| Message ID | 1526638022-4137-2-git-send-email-vladimir.murzin@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Vladimir, On 18/05/18 11:07, Vladimir Murzin wrote: > Common Not Private (CNP) is a feature of ARMv8.2 extension which > allows translation table entries to be shared between different PEs in > the same inner shareable domain, so the hardware can use this fact to > optimise the caching of such entries in the TLB. > > CNP occupies one bit in TTBRx_ELy and VTTBR_EL2, which advertises to > the hardware that the translation table entries pointed to by this > TTBR are the same as every PE in the same inner shareable domain for > which the equivalent TTBR also has CNP bit set. In case CNP bit is set > but TTBR does not point at the same translation table entries for a > given ASID and VMID, then the system is mis-configured, so the results > of translations are UNPREDICTABLE. > > For EL1 we postpone setting CNP till all cpus are up and rely on > cpufeature framework to 1) patch the code which is sensitive to CNP > and 2) update TTBR1_EL1 with CNP bit set. TTBR1_EL1 can be > reprogrammed as result of hibernation or cpuidle (via __enable_mmu). > cpuidle's path has been changed to restore CnP and for hibernation the > code has been changed to save raw TTBR1_EL1 and blindly restore it on > resume. > > For EL0 there are a few cases we need to care of changes in > TTBR0_EL1: > - a switch to idmap > - software emulated PAN (Nit: I don't think this has anything to do with EL0, its just the low half of the VA space in TTBR0...) > we rule out latter via Kconfig options and for the former we make > sure that CNP is set for non-zero ASIDs only. Reviewed-by: James Morse <james.morse@arm.com> > diff --git a/arch/arm64/include/asm/mmu_context.h b/arch/arm64/include/asm/mmu_context.h > index 39ec0b8..4f9e3ea 100644 > --- a/arch/arm64/include/asm/mmu_context.h > +++ b/arch/arm64/include/asm/mmu_context.h > @@ -149,6 +149,18 @@ static inline void cpu_replace_ttbr1(pgd_t *pgdp) > > phys_addr_t pgd_phys = virt_to_phys(pgdp); > > + if (system_supports_cnp() && !WARN_ON(pgdp != lm_alias(swapper_pg_dir))) { > + /* > + * cpu_replace_ttbr1() is used when there's a boot CPU > + * up (i.e. cpufeature framework is not up yet) and > + * latter only when we enable CNP via cpufeature's > + * enable() callback. > + * Also we rely on the cpu_hwcap bit being set before (Nit: stray whitespace) > + * calling the enable() function. > + */ > + pgd_phys |= TTBR_CNP_BIT; > + } > + > replace_phys = (void *)__pa_symbol(idmap_cpu_replace_ttbr1); > > cpu_install_idmap(); > diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c > index 9d1b06d..199e9dd 100644 > --- a/arch/arm64/kernel/cpufeature.c > +++ b/arch/arm64/kernel/cpufeature.c > @@ -858,6 +860,16 @@ static bool has_cache_dic(const struct arm64_cpu_capabilities *entry, > return read_sanitised_ftr_reg(SYS_CTR_EL0) & BIT(CTR_DIC_SHIFT); > } > > +static bool __maybe_unused > +has_useable_cnp(const struct arm64_cpu_capabilities *entry, int scope) > +{ > +#ifdef CONFIG_CRASH_DUMP > + if (elfcorehdr_size) > + return false; > +#endif It might be worth a comment why kdump is relevant, (here or in the commit message). Kdump isn't guaranteed to power-off all secondary CPUs, CNP may share TLB entries with a CPU stuck in the crashed kernel. I don't think we can't trust it even if we bring all CPUs into the new kernel as the DT may not reflect all the CPUs the crashed-kernel had. (kexec doesn't