diff options
Diffstat (limited to 'arch/x86/kernel/fpu/xstate.c')
| -rw-r--r-- | arch/x86/kernel/fpu/xstate.c | 602 |
1 files changed, 409 insertions, 193 deletions
diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c index 59e543b95a3c..48113c5193aa 100644 --- a/arch/x86/kernel/fpu/xstate.c +++ b/arch/x86/kernel/fpu/xstate.c @@ -8,21 +8,28 @@ #include <linux/compat.h> #include <linux/cpu.h> #include <linux/mman.h> +#include <linux/kvm_types.h> #include <linux/nospec.h> #include <linux/pkeys.h> #include <linux/seq_file.h> #include <linux/proc_fs.h> #include <linux/vmalloc.h> +#include <linux/coredump.h> +#include <linux/sort.h> #include <asm/fpu/api.h> #include <asm/fpu/regset.h> #include <asm/fpu/signal.h> #include <asm/fpu/xcr.h> +#include <asm/cpuid/api.h> +#include <asm/msr.h> #include <asm/tlbflush.h> #include <asm/prctl.h> #include <asm/elf.h> +#include <uapi/asm/elf.h> + #include "context.h" #include "internal.h" #include "legacy.h" @@ -39,26 +46,27 @@ */ static const char *xfeature_names[] = { - "x87 floating point registers" , - "SSE registers" , - "AVX registers" , - "MPX bounds registers" , - "MPX CSR" , - "AVX-512 opmask" , - "AVX-512 Hi256" , - "AVX-512 ZMM_Hi256" , - "Processor Trace (unused)" , + "x87 floating point registers", + "SSE registers", + "AVX registers", + "MPX bounds registers", + "MPX CSR", + "AVX-512 opmask", + "AVX-512 Hi256", + "AVX-512 ZMM_Hi256", + "Processor Trace (unused)", "Protection Keys User registers", "PASID state", - "unknown xstate feature" , - "unknown xstate feature" , - "unknown xstate feature" , - "unknown xstate feature" , - "unknown xstate feature" , - "unknown xstate feature" , - "AMX Tile config" , - "AMX Tile data" , - "unknown xstate feature" , + "Control-flow User registers", + "Control-flow Kernel registers (KVM only)", + "unknown xstate feature", + "unknown xstate feature", + "unknown xstate feature", + "unknown xstate feature", + "AMX Tile config", + "AMX Tile data", + "APX registers", + "unknown xstate feature", }; static unsigned short xsave_cpuid_features[] __initdata = { @@ -71,10 +79,13 @@ static unsigned short xsave_cpuid_features[] __initdata = { [XFEATURE_ZMM_Hi256] = X86_FEATURE_AVX512F, [XFEATURE_Hi16_ZMM] = X86_FEATURE_AVX512F, [XFEATURE_PT_UNIMPLEMENTED_SO_FAR] = X86_FEATURE_INTEL_PT, - [XFEATURE_PKRU] = X86_FEATURE_PKU, + [XFEATURE_PKRU] = X86_FEATURE_OSPKE, [XFEATURE_PASID] = X86_FEATURE_ENQCMD, + [XFEATURE_CET_USER] = X86_FEATURE_SHSTK, + [XFEATURE_CET_KERNEL] = X86_FEATURE_SHSTK, [XFEATURE_XTILE_CFG] = X86_FEATURE_AMX_TILE, [XFEATURE_XTILE_DATA] = X86_FEATURE_AMX_TILE, + [XFEATURE_APX] = X86_FEATURE_APX, }; static unsigned int xstate_offsets[XFEATURE_MAX] __ro_after_init = @@ -83,6 +94,31 @@ static unsigned int xstate_sizes[XFEATURE_MAX] __ro_after_init = { [ 0 ... XFEATURE_MAX - 1] = -1}; static unsigned int xstate_flags[XFEATURE_MAX] __ro_after_init; +/* + * Ordering of xstate components in uncompacted format: The xfeature + * number does not necessarily indicate its position in the XSAVE buffer. + * This array defines the traversal order of xstate features. + */ +static unsigned int xfeature_uncompact_order[XFEATURE_MAX] __ro_after_init = + { [ 0 ... XFEATURE_MAX - 1] = -1}; + +static inline unsigned int next_xfeature_order(unsigned int i, u64 mask) +{ + for (; xfeature_uncompact_order[i] != -1; i++) { + if (mask & BIT_ULL(xfeature_uncompact_order[i])) + break; + } + + return i; +} + +/* Iterate xstate features in uncompacted order: */ +#define for_each_extended_xfeature_in_order(i, mask) \ + for (i = 0; \ + i = next_xfeature_order(i, mask), \ + xfeature_uncompact_order[i] != -1; \ + i++) + #define XSTATE_FLAG_SUPERVISOR BIT(0) #define XSTATE_FLAG_ALIGNED64 BIT(1) @@ -177,10 +213,11 @@ void fpu__init_cpu_xstate(void) * Must happen after CR4 setup and before xsetbv() to allow KVM * lazy passthrough. Write independent of the dynamic state static * key as that does not work on the boot CPU. This also ensures - * that any stale state is wiped out from XFD. + * that any stale state is wiped out from XFD. Reset the per CPU + * xfd cache too. */ if (cpu_feature_enabled(X86_FEATURE_XFD)) - wrmsrl(MSR_IA32_XFD, init_fpstate.xfd); + xfd_set_state(init_fpstate.xfd); /* * XCR_XFEATURE_ENABLED_MASK (aka. XCR0) sets user features @@ -193,7 +230,7 @@ void fpu__init_cpu_xstate(void) * MSR_IA32_XSS sets supervisor states managed by XSAVES. */ if (boot_cpu_has(X86_FEATURE_XSAVES)) { - wrmsrl(MSR_IA32_XSS, xfeatures_mask_supervisor() | + wrmsrq(MSR_IA32_XSS, xfeatures_mask_supervisor() | xfeatures_mask_independent()); } } @@ -203,16 +240,20 @@ static bool xfeature_enabled(enum xfeature xfeature) return fpu_kernel_cfg.max_features & BIT_ULL(xfeature); } +static int compare_xstate_offsets(const void *xfeature1, const void *xfeature2) +{ + return xstate_offsets[*(unsigned int *)xfeature1] - + xstate_offsets[*(unsigned int *)xfeature2]; +} + /* * Record the offsets and sizes of various xstates contained - * in the XSAVE state memory layout. + * in the XSAVE state memory layout. Also, create an ordered + * list of xfeatures for handling out-of-order offsets. */ static void __init setup_xstate_cache(void) { - u32 eax, ebx, ecx, edx, i; - /* start at the beginning of the "extended state" */ - unsigned int last_good_offset = offsetof(struct xregs_state, - extended_state_area); + u32 eax, ebx, ecx, edx, xfeature, i = 0; /* * The FP xstates and SSE xstates are legacy states. They are always * in the fixed offsets in the xsave area in either compacted form @@ -226,39 +267,30 @@ static void __init setup_xstate_cache(void) xstate_sizes[XFEATURE_SSE] = sizeof_field(struct fxregs_state, xmm_space); - for_each_extended_xfeature(i, fpu_kernel_cfg.max_features) { - cpuid_count(XSTATE_CPUID, i, &eax, &ebx, &ecx, &edx); + for_each_extended_xfeature(xfeature, fpu_kernel_cfg.max_features) { + cpuid_count(CPUID_LEAF_XSTATE, xfeature, &eax, &ebx, &ecx, &edx); - xstate_sizes[i] = eax; - xstate_flags[i] = ecx; + xstate_sizes[xfeature] = eax; + xstate_flags[xfeature] = ecx; /* * If an xfeature is supervisor state, the offset in EBX is * invalid, leave it to -1. */ - if (xfeature_is_supervisor(i)) + if (xfeature_is_supervisor(xfeature)) continue; - xstate_offsets[i] = ebx; - - /* - * In our xstate size checks, we assume that the highest-numbered - * xstate feature has the highest offset in the buffer. Ensure - * it does. - */ - WARN_ONCE(last_good_offset > xstate_offsets[i], - "x86/fpu: misordered xstate at %d\n", last_good_offset); + xstate_offsets[xfeature] = ebx; - last_good_offset = xstate_offsets[i]; + /* Populate the list of xfeatures before sorting */ + xfeature_uncompact_order[i++] = xfeature; } -} -static void __init print_xstate_feature(u64 xstate_mask) -{ - const char *feature_name; - - if (cpu_has_xfeatures(xstate_mask, &feature_name)) - pr_info("x86/fpu: Supporting XSAVE feature 0x%03Lx: '%s'\n", xstate_mask, feature_name); + /* + * Sort xfeatures by their offsets to support out-of-order + * offsets in the uncompacted format. + */ + sort(xfeature_uncompact_order, i, sizeof(unsigned int), compare_xstate_offsets, NULL); } /* @@ -266,18 +298,15 @@ static void __init print_xstate_feature(u64 xstate_mask) */ static void __init print_xstate_features(void) { - print_xstate_feature(XFEATURE_MASK_FP); - print_xstate_feature(XFEATURE_MASK_SSE); - print_xstate_feature(XFEATURE_MASK_YMM); - print_xstate_feature(XFEATURE_MASK_BNDREGS); - print_xstate_feature(XFEATURE_MASK_BNDCSR); - print_xstate_feature(XFEATURE_MASK_OPMASK); - print_xstate_feature(XFEATURE_MASK_ZMM_Hi256); - print_xstate_feature(XFEATURE_MASK_Hi16_ZMM); - print_xstate_feature(XFEATURE_MASK_PKRU); - print_xstate_feature(XFEATURE_MASK_PASID); - print_xstate_feature(XFEATURE_MASK_XTILE_CFG); - print_xstate_feature(XFEATURE_MASK_XTILE_DATA); + int i; + + for (i = 0; i < XFEATURE_MAX; i++) { + u64 mask = BIT_ULL(i); + const char *name; + + if (cpu_has_xfeatures(mask, &name)) + pr_info("x86/fpu: Supporting XSAVE feature 0x%03Lx: '%s'\n", mask, name); + } } /* @@ -344,7 +373,10 @@ static __init void os_xrstor_booting(struct xregs_state *xstate) XFEATURE_MASK_BNDREGS | \ XFEATURE_MASK_BNDCSR | \ XFEATURE_MASK_PASID | \ - XFEATURE_MASK_XTILE) + XFEATURE_MASK_CET_USER | \ + XFEATURE_MASK_CET_KERNEL | \ + XFEATURE_MASK_XTILE | \ + XFEATURE_MASK_APX) /* * setup the xstate image representing the init state @@ -391,7 +423,7 @@ int xfeature_size(int xfeature_nr) u32 eax, ebx, ecx, edx; CHECK_XFEATURE(xfeature_nr); - cpuid_count(XSTATE_CPUID, xfeature_nr, &eax, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_XSTATE, xfeature_nr, &eax, &ebx, &ecx, &edx); return eax; } @@ -434,26 +466,27 @@ static void __init __xstate_dump_leaves(void) * just in case there are some goodies up there */ for (i = 0; i < XFEATURE_MAX + 10; i++) { - cpuid_count(XSTATE_CPUID, i, &eax, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_XSTATE, i, &eax, &ebx, &ecx, &edx); pr_warn("CPUID[%02x, %02x]: eax=%08x ebx=%08x ecx=%08x