| Message ID | 1555609298-10498-1-git-send-email-Dave.Martin@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | arm64: Expose SVE2 features for userspace | expand |
On Thu, Apr 18, 2019 at 06:41:38PM +0100, Dave Martin wrote: > This patch provides support for reporting the presence of SVE2 and > its optional features to userspace. > > This will also enable visibility of SVE2 for guests, when KVM > support for SVE-enabled guests is available. > > Signed-off-by: Dave Martin <Dave.Martin@arm.com> Cheers, I'll pick this up for 5.2. Will
On Tue, Apr 23, 2019 at 04:32:18PM +0100, Will Deacon wrote: > On Thu, Apr 18, 2019 at 06:41:38PM +0100, Dave Martin wrote: > > This patch provides support for reporting the presence of SVE2 and > > its optional features to userspace. > > > > This will also enable visibility of SVE2 for guests, when KVM > > support for SVE-enabled guests is available. > > > > Signed-off-by: Dave Martin <Dave.Martin@arm.com> > > Cheers, I'll pick this up for 5.2. Thanks (assuming no objection from anyone else) ---Dave
On Thu, Apr 18, 2019 at 06:41:38PM +0100, Dave Martin wrote: > This patch provides support for reporting the presence of SVE2 and > its optional features to userspace. > > This will also enable visibility of SVE2 for guests, when KVM > support for SVE-enabled guests is available. > > Signed-off-by: Dave Martin <Dave.Martin@arm.com> > --- > > This patch is based on > > git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git > f645d37e281f ("Merge remote-tracking branch 'arm64/for-next/mmiowb' into for-next/core") > > (for Andrew Murray's HWCAP2 support). > > > Tested with the Arm Fast Model SVE2 plugin. > > For KVM, tested with: > > git://linux-arm.org/linux-dm.git sve-kvm-fixes/v2/head > http://linux-arm.org/git?p=linux-dm.git;a=shortlog;h=refs/heads/sve-kvm-fixes/v2/head > > (which is also now merged to kvmarm/next). > > Documentation/arm64/cpu-feature-registers.txt | 16 ++++++++++++++++ > Documentation/arm64/elf_hwcaps.txt | 24 ++++++++++++++++++++++++ > Documentation/arm64/sve.txt | 17 +++++++++++++++++ > arch/arm64/Kconfig | 3 +++ > arch/arm64/include/asm/hwcap.h | 6 ++++++ > arch/arm64/include/asm/sysreg.h | 14 ++++++++++++++ > arch/arm64/include/uapi/asm/hwcap.h | 6 ++++++ > arch/arm64/kernel/cpufeature.c | 17 ++++++++++++++++- > arch/arm64/kernel/cpuinfo.c | 6 ++++++ > 9 files changed, 108 insertions(+), 1 deletion(-) > > diff --git a/Documentation/arm64/cpu-feature-registers.txt b/Documentation/arm64/cpu-feature-registers.txt > index d4b4dd1f..4e47a01 100644 > --- a/Documentation/arm64/cpu-feature-registers.txt > +++ b/Documentation/arm64/cpu-feature-registers.txt > @@ -209,6 +209,22 @@ infrastructure: > | AT | [35-32] | y | > x--------------------------------------------------x > > + 6) ID_AA64ZFR0_EL1 - SVE feature ID register 0 > + > + x--------------------------------------------------x > + | Name | bits | visible | > + |--------------------------------------------------| > + | SM4 | [43-40] | y | > + x--------------------------------------------------x > + | SHA3 | [35-32] | y | > + x--------------------------------------------------x > + | BitPerm | [19-16] | y | > + x--------------------------------------------------x > + | AES | [7-4] | y | > + x--------------------------------------------------x > + | SVEVer | [3-0] | y | > + x--------------------------------------------------x Trivial nit: for consistency with all the other tables, please only use 'x' at the corners, and use '|' at the edges. Otherwise, this looks good to me; sorry for the late style nit! Thanks, Mark.
