KVM: arm64: Errata management for VM Live migration
Subsystems (10)
v8 cover KVM: arm64: Errata management for VM Live migration
Hi,
v7-->v8
https://lore.kernel.org/kvmarm/[email protected]/
Changes:
-Fixed the compile error issue due to CONFIG_CORESIGHT_SOURCE_ETM4X=m
(Thanks to Sebastian for checking that config)- patch #4
-Added R-by tags.
This can be sanity tested by the Qemu branch here,
https://github.com/hisilicon/qemu/tree/v9.0-nv-rfcv4-vcpu-model-v2-target-cpu-errata
(branch based on Eric's/Connie's NV + custom CPU series)
Eg: to specify target impl CPUs,
-machine virt,.., x-target-impl-cpus=0xMIDR1:0xREVIDR1-0xMIDR2:REVIDR2
Please take a look and let me know your feedback.
Thanks,
Shameer
v6-->v7
https://lore.kernel.org/kvmarm/[email protected]/
Changes:
-Introduced KVM_REG_ARM_VENDOR_HYP_BMAP_2 register for new hypercall
support(patch #3)
-Added selftest for KVM_REG_ARM_VENDOR_HYP_BMAP_2(patch #5).
v5 --> v6
https://lore.kernel.org/kvmarm/[email protected]/
-Addressed comments from Oliver,
-Changed version to 32 bit to match SMCCC_VERSION -Don't advertise pKVM hypercalls to guests. -Moved hypercall invocation to kvm_guest.c v4 --> v5
https://lore.kernel.org/kvmarm/[email protected]/
-Addressed comments from Marc,
-Added an hypercall to retrieve version and number of supported target impl CPUs.
-Added a check for KVM hypercall services availability. -Removed R-by tags from Connie & Sebastian as patches 2 & 4 are now changed a bit. Please take another look.
v3 --> v4(Minor updates)
https://lore.kernel.org/kvmarm/[email protected]/
-Changed MIDR/REVIDR to 64 bits based on feedback from Connie and Marc(Patch #3).
-Added R-by tags from Sebastian (Thanks!).
RFC v2 --> v3
https://lore.kernel.org/kvmarm/[email protected]/
-Addressed comments from Oliver(Thanks!).
-Using implementation CPUs MIDR/REVIDR when it is set for _midr_range() functions(Patch 1 & 3)
-New hypercall for retrieving implementation CPUs(Patch #2). -Dropped RFC.
RFC v1 --> RFCv2:
https://lore.kernel.org/kvmarm/[email protected]/
-Introduced hypercalls to retrieve target CPUs info from user space VMM. see patch #1 for details.
-Patch #2 uses the hypercall to retrieve the target CPU info if any. -Use the target CPUs MIDR/REVIDR in errata enablement. See patch #3.
Background from v1:
On ARM64 platforms most of the errata workarounds are based on CPU MIDR/REVIDR values and a number of these workarounds need to be implemented by the Guest kernel as well. This creates a problem when Guest needs to be migrated to a platform that differs in these MIDR/REVIDR values even if the VMM can come up with a common minimum feature list for the Guest using the recently introduced "Writable ID registers" support.
(This is roughly based on a discussion I had with Marc and Oliver at KVM forum. Marc outlined his idea for a solution and this is an attempt to implement it. Thanks to both and I take all the blame if this is nowhere near what is intended/required)
Shameer Kolothum (6):
arm64: Modify _midr_range() functions to read MIDR/REVIDR internally
KVM: arm64: Introduce hypercall support for retrieving target
implementations
KVM: arm64: Introduce KVM_REG_ARM_VENDOR_HYP_BMAP_2
arm64: Make _midr_in_range_list() an exported function smccc/kvm_guest: Enable errata based on implementation CPUs
KVM: selftests: Add test for KVM_REG_ARM_VENDOR_HYP_BMAP_2
.../virt/kvm/arm/fw-pseudo-registers.rst | 15 +++- Documentation/virt/kvm/arm/hypercalls.rst | 59 +++++++++++++++ arch/arm64/include/asm/cputype.h | 40 +++++----- arch/arm64/include/asm/hypervisor.h | 1 + arch/arm64/include/asm/kvm_host.h | 3 +- arch/arm64/include/asm/mmu.h | 3 +- arch/arm64/include/uapi/asm/kvm.h | 12 +++ arch/arm64/kernel/cpu_errata.c | 73 ++++++++++++++++--- arch/arm64/kernel/cpufeature.c | 8 +- arch/arm64/kernel/image-vars.h | 1 + arch/arm64/kernel/proton-pack.c | 17 ++--- arch/arm64/kvm/hypercalls.c | 13 ++++ arch/arm64/kvm/vgic/vgic-v3.c | 2 +- drivers/clocksource/arm_arch_timer.c | 2 +- drivers/firmware/smccc/kvm_guest.c | 64 ++++++++++++++++ .../coresight/coresight-etm4x-core.c | 2 +- include/linux/arm-smccc.h | 15 ++++ tools/arch/arm64/include/uapi/asm/kvm.h | 12 +++ .../selftests/kvm/arm64/get-reg-list.c | 1 + .../testing/selftests/kvm/arm64/hypercalls.c | 46 +++++++++--- 20 files changed, 327 insertions(+), 62 deletions(-)
Signature · 1 line
2.47.0
PATCH v8 1/6 arm64: Modify _midr_range() functions to read MIDR/REVIDR internally
These changes lay the groundwork for adding support for guest kernels, allowing them to leverage target CPU implementations provided by the VMM.
No functional changes intended.
--- arch/arm64/include/asm/cputype.h | 28 +++++++++---------- arch/arm64/include/asm/mmu.h | 3 +- arch/arm64/kernel/cpu_errata.c | 23 ++++++++------- arch/arm64/kernel/cpufeature.c | 6 ++-- arch/arm64/kernel/proton-pack.c | 17 ++++++----- arch/arm64/kvm/vgic/vgic-v3.c | 2 +- drivers/clocksource/arm_arch_timer.c | 2 +- .../coresight/coresight-etm4x-core.c | 2 +- 8 files changed, 42 insertions(+), 41 deletions(-)
arch/arm64/include/asm/cputype.h +14 -14
@@ -231,6 +231,16 @@#define read_cpuid (reg) read_sysreg_s(SYS_ ## reg) +/* +* The CPU ID never changes at run time, so we might as well tell the +* compiler that it's constant. Use this function to read the CPU ID +* rather than directly reading processor_id or read_cpuid() directly. +*/ +static inline u32 __attribute_const__ read_cpuid_id(void) +{ +return read_cpuid (MIDR_EL1 ); +} +/* * Represent a range of MIDR values for a given CPU model and a * range of variant/revision values.