edx=%08x\n", - XSTATE_CPUID, i, eax, ebx, ecx, edx); + CPUID_LEAF_XSTATE, i, eax, ebx, ecx, edx); } } -#define XSTATE_WARN_ON(x) do { \ - if (WARN_ONCE(x, "XSAVE consistency problem, dumping leaves")) { \ +#define XSTATE_WARN_ON(x, fmt, ...) do { \ + if (WARN_ONCE(x, "XSAVE consistency problem: " fmt, ##__VA_ARGS__)) { \ __xstate_dump_leaves(); \ } \ } while (0) -#define XCHECK_SZ(sz, nr, nr_macro, __struct) do { \ - if ((nr == nr_macro) && \ - WARN_ONCE(sz != sizeof(__struct), \ - "%s: struct is %zu bytes, cpu state %d bytes\n", \ - __stringify(nr_macro), sizeof(__struct), sz)) { \ +#define XCHECK_SZ(sz, nr, __struct) ({ \ + if (WARN_ONCE(sz != sizeof(__struct), \ + "[%s]: struct is %zu bytes, cpu state %d bytes\n", \ + xfeature_names[nr], sizeof(__struct), sz)) { \ __xstate_dump_leaves(); \ } \ -} while (0) + true; \ +}) + /** * check_xtile_data_against_struct - Check tile data state size. @@ -476,7 +509,7 @@ static int __init check_xtile_data_against_struct(int size) * Check the maximum palette id: * eax: the highest numbered palette subleaf. */ - cpuid_count(TILE_CPUID, 0, &max_palid, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_TILE, 0, &max_palid, &ebx, &ecx, &edx); /* * Cross-check each tile size and find the maximum number of @@ -490,7 +523,7 @@ static int __init check_xtile_data_against_struct(int size) * eax[31:16]: bytes per title * ebx[31:16]: the max names (or max number of tiles) */ - cpuid_count(TILE_CPUID, palid, &eax, &ebx, &edx, &edx); + cpuid_count(CPUID_LEAF_TILE, palid, &eax, &ebx, &edx, &edx); tile_size = eax >> 16; max = ebx >> 16; @@ -527,50 +560,50 @@ static bool __init check_xstate_against_struct(int nr) * Ask the CPU for the size of the state. */ int sz = xfeature_size(nr); + /* * Match each CPU state with the corresponding software * structure. */ - XCHECK_SZ(sz, nr, XFEATURE_YMM, struct ymmh_struct); - XCHECK_SZ(sz, nr, XFEATURE_BNDREGS, struct mpx_bndreg_state); - XCHECK_SZ(sz, nr, XFEATURE_BNDCSR, struct mpx_bndcsr_state); - XCHECK_SZ(sz, nr, XFEATURE_OPMASK, struct avx_512_opmask_state); - XCHECK_SZ(sz, nr, XFEATURE_ZMM_Hi256, struct avx_512_zmm_uppers_state); - XCHECK_SZ(sz, nr, XFEATURE_Hi16_ZMM, struct avx_512_hi16_state); - XCHECK_SZ(sz, nr, XFEATURE_PKRU, struct pkru_state); - XCHECK_SZ(sz, nr, XFEATURE_PASID, struct ia32_pasid_state); - XCHECK_SZ(sz, nr, XFEATURE_XTILE_CFG, struct xtile_cfg); - - /* The tile data size varies between implementations. */ - if (nr == XFEATURE_XTILE_DATA) - check_xtile_data_against_struct(sz); - - /* - * Make *SURE* to add any feature numbers in below if - * there are "holes" in the xsave state component - * numbers. - */ - if ((nr < XFEATURE_YMM) || - (nr >= XFEATURE_MAX) || - (nr == XFEATURE_PT_UNIMPLEMENTED_SO_FAR) || - ((nr >= XFEATURE_RSRVD_COMP_11) && (nr <= XFEATURE_RSRVD_COMP_16))) { - WARN_ONCE(1, "no structure for xstate: %d\n", nr); - XSTATE_WARN_ON(1); + switch (nr) { + case XFEATURE_YMM: return XCHECK_SZ(sz, nr, struct ymmh_struct); + case XFEATURE_BNDREGS: return XCHECK_SZ(sz, nr, struct mpx_bndreg_state); + case XFEATURE_BNDCSR: return XCHECK_SZ(sz, nr, struct mpx_bndcsr_state); + case XFEATURE_OPMASK: return XCHECK_SZ(sz, nr, struct avx_512_opmask_state); + case XFEATURE_ZMM_Hi256: return XCHECK_SZ(sz, nr, struct avx_512_zmm_uppers_state); + case XFEATURE_Hi16_ZMM: return XCHECK_SZ(sz, nr, struct avx_512_hi16_state); + case XFEATURE_PKRU: return XCHECK_SZ(sz, nr, struct pkru_state); + case XFEATURE_PASID: return XCHECK_SZ(sz, nr, struct ia32_pasid_state); + case XFEATURE_XTILE_CFG: return XCHECK_SZ(sz, nr, struct xtile_cfg); + case XFEATURE_CET_USER: return XCHECK_SZ(sz, nr, struct cet_user_state); + case XFEATURE_CET_KERNEL: return XCHECK_SZ(sz, nr, struct cet_supervisor_state); + case XFEATURE_APX: return XCHECK_SZ(sz, nr, struct apx_state); + case XFEATURE_XTILE_DATA: check_xtile_data_against_struct(sz); return true; + default: + XSTATE_WARN_ON(1, "No structure for xstate: %d\n", nr); return false; } + return true; } static unsigned int xstate_calculate_size(u64 xfeatures, bool compacted) { unsigned int topmost = fls64(xfeatures) - 1; - unsigned int offset = xstate_offsets[topmost]; + unsigned int offset, i; if (topmost <= XFEATURE_SSE) return sizeof(struct xregs_state); - if (compacted) + if (compacted) { offset = xfeature_get_offset(xfeatures, topmost); + } else { + /* Walk through the xfeature order to pick the last */ + for_each_extended_xfeature_in_order(i, xfeatures) + topmost = xfeature_uncompact_order[i]; + offset = xstate_offsets[topmost]; + } + return