On Tue, Apr 23, 2019 at 05:28:52PM +0100, Mark Rutland wrote: > On Thu, Apr 18, 2019 at 06:41:38PM +0100, Dave Martin wrote: > > diff --git a/Documentation/arm64/cpu-feature-registers.txt b/Documentation/arm64/cpu-feature-registers.txt > > index d4b4dd1f..4e47a01 100644 > > --- a/Documentation/arm64/cpu-feature-registers.txt > > +++ b/Documentation/arm64/cpu-feature-registers.txt > > @@ -209,6 +209,22 @@ infrastructure: > > | AT | [35-32] | y | > > x--------------------------------------------------x > > > > + 6) ID_AA64ZFR0_EL1 - SVE feature ID register 0 > > + > > + x--------------------------------------------------x > > + | Name | bits | visible | > > + |--------------------------------------------------| > > + | SM4 | [43-40] | y | > > + x--------------------------------------------------x > > + | SHA3 | [35-32] | y | > > + x--------------------------------------------------x > > + | BitPerm | [19-16] | y | > > + x--------------------------------------------------x > > + | AES | [7-4] | y | > > + x--------------------------------------------------x > > + | SVEVer | [3-0] | y | > > + x--------------------------------------------------x > > Trivial nit: for consistency with all the other tables, please only use > 'x' at the corners, and use '|' at the edges. > > Otherwise, this looks good to me; sorry for the late style nit! I'll fix that up in my tree. Will
On Tue, Apr 23, 2019 at 05:30:17PM +0100, Will Deacon wrote: > On Tue, Apr 23, 2019 at 05:28:52PM +0100, Mark Rutland wrote: > > On Thu, Apr 18, 2019 at 06:41:38PM +0100, Dave Martin wrote: > > > diff --git a/Documentation/arm64/cpu-feature-registers.txt b/Documentation/arm64/cpu-feature-registers.txt > > > index d4b4dd1f..4e47a01 100644 > > > --- a/Documentation/arm64/cpu-feature-registers.txt > > > +++ b/Documentation/arm64/cpu-feature-registers.txt > > > @@ -209,6 +209,22 @@ infrastructure: > > > | AT | [35-32] | y | > > > x--------------------------------------------------x > > > > > > + 6) ID_AA64ZFR0_EL1 - SVE feature ID register 0 > > > + > > > + x--------------------------------------------------x > > > + | Name | bits | visible | > > > + |--------------------------------------------------| > > > + | SM4 | [43-40] | y | > > > + x--------------------------------------------------x > > > + | SHA3 | [35-32] | y | > > > + x--------------------------------------------------x > > > + | BitPerm | [19-16] | y | > > > + x--------------------------------------------------x > > > + | AES | [7-4] | y | > > > + x--------------------------------------------------x > > > + | SVEVer | [3-0] | y | > > > + x--------------------------------------------------x > > > > Trivial nit: for consistency with all the other tables, please only use > > 'x' at the corners, and use '|' at the edges. > > > > Otherwise, this looks good to me; sorry for the late style nit! > > I'll fix that up in my tree. Thanks. I didn't even spot the difference... Cheers ---Dave