have this problem as it always powers-off secondaries before kexec-ing) > + return has_cpuid_feature(entry, scope); > +} > + > #ifdef CONFIG_UNMAP_KERNEL_AT_EL0 > static int __kpti_forced; /* 0: not forced, >0: forced on, <0: forced off */ > > @@ -1642,6 +1667,11 @@ cpufeature_pan_not_uao(const struct arm64_cpu_capabilities *entry, int __unused) > return (cpus_have_const_cap(ARM64_HAS_PAN) && !cpus_have_const_cap(ARM64_HAS_UAO)); > } > > +static void __maybe_unused cpu_enable_cnp (struct arm64_cpu_capabilities const *cap) Nit: stray space between cpu_enable_cnp and the parameters. > +{ > + cpu_replace_ttbr1(lm_alias(swapper_pg_dir)); > +} > + > /* > * We emulate only the following system register space. > * Op0 = 0x3, CRn = 0x0, Op1 = 0x0, CRm = [0, 4 - 7] > diff --git a/arch/arm64/kernel/hibernate.c b/arch/arm64/kernel/hibernate.c > index 1ec5f28..ea27121 100644 > --- a/arch/arm64/kernel/hibernate.c > +++ b/arch/arm64/kernel/hibernate.c > @@ -125,7 +125,7 @@ int arch_hibernation_header_save(void *addr, unsigned int max_size) > return -EOVERFLOW; > > arch_hdr_invariants(&hdr->invariants); > - hdr->ttbr1_el1 = __pa_symbol(swapper_pg_dir); > + hdr->ttbr1_el1 = read_sysreg(ttbr1_el1); > hdr->reenter_kernel = _cpu_resume; > > /* We can't use __hyp_get_vectors() because kvm may still be loaded */ We also run__cpu_suspend_exit() from hibernate's restore, which will restore CNP so this isn't strictly necessary. There is some future-cleanup we can do here, if hibernate set the idmap when calling __cpu_suspend_exit(), that function could do the replace_ttbr1() without an extra install/uninstall of the idmap. Thanks, James
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index eb2cf49..cde5d75 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1125,6 +1125,19 @@ config ARM64_RAS_EXTN and access the new registers if the system supports the extension. Platform RAS features may additionally depend on firmware support. +config ARM64_CNP + bool "Enable support for Common Not Private (CNP) translations" + depends on ARM64_PAN || !ARM64_SW_TTBR0_PAN + help + Common Not Private (CNP) allows translation table entries to + be shared between different PEs in the same inner shareable + domain, so the hardware can use this fact to optimise the + caching of such entries in the TLB. + + Selecting this option allows the CNP feature to be detected + at runtime, and does not affect PEs that do not implement + this feature. + endmenu config ARM64_SVE diff --git a/arch/arm64/include/asm/cpucaps.h b/arch/arm64/include/asm/cpucaps.h index bc51b72..f474a61 100644 --- a/arch/arm64/include/asm/cpucaps.h +++ b/arch/arm64/include/asm/cpucaps.h @@ -48,7 +48,8 @@ #define ARM64_HAS_CACHE_IDC 27 #define ARM64_HAS_CACHE_DIC 28 #define ARM64_HW_DBM 29 +#define ARM64_HAS_CNP 30 -#define ARM64_NCAPS 30 +#define ARM64_NCAPS 31 #endif /* __ASM_CPUCAPS_H */ diff --git a/arch/arm64/include/asm/cpufeature.h b/arch/arm64/include/asm/cpufeature.h index 09b0f2a..1f7c056 100644 --- a/arch/arm64/include/asm/cpufeature.h +++ b/arch/arm64/include/asm/cpufeature.h @@ -510,6 +510,12 @@ static inline bool system_supports_sve(void) cpus_have_const_cap(ARM64_SVE); } +static inline bool system_supports_cnp(void) +{ + return IS_ENABLED(CONFIG_ARM64_CNP) && + cpus_have_const_cap(ARM64_HAS_CNP); +} + /* * Read the pseudo-ZCR used by cpufeatures to identify the supported SVE * vector length. diff --git a/arch/arm64/include/asm/mmu_context.h b/arch/arm64/include/asm/mmu_context.h