x86/cpufeature: Cleanup stuff more
Subsystems (7)
Borislav Petkov
maintainercover x86/cpufeature: Cleanup stuff more
From: Borislav Petkov <[email protected]>
Hi all,
this is the second submission. Changes from the previous one is that we're much more considerate so that we don't enlarge kernel .text and cause tinyconfig growth. Patch 4 adds a default-off config option which disables static_cpu_has* macros from being built and saves us ~1.6K .text. 0day report for this is:
i386-tinyconfig vmlinux size:
+-------+------+------+-----+---------------------------------------------------------------------------------+ | TOTAL | TEXT | DATA | BSS | | +-------+------+------+-----+---------------------------------------------------------------------------------+ | -1512 | -352 | 0 | 0 | df63d2e25f80 x86/cpu: Provide a config option to disable static_cpu_has | | +5 | 0 | 0 | 0 | 8b9e9097bf18 Merge remote-tracking branch 'tip/x86/cleanups' into cpufeature-v3 | | -1507 | -352 | 0 | 0 | ALL COMMITS | +-------+------+------+-----+---------------------------------------------------------------------------------+
The remaining cpu_has_xx uglies will be handled later, in a follow-up patchset.
Thanks.
Changelog:
v0:
so this should take care of cleaning up some aspects of our cpufeatures handling.
Patches should be pretty self-explanatory but let me send them out as an RFC - I might've missed something obvious of the sort "but but, you can't do that..."
Borislav Petkov (4):
x86/cpufeature: Move some of the scattered feature bits to
x86_capability
x86/cpufeature: Cleanup get_cpu_cap()
x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros x86/cpu: Provide a config option to disable static_cpu_has
arch/x86/Kconfig | 11 +++ arch/x86/crypto/chacha20_glue.c | 2 +- arch/x86/crypto/crc32c-intel_glue.c | 2 +- arch/x86/include/asm/cmpxchg_32.h | 2 +- arch/x86/include/asm/cmpxchg_64.h | 2 +- arch/x86/include/asm/cpufeature.h | 115 ++++++++++++++-------------- arch/x86/include/asm/xor_32.h | 2 +- arch/x86/kernel/cpu/amd.c | 4 +- arch/x86/kernel/cpu/centaur.c | 2 +- arch/x86/kernel/cpu/common.c | 52 +++++++------ arch/x86/kernel/cpu/intel.c | 3 +- arch/x86/kernel/cpu/intel_cacheinfo.c | 6 +- arch/x86/kernel/cpu/mtrr/generic.c | 2 +- arch/x86/kernel/cpu/mtrr/main.c | 2 +- arch/x86/kernel/cpu/perf_event_amd.c | 4 +- arch/x86/kernel/cpu/perf_event_amd_uncore.c | 11 +-- arch/x86/kernel/cpu/scattered.c | 20 ----- arch/x86/kernel/cpu/transmeta.c | 4 +- arch/x86/kernel/fpu/init.c | 4 +- arch/x86/kernel/hw_breakpoint.c | 6 +- arch/x86/kernel/smpboot.c | 2 +- arch/x86/kernel/vm86_32.c | 4 +- arch/x86/mm/setup_nx.c | 4 +- drivers/char/hw_random/via-rng.c | 5 +- drivers/crypto/padlock-aes.c | 2 +- drivers/crypto/padlock-sha.c | 2 +- drivers/iommu/intel_irq_remapping.c | 2 +- fs/btrfs/disk-io.c | 2 +- 28 files changed, 141 insertions(+), 138 deletions(-)
Signature · 1 line
2.3.5
Borislav Petkov
maintainerPATCH 1/4 x86/cpufeature: Move some of the scattered feature bits to x86_capability
From: Borislav Petkov <[email protected]>
Turn the CPUID leafs which are proper CPUID feature bit leafs into separate ->x86_capability words.
--- arch/x86/include/asm/cpufeature.h | 54 +++++++++++++++++++++++---------------- arch/x86/kernel/cpu/common.c | 5 ++++ arch/x86/kernel/cpu/scattered.c | 20 --------------- 3 files changed, 37 insertions(+), 42 deletions(-)
arch/x86/include/asm/cpufeature.h +32 -22
@@ -12,7 +12,7 @@#include <asm/disabled-features.h> #endif -#define NCAPINTS 14/* N 32-bit words worth of info */ +#define NCAPINTS 16/* N 32-bit words worth of info */ #define NBUGINTS 1/* N 32-bit bug flags */ /*
@@ -181,22 +181,17 @@/* * Auxiliary flags: Linux defined - For features scattered in various -* CPUID levels like 0x6, 0xA etc, word 7 +* CPUID levels like 0x6, 0xA etc, word 7. +* +* Reuse free bits when adding new feature flags! */ -#define X86_FEATURE_IDA ( 7*32+ 0)/* Intel Dynamic Acceleration */ -#define X86_FEATURE_ARAT ( 7*32+ 1)/* Always Running APIC Timer */ +#define X86_FEATURE_CPB ( 7*32+ 2)/* AMD Core Performance Boost */ #define X86_FEATURE_EPB ( 7*32+ 3)/* IA32_ENERGY_PERF_BIAS support */ -#define X86_FEATURE_PLN ( 7*32+ 5)/* Intel Power Limit Notification */ -#define X86_FEATURE_PTS ( 7*32+ 6)/* Intel Package Thermal Status */ -#define X86_FEATURE_DTHERM ( 7*32+ 7)/* Digital Thermal Sensor */ +#define X86_FEATURE_HW_PSTATE ( 7*32+ 8)/* AMD HW-PState */ #define X86_FEATURE_PROC_FEEDBACK ( 7*32+ 9)/* AMD ProcFeedbackInterface */ -#define X86_FEATURE_HWP ( 7*32+ 10)/* "hwp" Intel HWP */ -#define X86_FEATURE_HWP_NOTIFY ( 7*32+ 11)/* Intel HWP_NOTIFY */ -#define X86_FEATURE_HWP_ACT_WINDOW ( 7*32+ 12)/* Intel HWP_ACT_WINDOW */ -#define X86_FEATURE_HWP_EPP ( 7*32+13)/* Intel HWP_EPP */ -#define X86_FEATURE_HWP_PKG_REQ ( 7*32+14)/* Intel HWP_PKG_REQ */ +#define X86_FEATURE_INTEL_PT ( 7*32+15)/* Intel Processor Trace */ /* Virtualization flags: Linux defined, word 8 */
@@ -205,16 +200,7 @@#define X86_FEATURE_FLEXPRIORITY ( 8*32+ 2)/* Intel FlexPriority */ #define X86_FEATURE_EPT ( 8*32+ 3)/* Intel Extended Page Table */ #define X86_FEATURE_VPID ( 8*32+ 4)/* Intel Virtual Processor ID */ -#define X86_FEATURE_NPT ( 8*32+ 5)/* AMD Nested Page Table support */ -#define X86_FEATURE_LBRV ( 8*32+ 6)/* AMD LBR Virtualization support */ -#define X86_FEATURE_SVML ( 8*32+ 7)/* "svm_lock" AMD SVM locking MSR */ -#define X86_FEATURE_NRIPS ( 8*32+ 8)/* "nrip_save" AMD SVM next_rip save */ -#define X86_FEATURE_TSCRATEMSR ( 8*32+ 9)/* "tsc_scale" AMD TSC scaling support */ -#define X86_FEATURE_VMCBCLEAN ( 8*32+10)/* "vmcb_clean" AMD VMCB clean bits support */ -#define X86_FEATURE_FLUSHBYASID ( 8*32+11)/* AMD flush-by-ASID support */ -#define X86_FEATURE_DECODEASSISTS ( 8*32+12)/* AMD Decode Assists support */ -#define X86_FEATURE_PAUSEFILTER ( 8*32+13)/* AMD filtered pause intercept */ -#define X86_FEATURE_PFTHRESHOLD ( 8*32+14)/* AMD pause filter threshold */ +#define X86_FEATURE_VMMCALL ( 8*32+15)/* Prefer vmmcall to vmcall */
