i386: improve and correct inline asm memory constraints

Use "+m" rather than a combination of "=m" and "m" for improved clarity
and consistency.

This also fixes some inlines that incorrectly didn't tell the compiler
that they read the old value at all, potentially causing the compiler to
generate bogus code. It appear that all of those potential bugs were
hidden by the use of extra "volatile" specifiers on the data structures
in question, though.

Signed-off-by: Linus Torvalds <[e-mail]>

include/asm-i386/atomic.h

···
 {
 	__asm__ __volatile__(
 		LOCK_PREFIX "addl %1,%0"
-		:"=m" (v->counter)
-		:"ir" (i), "m" (v->counter));
+		:"+m" (v->counter)
+		:"ir" (i));
 }
 
 /**
···
 {
 	__asm__ __volatile__(
 		LOCK_PREFIX "subl %1,%0"
-		:"=m" (v->counter)
-		:"ir" (i), "m" (v->counter));
+		:"+m" (v->counter)
+		:"ir" (i));
 }
 
 /**
···
 
 	__asm__ __volatile__(
 		LOCK_PREFIX "subl %2,%0; sete %1"
-		:"=m" (v->counter), "=qm" (c)
-		:"ir" (i), "m" (v->counter) : "memory");
+		:"+m" (v->counter), "=qm" (c)
+		:"ir" (i) : "memory");
 	return c;
 }
 
···
 {
 	__asm__ __volatile__(
 		LOCK_PREFIX "incl %0"
-		:"=m" (v->counter)
-		:"m" (v->counter));
+		:"+m" (v->counter));
 }
 
 /**
···
 {
 	__asm__ __volatile__(
 		LOCK_PREFIX "decl %0"
-		:"=m" (v->counter)
-		:"m" (v->counter));
+		:"+m" (v->counter));
 }
 
 /**
···
 
 	__asm__ __volatile__(
 		LOCK_PREFIX "decl %0; sete %1"
-		:"=m" (v->counter), "=qm" (c)
-		:"m" (v->counter) : "memory");
+		:"+m" (v->counter), "=qm" (c)
+		: : "memory");
 	return c != 0;
 }
 
···
 
 	__asm__ __volatile__(
 		LOCK_PREFIX "incl %0; sete %1"
-		:"=m" (v->counter), "=qm" (c)
-		:"m" (v->counter) : "memory");
+		:"+m" (v->counter), "=qm" (c)
+		: : "memory");
 	return c != 0;
 }
 
···
 
 	__asm__ __volatile__(
 		LOCK_PREFIX "addl %2,%0; sets %1"
-		:"=m" (v->counter), "=qm" (c)
-		:"ir" (i), "m" (v->counter) : "memory");
+		:"+m" (v->counter), "=qm" (c)
+		:"ir" (i) : "memory");
 	return c;
 }
 

include/asm-i386/futex.h

···
 	.align	8\n\
 	.long	1b,3b\n\
 	.previous"						\
-	: "=r" (oldval), "=r" (ret), "=m" (*uaddr)		\
-	: "i" (-EFAULT), "m" (*uaddr), "0" (oparg), "1" (0))
+	: "=r" (oldval), "=r" (ret), "+m" (*uaddr)		\
+	: "i" (-EFAULT), "0" (oparg), "1" (0))
 
 #define __futex_atomic_op2(insn, ret, oldval, uaddr, oparg) \
   __asm__ __volatile (						\
···
 	.align	8\n\
 	.long	1b,4b,2b,4b\n\
 	.previous"						\
-	: "=&a" (oldval), "=&r" (ret), "=m" (*uaddr),		\
+	: "=&a" (oldval), "=&r" (ret), "+m" (*uaddr),		\
 	  "=&r" (tem)						\
-	: "r" (oparg), "i" (-EFAULT), "m" (*uaddr), "1" (0))
+	: "r" (oparg), "i" (-EFAULT), "1" (0))
 
 static inline int
 futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
···
 		"	.long   1b,3b				\n"
 		"	.previous				\n"
 
-		: "=a" (oldval), "=m" (*uaddr)
+		: "=a" (oldval), "+m" (*uaddr)
 		: "i" (-EFAULT), "r" (newval), "0" (oldval)
 		: "memory"
 	);

include/asm-i386/local.h

···
 {
 	__asm__ __volatile__(
 		"incl %0"
-		:"=m" (v->counter)
-		:"m" (v->counter));
+		:"+m" (v->counter));
 }
 
 static __inline__ void local_dec(local_t *v)
 {
 	__asm__ __volatile__(
 		"decl %0"
-		:"=m" (v->counter)
-		:"m" (v->counter));
+		:"+m" (v->counter));
 }
 
 static __inline__ void local_add(long i, local_t *v)
 {
 	__asm__ __volatile__(
 		"addl %1,%0"
-		:"=m" (v->counter)
-		:"ir" (i), "m" (v->counter));
+		:"+m" (v->counter)
+		:"ir" (i));
 }
 
 static __inline__ void local_sub(long i, local_t *v)
 {
 	__asm__ __volatile__(
 		"subl %1,%0"
-		:"=m" (v->counter)
-		:"ir" (i), "m" (v->counter));
+		:"+m" (v->counter)
+		:"ir" (i));
 }
 