@@ -266,31 +276,21 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } -static inline bool is_midr_in_range (u32 midr,struct midr_range const *range) +static inline bool is_midr_in_range (struct midr_range const *range) { -return midr_is_cpu_model_range (midr, range->model, +return midr_is_cpu_model_range (read_cpuid_id (), range->model, range->rv_min, range->rv_max); }static inline bool -is_midr_in_range_list (u32 midr,struct midr_range const *ranges) +is_midr_in_range_list (struct midr_range const *ranges) {while (ranges->model) -if (is_midr_in_range (midr, ranges++)) +if (is_midr_in_range (ranges++))return true;return false; } -/* -* The CPU ID never changes at run time, so we might as well tell the -* compiler that it's constant. Use this function to read the CPU ID -* rather than directly reading processor_id or read_cpuid() directly. -*/ -static inline u32 __attribute_const__ read_cpuid_id(void) -{ -return read_cpuid (MIDR_EL1 ); -} - static inline u64 __attribute_const__ read_cpuid_mpidr(void) {return read_cpuid (MPIDR_EL1 );
arch/arm64/include/asm/mmu.h +1 -2
@@ -101,8 +101,7 @@ static inline bool kaslr_requires_kpti(void)if (IS_ENABLED (CONFIG_CAVIUM_ERRATUM_27456 )) {extern const struct midr_range cavium_erratum_27456_cpus[]; -if (is_midr_in_range_list (read_cpuid_id (), - cavium_erratum_27456_cpus)) +if (is_midr_in_range_list (cavium_erratum_27456_cpus))return false; }
arch/arm64/kernel/cpu_errata.c +13 -10
@@ -15,30 +15,34 @@#include <asm/smp_plat.h> static bool __maybe_unused -is_affected_midr_range (const struct arm64_cpu_capabilities *entry,int scope) +__is_affected_midr_range (const struct arm64_cpu_capabilities *entry, +u32 midr,u32 revidr) {const struct arm64_midr_revidr *fix; -u32 midr =read_cpuid_id (), revidr; - -WARN_ON (scope !=SCOPE_LOCAL_CPU ||preemptible ()); -if (!is_midr_in_range (midr, &entry->midr_range)) +if (!is_midr_in_range (&entry->midr_range))return false; midr &=MIDR_REVISION_MASK |MIDR_VARIANT_MASK ; - revidr =read_cpuid (REVIDR_EL1 );for (fix = entry->fixed_revs; fix && fix->revidr_mask; fix++)if (midr == fix->midr_rv && (revidr & fix->revidr_mask))return false; -return true; } +static bool __maybe_unused +is_affected_midr_range (const struct arm64_cpu_capabilities *entry,int scope) +{ +WARN_ON (scope !=SCOPE_LOCAL_CPU ||preemptible ()); +return __is_affected_midr_range (entry,read_cpuid_id (), +read_cpuid (REVIDR_EL1 )); +} +static bool __maybe_unused is_affected_midr_range_list (const struct arm64_cpu_capabilities *entry,int scope) {WARN_ON (scope !=SCOPE_LOCAL_CPU ||preemptible ()); -return is_midr_in_range_list (read_cpuid_id (), entry->midr_range_list); +return is_midr_in_range_list (entry->midr_range_list); } static bool __maybe_unused
@@ -186,12 +190,11 @@ static bool __maybe_unused has_neoverse_n1_erratum_1542419(const struct arm64_cpu_capabilities *entry, int scope) { -u32 midr =read_cpuid_id (); bool has_dic = read_cpuid_cachetype() & BIT(CTR_EL0_DIC_SHIFT);const struct midr_range range =MIDR_ALL_VERSIONS (MIDR_NEOVERSE_N1 );WARN_ON (scope !=SCOPE_LOCAL_CPU ||preemptible ()); -return is_midr_in_range (midr, &range) && has_dic; +return is_midr_in_range (&range) && has_dic; }#ifdef CONFIG_ARM64_WORKAROUND_REPEAT_TLBI
arch/arm64/kernel/cpufeature.c +3 -3
@@ -1792,7 +1792,7 @@ static bool unmap_kernel_at_el0(const struct arm64_cpu_capabilities *entry,char const *str ="kpti command line option" ;bool meltdown_safe; - meltdown_safe =is_midr_in_range_list (read_cpuid_id (), kpti_safe_list); + meltdown_safe =is_midr_in_range_list (kpti_safe_list);/* Defer to CPU feature registers */ if (has_cpuid_feature (entry, scope))
@@ -1862,7 +1862,7 @@ static bool has_nv1(const struct arm64_cpu_capabilities *entry, int scope)return (__system_matches_cap (ARM64_HAS_NESTED_VIRT ) && !(has_cpuid_feature (entry, scope) || -is_midr_in_range_list (read_cpuid_id (), nv1_ni_list))); +is_midr_in_range_list (nv1_ni_list))); }#if defined(ID_AA64MMFR0_EL1_TGRAN_LPA2 ) && defined(ID_AA64MMFR0_EL1_TGRAN_2_SUPPORTED_LPA2 )
@@ -2045,7 +2045,7 @@ static bool cpu_has_broken_dbm(void) {}, }; -return is_midr_in_range_list (read_cpuid_id (), cpus); +return is_midr_in_range_list (cpus); }static bool cpu_can_use_dbm (const struct arm64_cpu_capabilities *cap)
arch/arm64/kernel/proton-pack.c +8 -9
@@ -172,7 +172,7 @@ static enum mitigation_state spectre_v2_get_cpu_hw_mitigation_state(void)return SPECTRE_UNAFFECTED ;/* Alternatively, we have a list of unaffected CPUs */ -if (is_midr_in_range_list (read_cpuid_id (), spectre_v2_safe_list)) +if (is_midr_in_range_list (spectre_v2_safe_list))return SPECTRE_UNAFFECTED ;return SPECTRE_VULNERABLE ;
@@ -331,7 +331,7 @@ bool has_spectre_v3a(const struct arm64_cpu_capabilities *entry, int scope) };WARN_ON (scope !=SCOPE_LOCAL_CPU ||preemptible ()); -return is_midr_in_range_list (read_cpuid_id (), spectre_v3a_unsafe_list); +return is_midr_in_range_list (spectre_v3a_unsafe_list); }void spectre_v3a_enable_mitigation (const struct arm64_cpu_capabilities *__unused)
@@ -475,7 +475,7 @@ static enum mitigation_state spectre_v4_get_cpu_hw_mitigation_state(void) { /* sentinel */ }, }; -if (is_midr_in_range_list (read_cpuid_id (), spectre_v4_safe_list)) +if (is_midr_in_range_list (spectre_v4_safe_list))return SPECTRE_UNAFFECTED ;/* CPU features are detected first */
@@ -878,13 +878,13 @@ u8 spectre_bhb_loop_affected(int scope) {}, }; -if (is_midr_in_range_list (read_cpuid_id (), spectre_bhb_k32_list)) +if (is_midr_in_range_list (spectre_bhb_k32_list)) k =32 ; -else if (is_midr_in_range_list (read_cpuid_id (), spectre_bhb_k24_list)) +else if (is_midr_in_range_list (spectre_bhb_k24_list)) k =24 ; -else if (is_midr_in_range_list (read_cpuid_id (), spectre_bhb_k11_list)) +else if (is_midr_in_range_list (spectre_bhb_k11_list)) k =11 ; -else if (is_midr_in_range_list (read_cpuid_id (), spectre_bhb_k8_list)) +else if (is_midr_in_range_list (spectre_bhb_k8_list)) k =8 ; max_bhb_k =max (max_bhb_k, k);
@@ -926,8 +926,7 @@ static bool is_spectre_bhb_fw_affected(int scope) MIDR_ALL_VERSIONS(MIDR_CORTEX_A75), {}, }; -bool cpu_in_list =is_midr_in_range_list (read_cpuid_id (), - spectre_bhb_firmware_mitigated_list); +bool cpu_in_list =is_midr_in_range_list (spectre_bhb_firmware_mitigated_list);if (scope !=SCOPE_LOCAL_CPU )return system_affected;
arch/arm64/kvm/vgic/vgic-v3.c +1 -1
@@ -633,7 +633,7 @@ static const struct midr_range broken_seis[] = {static bool vgic_v3_broken_seis (void ) {return ((kvm_vgic_global_state.ich_vtr_el2 &ICH_VTR_SEIS_MASK ) && -is_midr_in_range_list (read_cpuid_id (), broken_seis)); +is_midr_in_range_list (broken_seis)); }/**
drivers/clocksource/arm_arch_timer.c +1 -1
@@ -842,7 +842,7 @@ static u64 __arch_timer_check_delta(void) {}, }; -if (is_midr_in_range_list (read_cpuid_id (), broken_cval_midrs)) { +if (is_midr_in_range_list (broken_cval_midrs)) {pr_warn_once ("Broken CNTx_CVAL_EL1, using 31 bit TVAL instead.\n" );return CLOCKSOURCE_MASK (31 ); }
drivers/hwtracing/coresight/coresight-etm4x-core.c +1 -1
@@ -1216,7 +1216,7 @@ static void etm4_fixup_wrong_ccitmin(struct etmv4_drvdata *drvdata)* recorded value for 'drvdata->ccitmin' to workaround * this problem. */ -if (is_midr_in_range_list (read_cpuid_id (), etm_wrong_ccitmin_cpus)) { +if (is_midr_in_range_list (etm_wrong_ccitmin_cpus)) {if (drvdata->ccitmin ==256 ) drvdata->ccitmin =4 ; }
Signature · 1 line
2.47.0
Re: [PATCH v8 1/6] arm64: Modify _midr_range() functions to read MIDR/REVIDR internally
Quoted 10 lines · Shameer Kolothum
These changes lay the groundwork for adding support for guest kernels, allowing them to leverage target CPU implementations provided by the VMM.
No functional changes intended.
Suggested-by Oliver Upton <[email protected]>Reviewed-by Sebastian Ott <[email protected]>Reviewed-by Cornelia Huck <[email protected]>Signed-off-by Shameer Kolothum <[email protected]>
PATCH v8 2/6 KVM: arm64: Introduce hypercall support for retrieving target implementations
In reply to Shameer KolothumIf the Guest requires migration to multiple targets, these hypercalls will provide a way to retrieve the target CPU implementations from the user space VMM.
Subsequent patch will use this to enable the associated errata.