@@ -258,6 +244,30 @@/* AMD-defined CPU features, CPUID level 0x80000008 (ebx), word 13 */ #define X86_FEATURE_CLZERO (13*32+0)/* CLZERO instruction */ +/* Thermal and Power Management Leaf, CPUID level 0x00000006 (eax), word 14 */ +#define X86_FEATURE_DTHERM (14*32+ 0)/* Digital Thermal Sensor */ +#define X86_FEATURE_IDA (14*32+ 1)/* Intel Dynamic Acceleration */ +#define X86_FEATURE_ARAT (14*32+ 2)/* Always Running APIC Timer */ +#define X86_FEATURE_PLN (14*32+ 4)/* Intel Power Limit Notification */ +#define X86_FEATURE_PTS (14*32+ 6)/* Intel Package Thermal Status */ +#define X86_FEATURE_HWP (14*32+ 7)/* Intel Hardware P-states */ +#define X86_FEATURE_HWP_NOTIFY (14*32+ 8)/* HWP Notification */ +#define X86_FEATURE_HWP_ACT_WINDOW (14*32+ 9)/* HWP Activity Window */ +#define X86_FEATURE_HWP_EPP (14*32+10)/* HWP Energy Perf. Preference */ +#define X86_FEATURE_HWP_PKG_REQ (14*32+11)/* HWP Package Level Request */ + +/* AMD SVM Feature Identification, CPUID level 0x8000000a (edx), word 15 */ +#define X86_FEATURE_NPT (15*32+ 0)/* Nested Page Table support */ +#define X86_FEATURE_LBRV (15*32+ 1)/* LBR Virtualization support */ +#define X86_FEATURE_SVML (15*32+ 2)/* "svm_lock" SVM locking MSR */ +#define X86_FEATURE_NRIPS (15*32+ 3)/* "nrip_save" SVM next_rip save */ +#define X86_FEATURE_TSCRATEMSR (15*32+ 4)/* "tsc_scale" TSC scaling support */ +#define X86_FEATURE_VMCBCLEAN (15*32+ 5)/* "vmcb_clean" VMCB clean bits support */ +#define X86_FEATURE_FLUSHBYASID (15*32+ 6)/* flush-by-ASID support */ +#define X86_FEATURE_DECODEASSISTS (15*32+ 7)/* Decode Assists support */ +#define X86_FEATURE_PAUSEFILTER (15*32+10)/* filtered pause intercept */ +#define X86_FEATURE_PFTHRESHOLD (15*32+12)/* pause filter threshold */ +/* * BUG word(s) */
arch/x86/kernel/cpu/common.c +5 -0
@@ -618,6 +618,8 @@ void get_cpu_cap(struct cpuinfo_x86 *c)cpuid_count (0x00000007 ,0 , &eax, &ebx, &ecx, &edx); c->x86_capability[9 ] = ebx; + + c->x86_capability[14 ] =cpuid_eax (0x00000006 ); }/* Extended state features: level 0x0000000d */
@@ -679,6 +681,9 @@ void get_cpu_cap(struct cpuinfo_x86 *c)if (c->extended_cpuid_level >=0x80000007 ) c->x86_power =cpuid_edx (0x80000007 ); +if (c->extended_cpuid_level >=0x8000000a ) + c->x86_capability[15 ] =cpuid_edx (0x8000000a ); +init_scattered_cpuid_features (c); }
arch/x86/kernel/cpu/scattered.c +0 -20
@@ -31,32 +31,12 @@ void init_scattered_cpuid_features(struct cpuinfo_x86 *c)const struct cpuid_bit *cb;static const struct cpuid_bit cpuid_bits[] = { - {X86_FEATURE_DTHERM ,CR_EAX ,0 ,0x00000006 ,0 }, - {X86_FEATURE_IDA ,CR_EAX ,1 ,0x00000006 ,0 }, - {X86_FEATURE_ARAT ,CR_EAX ,2 ,0x00000006 ,0 }, - {X86_FEATURE_PLN ,CR_EAX ,4 ,0x00000006 ,0 }, - {X86_FEATURE_PTS ,CR_EAX ,6 ,0x00000006 ,0 }, - {X86_FEATURE_HWP ,CR_EAX ,7 ,0x00000006 ,0 }, - {X86_FEATURE_HWP_NOTIFY ,CR_EAX ,8 ,0x00000006 ,0 }, - {X86_FEATURE_HWP_ACT_WINDOW ,CR_EAX ,9 ,0x00000006 ,0 }, - {X86_FEATURE_HWP_EPP ,CR_EAX ,10 ,0x00000006 ,0 }, - {X86_FEATURE_HWP_PKG_REQ ,CR_EAX ,11 ,0x00000006 ,0 }, {X86_FEATURE_INTEL_PT ,CR_EBX ,25 ,0x00000007 ,0 }, {X86_FEATURE_APERFMPERF ,CR_ECX ,0 ,0x00000006 ,0 }, {X86_FEATURE_EPB ,CR_ECX ,3 ,0x00000006 ,0 }, {X86_FEATURE_HW_PSTATE ,CR_EDX ,7 ,0x80000007 ,0 }, {X86_FEATURE_CPB ,CR_EDX ,9 ,0x80000007 ,0 }, {X86_FEATURE_PROC_FEEDBACK ,CR_EDX ,11 ,0x80000007 ,0 }, - {X86_FEATURE_NPT ,CR_EDX ,0 ,0x8000000a ,0 }, - {X86_FEATURE_LBRV ,CR_EDX ,1 ,0x8000000a ,0 }, - {X86_FEATURE_SVML ,CR_EDX ,2 ,0x8000000a ,0 }, - {X86_FEATURE_NRIPS ,CR_EDX ,3 ,0x8000000a ,0 }, - {X86_FEATURE_TSCRATEMSR ,CR_EDX ,4 ,0x8000000a ,0 }, - {X86_FEATURE_VMCBCLEAN ,CR_EDX ,5 ,0x8000000a ,0 }, - {X86_FEATURE_FLUSHBYASID ,CR_EDX ,6 ,0x8000000a ,0 }, - {X86_FEATURE_DECODEASSISTS ,CR_EDX ,7 ,0x8000000a ,0 }, - {X86_FEATURE_PAUSEFILTER ,CR_EDX ,10 ,0x8000000a ,0 }, - {X86_FEATURE_PFTHRESHOLD ,CR_EDX ,12 ,0x8000000a ,0 }, {0 ,0 ,0 ,0 ,0 } };
Signature · 1 line
2.3.5
Borislav Petkov
maintainerPATCH 2/4 x86/cpufeature: Cleanup get_cpu_cap()
In reply to Borislav PetkovFrom: Borislav Petkov <[email protected]>
Add an enum for the ->x86_capability array indices and cleanup get_cpu_cap() by killing some redundant local vars.
--- arch/x86/include/asm/cpufeature.h | 20 +++++++++++++++++ arch/x86/kernel/cpu/centaur.c | 2 +- arch/x86/kernel/cpu/common.c | 47 ++++++++++++++++++--------------------- arch/x86/kernel/cpu/transmeta.c | 4 ++-- 4 files changed, 45 insertions(+), 28 deletions(-)
arch/x86/include/asm/cpufeature.h +20 -0
@@ -288,6 +288,26 @@#include <asm/asm.h> #include <linux/bitops.h> +enum cpuid_leafs +{ +CPUID_1_EDX =0 , +CPUID_8000_0001_EDX , +CPUID_8086_0001_EDX , +CPUID_LNX_1 , +CPUID_1_ECX , +CPUID_C000_0001_EDX , +CPUID_8000_0001_ECX , +CPUID_LNX_2 , +CPUID_LNX_3 , +CPUID_7_0_EBX , +CPUID_D_1_EAX , +CPUID_F_0_EDX , +CPUID_F_1_EDX , +CPUID_8000_0008_EBX , +CPUID_6_EAX , +CPUID_8000_000A_EDX , +}; +#ifdef CONFIG_X86_FEATURE_NAMES extern const char *const x86_cap_flags[NCAPINTS *32 ];extern const char *const x86_power_flags[32 ];
arch/x86/kernel/cpu/centaur.c +1 -1
@@ -43,7 +43,7 @@ static void init_c3(struct cpuinfo_x86 *c)/* store Centaur Extended Feature Flags as * word 5 of the CPU capability bit array */ - c->x86_capability[5 ] =cpuid_edx (0xC0000001 ); + c->x86_capability[CPUID_C000_0001_EDX ] =cpuid_edx (0xC0000001 ); }#ifdef CONFIG_X86_32 /* Cyrix III family needs CX8 & PGE explicitly enabled. */
arch/x86/kernel/cpu/common.c +22 -25
103 lines not on this page · raw message
arch/x86/kernel/cpu/transmeta.c +2 -2
@@ -12,7 +12,7 @@ static void early_init_transmeta(struct cpuinfo_x86 *c) xlvl =cpuid_eax (0x80860000 );if ((xlvl &0xffff0000 ) ==0x80860000 ) {if (xlvl >=0x80860001 ) - c->x86_capability[2 ] =cpuid_edx (0x80860001 ); + c->x86_capability[CPUID_8086_0001_EDX ] =cpuid_edx (0x80860001 ); } }
@@ -82,7 +82,7 @@ static void init_transmeta(struct cpuinfo_x86 *c)/* Unhide possibly hidden capability flags */ rdmsr (0x80860004 , cap_mask, uk);wrmsr (0x80860004 , ~0 , uk); - c->x86_capability[0 ] =cpuid_edx (0x00000001 ); + c->x86_capability[CPUID_1_EDX ] =cpuid_edx (0x00000001 );wrmsr (0x80860004 , cap_mask, uk);/* All Transmeta CPUs have a constant TSC */
Signature · 1 line
2.3.5
Borislav Petkov
maintainerPATCH 3/4 x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
In reply to Borislav PetkovFrom: Borislav Petkov <[email protected]>
Those are stupid and code should use static_cpu_has_safe() or boot_cpu_has() instead. Kill the least used and unused ones.
