = "REZ", > > > - > > > +[ 0] = "prechwt1", [ 1] = "avx512vbmi", > > > +[ 2] = "REZ", [ 3] = "pku", > > > +[ 4] = "ospke",[ 5] = "REZ", > > > +[ 6] = "avx512_vbmi2", [ 7] = "REZ", > > > +[ 8] = "gfni", [ 9] = "vaes", > > > +[10] = "vpclmulqdq", [11] = "avx512_vnni", > > > +[12] = "avx512_bitalg",[13] = "REZ", > > > [14] = "avx512_vpopcntdq", > > > > > > [15 ... 31] = "REZ", > > > diff --git a/xen/include/public/arch-x86/cpufeatureset.h > > > b/xen/include/public/arch-x86/cpufeatureset.h > > > index 0ee3ea3..bb24b79 100644 > > > --- a/xen/include/public/arch-x86/cpufeatureset.h > > > +++ b/xen/include/public/arch-x86/cpufeatureset.h > > > @@ -228,6 +228,12 @@ XEN_CPUFEATURE(AVX512VBMI,6*32+ 1) /*A AVX-512 > > > Vector Byte Manipulation Ins > > > XEN_CPUFEATURE(UMIP, 6*32+ 2) /*S User Mode Instruction > > > Prevention */ > > > XEN_CPUFEATURE(PKU, 6*32+ 3) /*H Protection Keys for > > > Userspace */ > > >
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
On Wed, Nov 01, 2017 at 03:29:16PM -0400, Konrad Rzeszutek Wilk wrote: > On Fri, Oct 27, 2017 at 10:18:04PM +0800, Yang Zhong wrote: > > Intel IceLake cpu has added new cpu features: AVX512VBMI2/GFNI/ > > VAES/AVX512VNNI/AVX512BITALG/VPCLMULQDQ. Those new cpu features > > need expose to guest.wq > > s/wq// Hello Konrad, Thanks for reviewing my patch, i will remove this .wq in next version,thanks! Regards, Yang > > > > The bit definition: > > CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512VBMI2 > > CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI > > CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES > > CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ > > CPUID.(EAX=7,ECX=0):ECX[bit 11] AVX512VNNI > > CPUID.(EAX=7,ECX=0):ECX[bit 12] AVX512_BITALG > > > > The release document ref below link: > > https://software.intel.com/sites/default/files/managed/c5/15/\ > > architecture-instruction-set-extensions-programming-reference.pdf > > Ah! Thank you! > > > > Signed-off-by: Yang Zhong> > --- > > docs/man/xl.cfg.pod.5.in| 3 ++- > > tools/libxl/libxl_cpuid.c | 6 ++ > > tools/misc/xen-cpuid.c | 13 +++-- > > xen/include/public/arch-x86/cpufeatureset.h | 6 ++ > > xen/tools/gen-cpuid.py | 3 ++- > > 5 files changed, 23 insertions(+), 8 deletions(-) > > > > diff --git a/docs/man/xl.cfg.pod.5.in b/docs/man/xl.cfg.pod.5.in > > index b7b91d8..d056768 100644 > > --- a/docs/man/xl.cfg.pod.5.in > > +++ b/docs/man/xl.cfg.pod.5.in > > @@ -1731,7 +1731,8 @@ perfctr_core perfctr_nb pge pku popcnt pse pse36 psn > > rdrand rdseed rdtscp rtm > > sha skinit smap smep smx ss sse sse2 sse3 sse4.1 sse4.2 sse4_1 sse4_2 sse4a > > ssse3 svm svm_decode svm_lbrv svm_npt svm_nrips svm_pausefilt svm_tscrate > > svm_vmcbclean syscall sysenter tbm tm tm2 topoext tsc tsc-deadline > > tsc_adjust > > -umip vme vmx wdt x2apic xop xsave xtpr > > +umip vme vmx wdt x2apic xop xsave xtpr