--- Documentation/virt/kvm/arm/hypercalls.rst | 59 +++++++++++++++++++++++ include/linux/arm-smccc.h | 15 ++++++ 2 files changed, 74 insertions(+)
Documentation/virt/kvm/arm/hypercalls.rst +59 -0
@@ -142,3 +142,62 @@ region is equal to the memory protection granule advertised by | | | +---------------------------------------------+ | | | | ``INVALID_PARAMETER (-3)`` | +---------------------+----------+----+---------------------------------------------+ + +``ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_VER_FUNC_ID`` +------------------------------------------------------- +Request the target CPU implementation version information and the number of target +implementations for the Guest VM. + ++---------------------+-------------------------------------------------------------+ +| Presence: | Optional; KVM/ARM64 Guests only | ++---------------------+-------------------------------------------------------------+ +| Calling convention: | HVC64 | ++---------------------+----------+--------------------------------------------------+ +| Function ID: | (uint32) | 0xC6000040 | ++---------------------+----------+--------------------------------------------------+ +| Arguments: | None | ++---------------------+----------+----+---------------------------------------------+ +| Return Values: | (int64) | R0 | ``SUCCESS (0)`` | +| | | +---------------------------------------------+ +| | | | ``NOT_SUPPORTED (-1)`` | +| +----------+----+---------------------------------------------+ +| | (uint64) | R1 | Bits [63:32] Reserved/Must be zero | +| | | +---------------------------------------------+ +| | | | Bits [31:16] Major version | +| | | +---------------------------------------------+ +| | | | Bits [15:0] Minor version | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | Number of target implementations | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | Reserved / Must be zero | ++---------------------+----------+----+---------------------------------------------+ + +``ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_CPUS_FUNC_ID`` +------------------------------------------------------- + +Request the target CPU implementation information for the Guest VM. The Guest kernel +will use this information to enable the associated errata. + ++---------------------+-------------------------------------------------------------+ +| Presence: | Optional; KVM/ARM64 Guests only | ++---------------------+-------------------------------------------------------------+ +| Calling convention: | HVC64 | ++---------------------+----------+--------------------------------------------------+ +| Function ID: | (uint32) | 0xC6000041 | ++---------------------+----------+----+---------------------------------------------+ +| Arguments: | (uint64) | R1 | selected implementation index | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | Reserved / Must be zero | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | Reserved / Must be zero | ++---------------------+----------+----+---------------------------------------------+ +| Return Values: | (int64) | R0 | ``SUCCESS (0)`` | +| | | +---------------------------------------------+ +| | | | ``INVALID_PARAMETER (-3)`` | +| +----------+----+---------------------------------------------+ +| | (uint64) | R1 | MIDR_EL1 of the selected implementation | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | REVIDR_EL1 of the selected implementation | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | AIDR_EL1 of the selected implementation | ++---------------------+----------+----+---------------------------------------------+
include/linux/arm-smccc.h +15 -0
@@ -179,6 +179,9 @@#define ARM_SMCCC_KVM_FUNC_PKVM_RESV_62 62#define ARM_SMCCC_KVM_FUNC_PKVM_RESV_63 63/* End of pKVM hypercall range */ +#define ARM_SMCCC_KVM_FUNC_DISCOVER_IMPL_VER 64 +#define ARM_SMCCC_KVM_FUNC_DISCOVER_IMPL_CPUS 65 +#define ARM_SMCCC_KVM_FUNC_FEATURES_2 127#define ARM_SMCCC_KVM_NUM_FUNCS 128
@@ -225,6 +228,18 @@ ARM_SMCCC_OWNER_VENDOR_HYP, \ ARM_SMCCC_KVM_FUNC_MMIO_GUARD) +#define ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_VER_FUNC_ID \ + ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, \ + ARM_SMCCC_SMC_64, \ + ARM_SMCCC_OWNER_VENDOR_HYP, \ + ARM_SMCCC_KVM_FUNC_DISCOVER_IMPL_VER) + +#define ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_CPUS_FUNC_ID \ + ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, \ + ARM_SMCCC_SMC_64, \ + ARM_SMCCC_OWNER_VENDOR_HYP, \ + ARM_SMCCC_KVM_FUNC_DISCOVER_IMPL_CPUS) +/* ptp_kvm counter type ID */ #define KVM_PTP_VIRT_COUNTER 0#define KVM_PTP_PHYS_COUNTER 1
Signature · 1 line
2.47.0
PATCH v8 3/6 KVM: arm64: Introduce KVM_REG_ARM_VENDOR_HYP_BMAP_2
In reply to Shameer KolothumThe vendor_hyp_bmap bitmap holds the information about the Vendor Hyp services available to the user space and can be get/set using {G, S}ET_ONE_REG interfaces. This is done using the pseudo-firmware bitmap register KVM_REG_ARM_VENDOR_HYP_BMAP.
At present, this bitmap is a 64 bit one and since the function numbers for newly added DISCOVER_IPML_* hypercalls are 64-65, introduce another pseudo-firmware bitmap register KVM_REG_ARM_VENDOR_HYP_BMAP_2.
--- .../virt/kvm/arm/fw-pseudo-registers.rst | 15 ++++++++++++++- arch/arm64/include/asm/kvm_host.h | 3 ++- arch/arm64/include/uapi/asm/kvm.h | 12 ++++++++++++ arch/arm64/kvm/hypercalls.c | 13 +++++++++++++ 4 files changed, 41 insertions(+), 2 deletions(-)
Documentation/virt/kvm/arm/fw-pseudo-registers.rst +14 -1
@@ -116,7 +116,7 @@ The pseudo-firmware bitmap register are as follows: ARM DEN0057A. * KVM_REG_ARM_VENDOR_HYP_BMAP: - Controls the bitmap of the Vendor specific Hypervisor Service Calls. + Controls the bitmap of the Vendor specific Hypervisor Service Calls[0-63]. The following bits are accepted:
@@ -127,6 +127,19 @@ The pseudo-firmware bitmap register are as follows: Bit-1: KVM_REG_ARM_VENDOR_HYP_BIT_PTP: The bit represents the Precision Time Protocol KVM service. +* KVM_REG_ARM_VENDOR_HYP_BMAP_2: + Controls the bitmap of the Vendor specific Hypervisor Service Calls[64-127]. + + The following bits are accepted: + + Bit-0: KVM_REG_ARM_VENDOR_HYP_BIT_DISCOVER_IMPL_VER + This represents the ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_VER_FUNC_ID + function-id. This is reset to 0. + + Bit-1: KVM_REG_ARM_VENDOR_HYP_BIT_DISCOVER_IMPL_CPUS + This represents the ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_CPUS_FUNC_ID + function-id. This is reset to 0. + Errors: ======= =============================================================
arch/arm64/include/asm/kvm_host.h +2 -1
@@ -237,7 +237,8 @@ struct kvm_arch_memory_slot {struct kvm_smccc_features {unsigned long std_bmap;unsigned long std_hyp_bmap; -unsigned long vendor_hyp_bmap; +unsigned long vendor_hyp_bmap;/* Function numbers 0-63 */ +unsigned long vendor_hyp_bmap_2;/* Function numbers 64-127 */ };typedef unsigned int pkvm_handle_t ;
arch/arm64/include/uapi/asm/kvm.h +12 -0
@@ -371,6 +371,7 @@ enum {#endif }; +/* Vendor hyper call function numbers 0-63 */ #define KVM_REG_ARM_VENDOR_HYP_BMAP KVM_REG_ARM_FW_FEAT_BMAP_REG(2)enum {
@@ -381,6 +382,17 @@ enum {#endif }; +/* Vendor hyper call function numbers 64-127 */ +#define KVM_REG_ARM_VENDOR_HYP_BMAP_2 KVM_REG_ARM_FW_FEAT_BMAP_REG(3) + +enum { +KVM_REG_ARM_VENDOR_HYP_BIT_DISCOVER_IMPL_VER =0 , +KVM_REG_ARM_VENDOR_HYP_BIT_DISCOVER_IMPL_CPUS =1 , +#ifdef __KERNEL__ +KVM_REG_ARM_VENDOR_HYP_BMAP_2_BIT_COUNT , +#endif +}; +/* Device Control API on vm fd */ #define KVM_ARM_VM_SMCCC_CTRL 0#define KVM_ARM_VM_SMCCC_FILTER 0
arch/arm64/kvm/hypercalls.c +13 -0
@@ -15,6 +15,8 @@ GENMASK(KVM_REG_ARM_STD_HYP_BMAP_BIT_COUNT - 1, 0)#define KVM_ARM_SMCCC_VENDOR_HYP_FEATURES \ GENMASK(KVM_REG_ARM_VENDOR_HYP_BMAP_BIT_COUNT - 1, 0) +#define KVM_ARM_SMCCC_VENDOR_HYP_FEATURES_2 \ + GENMASK(KVM_REG_ARM_VENDOR_HYP_BMAP_2_BIT_COUNT - 1, 0)static void kvm_ptp_get_time (struct kvm_vcpu *vcpu,u64 *val) {