The remaining ones need more careful inspection before a conversion can happen. On the TODO.
--- arch/x86/crypto/chacha20_glue.c | 2 +- arch/x86/crypto/crc32c-intel_glue.c | 2 +- arch/x86/include/asm/cmpxchg_32.h | 2 +- arch/x86/include/asm/cmpxchg_64.h | 2 +- arch/x86/include/asm/cpufeature.h | 37 ++++------------------------- arch/x86/include/asm/xor_32.h | 2 +- arch/x86/kernel/cpu/amd.c | 4 ++-- arch/x86/kernel/cpu/common.c | 4 +++- arch/x86/kernel/cpu/intel.c | 3 ++- arch/x86/kernel/cpu/intel_cacheinfo.c | 6 ++--- arch/x86/kernel/cpu/mtrr/generic.c | 2 +- arch/x86/kernel/cpu/mtrr/main.c | 2 +- arch/x86/kernel/cpu/perf_event_amd.c | 4 ++-- arch/x86/kernel/cpu/perf_event_amd_uncore.c | 11 +++++---- arch/x86/kernel/fpu/init.c | 4 ++-- arch/x86/kernel/hw_breakpoint.c | 6 +++-- arch/x86/kernel/smpboot.c | 2 +- arch/x86/kernel/vm86_32.c | 4 +++- arch/x86/mm/setup_nx.c | 4 ++-- drivers/char/hw_random/via-rng.c | 5 ++-- drivers/crypto/padlock-aes.c | 2 +- drivers/crypto/padlock-sha.c | 2 +- drivers/iommu/intel_irq_remapping.c | 2 +- fs/btrfs/disk-io.c | 2 +- 24 files changed, 48 insertions(+), 68 deletions(-)
arch/x86/crypto/chacha20_glue.c +1 -1
@@ -125,7 +125,7 @@ static struct crypto_alg alg = {static int __init chacha20_simd_mod_init (void ) { -if (!cpu_has_ssse3) +if (!boot_cpu_has (X86_FEATURE_SSSE3 ))return -ENODEV ;#ifdef CONFIG_AS_AVX2
arch/x86/crypto/crc32c-intel_glue.c +1 -1
@@ -257,7 +257,7 @@ static int __init crc32c_intel_mod_init(void)if (!x86_match_cpu (crc32c_cpu_id))return -ENODEV ;#ifdef CONFIG_X86_64 -if (cpu_has_pclmulqdq) { +if (boot_cpu_has (X86_FEATURE_PCLMULQDQ )) { alg.update = crc32c_pcl_intel_update; alg.finup = crc32c_pcl_intel_finup; alg.digest = crc32c_pcl_intel_digest;
arch/x86/include/asm/cmpxchg_32.h +1 -1
@@ -109,6 +109,6 @@ static inline u64 __cmpxchg64_local(volatile u64 *ptr, u64 old, u64 new)#endif -#define system_has_cmpxchg_double () cpu_has_cx8 +#define system_has_cmpxchg_double () boot_cpu_has(X86_FEATURE_CX8)#endif /* _ASM_X86_CMPXCHG_32_H */
arch/x86/include/asm/cmpxchg_64.h +1 -1
@@ -18,6 +18,6 @@ static inline void set_64bit(volatile u64 *ptr, u64 val)cmpxchg_local ((ptr), (o), (n)); \ }) -#define system_has_cmpxchg_double () cpu_has_cx16 +#define system_has_cmpxchg_double () boot_cpu_has(X86_FEATURE_CX16)#endif /* _ASM_X86_CMPXCHG_64_H */
arch/x86/include/asm/cpufeature.h +4 -33
@@ -385,58 +385,29 @@ extern const char * const x86_bug_flags[NBUGINTS*32]; } while (0)#define cpu_has_fpu boot_cpu_has(X86_FEATURE_FPU) -#define cpu_has_de boot_cpu_has(X86_FEATURE_DE)#define cpu_has_pse boot_cpu_has(X86_FEATURE_PSE)#define cpu_has_tsc boot_cpu_has(X86_FEATURE_TSC)#define cpu_has_pge boot_cpu_has(X86_FEATURE_PGE)#define cpu_has_apic boot_cpu_has(X86_FEATURE_APIC) -#define cpu_has_sep boot_cpu_has(X86_FEATURE_SEP) -#define cpu_has_mtrr boot_cpu_has(X86_FEATURE_MTRR) -#define cpu_has_mmx boot_cpu_has(X86_FEATURE_MMX)#define cpu_has_fxsr boot_cpu_has(X86_FEATURE_FXSR)#define cpu_has_xmm boot_cpu_has(X86_FEATURE_XMM)#define cpu_has_xmm2 boot_cpu_has(X86_FEATURE_XMM2) -#define cpu_has_xmm3 boot_cpu_has(X86_FEATURE_XMM3) -#define cpu_has_ssse3 boot_cpu_has(X86_FEATURE_SSSE3)#define cpu_has_aes boot_cpu_has(X86_FEATURE_AES)#define cpu_has_avx boot_cpu_has(X86_FEATURE_AVX)#define cpu_has_avx2 boot_cpu_has(X86_FEATURE_AVX2) -#define cpu_has_ht boot_cpu_has(X86_FEATURE_HT) -#define cpu_has_nx boot_cpu_has(X86_FEATURE_NX) -#define cpu_has_xstore boot_cpu_has(X86_FEATURE_XSTORE) -#define cpu_has_xstore_enabled boot_cpu_has(X86_FEATURE_XSTORE_EN) -#define cpu_has_xcrypt boot_cpu_has(X86_FEATURE_XCRYPT) -#define cpu_has_xcrypt_enabled boot_cpu_has(X86_FEATURE_XCRYPT_EN) -#define cpu_has_ace2 boot_cpu_has(X86_FEATURE_ACE2) -#define cpu_has_ace2_enabled boot_cpu_has(X86_FEATURE_ACE2_EN) -#define cpu_has_phe boot_cpu_has(X86_FEATURE_PHE) -#define cpu_has_phe_enabled boot_cpu_has(X86_FEATURE_PHE_EN) -#define cpu_has_pmm boot_cpu_has(X86_FEATURE_PMM) -#define cpu_has_pmm_enabled boot_cpu_has(X86_FEATURE_PMM_EN) -#define cpu_has_ds boot_cpu_has(X86_FEATURE_DS) -#define cpu_has_pebs boot_cpu_has(X86_FEATURE_PEBS)#define cpu_has_clflush boot_cpu_has(X86_FEATURE_CLFLUSH) -#define cpu_has_bts boot_cpu_has(X86_FEATURE_BTS)#define cpu_has_gbpages boot_cpu_has(X86_FEATURE_GBPAGES)#define cpu_has_arch_perfmon boot_cpu_has(X86_FEATURE_ARCH_PERFMON)#define cpu_has_pat boot_cpu_has(X86_FEATURE_PAT) -#define cpu_has_xmm4_1 boot_cpu_has(X86_FEATURE_XMM4_1) -#define cpu_has_xmm4_2 boot_cpu_has(X86_FEATURE_XMM4_2)#define cpu_has_x2apic boot_cpu_has(X86_FEATURE_X2APIC)#define cpu_has_xsave boot_cpu_has(X86_FEATURE_XSAVE) -#define cpu_has_xsaveopt boot_cpu_has(X86_FEATURE_XSAVEOPT)#define cpu_has_xsaves boot_cpu_has(X86_FEATURE_XSAVES)#define cpu_has_osxsave boot_cpu_has(X86_FEATURE_OSXSAVE)#define cpu_has_hypervisor boot_cpu_has(X86_FEATURE_HYPERVISOR) -#define cpu_has_pclmulqdq boot_cpu_has(X86_FEATURE_PCLMULQDQ) -#define cpu_has_perfctr_core boot_cpu_has(X86_FEATURE_PERFCTR_CORE) -#define cpu_has_perfctr_nb boot_cpu_has(X86_FEATURE_PERFCTR_NB) -#define cpu_has_perfctr_l2 boot_cpu_has(X86_FEATURE_PERFCTR_L2) -#define cpu_has_cx8 boot_cpu_has(X86_FEATURE_CX8) -#define cpu_has_cx16 boot_cpu_has(X86_FEATURE_CX16) -#define cpu_has_eager_fpu boot_cpu_has(X86_FEATURE_EAGER_FPU) -#define cpu_has_topoext boot_cpu_has(X86_FEATURE_TOPOEXT) -#define cpu_has_bpext boot_cpu_has(X86_FEATURE_BPEXT) +/* +* Do not add any more of those clumsy macros - use static_cpu_has_safe() for +* fast paths and boot_cpu_has() otherwise! +*/ #if __GNUC__ >=4 extern void warn_pre_alternatives (void );
arch/x86/include/asm/xor_32.h +1 -1
