summaryrefslogtreecommitdiff
path: root/arch/x86/include
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/include')
-rw-r--r--arch/x86/include/asm/cpufeature.h11
-rw-r--r--arch/x86/include/asm/fpu/internal.h7
-rw-r--r--arch/x86/include/asm/sync_bitops.h31
-rw-r--r--arch/x86/include/asm/uaccess.h3
4 files changed, 20 insertions, 32 deletions
diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h
index 0e56ff7e4848..1d337c51f7e6 100644
--- a/arch/x86/include/asm/cpufeature.h
+++ b/arch/x86/include/asm/cpufeature.h
@@ -156,11 +156,14 @@ extern void clear_cpu_cap(struct cpuinfo_x86 *c, unsigned int bit);
#else
/*
- * Static testing of CPU features. Used the same as boot_cpu_has().
- * These will statically patch the target code for additional
- * performance.
+ * Static testing of CPU features. Used the same as boot_cpu_has(). It
+ * statically patches the target code for additional performance. Use
+ * static_cpu_has() only in fast paths, where every cycle counts. Which
+ * means that the boot_cpu_has() variant is already fast enough for the
+ * majority of cases and you should stick to using it as it is generally
+ * only two instructions: a RIP-relative MOV and a TEST.
*/
-static __always_inline __pure bool _static_cpu_has(u16 bit)
+static __always_inline bool _static_cpu_has(u16 bit)
{
asm_volatile_goto("1: jmp 6f\n"
"2:\n"
diff --git a/arch/x86/include/asm/fpu/internal.h b/arch/x86/include/asm/fpu/internal.h
index fb04a3ded7dd..745a19d34f23 100644
--- a/arch/x86/include/asm/fpu/internal.h
+++ b/arch/x86/include/asm/fpu/internal.h
@@ -253,7 +253,7 @@ static inline void copy_xregs_to_kernel_booting(struct xregs_state *xstate)
WARN_ON(system_state != SYSTEM_BOOTING);
- if (static_cpu_has(X86_FEATURE_XSAVES))
+ if (boot_cpu_has(X86_FEATURE_XSAVES))
XSTATE_OP(XSAVES, xstate, lmask, hmask, err);
else
XSTATE_OP(XSAVE, xstate, lmask, hmask, err);
@@ -275,7 +275,7 @@ static inline void copy_kernel_to_xregs_booting(struct xregs_state *xstate)
WARN_ON(system_state != SYSTEM_BOOTING);
- if (static_cpu_has(X86_FEATURE_XSAVES))
+ if (boot_cpu_has(X86_FEATURE_XSAVES))
XSTATE_OP(XRSTORS, xstate, lmask, hmask, err);
else
XSTATE_OP(XRSTOR, xstate, lmask, hmask, err);
@@ -497,8 +497,7 @@ static inline void fpregs_activate(struct fpu *fpu)
* - switch_fpu_finish() restores the new state as
* necessary.
*/
-static inline void
-switch_fpu_prepare(struct fpu *old_fpu, int cpu)
+static inline void switch_fpu_prepare(struct fpu *old_fpu, int cpu)
{
if (static_cpu_has(X86_FEATURE_FPU) && old_fpu->initialized) {
if (!copy_fpregs_to_fpstate(old_fpu))
diff --git a/arch/x86/include/asm/sync_bitops.h b/arch/x86/include/asm/sync_bitops.h
index 2fe745356fb1..6d8d6bc183b7 100644
--- a/arch/x86/include/asm/sync_bitops.h
+++ b/arch/x86/include/asm/sync_bitops.h
@@ -14,6 +14,8 @@
* bit 0 is the LSB of addr; bit 32 is the LSB of (addr+1).