@@ -360,6 +362,8 @@ int kvm_smccc_call_handler(struct kvm_vcpu *vcpu)break ;case ARM_SMCCC_VENDOR_HYP_KVM_FEATURES_FUNC_ID : val[0 ] = smccc_feat->vendor_hyp_bmap; +/* Function numbers 2-63 are reserved for pKVM for now */ + val[2 ] = smccc_feat->vendor_hyp_bmap_2;break ;case ARM_SMCCC_VENDOR_HYP_KVM_PTP_FUNC_ID :kvm_ptp_get_time (vcpu, val);
@@ -387,6 +391,7 @@ static const u64 kvm_arm_fw_reg_ids[] = {KVM_REG_ARM_STD_BMAP ,KVM_REG_ARM_STD_HYP_BMAP ,KVM_REG_ARM_VENDOR_HYP_BMAP , + KVM_REG_ARM_VENDOR_HYP_BMAP_2, };void kvm_arm_init_hypercalls (struct kvm *kvm)
@@ -497,6 +502,9 @@ int kvm_arm_get_fw_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)case KVM_REG_ARM_VENDOR_HYP_BMAP : val =READ_ONCE (smccc_feat->vendor_hyp_bmap);break ; +case KVM_REG_ARM_VENDOR_HYP_BMAP_2 : + val =READ_ONCE (smccc_feat->vendor_hyp_bmap_2); +break ;default :return -ENOENT ; }
@@ -527,6 +535,10 @@ static int kvm_arm_set_fw_reg_bmap(struct kvm_vcpu *vcpu, u64 reg_id, u64 val) fw_reg_bmap = &smccc_feat->vendor_hyp_bmap; fw_reg_features =KVM_ARM_SMCCC_VENDOR_HYP_FEATURES ;break ; +case KVM_REG_ARM_VENDOR_HYP_BMAP_2 : + fw_reg_bmap = &smccc_feat->vendor_hyp_bmap_2; + fw_reg_features =KVM_ARM_SMCCC_VENDOR_HYP_FEATURES_2 ; +break ;default :return -ENOENT ; }
@@ -633,6 +645,7 @@ int kvm_arm_set_fw_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)case KVM_REG_ARM_STD_BMAP :case KVM_REG_ARM_STD_HYP_BMAP :case KVM_REG_ARM_VENDOR_HYP_BMAP : +case KVM_REG_ARM_VENDOR_HYP_BMAP_2 :return kvm_arm_set_fw_reg_bmap (vcpu, reg->id, val);default :return -ENOENT ;
Signature · 1 line
2.47.0
PATCH v8 4/6 arm64: Make _midr_in_range_list() an exported function
In reply to Shameer KolothumSubsequent patch will add target implementation CPU support and that will require _midr_in_range_list() to access new data. To avoid exporting the data make _midr_in_range_list() a normal function and export it.
No functional changes intended.
--- arch/arm64/include/asm/cputype.h | 15 +-------------- arch/arm64/kernel/cpu_errata.c | 15 +++++++++++++++ arch/arm64/kernel/image-vars.h | 1 + 3 files changed, 17 insertions(+), 14 deletions(-)
arch/arm64/include/asm/cputype.h +1 -14
@@ -276,20 +276,7 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } -static inline bool is_midr_in_range (struct midr_range const *range) -{ -return midr_is_cpu_model_range (read_cpuid_id (), range->model, - range->rv_min, range->rv_max); -} - -static inline bool -is_midr_in_range_list (struct midr_range const *ranges) -{ -while (ranges->model) -if (is_midr_in_range (ranges++)) -return true; -return false; -} +bool is_midr_in_range_list (struct midr_range const *ranges); static inline u64 __attribute_const__ read_cpuid_mpidr(void) {
arch/arm64/kernel/cpu_errata.c +15 -0
@@ -14,6 +14,21 @@#include <asm/kvm_asm.h> #include <asm/smp_plat.h> +static inline bool is_midr_in_range (struct midr_range const *range) +{ +return midr_is_cpu_model_range (read_cpuid_id (), range->model, + range->rv_min, range->rv_max); +} + +bool is_midr_in_range_list (struct midr_range const *ranges) +{ +while (ranges->model) +if (is_midr_in_range (ranges++)) +return true; +return false; +} +EXPORT_SYMBOL_GPL (is_midr_in_range_list); +static bool __maybe_unused __is_affected_midr_range (const struct arm64_cpu_capabilities *entry,u32 midr,u32 revidr)
arch/arm64/kernel/image-vars.h +1 -0
@@ -49,6 +49,7 @@ PROVIDE(__pi_arm64_sw_feature_override = arm64_sw_feature_override);PROVIDE (__pi_arm64_use_ng_mappings = arm64_use_ng_mappings);#ifdef CONFIG_CAVIUM_ERRATUM_27456 PROVIDE (__pi_cavium_erratum_27456_cpus = cavium_erratum_27456_cpus); +PROVIDE (__pi_is_midr_in_range_list = is_midr_in_range_list);#endif PROVIDE (__pi__ctype = _ctype);PROVIDE (__pi_memstart_offset_seed = memstart_offset_seed);
Signature · 1 line
2.47.0
Re: [PATCH v8 4/6] arm64: Make _midr_in_range_list() an exported function
Quoted 8 lines · Shameer Kolothum
Subsequent patch will add target implementation CPU support and that will require
_midr_in_range_list()to access new data. To avoid exporting the data make_midr_in_range_list()a normal function and export it.No functional changes intended.
Signed-off-by Shameer Kolothum <[email protected]>
Re: [BUG][PATCH v8 4/6] arm64: Make _midr_in_range_list() an exported function
In reply to Shameer KolothumHello,
I discovered that this patch breaks boot for some CPUs when building the default defconfig plus KASAN. This is still the case in v6.15-rc1 and rc2.
This patch marks is_midr_in_range_list as position independent but it isn't, breaking early boot when instrumented with KASAN and CONFIG_RANDOMIZE_BASE enabled.
The breaking usage seems to be in kaslr_requires_kpti() called in early_map_kernel().
My testing on an AMD Seattle board does crash, but newer machines implementing E0PD do not crash as they do not need to check MIDRs in kaslr_requires_kpti(). is_mdr_in_range_list did work in PI code previously because it was inline, which this patch changes.
Best regards,
Ada Couprie Diaz
Quoted 78 lines · Shameer Kolothum
Subsequent patch will add target implementation CPU support and that will require
_midr_in_range_list()to access new data. To avoid exporting the data make_midr_in_range_list()a normal function and export it.No functional changes intended.
Signed-off-by Shameer Kolothum <[email protected]>--- arch/arm64/include/asm/cputype.h | 15 +-------------- arch/arm64/kernel/cpu_errata.c | 15 +++++++++++++++ arch/arm64/kernel/image-vars.h | 1 + 3 files changed, 17 insertions(+), 14 deletions(-)arch/arm64/include/asm/cputype.h +1
-14@@ -276,20 +276,7 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } -static inline bool is_midr_in_range (struct midr_range const *range) -{ -return midr_is_cpu_model_range (read_cpuid_id (), range->model, - range->rv_min, range->rv_max); -} - -static inline bool -is_midr_in_range_list (struct midr_range const *ranges) -{ -while (ranges->model) -if (is_midr_in_range (ranges++)) -return true; -return false; -} +bool is_midr_in_range_list (struct midr_range const *ranges); static inline u64 __attribute_const__ read_cpuid_mpidr(void) {arch/arm64/kernel/cpu_errata.c +15
-0@@ -14,6 +14,21 @@#include <asm/kvm_asm.h> #include <asm/smp_plat.h> +static inline bool is_midr_in_range (struct midr_range const *range) +{ +return midr_is_cpu_model_range (read_cpuid_id (), range->model, + range->rv_min, range->rv_max); +} + +bool is_midr_in_range_list (struct midr_range const *ranges) +{ +while (ranges->model) +if (is_midr_in_range (ranges++)) +return true; +return false; +} +EXPORT_SYMBOL_GPL (is_midr_in_range_list); +static bool __maybe_unused __is_affected_midr_range (const struct arm64_cpu_capabilities *entry,u32 midr,u32 revidr)arch/arm64/kernel/image-vars.h +1
-0@@ -49,6 +49,7 @@ PROVIDE(__pi_arm64_sw_feature_override = arm64_sw_feature_override);PROVIDE (__pi_arm64_use_ng_mappings = arm64_use_ng_mappings);#ifdef CONFIG_CAVIUM_ERRATUM_27456 PROVIDE (__pi_cavium_erratum_27456_cpus = cavium_erratum_27456_cpus); +PROVIDE (__pi_is_midr_in_range_list = is_midr_in_range_list);#endif PROVIDE (__pi__ctype = _ctype);PROVIDE (__pi_memstart_offset_seed = memstart_offset_seed);
RE: [BUG][PATCH v8 4/6] arm64: Make _midr_in_range_list() an exported function
Hi,
Quoted 16 lines
-----Original Message-----
From: Ada Couprie Diaz <[email protected]>
Sent: Tuesday, April 15, 2025 11:58 AM
To: Shameerali Kolothum Thodi <[email protected]>; [email protected]; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; yuzenghui <[email protected]>; Wangzhou (B) <[email protected]>; jiangkunkun <[email protected]>; Jonathan Cameron <[email protected]>; Anthony Jebson <[email protected]>; [email protected]; Linuxarm <[email protected]>
Subject: Re: [BUG][PATCH v8 4/6] arm64: Make_midr_in_range_list()an exported function
Quoted 4 lines
Hello,
I discovered that this patch breaks boot for some CPUs when building the default defconfig plus KASAN. This is still the case in v6.15-rc1 and rc2.