avx512_vbmi2 gfni vaes vpclmulqdq > > +avx512_vnni avx512_bitalg > > > > > > The xend syntax is a list of values in the form of > > diff --git a/tools/libxl/libxl_cpuid.c b/tools/libxl/libxl_cpuid.c > > index e692b61..614991f 100644 > > --- a/tools/libxl/libxl_cpuid.c > > +++ b/tools/libxl/libxl_cpuid.c > > @@ -199,6 +199,12 @@ int libxl_cpuid_parse_config(libxl_cpuid_policy_list > > *cpuid, const char* str) > > {"umip", 0x0007, 0, CPUID_REG_ECX, 2, 1}, > > {"pku", 0x0007, 0, CPUID_REG_ECX, 3, 1}, > > {"ospke",0x0007, 0, CPUID_REG_ECX, 4, 1}, > > +{"avx512_vbmi2", 0x0007, 0, CPUID_REG_ECX, 6, 1}, > > +{"gfni", 0x0007, 0, CPUID_REG_ECX, 8, 1}, > > +{"vaes", 0x0007, 0, CPUID_REG_ECX, 9, 1}, > > +{"vpclmulqdq", 0x0007, 0, CPUID_REG_ECX, 10, 1}, > > +{"avx512_vnni", 0x0007, 0, CPUID_REG_ECX, 11, 1}, > > +{"avx512_bitalg",0x0007, 0, CPUID_REG_ECX, 12, 1}, > > > > {"avx512-4vnniw",0x0007, 0, CPUID_REG_EDX, 2, 1}, > > {"avx512-4fmaps",0x0007, 0, CPUID_REG_EDX, 3, 1}, > > diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c > > index 106be0f..985deea 100644 > > --- a/tools/misc/xen-cpuid.c > > +++ b/tools/misc/xen-cpuid.c > > @@ -120,12 +120,13 @@ static const char *str_Da1[32] = > > > > static const char *str_7c0[32] = > > { > > -[ 0] = "prechwt1", [ 1] = "avx512vbmi", > > -[ 2] = "REZ", [ 3] = "pku", > > -[ 4] = "ospke", > > - > > -[5 ... 13] = "REZ", > > - > > +[ 0] = "prechwt1", [ 1] = "avx512vbmi", > > +[ 2] = "REZ", [ 3] = "pku", > > +[ 4] = "ospke",[ 5] = "REZ", > > +[ 6] = "avx512_vbmi2", [ 7] = "REZ", > > +[ 8] = "gfni", [ 9] = "vaes", > > +[10] = "vpclmulqdq", [11] = "avx512_vnni", > > +[12] = "avx512_bitalg",[13] = "REZ", > > [14] = "avx512_vpopcntdq", > > > > [15 ... 31] = "REZ", > > diff --git a/xen/include/public/arch-x86/cpufeatureset.h > > b/xen/include/public/arch-x86/cpufeatureset.h > > index 0ee3ea3..bb24b79 100644 > > --- a/xen/include/public/arch-x86/cpufeatureset.h > > +++ b/xen/include/public/arch-x86/cpufeatureset.h > > @@ -228,6 +228,12 @@ XEN_CPUFEATURE(AVX512VBMI,6*32+ 1) /*A AVX-512 > > Vector Byte Manipulation Ins > > XEN_CPUFEATURE(UMIP, 6*32+ 2) /*S User Mode Instruction > > Prevention */ > > XEN_CPUFEATURE(PKU, 6*32+ 3) /*H Protection Keys for Userspace > > */ > > XEN_CPUFEATURE(OSPKE, 6*32+ 4) /*! OS Protection Keys Enable */ > > +XEN_CPUFEATURE(AVX512_VBMI2, 6*32+ 6) /*A addition AVX-512 VBMI > > Instructions */ > > +XEN_CPUFEATURE(GFNI, 6*32+ 8) /*A Galois Field New Instructions > > */ > > +XEN_CPUFEATURE(VAES, 6*32+ 9) /*A Vector AES instructions */ > > +XEN_CPUFEATURE(VPCLMULQDQ,6*32+ 10) /*A vector PCLMULQDQ instructions > > */ > >
Re: [Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