@@ -553,7 +553,7 @@ do { \if (cpu_has_xmm) { \xor_speed (&xor_block_pIII_sse); \xor_speed (&xor_block_sse_pf64); \ - }else if (cpu_has_mmx) { \ + }else if (static_cpu_has_safe (X86_FEATURE_MMX )) { \xor_speed (&xor_block_pII_mmx); \xor_speed (&xor_block_p5_mmx); \ }else { \
arch/x86/kernel/cpu/amd.c +2 -2
@@ -304,7 +304,7 @@ static void amd_get_topology(struct cpuinfo_x86 *c)int cpu =smp_processor_id ();/* get information required for multi-node processors */ -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) {u32 eax, ebx, ecx, edx;cpuid (0x8000001e , &eax, &ebx, &ecx, &edx);
@@ -922,7 +922,7 @@ static bool cpu_has_amd_erratum(struct cpuinfo_x86 *cpu, const int *erratum)void set_dr_addr_mask (unsigned long mask,int dr) { -if (!cpu_has_bpext) +if (!boot_cpu_has (X86_FEATURE_BPEXT ))return ;switch (dr) {
arch/x86/kernel/cpu/common.c +3 -1
@@ -1445,7 +1445,9 @@ void cpu_init(void)printk (KERN_INFO "Initializing CPU#%d\n" , cpu); -if (cpu_feature_enabled (X86_FEATURE_VME ) || cpu_has_tsc || cpu_has_de) +if (cpu_feature_enabled (X86_FEATURE_VME ) || + cpu_has_tsc || +boot_cpu_has (X86_FEATURE_DE ))cr4_clear_bits (X86_CR4_VME |X86_CR4_PVI |X86_CR4_TSD |X86_CR4_DE );load_current_idt ();
arch/x86/kernel/cpu/intel.c +2 -1
@@ -445,7 +445,8 @@ static void init_intel(struct cpuinfo_x86 *c)if (cpu_has_xmm2)set_cpu_cap (c,X86_FEATURE_LFENCE_RDTSC ); -if (cpu_has_ds) { + +if (boot_cpu_has (X86_FEATURE_DS )) {unsigned int l1;rdmsr (MSR_IA32_MISC_ENABLE , l1, l2);if (!(l1 & (1 <<11 )))
arch/x86/kernel/cpu/intel_cacheinfo.c +3 -3
@@ -591,7 +591,7 @@ cpuid4_cache_lookup_regs(int index, struct _cpuid4_info_regs *this_leaf)unsigned edx;if (boot_cpu_data.x86_vendor ==X86_VENDOR_AMD ) { -if (cpu_has_topoext) +if (boot_cpu_has (X86_FEATURE_TOPOEXT ))cpuid_count (0x8000001d , index, &eax.full, &ebx.full, &ecx.full, &edx);else
@@ -637,7 +637,7 @@ static int find_num_cache_leaves(struct cpuinfo_x86 *c)void init_amd_cacheinfo (struct cpuinfo_x86 *c) { -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) { num_cache_leaves =find_num_cache_leaves (c); }else if (c->extended_cpuid_level >=0x80000006 ) {if (cpuid_edx (0x80000006 ) &0xf000 )
@@ -809,7 +809,7 @@ static int __cache_amd_cpumap_setup(unsigned int cpu, int index,struct cacheinfo *this_leaf;int i, sibling; -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) {unsigned int apicid, nshared, first, last; this_leaf = this_cpu_ci->info_list + index;
arch/x86/kernel/cpu/mtrr/generic.c +1 -1
@@ -349,7 +349,7 @@ static void get_fixed_ranges(mtrr_type *frs)void mtrr_save_fixed_ranges (void *info) { -if (cpu_has_mtrr) +if (static_cpu_has_safe (X86_FEATURE_MTRR ))get_fixed_ranges (mtrr_state.fixed_ranges); }
arch/x86/kernel/cpu/mtrr/main.c +1 -1
@@ -682,7 +682,7 @@ void __init mtrr_bp_init(void) phys_addr =32 ; -if (cpu_has_mtrr) { +if (boot_cpu_has (X86_FEATURE_MTRR )) { mtrr_if = &generic_mtrr_ops; size_or_mask =SIZE_OR_MASK_BITS (36 ); size_and_mask =0x00f00000 ;
arch/x86/kernel/cpu/perf_event_amd.c +2 -2
@@ -160,7 +160,7 @@ static inline int amd_pmu_addr_offset(int index, bool eventsel)if (offset)return offset; -if (!cpu_has_perfctr_core) +if (!boot_cpu_has (X86_FEATURE_PERFCTR_CORE )) offset = index;else offset = index <<1 ;
@@ -652,7 +652,7 @@ static __initconst const struct x86_pmu amd_pmu = {static int __init amd_core_pmu_init (void ) { -if (!cpu_has_perfctr_core) +if (!boot_cpu_has (X86_FEATURE_PERFCTR_CORE ))return 0 ;switch (boot_cpu_data.x86) {
arch/x86/kernel/cpu/perf_event_amd_uncore.c +6 -5
@@ -523,10 +523,10 @@ static int __init amd_uncore_init(void)if (boot_cpu_data.x86_vendor !=X86_VENDOR_AMD ) goto fail_nodev; -if (!cpu_has_topoext) +if (!boot_cpu_has (X86_FEATURE_TOPOEXT )) goto fail_nodev; -if (cpu_has_perfctr_nb) { +if (boot_cpu_has (X86_FEATURE_PERFCTR_NB )) { amd_uncore_nb = alloc_percpu(struct amd_uncore *);if (!amd_uncore_nb) { ret = -ENOMEM ;
@@ -540,7 +540,7 @@ static int __init amd_uncore_init(void) ret =0 ; } -if (cpu_has_perfctr_l2) { +if (boot_cpu_has (X86_FEATURE_PERFCTR_L2 )) { amd_uncore_l2 = alloc_percpu(struct amd_uncore *);if (!amd_uncore_l2) { ret = -ENOMEM ;
@@ -583,10 +583,11 @@ fail_online:/* amd_uncore_nb/l2 should have been freed by cleanup_cpu_online */ amd_uncore_nb = amd_uncore_l2 =NULL ; -if (cpu_has_perfctr_l2) + +if (boot_cpu_has (X86_FEATURE_PERFCTR_L2 ))perf_pmu_unregister (&amd_l2_pmu); fail_l2: -if (cpu_has_perfctr_nb) +if (boot_cpu_has (X86_FEATURE_PERFCTR_NB ))perf_pmu_unregister (&amd_nb_pmu);if (amd_uncore_l2)free_percpu (amd_uncore_l2);
arch/x86/kernel/fpu/init.c +2 -2
@@ -12,7 +12,7 @@*/ static void fpu__init_cpu_ctx_switch (void ) { -if (!cpu_has_eager_fpu) +if (!boot_cpu_has (X86_FEATURE_EAGER_FPU ))stts ();else clts ();
@@ -287,7 +287,7 @@ static void __init fpu__init_system_ctx_switch(void)current_thread_info ()->status =0 ;/* Auto enable eagerfpu for xsaveopt */ -if (cpu_has_xsaveopt && eagerfpu !=DISABLE ) +if (boot_cpu_has (X86_FEATURE_XSAVEOPT ) && eagerfpu !=DISABLE ) eagerfpu =ENABLE ;if (xfeatures_mask &XFEATURE_MASK_EAGER ) {
arch/x86/kernel/hw_breakpoint.c +4 -2
@@ -300,6 +300,10 @@ static int arch_build_bp_info(struct perf_event *bp)return -EINVAL ;if (bp->attr.bp_addr & (bp->attr.bp_len -1 ))return -EINVAL ; + +if (!boot_cpu_has (X86_FEATURE_BPEXT )) +return -EOPNOTSUPP ; +/* * It's impossible to use a range breakpoint to fake out * user vs kernel detection because bp_len - 1 can't
@@ -307,8 +311,6 @@ static int arch_build_bp_info(struct perf_event *bp)* breakpoints, then we'll have to check for kprobe-blacklisted * addresses anywhere in the range. */ -if (!cpu_has_bpext) -return -EOPNOTSUPP ; info->mask = bp->attr.bp_len -1 ; info->len =X86_BREAKPOINT_LEN_1 ; }