offset + xstate_sizes[topmost]; } @@ -598,12 +631,13 @@ static bool __init paranoid_xstate_size_valid(unsigned int kernel_size) * XSAVES. */ if (!xsaves && xfeature_is_supervisor(i)) { - XSTATE_WARN_ON(1); + XSTATE_WARN_ON(1, "Got supervisor feature %d, but XSAVES not advertised\n", i); return false; } } size = xstate_calculate_size(fpu_kernel_cfg.max_features, compacted); - XSTATE_WARN_ON(size != kernel_size); + XSTATE_WARN_ON(size != kernel_size, + "size %u != kernel_size %u\n", size, kernel_size); return size == kernel_size; } @@ -633,7 +667,7 @@ static unsigned int __init get_compacted_size(void) * are no supervisor states, but XSAVEC still uses compacted * format. */ - cpuid_count(XSTATE_CPUID, 1, &eax, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_XSTATE, 1, &eax, &ebx, &ecx, &edx); return ebx; } @@ -650,7 +684,7 @@ static unsigned int __init get_xsave_compacted_size(void) return get_compacted_size(); /* Disable independent features. */ - wrmsrl(MSR_IA32_XSS, xfeatures_mask_supervisor()); + wrmsrq(MSR_IA32_XSS, xfeatures_mask_supervisor()); /* * Ask the hardware what size is required of the buffer. @@ -659,7 +693,7 @@ static unsigned int __init get_xsave_compacted_size(void) size = get_compacted_size(); /* Re-enable independent features so XSAVES will work on them again. */ - wrmsrl(MSR_IA32_XSS, xfeatures_mask_supervisor() | mask); + wrmsrq(MSR_IA32_XSS, xfeatures_mask_supervisor() | mask); return size; } @@ -674,7 +708,7 @@ static unsigned int __init get_xsave_size_user(void) * containing all the *user* state components * corresponding to bits currently set in XCR0. */ - cpuid_count(XSTATE_CPUID, 0, &eax, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_XSTATE, 0, &eax, &ebx, &ecx, &edx); return ebx; } @@ -713,6 +747,9 @@ static int __init init_xstate_size(void) fpu_user_cfg.default_size = xstate_calculate_size(fpu_user_cfg.default_features, false); + guest_default_cfg.size = + xstate_calculate_size(guest_default_cfg.features, compacted); + return 0; } @@ -722,6 +759,8 @@ static int __init init_xstate_size(void) */ static void __init fpu__init_disable_system_xstate(unsigned int legacy_size) { + pr_info("x86/fpu: XSAVE disabled\n"); + fpu_kernel_cfg.max_features = 0; cr4_clear_bits(X86_CR4_OSXSAVE); setup_clear_cpu_cap(X86_FEATURE_XSAVE); @@ -731,6 +770,7 @@ static void __init fpu__init_disable_system_xstate(unsigned int legacy_size) fpu_kernel_cfg.default_size = legacy_size; fpu_user_cfg.max_size = legacy_size; fpu_user_cfg.default_size = legacy_size; + guest_default_cfg.size = legacy_size; /* * Prevent enabling the static branch which enables writes to the @@ -738,7 +778,25 @@ static void __init fpu__init_disable_system_xstate(unsigned int legacy_size) */ init_fpstate.xfd = 0; - fpstate_reset(¤t->thread.fpu); + fpstate_reset(x86_task_fpu(current)); +} + +static u64 __init host_default_mask(void) +{ + /* + * Exclude dynamic features (require userspace opt-in) and features + * that are supported only for KVM guests. + */ + return ~((u64)XFEATURE_MASK_USER_DYNAMIC | XFEATURE_MASK_GUEST_SUPERVISOR); +} + +static u64 __init guest_default_mask(void) +{ + /* + * Exclude dynamic features, which require userspace opt-in even + * for KVM guests. + */ + return ~(u64)XFEATURE_MASK_USER_DYNAMIC; } /* @@ -763,21 +821,16 @@ void __init fpu__init_system_xstate(unsigned int legacy_size) return; } - if (boot_cpu_data.cpuid_level < XSTATE_CPUID) { - WARN_ON_FPU(1); - return; - } - /* * Find user xstates supported by the processor. */ - cpuid_count(XSTATE_CPUID, 0, &eax, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_XSTATE, 0, &eax, &ebx, &ecx, &edx); fpu_kernel_cfg.max_features = eax + ((u64)edx << 32); /* * Find supervisor xstates supported by the processor. */ - cpuid_count(XSTATE_CPUID, 1, &eax, &ebx, &ecx, &edx); + cpuid_count(CPUID_LEAF_XSTATE, 1, &eax, &ebx, &ecx, &edx); fpu_kernel_cfg.max_features |= ecx + ((u64)edx << 32); if ((fpu_kernel_cfg.max_features & XFEATURE_MASK_FPSSE) != XFEATURE_MASK_FPSSE) { @@ -791,6 +844,20 @@ void __init fpu__init_system_xstate(unsigned int legacy_size) goto out_disable; } + if (fpu_kernel_cfg.max_features & XFEATURE_MASK_APX && + fpu_kernel_cfg.max_features & (XFEATURE_MASK_BNDREGS | XFEATURE_MASK_BNDCSR)) { + /* + * This is a problematic CPU configuration where two + * conflicting state components are both enumerated. + */ + pr_err("x86/fpu: Both APX/MPX present in the CPU's xstate features: 0x%llx.