On Fri, Oct 27, 2017 at 10:18:04PM +0800, Yang Zhong wrote: > Intel IceLake cpu has added new cpu features: AVX512VBMI2/GFNI/ > VAES/AVX512VNNI/AVX512BITALG/VPCLMULQDQ. Those new cpu features > need expose to guest.wq s/wq// > > The bit definition: > CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512VBMI2 > CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI > CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES > CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ > CPUID.(EAX=7,ECX=0):ECX[bit 11] AVX512VNNI > CPUID.(EAX=7,ECX=0):ECX[bit 12] AVX512_BITALG > > The release document ref below link: > https://software.intel.com/sites/default/files/managed/c5/15/\ > architecture-instruction-set-extensions-programming-reference.pdf Ah! Thank you! > > Signed-off-by: Yang Zhong> --- > docs/man/xl.cfg.pod.5.in| 3 ++- > tools/libxl/libxl_cpuid.c | 6 ++ > tools/misc/xen-cpuid.c | 13 +++-- > xen/include/public/arch-x86/cpufeatureset.h | 6 ++ > xen/tools/gen-cpuid.py | 3 ++- > 5 files changed, 23 insertions(+), 8 deletions(-) > > diff --git a/docs/man/xl.cfg.pod.5.in b/docs/man/xl.cfg.pod.5.in > index b7b91d8..d056768 100644 > --- a/docs/man/xl.cfg.pod.5.in > +++ b/docs/man/xl.cfg.pod.5.in > @@ -1731,7 +1731,8 @@ perfctr_core perfctr_nb pge pku popcnt pse pse36 psn > rdrand rdseed rdtscp rtm > sha skinit smap smep smx ss sse sse2 sse3 sse4.1 sse4.2 sse4_1 sse4_2 sse4a > ssse3 svm svm_decode svm_lbrv svm_npt svm_nrips svm_pausefilt svm_tscrate > svm_vmcbclean syscall sysenter tbm tm tm2 topoext tsc tsc-deadline tsc_adjust > -umip vme vmx wdt x2apic xop xsave xtpr > +umip vme vmx wdt x2apic xop xsave xtpr avx512_vbmi2 gfni vaes vpclmulqdq > +avx512_vnni avx512_bitalg > > > The xend syntax is a list of values in the form of > diff --git a/tools/libxl/libxl_cpuid.c b/tools/libxl/libxl_cpuid.c > index e692b61..614991f 100644 > --- a/tools/libxl/libxl_cpuid.c > +++ b/tools/libxl/libxl_cpuid.c > @@ -199,6 +199,12 @@ int libxl_cpuid_parse_config(libxl_cpuid_policy_list > *cpuid, const char* str) > {"umip", 0x0007, 0, CPUID_REG_ECX, 2, 1}, > {"pku", 0x0007, 0, CPUID_REG_ECX, 3, 1}, > {"ospke",0x0007, 0, CPUID_REG_ECX, 4, 1}, > +{"avx512_vbmi2", 0x0007, 0, CPUID_REG_ECX, 6, 1}, > +{"gfni", 0x0007, 0, CPUID_REG_ECX, 8, 1}, > +{"vaes", 0x0007, 0, CPUID_REG_ECX, 9, 1}, > +{"vpclmulqdq", 0x0007, 0, CPUID_REG_ECX, 10, 1}, > +{"avx512_vnni", 0x0007, 0, CPUID_REG_ECX, 11, 1}, > +{"avx512_bitalg",0x0007, 0, CPUID_REG_ECX, 12, 1}, > > {"avx512-4vnniw",0x0007, 0, CPUID_REG_EDX, 2, 1}, > {"avx512-4fmaps",0x0007, 0, CPUID_REG_EDX, 3, 1}, > diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c > index 106be0f..985deea 100644 > --- a/tools/misc/xen-cpuid.c > +++ b/tools/misc/xen-cpuid.c > @@ -120,12 +120,13 @@ static const char *str_Da1[32] = > > static const char *str_7c0[32] = > { > -[ 0] = "prechwt1", [ 1] = "avx512vbmi", > -[ 2] = "REZ", [ 3] = "pku", > -[ 4] = "ospke", > - > -[5 ... 