diff --git a/Documentation/arm64/cpu-feature-registers.txt b/Documentation/arm64/cpu-feature-registers.txt index d4b4dd1f..4e47a01 100644 --- a/Documentation/arm64/cpu-feature-registers.txt +++ b/Documentation/arm64/cpu-feature-registers.txt @@ -209,6 +209,22 @@ infrastructure: | AT | [35-32] | y | x--------------------------------------------------x + 6) ID_AA64ZFR0_EL1 - SVE feature ID register 0 + + x--------------------------------------------------x + | Name | bits | visible | + |--------------------------------------------------| + | SM4 | [43-40] | y | + x--------------------------------------------------x + | SHA3 | [35-32] | y | + x--------------------------------------------------x + | BitPerm | [19-16] | y | + x--------------------------------------------------x + | AES | [7-4] | y | + x--------------------------------------------------x + | SVEVer | [3-0] | y | + x--------------------------------------------------x + Appendix I: Example --------------------------- diff --git a/Documentation/arm64/elf_hwcaps.txt b/Documentation/arm64/elf_hwcaps.txt index 55431fd..b73a251 100644 --- a/Documentation/arm64/elf_hwcaps.txt +++ b/Documentation/arm64/elf_hwcaps.txt @@ -163,6 +163,30 @@ HWCAP_SVE Functionality implied by ID_AA64PFR0_EL1.SVE == 0b0001. +HWCAP2_SVE2 + + Functionality implied by ID_AA64ZFR0_EL1.SVEVer == 0b0001. + +HWCAP2_SVEAES + + Functionality implied by ID_AA64ZFR0_EL1.AES == 0b0001. + +HWCAP2_SVEPMULL + + Functionality implied by ID_AA64ZFR0_EL1.AES == 0b0010. + +HWCAP2_SVEBITPERM + + Functionality implied by ID_AA64ZFR0_EL1.BitPerm == 0b0001. + +HWCAP2_SVESHA3 + + Functionality implied by ID_AA64ZFR0_EL1.SHA3 == 0b0001. + +HWCAP2_SVESM4 + + Functionality implied by ID_AA64ZFR0_EL1.SM4 == 0b0001. + HWCAP_ASIMDFHM Functionality implied by ID_AA64ISAR0_EL1.FHM == 0b0001. diff --git a/Documentation/arm64/sve.txt b/Documentation/arm64/sve.txt index 7169a0e..9940e92 100644 --- a/Documentation/arm64/sve.txt +++ b/Documentation/arm64/sve.txt @@ -34,6 +34,23 @@ model features for SVE is included in Appendix A. following sections: software that needs to verify that those interfaces are present must check for HWCAP_SVE instead. +* On hardware that supports the SVE2 extensions, HWCAP2_SVE2 will also + be reported in the AT_HWCAP2 aux vector entry. In addition to this, + optional extensions to SVE2 may be reported by the presence of: + + HWCAP2_SVE2 + HWCAP2_SVEAES + HWCAP2_SVEPMULL + HWCAP2_SVEBITPERM + HWCAP2_SVESHA3 + HWCAP2_SVESM4 + + This list may be extended over time as the SVE architecture evolves. + + These extensions are also reported via the CPU ID register ID_AA64ZFR0_EL1, + which userspace can read using an MRS instruction. See elf_hwcaps.txt and + cpu-feature-registers.txt for details. + * Debuggers should restrict themselves to interacting with the target via the NT_ARM_SVE regset. The recommended way of detecting support for this regset is to connect to a target process first and then attempt a diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c383625..4d79446 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1322,6 +1322,9 @@ config ARM64_SVE To enable use of this extension on CPUs that implement it, say Y. + On CPUs that support the SVE2 extensions, this option will enable + those too. + Note that for architectural reasons, firmware _must_ implement SVE support when running