index 39ec0b8..4f9e3ea 100644 --- a/arch/arm64/include/asm/mmu_context.h +++ b/arch/arm64/include/asm/mmu_context.h @@ -149,6 +149,18 @@ static inline void cpu_replace_ttbr1(pgd_t *pgdp) phys_addr_t pgd_phys = virt_to_phys(pgdp); + if (system_supports_cnp() && !WARN_ON(pgdp != lm_alias(swapper_pg_dir))) { + /* + * cpu_replace_ttbr1() is used when there's a boot CPU + * up (i.e. cpufeature framework is not up yet) and + * latter only when we enable CNP via cpufeature's + * enable() callback. + * Also we rely on the cpu_hwcap bit being set before + * calling the enable() function. + */ + pgd_phys |= TTBR_CNP_BIT; + } + replace_phys = (void *)__pa_symbol(idmap_cpu_replace_ttbr1); cpu_install_idmap(); diff --git a/arch/arm64/include/asm/pgtable-hwdef.h b/arch/arm64/include/asm/pgtable-hwdef.h index fd208ea..1d7d8da 100644 --- a/arch/arm64/include/asm/pgtable-hwdef.h +++ b/arch/arm64/include/asm/pgtable-hwdef.h @@ -211,6 +211,8 @@ #define PHYS_MASK_SHIFT (CONFIG_ARM64_PA_BITS) #define PHYS_MASK ((UL(1) << PHYS_MASK_SHIFT) - 1) +#define TTBR_CNP_BIT (UL(1) << 0) + /* * TCR flags. */ diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c index 9d1b06d..199e9dd 100644 --- a/arch/arm64/kernel/cpufeature.c +++ b/arch/arm64/kernel/cpufeature.c @@ -20,6 +20,7 @@ #include <linux/bsearch.h> #include <linux/cpumask.h> +#include <linux/crash_dump.h> #include <linux/sort.h> #include <linux/stop_machine.h> #include <linux/types.h> @@ -117,6 +118,7 @@ EXPORT_SYMBOL(cpu_hwcap_keys); static bool __maybe_unused cpufeature_pan_not_uao(const struct arm64_cpu_capabilities *entry, int __unused); +static void cpu_enable_cnp(struct arm64_cpu_capabilities const *cap); /* * NOTE: Any changes to the visibility of features should be kept in @@ -858,6 +860,16 @@ static bool has_cache_dic(const struct arm64_cpu_capabilities *entry, return read_sanitised_ftr_reg(SYS_CTR_EL0) & BIT(CTR_DIC_SHIFT); } +static bool __maybe_unused +has_useable_cnp(const struct arm64_cpu_capabilities *entry, int scope) +{ +#ifdef CONFIG_CRASH_DUMP + if (elfcorehdr_size) + return false; +#endif + return has_cpuid_feature(entry, scope); +} + #ifdef CONFIG_UNMAP_KERNEL_AT_EL0 static int __kpti_forced; /* 0: not forced, >0: forced on, <0: forced off */ @@ -1202,6 +1214,19 @@ static const struct arm64_cpu_capabilities arm64_features[] = { .cpu_enable = cpu_enable_hw_dbm, }, #endif +#ifdef CONFIG_ARM64_CNP + { + .desc = "Common not Private translations", + .capability = ARM64_HAS_CNP, + .type = ARM64_CPUCAP_SYSTEM_FEATURE, + .matches = has_useable_cnp, + .sys_reg = SYS_ID_AA64MMFR2_EL1, + .sign = FTR_UNSIGNED, + .field_pos = ID_AA64MMFR2_CNP_SHIFT, + .min_field_value = 1, + .cpu_enable = cpu_enable_cnp, + }, +#endif {}, }; @@ -1642,6 +1667,11 @@ cpufeature_pan_not_uao(const struct arm64_cpu_capabilities *entry, int __unused) return (cpus_have_const_cap(ARM64_HAS_PAN) && !cpus_have_const_cap(ARM64_HAS_UAO)); } +static void __maybe_unused cpu_enable_cnp (struct arm64_cpu_capabilities const *cap) +{ + cpu_replace_ttbr1(lm_alias(swapper_pg_dir)); +} + /* * We emulate only the following system register space. * Op0 = 0x3, CRn = 0x0, Op1 = 0x0, CRm = [0, 4 - 7] diff --git a/arch/arm64/kernel/hibernate.c b/arch/arm64/kernel/hibernate.c index 1ec5f28..ea27121 100644 --- a/arch/arm64/kernel/hibernate.c +++ b/arch/arm64/kernel/hibernate.c @@ -125,7 +125,7 @@ int arch_hibernation_header_save(void *addr, unsigned int max_size) return -EOVERFLOW; arch_hdr_invariants(&hdr->invariants); - hdr->ttbr1_el1 = __pa_symbol(swapper_pg_dir); + hdr->ttbr1_el1 = read_sysreg(ttbr1_el1); hdr->reenter_kernel = _cpu_resume; /* We can't use __hyp_get_vectors() because kvm may still be loaded */ diff --git a/arch/arm64/kernel/suspend.c b/arch/arm64/kernel/suspend.c index a307b9e..9576643 100644 --- a/arch/arm64/kernel/suspend.c +++ b/arch/arm64/kernel/suspend.c @@ -48,6 +48,10 @@ void notrace __cpu_suspend_exit(void) */ cpu_uninstall_idmap(); + /* Restore CnP bit in TTBR1_EL1 */ + if (system_supports_cnp()) + cpu_replace_ttbr1(lm_alias(swapper_pg_dir)); + /* * PSTATE was not saved over suspend/resume, re-enable any detected * features that might not have been set correctly. diff --git a/arch/arm64/mm/context.c b/arch/arm64/mm/context.c index 301417a..3c1b2c1 100644 --- a/arch/arm64/mm/context.c +++ b/arch/arm64/mm/context.c @@ -196,6 +196,9 @@ void check_and_switch_context(struct mm_struct *mm, unsigned int cpu) unsigned long flags; u64 asid, old_active_asid; + if (system_supports_cnp()) + cpu_set_reserved_ttbr0(); + asid = atomic64_read(&mm->context.id); /* diff --git a/arch/arm64/mm/proc.S b/arch/arm64/mm/proc.S index 5f9a73a..d46d0ca 100644 --- a/arch/arm64/mm/proc.S +++ b/arch/arm64/mm/proc.S @@ -160,6 +160,12 @@ ENTRY(cpu_do_switch_mm) mrs x2, ttbr1_el1 mmid x1, x1 // get mm->context.id phys_to_ttbr x3, x0 + +alternative_if ARM64_HAS_CNP + cbz x1, 1f // skip CNP for reserved ASID + orr x3, x3, #TTBR_CNP_BIT +1: +alternative_else_nop_endif #ifdef CONFIG_ARM64_SW_TTBR0_PAN bfi x3, x1, #48, #16 // set the ASID field in TTBR0 #endif
Common Not Private (CNP) is a feature of ARMv8.2 extension which allows translation table entries to be shared between different PEs in the same inner shareable domain, so the hardware can use this fact to optimise the caching of such entries in the TLB. CNP occupies one bit in TTBRx_ELy and VTTBR_EL2, which advertises to the hardware that the translation table entries pointed to by this TTBR are the same as every PE in the same inner shareable domain for which the equivalent TTBR also has CNP bit set. In case CNP bit is set but TTBR does not point at the same translation table entries for a given ASID and VMID, then the system is mis-configured, so the results of translations are UNPREDICTABLE. For EL1 we postpone setting CNP till all cpus are up and rely on cpufeature framework to 1) patch the code which is sensitive to CNP and 2) update TTBR1_EL1 with CNP bit set. TTBR1_EL1 can be reprogrammed as result of hibernation or cpuidle (via __enable_mmu). cpuidle's path has been changed to restore CnP and for hibernation the code has been changed to save raw TTBR1_EL1 and blindly restore it on resume. For EL0 there are a few cases we need to care of changes in TTBR0_EL1: - a switch to idmap - software emulated PAN we rule out latter via Kconfig options and for the former we make sure that CNP is set for non-zero ASIDs only. Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com> --- arch/arm64/Kconfig | 13 +++++++++++++ arch/arm64/include/asm/cpucaps.h | 3 ++- arch/arm64/include/asm/cpufeature.h | 6 ++++++ arch/arm64/include/asm/mmu_context.h | 12 ++++++++++++ arch/arm64/include/asm/pgtable-hwdef.h | 2 ++ arch/arm64/kernel/cpufeature.c | 30 ++++++++++++++++++++++++++++++ arch/arm64/kernel/hibernate.c | 2 +- arch/arm64/kernel/suspend.c | 4 ++++ arch/arm64/mm/context.c | 3 +++ arch/arm64/mm/proc.S | 6 ++++++ 10 files changed, 79 insertions(+), 2 deletions(-)