Thanks for testing and reporting this.
Quoted 4 lines
This patch marks
is_midr_in_range_listas position independent but it isn't, breaking early boot when instrumented with KASAN andCONFIG_RANDOMIZE_BASEenabled.
I am not sure what else is needed to make this position independent other than,
PROVIDE(__pi_is_midr_in_range_list = is_midr_in_range_list);
May be since this is now a text session that we are making PI, it is probably missing some other bits. My understanding of that part of the code is very limited.
Marc/Oliver,
Any thoughts?
Quoted 8 lines
The breaking usage seems to be in
kaslr_requires_kpti()called inearly_map_kernel().
My testing on an AMD Seattle board does crash, but newer machines implementing E0PD do not crash as they do not need to check MIDRs inkaslr_requires_kpti().is_mdr_in_range_listdid work in PI code previously because it wasinline, which this patch changes.
Yeah. The change from inline happened after the v7[0]. May we could go back to that so that only data needs to be PI if fixing above is invasive.
I just tested with the v7 version and it boots with KASAN and CONFIG_RANDOMIZE_BASE.
Thanks,
Shameer
[0] https://lore.kernel.org/kvmarm/[email protected]/
Quoted 90 lines
Best regards,
Ada Couprie DiazQuoted 8 lines · Shameer Kolothum
Subsequent patch will add target implementation CPU support and that will require
_midr_in_range_list()to access new data. To avoid exporting the data make_midr_in_range_list()a normal function and export it.No functional changes intended.
Signed-off-by Shameer KolothumReview
Quoted 7 lines
--- arch/arm64/include/asm/cputype.h | 15 +-------------- arch/arm64/kernel/cpu_errata.c | 15 +++++++++++++++ arch/arm64/kernel/image-vars.h | 1 + 3 files changed, 17 insertions(+), 14 deletions(-)(quoted diff) +0
-0
b/arch/arm64/include/asm/cputype.harch/arm64/include/asm/cputype.h static inline bool midr_is_cpu_model_range(u32
Quoted 4 lines
index
2a76f0e30006..ccb4a155d118100644arch/arm64/include/asm/cputype.h +0
-0@@ -276,20 +276,7 @@ static inline bool midr_is_cpu_model_range(u32midr, u32 model, u32
rv_min,Review
Quoted 22 lines
arch/arm64/include/asm/cputype.h +1
-14return _model == model && rv >= rv_min && rv <= rv_max; } -static inline bool is_midr_in_range (struct midr_range const *range) -{ -return midr_is_cpu_model_range (read_cpuid_id (), range->model, - range->rv_min, range->rv_max); -} - -static inline bool -is_midr_in_range_list (struct midr_range const *ranges) -{ -while (ranges->model) -if (is_midr_in_range (ranges++)) -return true; -return false; -} +bool is_midr_in_range_list (struct midr_range const *ranges); static inline u64 __attribute_const__ read_cpuid_mpidr(void) {(quoted diff) +0
-0
b/arch/arm64/kernel/cpu_errata.carch/arm64/kernel/image-
Quoted 26 lines
index
99b55893fc4e..1f51cf6378c5100644arch/arm64/kernel/cpu_errata.c +15
-0@@ -14,6 +14,21 @@#include <asm/kvm_asm.h> #include <asm/smp_plat.h> +static inline bool is_midr_in_range (struct midr_range const *range) +{ +return midr_is_cpu_model_range (read_cpuid_id (), range->model, + range->rv_min, range->rv_max); +} + +bool is_midr_in_range_list (struct midr_range const *ranges) +{ +while (ranges->model) +if (is_midr_in_range (ranges++)) +return true; +return false; +} +EXPORT_SYMBOL_GPL (is_midr_in_range_list); +static bool __maybe_unused __is_affected_midr_range (const struct arm64_cpu_capabilities *entry,u32 midr,u32 revidr)arch/arm64/kernel/image- renamed from arch/arm64/kernel/image-vars.h +0
-0
vars.harch/arm64/kernel/image-vars.h PROVIDE(__pi_arm64_sw_feature_override =
Quoted 4 lines
index
ef3a69cc398e..de3d081e5a57100644arch/arm64/kernel/image-vars.h +0
-0@@ -49,6 +49,7 @@ PROVIDE(__pi_arm64_sw_feature_override =
arm64_sw_feature_override);arch/arm64/kernel/image-vars.h PROVIDE(__pi_arm64_sw_feature_override = arm64_sw_feature_override);
Quoted 3 lines
arch/arm64/kernel/image-vars.h +0
-0PROVIDE (__pi_arm64_use_ng_mappings = arm64_use_ng_mappings);#ifdef CONFIG_CAVIUM_ERRATUM_27456 PROVIDE(__pi_cavium_erratum_27456_cpus =
cavium_erratum_27456_cpus);Quoted 4 lines
arch/arm64/kernel/image-vars.h +1
-0+PROVIDE (__pi_is_midr_in_range_list = is_midr_in_range_list);#endif PROVIDE (__pi__ctype = _ctype);PROVIDE (__pi_memstart_offset_seed = memstart_offset_seed);
Marc Zyngier
maintainerRe: [BUG][PATCH v8 4/6] arm64: Make _midr_in_range_list() an exported function
In reply to Ada Couprie DiazQuoted 11 lines · Ada Couprie Diaz
Hello,
I discovered that this patch breaks boot for some CPUs when building the default defconfig plus KASAN. This is still the case in v6.15-rc1 and rc2.
This patch marks
is_midr_in_range_listas position independent but it isn't, breaking early boot when instrumented with KASAN andCONFIG_RANDOMIZE_BASEenabled.
Quoted 7 lines
The breaking usage seems to be in
kaslr_requires_kpti()called inearly_map_kernel().
My testing on an AMD Seattle board does crash, but newer machines implementing E0PD do not crash as they do not need to check MIDRs inkaslr_requires_kpti().is_mdr_in_range_listdid work in PI code previously because it wasinline, which this patch changes.
OK, this is much more of a pain than I thought.
I tried bringing the various helpers into the PI section, but it ended up being extremely ugly.
More importantly, this is something that is pretty much at odds with the whole idea of the MIDR override -- it happens way earlier than we can populate the table.
The thing is, the only reason we need to do this is that we need to support the Cavium SEFAC (Sorry Excuse For A Computer) that cannot run with KPTI.
I can restore harmony with the following hack. But maybe we should just prune TX from the kernel and be done with this contraption.