on SVE capable hardware. The required support is present in: diff --git a/arch/arm64/include/asm/hwcap.h b/arch/arm64/include/asm/hwcap.h index f78c86c..b4bfb66 100644 --- a/arch/arm64/include/asm/hwcap.h +++ b/arch/arm64/include/asm/hwcap.h @@ -89,6 +89,12 @@ #define __khwcap2_feature(x) (const_ilog2(HWCAP2_ ## x) + 32) #define KERNEL_HWCAP_DCPODP __khwcap2_feature(DCPODP) +#define KERNEL_HWCAP_SVE2 __khwcap2_feature(SVE2) +#define KERNEL_HWCAP_SVEAES __khwcap2_feature(SVEAES) +#define KERNEL_HWCAP_SVEPMULL __khwcap2_feature(SVEPMULL) +#define KERNEL_HWCAP_SVEBITPERM __khwcap2_feature(SVEBITPERM) +#define KERNEL_HWCAP_SVESHA3 __khwcap2_feature(SVESHA3) +#define KERNEL_HWCAP_SVESM4 __khwcap2_feature(SVESM4) /* * This yields a mask that user programs can use to figure out what diff --git a/arch/arm64/include/asm/sysreg.h b/arch/arm64/include/asm/sysreg.h index 5b267de..29c6559 100644 --- a/arch/arm64/include/asm/sysreg.h +++ b/arch/arm64/include/asm/sysreg.h @@ -606,6 +606,20 @@ #define ID_AA64PFR1_SSBS_PSTATE_ONLY 1 #define ID_AA64PFR1_SSBS_PSTATE_INSNS 2 +/* id_aa64zfr0 */ +#define ID_AA64ZFR0_SM4_SHIFT 40 +#define ID_AA64ZFR0_SHA3_SHIFT 32 +#define ID_AA64ZFR0_BITPERM_SHIFT 16 +#define ID_AA64ZFR0_AES_SHIFT 4 +#define ID_AA64ZFR0_SVEVER_SHIFT 0 + +#define ID_AA64ZFR0_SM4 0x1 +#define ID_AA64ZFR0_SHA3 0x1 +#define ID_AA64ZFR0_BITPERM 0x1 +#define ID_AA64ZFR0_AES 0x1 +#define ID_AA64ZFR0_AES_PMULL 0x2 +#define ID_AA64ZFR0_SVEVER_SVE2 0x1 + /* id_aa64mmfr0 */ #define ID_AA64MMFR0_TGRAN4_SHIFT 28 #define ID_AA64MMFR0_TGRAN64_SHIFT 24 diff --git a/arch/arm64/include/uapi/asm/hwcap.h b/arch/arm64/include/uapi/asm/hwcap.h index d64af39..1a772b1 100644 --- a/arch/arm64/include/uapi/asm/hwcap.h +++ b/arch/arm64/include/uapi/asm/hwcap.h @@ -57,5 +57,11 @@ * HWCAP2 flags - for AT_HWCAP2 */ #define HWCAP2_DCPODP (1 << 0) +#define HWCAP2_SVE2 (1 << 1) +#define HWCAP2_SVEAES (1 << 2) +#define HWCAP2_SVEPMULL (1 << 3) +#define HWCAP2_SVEBITPERM (1 << 4) +#define HWCAP2_SVESHA3 (1 << 5) +#define HWCAP2_SVESM4 (1 << 6) #endif /* _UAPI__ASM_HWCAP_H */ diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c index 9d18e45..d856c55 100644 --- a/arch/arm64/kernel/cpufeature.c +++ b/arch/arm64/kernel/cpufeature.c @@ -184,6 +184,15 @@ static const struct arm64_ftr_bits ftr_id_aa64pfr1[] = { ARM64_FTR_END, }; +static const struct arm64_ftr_bits ftr_id_aa64zfr0[] = { + ARM64_FTR_BITS(FTR_VISIBLE, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64ZFR0_SM4_SHIFT, 4, 0), + ARM64_FTR_BITS(FTR_VISIBLE, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64ZFR0_SHA3_SHIFT, 4, 0), + ARM64_FTR_BITS(FTR_VISIBLE, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64ZFR0_BITPERM_SHIFT, 4, 0), + ARM64_FTR_BITS(FTR_VISIBLE, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64ZFR0_AES_SHIFT, 4, 0), + ARM64_FTR_BITS(FTR_VISIBLE, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64ZFR0_SVEVER_SHIFT, 4, 0), + ARM64_FTR_END, +}; + static const struct arm64_ftr_bits ftr_id_aa64mmfr0[] = { S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI), S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI), @@ -392,7 +401,7 @@ static const