*/
+#include <asm/rmwcc.h>
+
#define ADDR (*(volatile long *)addr)
/**
@@ -29,7 +31,7 @@
*/
static inline void sync_set_bit(long nr, volatile unsigned long *addr)
{
- asm volatile("lock; bts %1,%0"
+ asm volatile("lock; " __ASM_SIZE(bts) " %1,%0"
: "+m" (ADDR)
: "Ir" (nr)
: "memory");
@@ -47,7 +49,7 @@ static inline void sync_set_bit(long nr, volatile unsigned long *addr)
*/
static inline void sync_clear_bit(long nr, volatile unsigned long *addr)
{
- asm volatile("lock; btr %1,%0"
+ asm volatile("lock; " __ASM_SIZE(btr) " %1,%0"
: "+m" (ADDR)
: "Ir" (nr)
: "memory");
@@ -64,7 +66,7 @@ static inline void sync_clear_bit(long nr, volatile unsigned long *addr)
*/
static inline void sync_change_bit(long nr, volatile unsigned long *addr)
{
- asm volatile("lock; btc %1,%0"
+ asm volatile("lock; " __ASM_SIZE(btc) " %1,%0"
: "+m" (ADDR)
: "Ir" (nr)
: "memory");
@@ -78,14 +80,9 @@ static inline void sync_change_bit(long nr, volatile unsigned long *addr)
* This operation is atomic and cannot be reordered.
* It also implies a memory barrier.
*/
-static inline int sync_test_and_set_bit(long nr, volatile unsigned long *addr)
+static inline bool sync_test_and_set_bit(long nr, volatile unsigned long *addr)
{
- unsigned char oldbit;
-
- asm volatile("lock; bts %2,%1\n\tsetc %0"
- : "=qm" (oldbit), "+m" (ADDR)
- : "Ir" (nr) : "memory");
- return oldbit;
+ return GEN_BINARY_RMWcc("lock; " __ASM_SIZE(bts), *addr, c, "Ir", nr);
}
/**
@@ -98,12 +95,7 @@ static inline int sync_test_and_set_bit(long nr, volatile unsigned long *addr)
*/
static inline int sync_test_and_clear_bit(long nr, volatile unsigned long *addr)
{
- unsigned char oldbit;
-
- asm volatile("lock; btr %2,%1\n\tsetc %0"
- : "=qm" (oldbit), "+m" (ADDR)
- : "Ir" (nr) : "memory");
- return oldbit;
+ return GEN_BINARY_RMWcc("lock; " __ASM_SIZE(btr), *addr, c, "Ir", nr);
}
/**
@@ -116,12 +108,7 @@ static inline int sync_test_and_clear_bit(long nr, volatile unsigned long *addr)
*/
static inline int sync_test_and_change_bit(long nr, volatile unsigned long *addr)
{
- unsigned char oldbit;
-
- asm volatile("lock; btc %2,%1\n\tsetc %0"
- : "=qm" (oldbit), "+m" (ADDR)
- : "Ir" (nr) : "memory");
- return oldbit;
+ return GEN_BINARY_RMWcc("lock; " __ASM_SIZE(btc), *addr, c, "Ir", nr);
}
#define sync_test_bit(nr, addr) test_bit(nr, addr)
diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h
index bb21913885a3..c82abd6e4ca3 100644
--- a/arch/x86/include/asm/uaccess.h
+++ b/arch/x86/include/asm/uaccess.h
@@ -586,7 +586,6 @@ extern void __cmpxchg_wrong_size(void)
#define __user_atomic_cmpxchg_inatomic(uval, ptr, old, new, size) \
({ \
int __ret = 0; \
- __typeof__(ptr) __uval = (uval); \
__typeof__(*(ptr)) __old = (old); \
__typeof__(*(ptr)) __new = (new); \
__uaccess_begin_nospec(); \
@@ -662,7 +661,7 @@ extern void __cmpxchg_wrong_size(void)
__cmpxchg_wrong_size(); \
} \
__uaccess_end(); \
- *__uval = __old; \
+ *(uval) = __old; \
__ret; \
})
generated by cgit v1.2.3 (git 2.39.0) at 2024-11-29 14:28:33 +0300