M.
arch/arm64/include/asm/cputype.h +14 -0
@@ -292,6 +292,20 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } +static inline bool __is_midr_in_range (struct midr_range const *range) +{ +return midr_is_cpu_model_range (read_cpuid_id (), range->model, + range->rv_min, range->rv_max); +} + +static inline bool __is_midr_in_range_list (struct midr_range const *ranges) +{ +while (ranges->model) +if (__is_midr_in_range (ranges++)) +return true; +return false; +} +struct target_impl_cpu {u64 midr;u64 revidr;
arch/arm64/include/asm/mmu.h +29 -6
@@ -72,6 +72,17 @@ extern void create_pgd_mapping(struct mm_struct *mm, phys_addr_t phys,extern void *fixmap_remap_fdt (phys_addr_t dt_phys,int *size,pgprot_t prot);extern void mark_linear_text_alias_ro (void ); +static inline bool cpu_has_e0pd (void ) +{ +if (IS_ENABLED (CONFIG_ARM64_E0PD )) { +u64 mmfr2 =read_sysreg_s (SYS_ID_AA64MMFR2_EL1 ); +return (cpuid_feature_extract_unsigned_field (mmfr2, +ID_AA64MMFR2_EL1_E0PD_SHIFT )); + } + +return false; +} +/* * This check is triggered during the early boot before the cpufeature * is initialised. Checking the status on the local CPU allows the boot
@@ -87,12 +98,8 @@ static inline bool kaslr_requires_kpti(void)* E0PD does a similar job to KPTI so can be used instead * where available. */ -if (IS_ENABLED (CONFIG_ARM64_E0PD )) { -u64 mmfr2 =read_sysreg_s (SYS_ID_AA64MMFR2_EL1 ); -if (cpuid_feature_extract_unsigned_field (mmfr2, -ID_AA64MMFR2_EL1_E0PD_SHIFT )) -return false; - } +if (cpu_has_e0pd ()) +return false;/* * Systems affected by Cavium erratum 24756 are incompatible
@@ -108,5 +115,21 @@ static inline bool kaslr_requires_kpti(void)return true; } +/* Same as the above, but limited to the local CPU, ignoring the MIDR list */ +static inline bool __kaslr_requires_kpti (void ) +{ +if (cpu_has_e0pd ()) +return false; + +if (IS_ENABLED (CONFIG_CAVIUM_ERRATUM_27456 )) { +extern const struct midr_range cavium_erratum_27456_cpus[]; + +if (__is_midr_in_range_list (cavium_erratum_27456_cpus)) +return false; + } + +return true; +} +#endif /* !__ASSEMBLY__ */ #endif
arch/arm64/kernel/cpu_errata.c +1 -2
@@ -32,8 +32,7 @@ static inline bool is_midr_in_range(struct midr_range const *range)int i;if (!target_impl_cpu_num) -return midr_is_cpu_model_range (read_cpuid_id (), range->model, - range->rv_min, range->rv_max); +return __is_midr_in_range (range);for (i =0 ; i < target_impl_cpu_num; i++) { if (midr_is_cpu_model_range(target_impl_cpus[i].midr,
arch/arm64/kernel/pi/map_kernel.c +1 -1
@@ -245,7 +245,7 @@ asmlinkage void __init early_map_kernel(u64 boot_status, void *fdt)if (IS_ENABLED (CONFIG_RANDOMIZE_BASE )) {u64 kaslr_seed =kaslr_early_init (fdt, chosen); -if (kaslr_seed &&kaslr_requires_kpti ()) +if (kaslr_seed &&__kaslr_requires_kpti ()) arm64_use_ng_mappings = true; kaslr_offset |= kaslr_seed & ~(MIN_KIMG_ALIGN -1 );
Signature · 1 line
Without deviation from the norm, progress is not possible.
Catalin Marinas
maintainerRe: [BUG][PATCH v8 4/6] arm64: Make _midr_in_range_list() an exported function
Quoted 27 lines · Marc Zyngier
Quoted 8 lines · Ada Couprie Diaz
I discovered that this patch breaks boot for some CPUs when building the default defconfig plus KASAN. This is still the case in v6.15-rc1 and rc2.
This patch marks
is_midr_in_range_listas position independent but it isn't, breaking early boot when instrumented with KASAN andCONFIG_RANDOMIZE_BASEenabled.Quoted 7 lines
The breaking usage seems to be in
kaslr_requires_kpti()called inearly_map_kernel().
My testing on an AMD Seattle board does crash, but newer machines implementing E0PD do not crash as they do not need to check MIDRs inkaslr_requires_kpti().is_mdr_in_range_listdid work in PI code previously because it wasinline, which this patch changes.OK, this is much more of a pain than I thought.
I tried bringing the various helpers into the PI section, but it ended up being extremely ugly.
More importantly, this is something that is pretty much at odds with the whole idea of the MIDR override -- it happens way earlier than we can populate the table.
Quoted 6 lines
The thing is, the only reason we need to do this is that we need to support the Cavium SEFAC (Sorry Excuse For A Computer) that cannot run with KPTI.
I can restore harmony with the following hack. But maybe we should just prune TX from the kernel and be done with this contraption.
That works for me. Well, we can still keep it around but panic on boot if kpti is enabled on this platform. Just don't bother checking it via early_map_kernel() from the PI code.
Signature · 1 line
Catalin
Marc Zyngier
maintainerRe: [BUG][PATCH v8 4/6] arm64: Make _midr_in_range_list() an exported function
Quoted 33 lines · Catalin Marinas
Quoted 31 lines · Marc Zyngier
Quoted 8 lines · Ada Couprie Diaz
I discovered that this patch breaks boot for some CPUs when building the default defconfig plus KASAN. This is still the case in v6.15-rc1 and rc2.
This patch marks
is_midr_in_range_listas position independent but it isn't, breaking early boot when instrumented with KASAN andCONFIG_RANDOMIZE_BASEenabled.Quoted 7 lines
The breaking usage seems to be in
kaslr_requires_kpti()called inearly_map_kernel().
My testing on an AMD Seattle board does crash, but newer machines implementing E0PD do not crash as they do not need to check MIDRs inkaslr_requires_kpti().is_mdr_in_range_listdid work in PI code previously because it wasinline, which this patch changes.OK, this is much more of a pain than I thought.
I tried bringing the various helpers into the PI section, but it ended up being extremely ugly.
More importantly, this is something that is pretty much at odds with the whole idea of the MIDR override -- it happens way earlier than we can populate the table.
The thing is, the only reason we need to do this is that we need to support the Cavium SEFAC (Sorry Excuse For A Computer) that cannot run with KPTI.
Quoted 6 lines
Quoted 2 lines
I can restore harmony with the following hack. But maybe we should just prune TX from the kernel and be done with this contraption.
That works for me. Well, we can still keep it around but panic on boot if kpti is enabled on this platform. Just don't bother checking it via
early_map_kernel()from the PI code.
early_map_kernel() immediately builds nG page tables, and that's a strong guarantee to catch fire on this machine. Which means that distro kernels (which all have KPTI enabled by default) would explode.
The only way to avoid it would be to mandate that users of these machines have nokaslr on the command-line. I'm happy to do so on my box, but I'm hardly representative of arm64 users...
M.
Signature · 1 line
Without deviation from the norm, progress is not possible.
PATCH v8 5/6 smccc/kvm_guest: Enable errata based on implementation CPUs
In reply to Shameer KolothumRetrieve any migration target implementation CPUs using the hypercall and enable associated errata.
--- arch/arm64/include/asm/cputype.h | 7 ++++ arch/arm64/include/asm/hypervisor.h | 1 + arch/arm64/kernel/cpu_errata.c | 45 +++++++++++++++++--- arch/arm64/kernel/cpufeature.c | 2 + drivers/firmware/smccc/kvm_guest.c | 64 +++++++++++++++++++++++++++++ 5 files changed, 114 insertions(+), 5 deletions(-)
arch/arm64/include/asm/cputype.h +7 -0
@@ -276,6 +276,13 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } +struct target_impl_cpu { +u64 midr; +u64 revidr; +u64 aidr; +}; + +bool cpu_errata_set_target_impl (u64 num,void *impl_cpus);bool is_midr_in_range_list (struct midr_range const *ranges); static inline u64 __attribute_const__ read_cpuid_mpidr(void)
arch/arm64/include/asm/hypervisor.h +1 -0
@@ -6,6 +6,7 @@void kvm_init_hyp_services (void );bool kvm_arm_hyp_service_available (u32 func_id); +void kvm_arm_target_impl_cpu_init (void );#ifdef CONFIG_ARM_PKVM_GUEST void pkvm_init_hyp_services (void );
arch/arm64/kernel/cpu_errata.c +40 -5
59 lines not on this page · raw message
arch/arm64/kernel/cpufeature.c +2 -0
@@ -86,6 +86,7 @@#include <asm/kvm_host.h> #include <asm/mmu_context.h> #include <asm/mte.h> +#include <asm/hypervisor.h> #include <asm/processor.h> #include <asm/smp.h> #include <asm/sysreg.h>
@@ -3679,6 +3680,7 @@ unsigned long cpu_get_elf_hwcap3(void)static void __init setup_boot_cpu_capabilities (void ) { +kvm_arm_target_impl_cpu_init ();/* * The boot CPU's feature register values have been recorded. Detect * boot cpucaps and local cpucaps for the boot CPU, then enable and
drivers/firmware/smccc/kvm_guest.c +64 -0
75 lines not on this page · raw message
Signature · 1 line
2.47.0
Re: [PATCH v8 5/6] smccc/kvm_guest: Enable errata based on implementation CPUs
Quoted 6 lines · Shameer Kolothum
Retrieve any migration target implementation CPUs using the hypercall and enable associated errata.