13] = "REZ", > - > +[ 0] = "prechwt1", [ 1] = "avx512vbmi", > +[ 2] = "REZ", [ 3] = "pku", > +[ 4] = "ospke",[ 5] = "REZ", > +[ 6] = "avx512_vbmi2", [ 7] = "REZ", > +[ 8] = "gfni", [ 9] = "vaes", > +[10] = "vpclmulqdq", [11] = "avx512_vnni", > +[12] = "avx512_bitalg",[13] = "REZ", > [14] = "avx512_vpopcntdq", > > [15 ... 31] = "REZ", > diff --git a/xen/include/public/arch-x86/cpufeatureset.h > b/xen/include/public/arch-x86/cpufeatureset.h > index 0ee3ea3..bb24b79 100644 > --- a/xen/include/public/arch-x86/cpufeatureset.h > +++ b/xen/include/public/arch-x86/cpufeatureset.h > @@ -228,6 +228,12 @@ XEN_CPUFEATURE(AVX512VBMI,6*32+ 1) /*A AVX-512 > Vector Byte Manipulation Ins > XEN_CPUFEATURE(UMIP, 6*32+ 2) /*S User Mode Instruction Prevention > */ > XEN_CPUFEATURE(PKU, 6*32+ 3) /*H Protection Keys for Userspace */ > XEN_CPUFEATURE(OSPKE, 6*32+ 4) /*! OS Protection Keys Enable */ > +XEN_CPUFEATURE(AVX512_VBMI2, 6*32+ 6) /*A addition AVX-512 VBMI > Instructions */ > +XEN_CPUFEATURE(GFNI, 6*32+ 8) /*A Galois Field New Instructions */ > +XEN_CPUFEATURE(VAES, 6*32+ 9) /*A Vector AES instructions */ > +XEN_CPUFEATURE(VPCLMULQDQ,6*32+ 10) /*A vector PCLMULQDQ instructions */ > +XEN_CPUFEATURE(AVX512_VNNI, 6*32+ 11) /*A Vector Neural Network > Instructions */ > +XEN_CPUFEATURE(AVX512_BITALG, 6*32+ 12) /*A support for VPOPCNT[B,W] and > VPSHUFBITQMB*/ > XEN_CPUFEATURE(AVX512_VPOPCNTDQ, 6*32+14) /*A POPCNT for vectors of DW/QW */ > XEN_CPUFEATURE(RDPID, 6*32+22) /*A RDPID instruction */ > > diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py > index 9ec4486..be8df48
[Xen-devel] [PATCH] x86/cpuid: Enable new SSE/AVX/AVX512 cpu features
Intel IceLake cpu has added new cpu features: AVX512VBMI2/GFNI/ VAES/AVX512VNNI/AVX512BITALG/VPCLMULQDQ. Those new cpu features need expose to guest.wq The bit definition: CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512VBMI2 CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ CPUID.(EAX=7,ECX=0):ECX[bit 11] AVX512VNNI CPUID.(EAX=7,ECX=0):ECX[bit 12] AVX512_BITALG The release document ref below link: https://software.intel.com/sites/default/files/managed/c5/15/\ architecture-instruction-set-extensions-programming-reference.pdf Signed-off-by: Yang Zhong--- docs/man/xl.cfg.pod.5.in| 3 ++- tools/libxl/libxl_cpuid.c | 6 ++ tools/misc/xen-cpuid.c | 13 +++-- xen/include/public/arch-x86/cpufeatureset.h | 6 ++ xen/tools/gen-cpuid.py | 3 ++- 5 files changed, 23 insertions(+), 8 deletions(-) diff --git a/docs/man/xl.cfg.pod.5.in b/docs/man/xl.cfg.pod.5.in index b7b91d8..d056768 100644 --- a/docs/man/xl.cfg.pod.5.in +++ b/docs/man/xl.cfg.pod.5.in @@ -1731,7 +1731,8 @@ perfctr_core perfctr_nb pge pku popcnt pse pse36 psn rdrand rdseed rdtscp rtm sha skinit smap smep smx ss sse sse2 sse3 sse4.1 sse4.2 sse4_1 sse4_2 sse4a ssse3 svm svm_decode svm_lbrv svm_npt svm_nrips svm_pausefilt svm_tscrate svm_vmcbclean syscall sysenter tbm tm tm2 topoext tsc tsc-deadline tsc_adjust -umip vme vmx wdt x2apic xop xsave xtpr +umip vme vmx wdt