arch/x86/kernel/smpboot.c +1 -1
@@ -304,7 +304,7 @@ do { \static bool match_smt (struct cpuinfo_x86 *c,struct cpuinfo_x86 *o) { -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) {int cpu1 = c->cpu_index, cpu2 = o->cpu_index; if (c->phys_proc_id == o->phys_proc_id &&
arch/x86/kernel/vm86_32.c +3 -1
@@ -357,8 +357,10 @@ static long do_sys_vm86(struct vm86plus_struct __user *user_vm86, bool plus) tss = &per_cpu (cpu_tss,get_cpu ());/* make room for real-mode segments */ tsk->thread.sp0 +=16 ; -if (cpu_has_sep) + +if (static_cpu_has_safe (X86_FEATURE_SEP )) tsk->thread.sysenter_cs =0 ; +load_sp0 (tss, &tsk->thread);put_cpu ();
arch/x86/mm/setup_nx.c +2 -2
@@ -31,7 +31,7 @@ early_param("noexec", noexec_setup);void x86_configure_nx (void ) { -if (cpu_has_nx && !disable_nx) +if (boot_cpu_has (X86_FEATURE_NX ) && !disable_nx) __supported_pte_mask |= _PAGE_NX;else __supported_pte_mask &= ~_PAGE_NX;
@@ -39,7 +39,7 @@ void x86_configure_nx(void)void __init x86_report_nx (void ) { -if (!cpu_has_nx) { +if (!boot_cpu_has (X86_FEATURE_NX )) {printk (KERN_NOTICE "Notice: NX (Execute Disable) protection " "missing in CPU!\n" ); }else {
drivers/char/hw_random/via-rng.c +3 -2
@@ -140,7 +140,7 @@ static int via_rng_init(struct hwrng *rng)* RNG configuration like it used to be the case in this * register */ if ((c->x86 ==6 ) && (c->x86_model >=0x0f )) { -if (!cpu_has_xstore_enabled) { +if (!boot_cpu_has (X86_FEATURE_XSTORE_EN )) {pr_err (PFX "can't enable hardware RNG " "if XSTORE is not enabled\n" );return -ENODEV ;
@@ -200,8 +200,9 @@ static int __init mod_init(void) {int err; -if (!cpu_has_xstore) +if (!boot_cpu_has (X86_FEATURE_XSTORE ))return -ENODEV ; +pr_info ("VIA RNG detected\n" ); err =hwrng_register (&via_rng);if (err) {
drivers/crypto/padlock-aes.c +1 -1
@@ -515,7 +515,7 @@ static int __init padlock_init(void)if (!x86_match_cpu (padlock_cpu_id))return -ENODEV ; -if (!cpu_has_xcrypt_enabled) { +if (!boot_cpu_has (X86_FEATURE_XCRYPT_EN )) { printk(KERN_NOTICE PFX "VIA PadLock detected, but not enabled. Hmm, strange...\n");return -ENODEV ; }
drivers/crypto/padlock-sha.c +1 -1
@@ -540,7 +540,7 @@ static int __init padlock_init(void)struct shash_alg *sha1;struct shash_alg *sha256; -if (!x86_match_cpu (padlock_sha_ids) || !cpu_has_phe_enabled) +if (!x86_match_cpu (padlock_sha_ids) || !boot_cpu_has (X86_FEATURE_PHE_EN ))return -ENODEV ;/* Register the newly added algorithm module if on *
drivers/iommu/intel_irq_remapping.c +1 -1
@@ -753,7 +753,7 @@ static inline void set_irq_posting_cap(void)* should have X86_FEATURE_CX16 support, this has been confirmed * with Intel hardware guys. */ -if ( cpu_has_cx16 ) +if (boot_cpu_has (X86_FEATURE_CX16 )) intel_irq_remap_ops.capability |=1 <<IRQ_POSTING_CAP ;for_each_iommu (iommu, drhd)
fs/btrfs/disk-io.c +1 -1
@@ -923,7 +923,7 @@ static int check_async_write(struct inode *inode, unsigned long bio_flags)if (bio_flags &EXTENT_BIO_TREE_LOG )return 0 ;#ifdef CONFIG_X86 -if (cpu_has_xmm4_2) +if (static_cpu_has_safe (X86_FEATURE_XMM4_2 ))return 0 ;#endif return 1 ;
Signature · 1 line
2.3.5
Re: [PATCH 3/4] x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
arch/x86/include/asm/xor_32.h do { \ · line 556
Quoted 12 lines · Borislav Petkov
#if
__GNUC__>= 4
extern voidwarn_pre_alternatives(void);arch/x86/include/asm/xor_32.h +1
-1@@ -553,7 +553,7 @@ do { \if (cpu_has_xmm) { \xor_speed (&xor_block_pIII_sse); \xor_speed (&xor_block_sse_pf64); \ - }else if (cpu_has_mmx) { \ + }else if (static_cpu_has_safe (X86_FEATURE_MMX )) { \
This one can be boot_cpu_has. It's only used in crypto init for testing the fastest xor algo.
arch/x86/kernel/cpu/mtrr/generic.c static void get_fixed_ranges(mtrr_type *frs) · line 353
Quoted 11 lines
arch/x86/kernel/cpu/mtrr/generic.c +1
-1@@ -349,7 +349,7 @@ static void get_fixed_ranges(mtrr_type *frs)void mtrr_save_fixed_ranges (void *info) { -if (cpu_has_mtrr) +if (static_cpu_has_safe (X86_FEATURE_MTRR ))get_fixed_ranges (mtrr_state.fixed_ranges);
This one is not really hotpath either. Its called from cpu_up and from hibernation code.
Thanks,
tglx
Borislav Petkov
maintainerPATCH 3/4 x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
Quoted 23 lines · Thomas Gleixner
arch/x86/include/asm/xor_32.h do { \ · line 556
Quoted 12 lines · Borislav Petkov
#if
__GNUC__>= 4
extern voidwarn_pre_alternatives(void);arch/x86/include/asm/xor_32.h +1
-1@@ -553,7 +553,7 @@ do { \if (cpu_has_xmm) { \xor_speed (&xor_block_pIII_sse); \xor_speed (&xor_block_sse_pf64); \ - }else if (cpu_has_mmx) { \ + }else if (static_cpu_has_safe (X86_FEATURE_MMX )) { \This one can be
boot_cpu_has. It's only used in crypto init for testing the fastest xor algo.Quoted 6 lines
arch/x86/kernel/cpu/mtrr/generic.c +0
-0@@ -349,7 +349,7 @@ static void get_fixed_ranges(mtrr_type *frs)
Quoted 8 lines
arch/x86/kernel/cpu/mtrr/generic.c static void get_fixed_ranges(mtrr_type *frs) · line 353
Quoted 5 lines
arch/x86/kernel/cpu/mtrr/generic.c +1
-1void mtrr_save_fixed_ranges (void *info) { -if (cpu_has_mtrr) +if (static_cpu_has_safe (X86_FEATURE_MTRR ))get_fixed_ranges (mtrr_state.fixed_ranges);This one is not really hotpath either. Its called from
cpu_upand from hibernation code.
---
From: Borislav Petkov <[email protected]>
Date: Mon, 9 Nov 2015 10:38:45 +0100
Subject: [PATCH -v1.1 3/4] x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
Those are stupid and code should use static_cpu_has_safe() or boot_cpu_has() instead. Kill the least used and unused ones.
The remaining ones need more careful inspection before a conversion can happen. On the TODO.