\n", + fpu_kernel_cfg.max_features); + goto out_disable; + } + + fpu_kernel_cfg.independent_features = fpu_kernel_cfg.max_features & + XFEATURE_MASK_INDEPENDENT; + /* * Clear XSAVE features that are disabled in the normal CPUID. */ @@ -814,12 +881,13 @@ void __init fpu__init_system_xstate(unsigned int legacy_size) fpu_user_cfg.max_features = fpu_kernel_cfg.max_features; fpu_user_cfg.max_features &= XFEATURE_MASK_USER_SUPPORTED; - /* Clean out dynamic features from default */ - fpu_kernel_cfg.default_features = fpu_kernel_cfg.max_features; - fpu_kernel_cfg.default_features &= ~XFEATURE_MASK_USER_DYNAMIC; - - fpu_user_cfg.default_features = fpu_user_cfg.max_features; - fpu_user_cfg.default_features &= ~XFEATURE_MASK_USER_DYNAMIC; + /* + * Now, given maximum feature set, determine default values by + * applying default masks. + */ + fpu_kernel_cfg.default_features = fpu_kernel_cfg.max_features & host_default_mask(); + fpu_user_cfg.default_features = fpu_user_cfg.max_features & host_default_mask(); + guest_default_cfg.features = fpu_kernel_cfg.max_features & guest_default_mask(); /* Store it for paranoia check at the end */ xfeatures = fpu_kernel_cfg.max_features; @@ -847,9 +915,6 @@ void __init fpu__init_system_xstate(unsigned int legacy_size) if (err) goto out_disable; - /* Reset the state for the current task */ - fpstate_reset(¤t->thread.fpu); - /* * Update info used for ptrace frames; use standard-format size and no * supervisor xstates: @@ -865,7 +930,7 @@ void __init fpu__init_system_xstate(unsigned int legacy_size) init_fpstate.xfeatures = fpu_kernel_cfg.default_features; if (init_fpstate.size > sizeof(init_fpstate.regs)) { - pr_warn("x86/fpu: init_fpstate buffer too small (%zu < %d), disabling XSAVE\n", + pr_warn("x86/fpu: init_fpstate buffer too small (%zu < %d)\n", sizeof(init_fpstate.regs), init_fpstate.size); goto out_disable; } @@ -877,11 +942,18 @@ void __init fpu__init_system_xstate(unsigned int legacy_size) * xfeatures mask. */ if (xfeatures != fpu_kernel_cfg.max_features) { - pr_err("x86/fpu: xfeatures modified from 0x%016llx to 0x%016llx during init, disabling XSAVE\n", + pr_err("x86/fpu: xfeatures modified from 0x%016llx to 0x%016llx during init\n", xfeatures, fpu_kernel_cfg.max_features); goto out_disable; } + /* + * CPU capabilities initialization runs before FPU init. So + * X86_FEATURE_OSXSAVE is not set. Now that XSAVE is completely + * functional, set the feature bit so depending code works. + */ + setup_force_cpu_cap(X86_FEATURE_OSXSAVE); + print_xstate_offset_size(); pr_info("x86/fpu: Enabled xstate features 0x%llx, context size is %d bytes, using '%s' format.\n", fpu_kernel_cfg.max_features, @@ -910,12 +982,12 @@ void fpu__resume_cpu(void) * of XSAVES and MSR_IA32_XSS. */ if (cpu_feature_enabled(X86_FEATURE_XSAVES)) { - wrmsrl(MSR_IA32_XSS, xfeatures_mask_supervisor() | + wrmsrq(MSR_IA32_XSS, xfeatures_mask_supervisor() | xfeatures_mask_independent()); } if (fpu_state_size_dynamic()) - wrmsrl(MSR_IA32_XFD, current->thread.fpu.fpstate->xfd); + wrmsrq(MSR_IA32_XFD, x86_task_fpu(current)->fpstate->xfd); } /* @@ -987,6 +1059,20 @@ void *get_xsave_addr(struct xregs_state *xsave, int xfeature_nr) return __raw_xsave_addr(xsave, xfeature_nr); } +EXPORT_SYMBOL_FOR_KVM(get_xsave_addr); + +/* + * Given an xstate feature nr, calculate where in the xsave buffer the state is. + * The xsave buffer should be in standard format, not compacted (e.g. user mode + * signal frames). + */ +void __user *get_xsave_addr_user(struct xregs_state __user *xsave, int xfeature_nr) +{ + if (WARN_ON_ONCE(!xfeature_enabled(xfeature_nr))) + return NULL; + + return (void __user *)xsave + xstate_offsets[xfeature_nr]; +} #ifdef CONFIG_ARCH_HAS_PKEYS @@ -1046,6 +1132,7 @@ static void copy_feature(bool from_xstate, struct membuf *to, void *xstate, * __copy_xstate_to_uabi_buf - Copy kernel saved xstate to a UABI buffer * @to: membuf descriptor * @fpstate: The fpstate buffer from which to copy + * @xfeatures: The mask of xfeatures to save (XSAVE mode only) * @pkru_val: The PKRU value to store in the PKRU component * @copy_mode: The requested copy mode * @@ -1056,15 +1143,15 @@ static void copy_feature(bool from_xstate, struct membuf *to, void *xstate, * It supports partial copy but @to.pos always starts from zero. */ void __copy_xstate_to_uabi_buf(struct membuf to, struct fpstate *fpstate, - u32 pkru_val, enum xstate_copy_mode copy_mode) + u64 xfeatures, u32 pkru_val, + enum xstate_copy_mode copy_mode) { const unsigned int off_mxcsr = offsetof(struct fxregs_state, mxcsr); struct xregs_state *xinit = &init_fpstate.regs.xsave; struct xregs_state *xsave = &fpstate->regs.xsave; + unsigned int zerofrom, i, xfeature; struct xstate_header header; - unsigned int zerofrom; u64 mask; - int i; memset(&header, 0, sizeof(header)); header.xfeatures = xsave->header.xfeatures; @@ -1080,7 +1167,7 @@ void __copy_xstate_to_uabi_buf(struct membuf to, struct fpstate *fpstate, break; case XSTATE_COPY_XSAVE: - header.xfeatures &= fpstate->user_xfeatures; + header.xfeatures &= fpstate->user_xfeatures & xfeatures; break; } @@ -1118,31 +1205,31 @@ void __copy_xstate_to_uabi_buf(struct membuf to, struct fpstate *fpstate, zerofrom = offsetof(struct xregs_state, extended_state_area); /* - * The ptrace buffer is in non-compacted XSAVE format. In - * non-compacted format disabled features still occupy state space, - * but there is no state to copy from in the compacted - * init_fpstate. The gap tracking will zero these states. - */ - mask = fpstate->user_xfeatures; - - /* - * Dynamic features are not present in init_fpstate. When they are - * in an all zeros init state, remove those from 'mask' to zero - * those features in the user buffer instead of retrieving them - * from init_fpstate. + * This 'mask' indicates which states to copy from fpstate. + * Those extended states that are not present in fpstate are + * either disabled or initialized: + * + * In non-compacted format, disabled features still occupy + * state space but there is no state to copy from in the + * compacted init_fpstate. The gap tracking will zero these + * states. + * + * The extended features have an all zeroes init state. Thus, + * remove them from 'mask' to zero those features in the user + * buffer instead of retrieving them from init_fpstate. */ - if (fpu_state_size_dynamic()) - mask &= (header.xfeatures | xinit->header.xcomp_bv); + mask = header.xfeatures; - for_each_extended_xfeature(i, mask) { + for_each_extended_xfeature_in_order(i, mask) { + xfeature = xfeature_uncompact_order[i]; /* * If there was a feature or alignment gap, zero the space * in the destination buffer. */ - if (zerofrom < xstate_offsets[i]) - membuf_zero(&to, xstate_offsets[i] - zerofrom); + if (zerofrom < xstate_offsets[xfeature]) + membuf_zero(&to, xstate_offsets[xfeature] - zerofrom); - if (i == XFEATURE_PKRU) { + if (xfeature == XFEATURE_PKRU) { struct pkru_state pkru = {0}; /* * PKRU is not necessarily up to date in the @@ -1151,16 +1238,15 @@ void __copy_xstate_to_uabi_buf(struct membuf to, struct fpstate *fpstate, pkru.pkru = pkru_val; membuf_write(&to, &pkru, sizeof(pkru)); } else { - copy_feature(header.xfeatures & BIT_ULL(i), &to, - __raw_xsave_addr(xsave, i), - __raw_xsave_addr(xinit, i), - xstate_sizes[i]); + membuf_write(&to, + __raw_xsave_addr(xsave, xfeature), + xstate_sizes[xfeature]); } /* * Keep track of the last copied state in the non-compacted * target buffer for gap zeroing. */ - zerofrom = xstate_offsets[i] + xstate_sizes[i]; + zerofrom = xstate_offsets[xfeature] + xstate_sizes[xfeature]; } out: @@ -1183,7 +1269,8 @@ out: void copy_xstate_to_uabi_buf(struct membuf to, struct task_struct *tsk, enum xstate_copy_mode copy_mode) { - __copy_xstate_to_uabi_buf(to, tsk->thread.fpu.fpstate, + __copy_xstate_to_uabi_buf(to, x86_task_fpu(tsk)->fpstate, + x86_task_fpu(tsk)->fpstate->user_xfeatures, tsk->thread.pkru, copy_mode); } @@ -1200,8 +1287,36 @@ static int copy_from_buffer(void *dst, unsigned int offset, unsigned int size, } +/** + * copy_uabi_to_xstate - Copy a UABI format buffer to the kernel xstate + * @fpstate: The fpstate buffer to copy to + * @kbuf: The UABI format buffer, if it comes from the kernel + * @ubuf: The UABI format buffer, if it comes from userspace + * @pkru: The location to write the PKRU value to + * + * Converts from the UABI format into the kernel internal hardware + * dependent format. + * + * This function ultimately has three different callers with distinct PKRU + * behavior. + * 1. When called from sigreturn the PKRU register will be restored from + * @fpstate via an XRSTOR. Correctly copying the UABI format buffer to + * @fpstate is sufficient to cover this case, but the caller will also + * pass a pointer to the thread_struct's pkru field in @pkru and updating + * it is harmless. + * 2. When called from ptrace the PKRU register will be restored from the + * thread_struct's pkru field. A pointer to that is passed in @pkru. + * The kernel will restore it manually, so the XRSTOR behavior that resets + * the PKRU register to the hardware init value (0) if the corresponding + * xfeatures bit is not set is emulated here. + * 3. When called from KVM the PKRU register will be restored from the vcpu's + * pkru field. A pointer to that is passed in @pkru. KVM hasn't used + * XRSTOR and hasn't had the PKRU resetting behavior described above. To + * preserve that KVM behavior, it passes NULL for @pkru if the xfeatures + * bit is not set. + */ static int copy_uabi_to_xstate(struct fpstate *fpstate, const void *kbuf, - const void __user *ubuf) + const void __user *ubuf, u32 *pkru) { struct xregs_state *xsave = &fpstate->regs.xsave; unsigned int offset, size; @@ -1250,6 +1365,20 @@ static int copy_uabi_to_xstate(struct fpstate *fpstate, const void *kbuf, } } + if (hdr.xfeatures & XFEATURE_MASK_PKRU) { + struct pkru_state *xpkru; + + xpkru = __raw_xsave_addr(xsave, XFEATURE_PKRU); + *pkru = xpkru->pkru; + } else { + /* + * KVM may pass NULL here to indicate that it does not need + * PKRU updated. + */ + if (pkru) + *pkru = 0; + } + /* * The state that came in from userspace was user-state only. * Mask all the user states out of 'xfeatures': @@ -1268,9 +1397,9 @@ static int copy_uabi_to_xstate(struct fpstate *fpstate, const void *kbuf, * Convert from a ptrace standard-format kernel buffer to kernel XSAVE[S] * format and copy to the target thread. Used by ptrace and KVM. */ -int copy_uabi_from_kernel_to_xstate(struct fpstate *fpstate, const void *kbuf) +int copy_uabi_from_kernel_to_xstate(struct fpstate *fpstate, const void *kbuf, u32 *pkru) { - return copy_uabi_to_xstate(fpstate, kbuf, NULL); + return copy_uabi_to_xstate(fpstate, kbuf, NULL, pkru); } /* @@ -1278,10 +1407,10 @@ int copy_uabi_from_kernel_to_xstate(struct fpstate *fpstate, const void *kbuf) * XSAVE[S] format and copy to the target thread. This is called from the * sigreturn() and rt_sigreturn() system calls. */ -int copy_sigframe_from_user_to_xstate(struct fpstate *fpstate, +int copy_sigframe_from_user_to_xstate(struct task_struct *tsk, const void __user *ubuf) { - return copy_uabi_to_xstate(fpstate, NULL, ubuf); + return copy_uabi_to_xstate(x86_task_fpu(tsk)->fpstate, NULL, ubuf, &tsk->thread.pkru); } static bool validate_independent_components(u64 mask) @@ -1347,14 +1476,14 @@ void xrstors(struct xregs_state *xstate, u64 mask) } #if IS_ENABLED(CONFIG_KVM) -void fpstate_clear_xstate_component(struct fpstate *fps, unsigned int xfeature) +void fpstate_clear_xstate_component(struct fpstate *fpstate, unsigned int xfeature) { - void *addr = get_xsave_addr(&fps->regs.xsave, xfeature); + void *addr = get_xsave_addr(&fpstate->regs.xsave, xfeature); if (addr) memset(addr, 0, xstate_sizes[xfeature]); } -EXPORT_SYMBOL_GPL(fpstate_clear_xstate_component); +EXPORT_SYMBOL_FOR_KVM(fpstate_clear_xstate_component); #endif #ifdef CONFIG_X86_64 @@ -1375,7 +1504,7 @@ static bool xstate_op_valid(struct fpstate *fpstate, u64 mask, bool rstor) * The XFD MSR does not match fpstate->xfd. That's invalid when * the passed in fpstate is current's fpstate. */ - if (fpstate->xfd == current->thread.fpu.fpstate->xfd) + if (fpstate->xfd == x86_task_fpu(current)->fpstate->xfd) return false; /* @@ -1387,8 +1516,8 @@ static bool xstate_op_valid(struct fpstate *fpstate, u64 mask, bool rstor) return rstor; /* - * XSAVE(S): clone(), fpu_swap_kvm_fpu() - * XRSTORS(S): fpu_swap_kvm_fpu() + * XSAVE(S): clone(), fpu_swap_kvm_fpstate() + * XRSTORS(S): fpu_swap_kvm_fpstate() */ /* @@ -1452,7 +1581,7 @@ void fpstate_free(struct fpu *fpu) static int fpstate_realloc(u64 xfeatures, unsigned int ksize, unsigned int usize, struct fpu_guest *guest_fpu) { - struct fpu *fpu = ¤t->thread.fpu; + struct fpu *fpu = x86_task_fpu(current); struct fpstate *curfps, *newfps = NULL; unsigned int fpsize; bool in_use; @@ -1493,10 +1622,7 @@ static int fpstate_realloc(u64 xfeatures, unsigned int ksize, fpregs_restore_userregs(); newfps->xfeatures = curfps->xfeatures | xfeatures; - - if (!guest_fpu) - newfps->user_xfeatures = curfps->user_xfeatures | xfeatures; - + newfps->user_xfeatures = curfps->user_xfeatures | xfeatures; newfps->xfd = curfps->xfd & ~xfeatures; /* Do the final updates within the locked region */ @@ -1548,7 +1674,7 @@ static int __xstate_request_perm(u64 permitted, u64 requested, bool guest) * AVX512. */ bool compacted = cpu_feature_enabled(X86_FEATURE_XCOMPACTED); - struct fpu *fpu = ¤t->group_leader->thread.fpu; + struct fpu *fpu = x86_task_fpu(current->group_leader); struct fpu_state_perm *perm; unsigned int ksize, usize; u64 mask; @@ -1558,16 +1684,20 @@ static int __xstate_request_perm(u64 permitted, u64 requested, bool guest) if ((permitted & requested) == requested) return 0; - /* Calculate the resulting kernel state size */ + /* + * Calculate the resulting kernel state size. Note, @permitted also + * contains supervisor xfeatures even though supervisor are always + * permitted for kernel and guest FPUs, and never permitted for user + * FPUs. + */ mask = permitted | requested; - /* Take supervisor states into account on the host */ - if (!guest) - mask |= xfeatures_mask_supervisor(); ksize = xstate_calculate_size(mask, compacted); - /* Calculate the resulting user state size */ - mask &= XFEATURE_MASK_USER_SUPPORTED; - usize = xstate_calculate_size(mask, false); + /* + * Calculate the resulting user state size. Take care not to clobber + * the supervisor xfeatures in the new mask! + */ + usize = xstate_calculate_size(mask & XFEATURE_MASK_USER_SUPPORTED, false); if (!guest) { ret = validate_sigaltstack(usize); @@ -1651,7 +1781,7 @@ int __xfd_enable_feature(u64 xfd_err, struct fpu_guest *guest_fpu) return -EPERM; } - fpu = ¤t->group_leader->thread.fpu; + fpu = x86_task_fpu(current->group_leader); perm = guest_fpu ? &fpu->guest_perm : &fpu->perm; ksize = perm->__state_size; usize = perm->__user_state_size; @@ -1689,11 +1819,10 @@ u64 xstate_get_guest_group_perm(void) { return xstate_get_group_perm(true); } -EXPORT_SYMBOL_GPL(xstate_get_guest_group_perm); +EXPORT_SYMBOL_FOR_KVM(xstate_get_guest_group_perm); /** * fpu_xstate_prctl - xstate permission operations - * @tsk: Redundant pointer to current * @option: A subfunction of arch_prctl() * @arg2: option argument * Return: 0 if successful; otherwise, an error code @@ -1753,19 +1882,20 @@ long fpu_xstate_prctl(int option, unsigned long arg2) #ifdef CONFIG_PROC_PID_ARCH_STATUS /* * Report the amount of time elapsed in millisecond since last AVX512 - * use in the task. + * use in the task. Report -1 if no AVX-512 usage. */ static void avx512_status(struct seq_file *m, struct task_struct *task) { - unsigned long timestamp = READ_ONCE(task->thread.fpu.avx512_timestamp); - long delta; + unsigned long timestamp; + long delta = -1; - if (!timestamp) { - /* - * Report -1 if no AVX512 usage - */ - delta = -1; - } else { + /* AVX-512 usage is not tracked for kernel threads. Don't report anything. */ + if (task->flags & (PF_KTHREAD | PF_USER_WORKER)) + return; + + timestamp = READ_ONCE(x86_task_fpu(task)->avx512_timestamp); + + if (timestamp) { delta = (long)(jiffies - timestamp); /* * Cap to LONG_MAX if time difference > LONG_MAX @@ -1794,3 +1924,89 @@ int proc_pid_arch_status(struct seq_file *m, struct pid_namespace *ns, return 0; } #endif /* CONFIG_PROC_PID_ARCH_STATUS */ + +#ifdef CONFIG_COREDUMP +static const char owner_name[] = "LINUX"; + +/* + * Dump type, size, offset and flag values for every xfeature that is present. + */ +static int dump_xsave_layout_desc(struct coredump_params *cprm) +{ + int num_records = 0; + int i; + + for_each_extended_xfeature(i, fpu_user_cfg.max_features) { + struct x86_xfeat_component xc = { + .type = i, + .size = xstate_sizes[i], + .offset = xstate_offsets[i], + /* reserved for future use */ + .flags = 0, + }; + + if (!dump_emit(cprm, &xc, sizeof(xc))) + return 0; + + num_records++; + } + return num_records; +} + +static u32 get_xsave_desc_size(void) +{ + u32 cnt = 0; + u32 i; + + for_each_extended_xfeature(i, fpu_user_cfg.max_features) + cnt++; + + return cnt * (sizeof(struct x86_xfeat_component)); +} + +int elf_coredump_extra_notes_write(struct coredump_params *cprm) +{ + int num_records = 0; + struct elf_note en; + + if (!fpu_user_cfg.max_features) + return 0; + + en.n_namesz = sizeof(owner_name); + en.n_descsz = get_xsave_desc_size(); + en.n_type = NT_X86_XSAVE_LAYOUT; + + if (!dump_emit(cprm, &en, sizeof(en))) + return 1; + if (!dump_emit(cprm, owner_name, en.n_namesz)) + return 1; + if (!dump_align(cprm, 4)) + return 1; + + num_records = dump_xsave_layout_desc(cprm); + if (!num_records) + return 1; + + /* Total size should be equal to the number of records */ + if ((sizeof(struct x86_xfeat_component) * num_records) != en.n_descsz) + return 1; + + return 0; +} + +int elf_coredump_extra_notes_size(void) +{ + int size; + + if (!fpu_user_cfg.max_features) + return 0; + + /* .note header */ + size = sizeof(struct elf_note); + /* Name plus alignment to 4 bytes */ + size += roundup(sizeof(owner_name), 4); + size += get_xsave_desc_size(); + + return size; +} +#endif /* CONFIG_COREDUMP */ |