Reviewed-by Cornelia Huck <[email protected]>Reviewed-by Sebastian Ott <[email protected]>Signed-off-by Shameer Kolothum <[email protected]>
I can see ways in which this can go wrong but, in the absence of a better idea:
Re: [PATCH v8 5/6] smccc/kvm_guest: Enable errata based on implementation CPUs
In reply to Shameer KolothumQuoted 97 lines · Shameer Kolothum
Retrieve any migration target implementation CPUs using the hypercall and enable associated errata.
Reviewed-by Cornelia Huck <[email protected]>Reviewed-by Sebastian Ott <[email protected]>Signed-off-by Shameer Kolothum <[email protected]>---
arch/arm64/include/asm/cputype.h | 7 ++++ arch/arm64/include/asm/hypervisor.h | 1 + arch/arm64/kernel/cpu_errata.c | 45 +++++++++++++++++--- arch/arm64/kernel/cpufeature.c | 2 + drivers/firmware/smccc/kvm_guest.c | 64 +++++++++++++++++++++++++++++ 5 files changed, 114 insertions(+), 5 deletions(-)arch/arm64/include/asm/cputype.h +7
-0@@ -276,6 +276,13 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } +struct target_impl_cpu { +u64 midr; +u64 revidr; +u64 aidr; +}; + +bool cpu_errata_set_target_impl (u64 num,void *impl_cpus);bool is_midr_in_range_list (struct midr_range const *ranges); static inline u64 __attribute_const__ read_cpuid_mpidr(void)arch/arm64/include/asm/hypervisor.h +1
-0@@ -6,6 +6,7 @@void kvm_init_hyp_services (void );bool kvm_arm_hyp_service_available (u32 func_id); +void kvm_arm_target_impl_cpu_init (void );#ifdef CONFIG_ARM_PKVM_GUEST void pkvm_init_hyp_services (void );arch/arm64/kernel/cpu_errata.c +26
-2@@ -14,10 +14,34 @@#include <asm/kvm_asm.h> #include <asm/smp_plat.h> +static u64 target_impl_cpu_num; +static struct target_impl_cpu *target_impl_cpus; + +bool cpu_errata_set_target_impl (u64 num,void *impl_cpus) +{ +if (target_impl_cpu_num || !num || !impl_cpus) +return false; + + target_impl_cpu_num = num; + target_impl_cpus = impl_cpus; +return true; +} +static inline bool is_midr_in_range (struct midr_range const *range) { -return midr_is_cpu_model_range (read_cpuid_id (), range->model, - range->rv_min, range->rv_max); +int i; + +if (!target_impl_cpu_num) +return midr_is_cpu_model_range (read_cpuid_id (), range->model, + range->rv_min, range->rv_max); + +for (i =0 ; i < target_impl_cpu_num; i++) { +if (midr_is_cpu_model_range (target_impl_cpus[i].midr, + range->model, + range->rv_min, range->rv_max)) +return true; + } +return false; }bool is_midr_in_range_list (struct midr_range const *ranges)@@ -47,9 +71,20 @@ __is_affected_midr_range(const struct arm64_cpu_capabilities *entry,static bool
__maybe_unusedis_affected_midr_range(const struct arm64_cpu_capabilities *entry, int scope){
-WARN_ON(scope != SCOPE_LOCAL_CPU || preemptible());
- return __is_affected_midr_range(entry,read_cpuid_id(),
-read_cpuid(REVIDR_EL1));
+ int i;
arch/arm64/kernel/cpu_errata.c __is_affected_midr_range(const struct arm64_cpu_capabilities *entry, · line 84
Quoted 10 lines
arch/arm64/kernel/cpu_errata.c +10
-0+ +if (!target_impl_cpu_num) { +WARN_ON (scope !=SCOPE_LOCAL_CPU ||preemptible ()); +return __is_affected_midr_range (entry,read_cpuid_id (), +read_cpuid (REVIDR_EL1 )); + } + +for (i =0 ; i < target_impl_cpu_num; i++) { +if (__is_affected_midr_range (entry, target_impl_cpus[i].midr, + target_impl_cpus[i].midr))
Typo? Shouldn't the second arg be target_impl_cpus[i].revidr?
Quoted 111 lines
+ return true;
+ }
+ return false;
}static bool
__maybe_unusedarch/arm64/kernel/cpufeature.c +2
-0@@ -86,6 +86,7 @@#include <asm/kvm_host.h> #include <asm/mmu_context.h> #include <asm/mte.h> +#include <asm/hypervisor.h> #include <asm/processor.h> #include <asm/smp.h> #include <asm/sysreg.h> @@ -3679,6 +3680,7 @@ unsigned long cpu_get_elf_hwcap3(void)static void __init setup_boot_cpu_capabilities (void ) { +kvm_arm_target_impl_cpu_init ();/* * The boot CPU's feature register values have been recorded. Detect * boot cpucaps and local cpucaps for the boot CPU, then enable and drivers/firmware/smccc/kvm_guest.c +64
-0@@ -6,8 +6,11 @@#include <linux/bitmap.h> #include <linux/cache.h> #include <linux/kernel.h> +#include <linux/memblock.h> #include <linux/string.h> +#include <uapi/linux/psci.h> +#include <asm/hypervisor.h> static DECLARE_BITMAP(__kvm_arm_hyp_services, ARM_SMCCC_KVM_NUM_FUNCS) __ro_after_init = { };@@ -51,3 +54,64 @@ bool kvm_arm_hyp_service_available(u32 func_id)return test_bit (func_id, __kvm_arm_hyp_services); }EXPORT_SYMBOL_GPL (kvm_arm_hyp_service_available); + +void __init kvm_arm_target_impl_cpu_init (void ) +{ +int i; +u32 ver; +u64 max_cpus; +struct arm_smccc_res res; +struct target_impl_cpu *target; + +if (!kvm_arm_hyp_service_available (ARM_SMCCC_KVM_FUNC_DISCOVER_IMPL_VER ) || + !kvm_arm_hyp_service_available (ARM_SMCCC_KVM_FUNC_DISCOVER_IMPL_CPUS )) +return ; + +arm_smccc_1_1_invoke (ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_VER_FUNC_ID , +0 , &res); +if (res.a0 !=SMCCC_RET_SUCCESS ) +return ; + +/* Version info is in lower 32 bits and is in SMMCCC_VERSION format */ + ver =lower_32_bits (res.a1); +if (PSCI_VERSION_MAJOR (ver) !=1 ) { +pr_warn ("Unsupported target CPU implementation version v%d.%d\n" , +PSCI_VERSION_MAJOR (ver),PSCI_VERSION_MINOR (ver)); +return ; + } + +if (!res.a2) { +pr_warn ("No target implementation CPUs specified\n" ); +return ; + } + + max_cpus = res.a2; + target = memblock_alloc(sizeof(*target) * max_cpus, __alignof__(*target)); +if (!target) { +pr_warn ("Not enough memory for struct target_impl_cpu\n" ); +return ; + } + +for (i =0 ; i < max_cpus; i++) { +arm_smccc_1_1_invoke (ARM_SMCCC_VENDOR_HYP_KVM_DISCOVER_IMPL_CPUS_FUNC_ID , + i, &res); +if (res.a0 !=SMCCC_RET_SUCCESS ) { +pr_warn ("Discovering target implementation CPUs failed\n" ); + goto mem_free; + } + target[i].midr = res.a1; + target[i].revidr = res.a2; + target[i].aidr = res.a3; + }; + +if (!cpu_errata_set_target_impl (max_cpus, target)) { +pr_warn ("Failed to set target implementation CPUs\n" ); + goto mem_free; + } + +pr_info ("Number of target implementation CPUs is %lld\n" , max_cpus); +return ; + +mem_free: +memblock_free (target,sizeof (*target) * max_cpus); +}Signature · 1 line
2.47.0
Marc Zyngier
maintainerRe: [PATCH v8 5/6] smccc/kvm_guest: Enable errata based on implementation CPUs
Quoted 101 lines · Khushit Shah
Quoted 101 lines · Shameer Kolothum
Retrieve any migration target implementation CPUs using the hypercall and enable associated errata.