--- arch/x86/crypto/chacha20_glue.c | 2 +- arch/x86/crypto/crc32c-intel_glue.c | 2 +- arch/x86/include/asm/cmpxchg_32.h | 2 +- arch/x86/include/asm/cmpxchg_64.h | 2 +- arch/x86/include/asm/cpufeature.h | 37 ++++------------------------- arch/x86/include/asm/xor_32.h | 2 +- arch/x86/kernel/cpu/amd.c | 4 ++-- arch/x86/kernel/cpu/common.c | 4 +++- arch/x86/kernel/cpu/intel.c | 3 ++- arch/x86/kernel/cpu/intel_cacheinfo.c | 6 ++--- arch/x86/kernel/cpu/mtrr/generic.c | 2 +- arch/x86/kernel/cpu/mtrr/main.c | 2 +- arch/x86/kernel/cpu/perf_event_amd.c | 4 ++-- arch/x86/kernel/cpu/perf_event_amd_uncore.c | 11 +++++---- arch/x86/kernel/fpu/init.c | 4 ++-- arch/x86/kernel/hw_breakpoint.c | 6 +++-- arch/x86/kernel/smpboot.c | 2 +- arch/x86/kernel/vm86_32.c | 4 +++- arch/x86/mm/setup_nx.c | 4 ++-- drivers/char/hw_random/via-rng.c | 5 ++-- drivers/crypto/padlock-aes.c | 2 +- drivers/crypto/padlock-sha.c | 2 +- drivers/iommu/intel_irq_remapping.c | 2 +- fs/btrfs/disk-io.c | 2 +- 24 files changed, 48 insertions(+), 68 deletions(-)
arch/x86/crypto/chacha20_glue.c +1 -1
@@ -125,7 +125,7 @@ static struct crypto_alg alg = {static int __init chacha20_simd_mod_init (void ) { -if (!cpu_has_ssse3) +if (!boot_cpu_has (X86_FEATURE_SSSE3 ))return -ENODEV ;#ifdef CONFIG_AS_AVX2
arch/x86/crypto/crc32c-intel_glue.c +1 -1
@@ -257,7 +257,7 @@ static int __init crc32c_intel_mod_init(void)if (!x86_match_cpu (crc32c_cpu_id))return -ENODEV ;#ifdef CONFIG_X86_64 -if (cpu_has_pclmulqdq) { +if (boot_cpu_has (X86_FEATURE_PCLMULQDQ )) { alg.update = crc32c_pcl_intel_update; alg.finup = crc32c_pcl_intel_finup; alg.digest = crc32c_pcl_intel_digest;
arch/x86/include/asm/cmpxchg_32.h +1 -1
@@ -109,6 +109,6 @@ static inline u64 __cmpxchg64_local(volatile u64 *ptr, u64 old, u64 new)#endif -#define system_has_cmpxchg_double () cpu_has_cx8 +#define system_has_cmpxchg_double () boot_cpu_has(X86_FEATURE_CX8)#endif /* _ASM_X86_CMPXCHG_32_H */
arch/x86/include/asm/cmpxchg_64.h +1 -1
@@ -18,6 +18,6 @@ static inline void set_64bit(volatile u64 *ptr, u64 val)cmpxchg_local ((ptr), (o), (n)); \ }) -#define system_has_cmpxchg_double () cpu_has_cx16 +#define system_has_cmpxchg_double () boot_cpu_has(X86_FEATURE_CX16)#endif /* _ASM_X86_CMPXCHG_64_H */
arch/x86/include/asm/cpufeature.h +4 -33
@@ -385,58 +385,29 @@ extern const char * const x86_bug_flags[NBUGINTS*32]; } while (0)#define cpu_has_fpu boot_cpu_has(X86_FEATURE_FPU) -#define cpu_has_de boot_cpu_has(X86_FEATURE_DE)#define cpu_has_pse boot_cpu_has(X86_FEATURE_PSE)#define cpu_has_tsc boot_cpu_has(X86_FEATURE_TSC)#define cpu_has_pge boot_cpu_has(X86_FEATURE_PGE)#define cpu_has_apic boot_cpu_has(X86_FEATURE_APIC) -#define cpu_has_sep boot_cpu_has(X86_FEATURE_SEP) -#define cpu_has_mtrr boot_cpu_has(X86_FEATURE_MTRR) -#define cpu_has_mmx boot_cpu_has(X86_FEATURE_MMX)#define cpu_has_fxsr boot_cpu_has(X86_FEATURE_FXSR)#define cpu_has_xmm boot_cpu_has(X86_FEATURE_XMM)#define cpu_has_xmm2 boot_cpu_has(X86_FEATURE_XMM2) -#define cpu_has_xmm3 boot_cpu_has(X86_FEATURE_XMM3) -#define cpu_has_ssse3 boot_cpu_has(X86_FEATURE_SSSE3)#define cpu_has_aes boot_cpu_has(X86_FEATURE_AES)#define cpu_has_avx boot_cpu_has(X86_FEATURE_AVX)#define cpu_has_avx2 boot_cpu_has(X86_FEATURE_AVX2) -#define cpu_has_ht boot_cpu_has(X86_FEATURE_HT) -#define cpu_has_nx boot_cpu_has(X86_FEATURE_NX) -#define cpu_has_xstore boot_cpu_has(X86_FEATURE_XSTORE) -#define cpu_has_xstore_enabled boot_cpu_has(X86_FEATURE_XSTORE_EN) -#define cpu_has_xcrypt boot_cpu_has(X86_FEATURE_XCRYPT) -#define cpu_has_xcrypt_enabled boot_cpu_has(X86_FEATURE_XCRYPT_EN) -#define cpu_has_ace2 boot_cpu_has(X86_FEATURE_ACE2) -#define cpu_has_ace2_enabled boot_cpu_has(X86_FEATURE_ACE2_EN) -#define cpu_has_phe boot_cpu_has(X86_FEATURE_PHE) -#define cpu_has_phe_enabled boot_cpu_has(X86_FEATURE_PHE_EN) -#define cpu_has_pmm boot_cpu_has(X86_FEATURE_PMM) -#define cpu_has_pmm_enabled boot_cpu_has(X86_FEATURE_PMM_EN) -#define cpu_has_ds boot_cpu_has(X86_FEATURE_DS) -#define cpu_has_pebs boot_cpu_has(X86_FEATURE_PEBS)#define cpu_has_clflush boot_cpu_has(X86_FEATURE_CLFLUSH) -#define cpu_has_bts boot_cpu_has(X86_FEATURE_BTS)#define cpu_has_gbpages boot_cpu_has(X86_FEATURE_GBPAGES)#define cpu_has_arch_perfmon boot_cpu_has(X86_FEATURE_ARCH_PERFMON)#define cpu_has_pat boot_cpu_has(X86_FEATURE_PAT) -#define cpu_has_xmm4_1 boot_cpu_has(X86_FEATURE_XMM4_1) -#define cpu_has_xmm4_2 boot_cpu_has(X86_FEATURE_XMM4_2)#define cpu_has_x2apic boot_cpu_has(X86_FEATURE_X2APIC)#define cpu_has_xsave boot_cpu_has(X86_FEATURE_XSAVE) -#define cpu_has_xsaveopt boot_cpu_has(X86_FEATURE_XSAVEOPT)#define cpu_has_xsaves boot_cpu_has(X86_FEATURE_XSAVES)#define cpu_has_osxsave boot_cpu_has(X86_FEATURE_OSXSAVE)#define cpu_has_hypervisor boot_cpu_has(X86_FEATURE_HYPERVISOR) -#define cpu_has_pclmulqdq boot_cpu_has(X86_FEATURE_PCLMULQDQ) -#define cpu_has_perfctr_core boot_cpu_has(X86_FEATURE_PERFCTR_CORE) -#define cpu_has_perfctr_nb boot_cpu_has(X86_FEATURE_PERFCTR_NB) -#define cpu_has_perfctr_l2 boot_cpu_has(X86_FEATURE_PERFCTR_L2) -#define cpu_has_cx8 boot_cpu_has(X86_FEATURE_CX8) -#define cpu_has_cx16 boot_cpu_has(X86_FEATURE_CX16) -#define cpu_has_eager_fpu boot_cpu_has(X86_FEATURE_EAGER_FPU) -#define cpu_has_topoext boot_cpu_has(X86_FEATURE_TOPOEXT) -#define cpu_has_bpext boot_cpu_has(X86_FEATURE_BPEXT) +/* +* Do not add any more of those clumsy macros - use static_cpu_has_safe() for +* fast paths and boot_cpu_has() otherwise! +*/ #if __GNUC__ >=4 extern void warn_pre_alternatives (void );
arch/x86/include/asm/xor_32.h +1 -1
@@ -553,7 +553,7 @@ do { \if (cpu_has_xmm) { \xor_speed (&xor_block_pIII_sse); \xor_speed (&xor_block_sse_pf64); \ - }else if (cpu_has_mmx) { \ + }else if (boot_cpu_has (X86_FEATURE_MMX )) { \xor_speed (&xor_block_pII_mmx); \xor_speed (&xor_block_p5_mmx); \ }else { \
arch/x86/kernel/cpu/amd.c +2 -2
@@ -304,7 +304,7 @@ static void amd_get_topology(struct cpuinfo_x86 *c)int cpu =smp_processor_id ();/* get information required for multi-node processors */ -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) {u32 eax, ebx, ecx, edx;cpuid (0x8000001e , &eax, &ebx, &ecx, &edx);
@@ -922,7 +922,7 @@ static bool cpu_has_amd_erratum(struct cpuinfo_x86 *cpu, const int *erratum)void set_dr_addr_mask (unsigned long mask,int dr) { -if (!cpu_has_bpext) +if (!boot_cpu_has (X86_FEATURE_BPEXT ))return ;switch (dr) {
arch/x86/kernel/cpu/common.c +3 -1
@@ -1445,7 +1445,9 @@ void cpu_init(void)printk (KERN_INFO "Initializing CPU#%d\n" , cpu); -if (cpu_feature_enabled (X86_FEATURE_VME ) || cpu_has_tsc || cpu_has_de) +if (cpu_feature_enabled (X86_FEATURE_VME ) || + cpu_has_tsc || +boot_cpu_has (X86_FEATURE_DE ))cr4_clear_bits (X86_CR4_VME |X86_CR4_PVI |X86_CR4_TSD |X86_CR4_DE );load_current_idt ();
arch/x86/kernel/cpu/intel.c +2 -1
@@ -445,7 +445,8 @@ static void init_intel(struct cpuinfo_x86 *c)if (cpu_has_xmm2)set_cpu_cap (c,X86_FEATURE_LFENCE_RDTSC ); -if (cpu_has_ds) { + +if (boot_cpu_has (X86_FEATURE_DS )) {unsigned int l1;rdmsr (MSR_IA32_MISC_ENABLE , l1, l2);if (!(l1 & (1 <<11 )))
arch/x86/kernel/cpu/intel_cacheinfo.c +3 -3