x2apic xop xsave xtpr avx512_vbmi2 gfni vaes vpclmulqdq +avx512_vnni avx512_bitalg The xend syntax is a list of values in the form of diff --git a/tools/libxl/libxl_cpuid.c b/tools/libxl/libxl_cpuid.c index e692b61..614991f 100644 --- a/tools/libxl/libxl_cpuid.c +++ b/tools/libxl/libxl_cpuid.c @@ -199,6 +199,12 @@ int libxl_cpuid_parse_config(libxl_cpuid_policy_list *cpuid, const char* str) {"umip", 0x0007, 0, CPUID_REG_ECX, 2, 1}, {"pku", 0x0007, 0, CPUID_REG_ECX, 3, 1}, {"ospke",0x0007, 0, CPUID_REG_ECX, 4, 1}, +{"avx512_vbmi2", 0x0007, 0, CPUID_REG_ECX, 6, 1}, +{"gfni", 0x0007, 0, CPUID_REG_ECX, 8, 1}, +{"vaes", 0x0007, 0, CPUID_REG_ECX, 9, 1}, +{"vpclmulqdq", 0x0007, 0, CPUID_REG_ECX, 10, 1}, +{"avx512_vnni", 0x0007, 0, CPUID_REG_ECX, 11, 1}, +{"avx512_bitalg",0x0007, 0, CPUID_REG_ECX, 12, 1}, {"avx512-4vnniw",0x0007, 0, CPUID_REG_EDX, 2, 1}, {"avx512-4fmaps",0x0007, 0, CPUID_REG_EDX, 3, 1}, diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index 106be0f..985deea 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -120,12 +120,13 @@ static const char *str_Da1[32] = static const char *str_7c0[32] = { -[ 0] = "prechwt1", [ 1] = "avx512vbmi", -[ 2] = "REZ", [ 3] = "pku", -[ 4] = "ospke", - -[5 ... 13] = "REZ", - +[ 0] = "prechwt1", [ 1] = "avx512vbmi", +[ 2] = "REZ", [ 3] = "pku", +[ 4] = "ospke",[ 5] = "REZ", +[ 6] = "avx512_vbmi2", [ 7] = "REZ", +[ 8] = "gfni", [ 9] = "vaes", +[10] = "vpclmulqdq", [11] = "avx512_vnni", +[12] = "avx512_bitalg",[13] = "REZ", [14] = "avx512_vpopcntdq", [15 ... 31] = "REZ", diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 0ee3ea3..bb24b79 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -228,6 +228,12 @@ XEN_CPUFEATURE(AVX512VBMI,6*32+ 1) /*A AVX-512 Vector Byte Manipulation Ins XEN_CPUFEATURE(UMIP, 6*32+ 2) /*S User Mode Instruction Prevention */ XEN_CPUFEATURE(PKU, 6*32+ 3) /*H Protection Keys for Userspace */ XEN_CPUFEATURE(OSPKE, 6*32+ 4) /*! OS Protection Keys Enable */ +XEN_CPUFEATURE(AVX512_VBMI2, 6*32+ 6) /*A addition AVX-512 VBMI Instructions */ +XEN_CPUFEATURE(GFNI, 6*32+ 8) /*A Galois Field New Instructions */ +XEN_CPUFEATURE(VAES, 6*32+ 9) /*A Vector AES instructions */ +XEN_CPUFEATURE(VPCLMULQDQ,6*32+ 10) /*A vector PCLMULQDQ instructions */ +XEN_CPUFEATURE(AVX512_VNNI, 6*32+ 11) /*A Vector Neural Network Instructions */ +XEN_CPUFEATURE(AVX512_BITALG, 6*32+ 12) /*A support for VPOPCNT[B,W] and VPSHUFBITQMB*/ XEN_CPUFEATURE(AVX512_VPOPCNTDQ, 6*32+14) /*A POPCNT for vectors of DW/QW */ XEN_CPUFEATURE(RDPID, 6*32+22) /*A RDPID instruction */ diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py index 9ec4486..be8df48 100755 --- a/xen/tools/gen-cpuid.py +++ b/xen/tools/gen-cpuid.py @@ -255,7 +255,8 @@ def crunch_numbers(state): # top of AVX512F AVX512F: [AVX512DQ, AVX512IFMA, AVX512PF, AVX512ER, AVX512CD, AVX512BW, AVX512VL, AVX512VBMI, AVX512_4VNNIW, -