struct __ftr_reg_entry { /* Op1 = 0, CRn = 0, CRm = 4 */ ARM64_FTR_REG(SYS_ID_AA64PFR0_EL1, ftr_id_aa64pfr0), ARM64_FTR_REG(SYS_ID_AA64PFR1_EL1, ftr_id_aa64pfr1), - ARM64_FTR_REG(SYS_ID_AA64ZFR0_EL1, ftr_raz), + ARM64_FTR_REG(SYS_ID_AA64ZFR0_EL1, ftr_id_aa64zfr0), /* Op1 = 0, CRn = 0, CRm = 5 */ ARM64_FTR_REG(SYS_ID_AA64DFR0_EL1, ftr_id_aa64dfr0), @@ -1610,6 +1619,12 @@ static const struct arm64_cpu_capabilities arm64_elf_hwcaps[] = { HWCAP_CAP(SYS_ID_AA64MMFR2_EL1, ID_AA64MMFR2_AT_SHIFT, FTR_UNSIGNED, 1, CAP_HWCAP, KERNEL_HWCAP_USCAT), #ifdef CONFIG_ARM64_SVE HWCAP_CAP(SYS_ID_AA64PFR0_EL1, ID_AA64PFR0_SVE_SHIFT, FTR_UNSIGNED, ID_AA64PFR0_SVE, CAP_HWCAP, KERNEL_HWCAP_SVE), + HWCAP_CAP(SYS_ID_AA64ZFR0_EL1, ID_AA64ZFR0_SVEVER_SHIFT, FTR_UNSIGNED, ID_AA64ZFR0_SVEVER_SVE2, CAP_HWCAP, KERNEL_HWCAP_SVE2), + HWCAP_CAP(SYS_ID_AA64ZFR0_EL1, ID_AA64ZFR0_AES_SHIFT, FTR_UNSIGNED, ID_AA64ZFR0_AES, CAP_HWCAP, KERNEL_HWCAP_SVEAES), + HWCAP_CAP(SYS_ID_AA64ZFR0_EL1, ID_AA64ZFR0_AES_SHIFT, FTR_UNSIGNED, ID_AA64ZFR0_AES_PMULL, CAP_HWCAP, KERNEL_HWCAP_SVEPMULL), + HWCAP_CAP(SYS_ID_AA64ZFR0_EL1, ID_AA64ZFR0_BITPERM_SHIFT, FTR_UNSIGNED, ID_AA64ZFR0_BITPERM, CAP_HWCAP, KERNEL_HWCAP_SVEBITPERM), + HWCAP_CAP(SYS_ID_AA64ZFR0_EL1, ID_AA64ZFR0_SHA3_SHIFT, FTR_UNSIGNED, ID_AA64ZFR0_SHA3, CAP_HWCAP, KERNEL_HWCAP_SVESHA3), + HWCAP_CAP(SYS_ID_AA64ZFR0_EL1, ID_AA64ZFR0_SM4_SHIFT, FTR_UNSIGNED, ID_AA64ZFR0_SM4, CAP_HWCAP, KERNEL_HWCAP_SVESM4), #endif HWCAP_CAP(SYS_ID_AA64PFR1_EL1, ID_AA64PFR1_SSBS_SHIFT, FTR_UNSIGNED, ID_AA64PFR1_SSBS_PSTATE_INSNS, CAP_HWCAP, KERNEL_HWCAP_SSBS), #ifdef CONFIG_ARM64_PTR_AUTH diff --git a/arch/arm64/kernel/cpuinfo.c b/arch/arm64/kernel/cpuinfo.c index 093ca53..f6f7936 100644 --- a/arch/arm64/kernel/cpuinfo.c +++ b/arch/arm64/kernel/cpuinfo.c @@ -86,6 +86,12 @@ static const char *const hwcap_str[] = { "paca", "pacg", "dcpodp", + "sve2", + "sveaes", + "svepmull", + "svebitperm", + "svesha3", + "svesm4", NULL };
This patch provides support for reporting the presence of SVE2 and its optional features to userspace. This will also enable visibility of SVE2 for guests, when KVM support for SVE-enabled guests is available. Signed-off-by: Dave Martin <Dave.Martin@arm.com> --- This patch is based on git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git f645d37e281f ("Merge remote-tracking branch 'arm64/for-next/mmiowb' into for-next/core") (for Andrew Murray's HWCAP2 support). Tested with the Arm Fast Model SVE2 plugin. For KVM, tested with: git://linux-arm.org/linux-dm.git sve-kvm-fixes/v2/head http://linux-arm.org/git?p=linux-dm.git;a=shortlog;h=refs/heads/sve-kvm-fixes/v2/head (which is also now merged to kvmarm/next). Documentation/arm64/cpu-feature-registers.txt | 16 ++++++++++++++++ Documentation/arm64/elf_hwcaps.txt | 24 ++++++++++++++++++++++++ Documentation/arm64/sve.txt | 17 +++++++++++++++++ arch/arm64/Kconfig | 3 +++ arch/arm64/include/asm/hwcap.h | 6 ++++++ arch/arm64/include/asm/sysreg.h | 14 ++++++++++++++ arch/arm64/include/uapi/asm/hwcap.h | 6 ++++++ arch/arm64/kernel/cpufeature.c | 17 ++++++++++++++++- arch/arm64/kernel/cpuinfo.c | 6 ++++++ 9 files changed, 108 insertions(+), 1 deletion(-)