@@ -591,7 +591,7 @@ cpuid4_cache_lookup_regs(int index, struct _cpuid4_info_regs *this_leaf)unsigned edx;if (boot_cpu_data.x86_vendor ==X86_VENDOR_AMD ) { -if (cpu_has_topoext) +if (boot_cpu_has (X86_FEATURE_TOPOEXT ))cpuid_count (0x8000001d , index, &eax.full, &ebx.full, &ecx.full, &edx);else
@@ -637,7 +637,7 @@ static int find_num_cache_leaves(struct cpuinfo_x86 *c)void init_amd_cacheinfo (struct cpuinfo_x86 *c) { -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) { num_cache_leaves =find_num_cache_leaves (c); }else if (c->extended_cpuid_level >=0x80000006 ) {if (cpuid_edx (0x80000006 ) &0xf000 )
@@ -809,7 +809,7 @@ static int __cache_amd_cpumap_setup(unsigned int cpu, int index,struct cacheinfo *this_leaf;int i, sibling; -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) {unsigned int apicid, nshared, first, last; this_leaf = this_cpu_ci->info_list + index;
arch/x86/kernel/cpu/mtrr/generic.c +1 -1
@@ -349,7 +349,7 @@ static void get_fixed_ranges(mtrr_type *frs)void mtrr_save_fixed_ranges (void *info) { -if (cpu_has_mtrr) +if (boot_cpu_has (X86_FEATURE_MTRR ))get_fixed_ranges (mtrr_state.fixed_ranges); }
arch/x86/kernel/cpu/mtrr/main.c +1 -1
@@ -682,7 +682,7 @@ void __init mtrr_bp_init(void) phys_addr =32 ; -if (cpu_has_mtrr) { +if (boot_cpu_has (X86_FEATURE_MTRR )) { mtrr_if = &generic_mtrr_ops; size_or_mask =SIZE_OR_MASK_BITS (36 ); size_and_mask =0x00f00000 ;
arch/x86/kernel/cpu/perf_event_amd.c +2 -2
@@ -160,7 +160,7 @@ static inline int amd_pmu_addr_offset(int index, bool eventsel)if (offset)return offset; -if (!cpu_has_perfctr_core) +if (!boot_cpu_has (X86_FEATURE_PERFCTR_CORE )) offset = index;else offset = index <<1 ;
@@ -652,7 +652,7 @@ static __initconst const struct x86_pmu amd_pmu = {static int __init amd_core_pmu_init (void ) { -if (!cpu_has_perfctr_core) +if (!boot_cpu_has (X86_FEATURE_PERFCTR_CORE ))return 0 ;switch (boot_cpu_data.x86) {
arch/x86/kernel/cpu/perf_event_amd_uncore.c +6 -5
@@ -523,10 +523,10 @@ static int __init amd_uncore_init(void)if (boot_cpu_data.x86_vendor !=X86_VENDOR_AMD ) goto fail_nodev; -if (!cpu_has_topoext) +if (!boot_cpu_has (X86_FEATURE_TOPOEXT )) goto fail_nodev; -if (cpu_has_perfctr_nb) { +if (boot_cpu_has (X86_FEATURE_PERFCTR_NB )) { amd_uncore_nb = alloc_percpu(struct amd_uncore *);if (!amd_uncore_nb) { ret = -ENOMEM ;
@@ -540,7 +540,7 @@ static int __init amd_uncore_init(void) ret =0 ; } -if (cpu_has_perfctr_l2) { +if (boot_cpu_has (X86_FEATURE_PERFCTR_L2 )) { amd_uncore_l2 = alloc_percpu(struct amd_uncore *);if (!amd_uncore_l2) { ret = -ENOMEM ;
@@ -583,10 +583,11 @@ fail_online:/* amd_uncore_nb/l2 should have been freed by cleanup_cpu_online */ amd_uncore_nb = amd_uncore_l2 =NULL ; -if (cpu_has_perfctr_l2) + +if (boot_cpu_has (X86_FEATURE_PERFCTR_L2 ))perf_pmu_unregister (&amd_l2_pmu); fail_l2: -if (cpu_has_perfctr_nb) +if (boot_cpu_has (X86_FEATURE_PERFCTR_NB ))perf_pmu_unregister (&amd_nb_pmu);if (amd_uncore_l2)free_percpu (amd_uncore_l2);
arch/x86/kernel/fpu/init.c +2 -2
@@ -12,7 +12,7 @@*/ static void fpu__init_cpu_ctx_switch (void ) { -if (!cpu_has_eager_fpu) +if (!boot_cpu_has (X86_FEATURE_EAGER_FPU ))stts ();else clts ();
@@ -287,7 +287,7 @@ static void __init fpu__init_system_ctx_switch(void)current_thread_info ()->status =0 ;/* Auto enable eagerfpu for xsaveopt */ -if (cpu_has_xsaveopt && eagerfpu !=DISABLE ) +if (boot_cpu_has (X86_FEATURE_XSAVEOPT ) && eagerfpu !=DISABLE ) eagerfpu =ENABLE ;if (xfeatures_mask &XFEATURE_MASK_EAGER ) {
arch/x86/kernel/hw_breakpoint.c +4 -2
@@ -300,6 +300,10 @@ static int arch_build_bp_info(struct perf_event *bp)return -EINVAL ;if (bp->attr.bp_addr & (bp->attr.bp_len -1 ))return -EINVAL ; + +if (!boot_cpu_has (X86_FEATURE_BPEXT )) +return -EOPNOTSUPP ; +/* * It's impossible to use a range breakpoint to fake out * user vs kernel detection because bp_len - 1 can't
@@ -307,8 +311,6 @@ static int arch_build_bp_info(struct perf_event *bp)* breakpoints, then we'll have to check for kprobe-blacklisted * addresses anywhere in the range. */ -if (!cpu_has_bpext) -return -EOPNOTSUPP ; info->mask = bp->attr.bp_len -1 ; info->len =X86_BREAKPOINT_LEN_1 ; }
arch/x86/kernel/smpboot.c +1 -1
@@ -304,7 +304,7 @@ do { \static bool match_smt (struct cpuinfo_x86 *c,struct cpuinfo_x86 *o) { -if (cpu_has_topoext) { +if (boot_cpu_has (X86_FEATURE_TOPOEXT )) {int cpu1 = c->cpu_index, cpu2 = o->cpu_index; if (c->phys_proc_id == o->phys_proc_id &&
arch/x86/kernel/vm86_32.c +3 -1
@@ -357,8 +357,10 @@ static long do_sys_vm86(struct vm86plus_struct __user *user_vm86, bool plus) tss = &per_cpu (cpu_tss,get_cpu ());/* make room for real-mode segments */ tsk->thread.sp0 +=16 ; -if (cpu_has_sep) + +if (static_cpu_has_safe (X86_FEATURE_SEP )) tsk->thread.sysenter_cs =0 ; +load_sp0 (tss, &tsk->thread);put_cpu ();
arch/x86/mm/setup_nx.c +2 -2
@@ -31,7 +31,7 @@ early_param("noexec", noexec_setup);void x86_configure_nx (void ) { -if (cpu_has_nx && !disable_nx) +if (boot_cpu_has (X86_FEATURE_NX ) && !disable_nx) __supported_pte_mask |= _PAGE_NX;else __supported_pte_mask &= ~_PAGE_NX;
@@ -39,7 +39,7 @@ void x86_configure_nx(void)void __init x86_report_nx (void ) { -if (!cpu_has_nx) { +if (!boot_cpu_has (X86_FEATURE_NX )) {printk (KERN_NOTICE "Notice: NX (Execute Disable) protection " "missing in CPU!\n" ); }else {
drivers/char/hw_random/via-rng.c +3 -2
@@ -140,7 +140,7 @@ static int via_rng_init(struct hwrng *rng)* RNG configuration like it used to be the case in this * register */ if ((c->x86 ==6 ) && (c->x86_model >=0x0f )) { -if (!cpu_has_xstore_enabled) { +if (!boot_cpu_has (X86_FEATURE_XSTORE_EN )) {pr_err (PFX "can't enable hardware RNG " "if XSTORE is not enabled\n" );return -ENODEV ;
@@ -200,8 +200,9 @@ static int __init mod_init(void) {int err; -if (!cpu_has_xstore) +if (!boot_cpu_has (X86_FEATURE_XSTORE ))return -ENODEV ; +pr_info ("VIA RNG detected\n" ); err =hwrng_register (&via_rng);if (err) {
drivers/crypto/padlock-aes.c +1 -1
@@ -515,7 +515,7 @@ static int __init padlock_init(void)if (!x86_match_cpu (padlock_cpu_id))return -ENODEV ; -if (!cpu_has_xcrypt_enabled) { +if (!boot_cpu_has (X86_FEATURE_XCRYPT_EN )) { printk(KERN_NOTICE PFX "VIA PadLock detected, but not enabled. Hmm, strange...\n");return -ENODEV ; }
drivers/crypto/padlock-sha.c +1 -1
@@ -540,7 +540,7 @@ static int __init padlock_init(void)struct shash_alg *sha1;struct shash_alg *sha256; -if (!x86_match_cpu (padlock_sha_ids) || !cpu_has_phe_enabled) +if (!x86_match_cpu (padlock_sha_ids) || !boot_cpu_has (X86_FEATURE_PHE_EN ))return -ENODEV ;/* Register the newly added algorithm module if on *
drivers/iommu/intel_irq_remapping.c +1 -1
@@ -753,7 +753,7 @@ static inline void set_irq_posting_cap(void)* should have X86_FEATURE_CX16 support, this has been confirmed * with Intel hardware guys. */ -if ( cpu_has_cx16 ) +if (boot_cpu_has (X86_FEATURE_CX16 )) intel_irq_remap_ops.capability |=1 <<IRQ_POSTING_CAP ;for_each_iommu (iommu, drhd)
fs/btrfs/disk-io.c +1 -1
@@ -923,7 +923,7 @@ static int check_async_write(struct inode *inode, unsigned long bio_flags)if (bio_flags &EXTENT_BIO_TREE_LOG )return 0 ;#ifdef CONFIG_X86 -if (cpu_has_xmm4_2) +if (static_cpu_has_safe (X86_FEATURE_XMM4_2 ))return 0 ;#endif return 1 ;
Signature · 7 lines
2.3.5
--
Regards/Gruss,
Boris.