Reviewed-by Cornelia Huck <[email protected]>Reviewed-by Sebastian Ott <[email protected]>Signed-off-by Shameer Kolothum <[email protected]>---
arch/arm64/include/asm/cputype.h | 7 ++++ arch/arm64/include/asm/hypervisor.h | 1 + arch/arm64/kernel/cpu_errata.c | 45 +++++++++++++++++--- arch/arm64/kernel/cpufeature.c | 2 + drivers/firmware/smccc/kvm_guest.c | 64 +++++++++++++++++++++++++++++ 5 files changed, 114 insertions(+), 5 deletions(-)arch/arm64/include/asm/cputype.h +7
-0@@ -276,6 +276,13 @@ static inline bool midr_is_cpu_model_range(u32 midr, u32 model, u32 rv_min,return _model == model && rv >= rv_min && rv <= rv_max; } +struct target_impl_cpu { +u64 midr; +u64 revidr; +u64 aidr; +}; + +bool cpu_errata_set_target_impl (u64 num,void *impl_cpus);bool is_midr_in_range_list (struct midr_range const *ranges); static inline u64 __attribute_const__ read_cpuid_mpidr(void)arch/arm64/include/asm/hypervisor.h +1
-0@@ -6,6 +6,7 @@void kvm_init_hyp_services (void );bool kvm_arm_hyp_service_available (u32 func_id); +void kvm_arm_target_impl_cpu_init (void );#ifdef CONFIG_ARM_PKVM_GUEST void pkvm_init_hyp_services (void );arch/arm64/kernel/cpu_errata.c +26
-2@@ -14,10 +14,34 @@#include <asm/kvm_asm.h> #include <asm/smp_plat.h> +static u64 target_impl_cpu_num; +static struct target_impl_cpu *target_impl_cpus; + +bool cpu_errata_set_target_impl (u64 num,void *impl_cpus) +{ +if (target_impl_cpu_num || !num || !impl_cpus) +return false; + + target_impl_cpu_num = num; + target_impl_cpus = impl_cpus; +return true; +} +static inline bool is_midr_in_range (struct midr_range const *range) { -return midr_is_cpu_model_range (read_cpuid_id (), range->model, - range->rv_min, range->rv_max); +int i; + +if (!target_impl_cpu_num) +return midr_is_cpu_model_range (read_cpuid_id (), range->model, + range->rv_min, range->rv_max); + +for (i =0 ; i < target_impl_cpu_num; i++) { +if (midr_is_cpu_model_range (target_impl_cpus[i].midr, + range->model, + range->rv_min, range->rv_max)) +return true; + } +return false; }bool is_midr_in_range_list (struct midr_range const *ranges)@@ -47,9 +71,20 @@ __is_affected_midr_range(const struct arm64_cpu_capabilities *entry,static bool
__maybe_unusedis_affected_midr_range(const struct arm64_cpu_capabilities *entry, int scope){
-WARN_ON(scope != SCOPE_LOCAL_CPU || preemptible());
- return __is_affected_midr_range(entry,read_cpuid_id(),
-read_cpuid(REVIDR_EL1));
+ int i;
+
+ if (!target_impl_cpu_num) {
+WARN_ON(scope != SCOPE_LOCAL_CPU || preemptible());
Quoted 8 lines
arch/arm64/kernel/cpu_errata.c __is_affected_midr_range(const struct arm64_cpu_capabilities *entry, · line 84
Quoted 7 lines
+ return __is_affected_midr_range(entry,
read_cpuid_id(),
+read_cpuid(REVIDR_EL1));
+ }arch/arm64/kernel/cpu_errata.c +4
-0+ +for (i =0 ; i < target_impl_cpu_num; i++) { +if (__is_affected_midr_range (entry, target_impl_cpus[i].midr, + target_impl_cpus[i].midr))Typo? Shouldn't the second arg be target_impl_cpus[i].revidr?
Most probably. Please send a fix.
M.
Signature · 1 line
Jazz isn't dead. It just smells funny.
PATCH v8 6/6 KVM: selftests: Add test for KVM_REG_ARM_VENDOR_HYP_BMAP_2
In reply to Shameer KolothumOne difference here with other pseudo-firmware bitmap registers is that the default/reset value for the supported hypercall function-ids is 0 at present. Hence, modify the test accordingly.
--- tools/arch/arm64/include/uapi/asm/kvm.h | 12 +++++ .../selftests/kvm/arm64/get-reg-list.c | 1 + .../testing/selftests/kvm/arm64/hypercalls.c | 46 +++++++++++++++---- 3 files changed, 49 insertions(+), 10 deletions(-)
tools/arch/arm64/include/uapi/asm/kvm.h +12 -0
@@ -374,6 +374,7 @@ enum {#endif }; +/* Vendor hyper call function numbers 0-63 */ #define KVM_REG_ARM_VENDOR_HYP_BMAP KVM_REG_ARM_FW_FEAT_BMAP_REG(2)enum {
@@ -384,6 +385,17 @@ enum {#endif }; +/* Vendor hyper call function numbers 64-127 */ +#define KVM_REG_ARM_VENDOR_HYP_BMAP_2 KVM_REG_ARM_FW_FEAT_BMAP_REG(3) + +enum { +KVM_REG_ARM_VENDOR_HYP_BIT_DISCOVER_IMPL_VER =0 , +KVM_REG_ARM_VENDOR_HYP_BIT_DISCOVER_IMPL_CPUS =1 , +#ifdef __KERNEL__ +KVM_REG_ARM_VENDOR_HYP_BMAP_2_BIT_COUNT , +#endif +}; +/* Device Control API on vm fd */ #define KVM_ARM_VM_SMCCC_CTRL 0#define KVM_ARM_VM_SMCCC_FILTER 0
tools/testing/selftests/kvm/arm64/get-reg-list.c +1 -0
@@ -332,6 +332,7 @@ static __u64 base_regs[] = {KVM_REG_ARM_FW_FEAT_BMAP_REG (0 ),/* KVM_REG_ARM_STD_BMAP */ KVM_REG_ARM_FW_FEAT_BMAP_REG (1 ),/* KVM_REG_ARM_STD_HYP_BMAP */ KVM_REG_ARM_FW_FEAT_BMAP_REG (2 ),/* KVM_REG_ARM_VENDOR_HYP_BMAP */ +KVM_REG_ARM_FW_FEAT_BMAP_REG (3 ),/* KVM_REG_ARM_VENDOR_HYP_BMAP_2 */ ARM64_SYS_REG (3 ,3 ,14 ,3 ,1 ),/* CNTV_CTL_EL0 */ ARM64_SYS_REG (3 ,3 ,14 ,3 ,2 ),/* CNTV_CVAL_EL0 */ ARM64_SYS_REG(3, 3, 14, 0, 2),
tools/testing/selftests/kvm/arm64/hypercalls.c +36 -10
80 lines not on this page · raw message
Signature · 1 line
2.47.0
Oliver Upton
AppliedQuoted 9 lines · Shameer Kolothum
v7-->v8
https://lore.kernel.org/kvmarm/[email protected]/Changes:
-Fixed the compile error issue due toCONFIG_CORESIGHT_SOURCE_ETM4X=m
(Thanks to Sebastian for checking that config)- patch #4
-Added R-by tags.[...]
The feature branch for all of this can be found here should the need arise.
git://git.kernel.org/pub/scm/linux/kernel/git/oupton/linux.git kvm-arm64/pv-cpuid
Applied to next, thanks!
[1/6] arm64: Modify _midr_range() functions to read MIDR/REVIDR internally
https://git.kernel.org/kvmarm/kvmarm/c/e3121298c7fc [2/6] KVM: arm64: Introduce hypercall support for retrieving target implementations https://git.kernel.org/kvmarm/kvmarm/c/57e5cc9b8a39 [3/6] KVM: arm64: Introduce KVM_REG_ARM_VENDOR_HYP_BMAP_2 https://git.kernel.org/kvmarm/kvmarm/c/c0000e58c74e [4/6] arm64: Make _midr_in_range_list() an exported function https://git.kernel.org/kvmarm/kvmarm/c/c8c2647e69be [5/6] smccc/kvm_guest: Enable errata based on implementation CPUs https://git.kernel.org/kvmarm/kvmarm/c/86edf6bdcf05 [6/6] KVM: selftests: Add test for KVM_REG_ARM_VENDOR_HYP_BMAP_2 https://git.kernel.org/kvmarm/kvmarm/c/f69656656fa7
--
Best,
Oliver