ECO tip #101: Trim your mails when you reply.Re: [PATCH 3/4] x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
In reply to Borislav PetkovQuoted 1 line · Borislav Petkov
-#define
cpu_has_htboot_cpu_has(X86_FEATURE_HT)
tip/arch/x86/include/asm/smp.h: In function ‘cpu_has_ht_siblings’:tip/arch/x86/include/asm/smp.h:28:17: error: ‘cpu_has_ht’ undeclared (first use in this function)
has_siblings = cpu_has_ht && smp_num_siblings > 1;
Borislav Petkov
maintainerRe: [PATCH 3/4] x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
Quoted 6 lines · Thomas Gleixner
Quoted 1 line · Borislav Petkov
-#define
cpu_has_htboot_cpu_has(X86_FEATURE_HT)
tip/arch/x86/include/asm/smp.h: In function ‘cpu_has_ht_siblings’:tip/arch/x86/include/asm/smp.h:28:17: error: ‘cpu_has_ht’ undeclared (first use in this function)
has_siblings=cpu_has_ht&&smp_num_siblings> 1;
Yeah, this series needs to go ontop of
ed29210cd6a6 ("x86: Remove unused function cpu_has_ht_siblings()")
which is in tip/x86/cleanups.
BTW, thanks for taking a look, I'll send an updated one removing the static_cpu_has_safe() calls as you suggested in your other mail.
Signature · 4 lines
Regards/Gruss,
Boris.
ECO tip #101: Trim your mails when you reply.Borislav Petkov
maintainerPATCH 4/4 x86/cpu: Provide a config option to disable static_cpu_has
In reply to Borislav PetkovFrom: Borislav Petkov <[email protected]>
This brings .text savings of about ~1.6K when building a tinyconfig. It is off by default so nothing changes for the default.
Kconfig help text from Josh.
--- arch/x86/Kconfig | 11 +++++++++++ arch/x86/include/asm/cpufeature.h | 2 +- 2 files changed, 12 insertions(+), 1 deletion(-)
arch/x86/Kconfig +11 -0
@@ -349,6 +349,17 @@ config X86_FEATURE_NAMES If in doubt, say Y. +config X86_FAST_FEATURE_TESTS + bool "Fast CPU feature tests" if EMBEDDED + default y + ---help--- + Some fast-paths in the kernel depend on the capabilities of the CPU. + Say Y here for the kernel to patch in the appropriate code at runtime + based on the capabilities of the CPU. The infrastructure for patching + code at runtime takes up some additional space; space-constrained + embedded systems may wish to say N here to produce smaller, slightly + slower code. + config X86_X2APIC bool "Support x2apic" depends on X86_LOCAL_APIC && X86_64 && (IRQ_REMAP || HYPERVISOR_GUEST)
arch/x86/include/asm/cpufeature.h +1 -1
@@ -409,7 +409,7 @@ extern const char * const x86_bug_flags[NBUGINTS*32];* fast paths and boot_cpu_has() otherwise! */ -#if __GNUC__ >=4 +#if __GNUC__ >=4 && defined(CONFIG_X86_FAST_FEATURE_TESTS )extern void warn_pre_alternatives (void );extern bool __static_cpu_has_safe (u16 bit);
Signature · 1 line
2.3.5
Re: [PATCH 4/4] x86/cpu: Provide a config option to disable static_cpu_has
Quoted 9 lines · Borislav Petkov
From: Borislav Petkov <[email protected]>
This brings .text savings of about ~1.6K when building a tinyconfig. It is off by default so nothing changes for the default.
Kconfighelp text from Josh.Signed-off-by Borislav Petkov <[email protected]>Cc Josh Triplett <[email protected]>
Assuming this passes the kernel build robot, then:
Quoted 42 lines
arch/x86/Kconfig | 11 +++++++++++ arch/x86/include/asm/cpufeature.h | 2 +- 2 files changed, 12 insertions(+), 1 deletion(-)arch/x86/Kconfig +11
-0@@ -349,6 +349,17 @@ config X86_FEATURE_NAMES If in doubt, say Y. +config X86_FAST_FEATURE_TESTS + bool "Fast CPU feature tests" if EMBEDDED + default y + ---help--- + Some fast-paths in the kernel depend on the capabilities of the CPU. + Say Y here for the kernel to patch in the appropriate code at runtime + based on the capabilities of the CPU. The infrastructure for patching + code at runtime takes up some additional space; space-constrained + embedded systems may wish to say N here to produce smaller, slightly + slower code. + config X86_X2APIC bool "Support x2apic" depends on X86_LOCAL_APIC && X86_64 && (IRQ_REMAP || HYPERVISOR_GUEST)arch/x86/include/asm/cpufeature.h +1
-1@@ -409,7 +409,7 @@ extern const char * const x86_bug_flags[NBUGINTS*32];* fast paths and boot_cpu_has() otherwise! */ -#if __GNUC__ >=4 +#if __GNUC__ >=4 && defined(CONFIG_X86_FAST_FEATURE_TESTS )extern void warn_pre_alternatives (void );extern bool __static_cpu_has_safe (u16 bit);Signature · 1 line
2.3.5
Borislav Petkov
maintainerRe: [PATCH 4/4] x86/cpu: Provide a config option to disable static_cpu_has
Quoted 1 line · Josh Triplett
Assuming this passes the kernel build robot,
Yeah, it does:
https://git.kernel.org/pub/scm/linux/kernel/git/bp/bp cpufeature-v3 df63d2e25f803a083c7c994aeb517bd8464cd970 x86/cpu: Provide a config option to disable static_cpu_has
i386-tinyconfig vmlinux size:
+-------+------+------+-----+---------------------------------------------------------------------------------+ | TOTAL | TEXT | DATA | BSS | | +-------+------+------+-----+---------------------------------------------------------------------------------+ | -1512 | -352 | 0 | 0 | df63d2e25f80 x86/cpu: Provide a config option to disable static_cpu_has | | +5 | 0 | 0 | 0 | 8b9e9097bf18 Merge remote-tracking branch 'tip/x86/cleanups' into cpufeature-v3 | | -1507 | -352 | 0 | 0 | ALL COMMITS | +-------+------+------+-----+---------------------------------------------------------------------------------+
elapsed time: 44m
configs tested: 106
...
Quoted 2 lines
then:
Reviewed-by Josh Triplett <[email protected]>
Thanks!
Signature · 4 lines
Regards/Gruss,
Boris.
ECO tip #101: Trim your mails when you reply.