 /* On x86, these are no better than the atomic variants. */

include/asm-i386/posix_types.h

···
 #undef	__FD_SET
 #define __FD_SET(fd,fdsetp) \
 		__asm__ __volatile__("btsl %1,%0": \
-			"=m" (*(__kernel_fd_set *) (fdsetp)):"r" ((int) (fd)))
+			"+m" (*(__kernel_fd_set *) (fdsetp)):"r" ((int) (fd)))
 
 #undef	__FD_CLR
 #define __FD_CLR(fd,fdsetp) \
 		__asm__ __volatile__("btrl %1,%0": \
-			"=m" (*(__kernel_fd_set *) (fdsetp)):"r" ((int) (fd)))
+			"+m" (*(__kernel_fd_set *) (fdsetp)):"r" ((int) (fd)))
 
 #undef	__FD_ISSET
 #define __FD_ISSET(fd,fdsetp) (__extension__ ({ \

include/asm-i386/rwlock.h

···
 			"popl %%eax\n\t" \
 			"1:\n", \
 			"subl $1,%0\n\t", \
-			"=m" (*(volatile int *)rw) : : "memory")
+			"+m" (*(volatile int *)rw) : : "memory")
 
 #define __build_read_lock(rw, helper)	do { \
 						if (__builtin_constant_p(rw)) \
···
 			"popl %%eax\n\t" \
 			"1:\n", \
 			"subl $" RW_LOCK_BIAS_STR ",%0\n\t", \
-			"=m" (*(volatile int *)rw) : : "memory")
+			"+m" (*(volatile int *)rw) : : "memory")
 
 #define __build_write_lock(rw, helper)	do { \
 						if (__builtin_constant_p(rw)) \

include/asm-i386/rwsem.h

···
 		"  jmp       1b\n"
 		LOCK_SECTION_END
 		"# ending down_read\n\t"
-		: "=m"(sem->count)
-		: "a"(sem), "m"(sem->count)
+		: "+m" (sem->count)
+		: "a" (sem)
 		: "memory", "cc");
 }
 
···
 		"  jnz	     1b\n\t"
 		"2:\n\t"
 		"# ending __down_read_trylock\n\t"
-		: "+m"(sem->count), "=&a"(result), "=&r"(tmp)
-		: "i"(RWSEM_ACTIVE_READ_BIAS)
+		: "+m" (sem->count), "=&a" (result), "=&r" (tmp)
+		: "i" (RWSEM_ACTIVE_READ_BIAS)
 		: "memory", "cc");
 	return result>=0 ? 1 : 0;
 }
···
 		"  jmp       1b\n"
 		LOCK_SECTION_END
 		"# ending down_write"
-		: "=m"(sem->count), "=d"(tmp)
-		: "a"(sem), "1"(tmp), "m"(sem->count)
+		: "+m" (sem->count), "=d" (tmp)
+		: "a" (sem), "1" (tmp)
 		: "memory", "cc");
 }
 
···
 		"  jmp       1b\n"
 		LOCK_SECTION_END
 		"# ending __up_read\n"
-		: "=m"(sem->count), "=d"(tmp)
-		: "a"(sem), "1"(tmp), "m"(sem->count)
+		: "+m" (sem->count), "=d" (tmp)
+		: "a" (sem), "1" (tmp)
 		: "memory", "cc");
 }
 
···
 		"  jmp       1b\n"
 		LOCK_SECTION_END
 		"# ending __up_write\n"
-		: "=m"(sem->count)
-		: "a"(sem), "i"(-RWSEM_ACTIVE_WRITE_BIAS), "m"(sem->count)
+		: "+m" (sem->count)
+		: "a" (sem), "i" (-RWSEM_ACTIVE_WRITE_BIAS)
 		: "memory", "cc", "edx");
 }
 
···
 		"  jmp       1b\n"
 		LOCK_SECTION_END
 		"# ending __downgrade_write\n"
-		: "=m"(sem->count)
-		: "a"(sem), "i"(-RWSEM_WAITING_BIAS), "m"(sem->count)
+		: "+m" (sem->count)
+		: "a" (sem), "i" (-RWSEM_WAITING_BIAS)
 		: "memory", "cc");
 }
 
···
 {
 	__asm__ __volatile__(
 LOCK_PREFIX	"addl %1,%0"
-		: "=m"(sem->count)
-		: "ir"(delta), "m"(sem->count));
+		: "+m" (sem->count)
+		: "ir" (delta));
 }
 
 /*
···
 	int tmp = delta;
 
 	__asm__ __volatile__(
-LOCK_PREFIX	"xadd %0,(%2)"
-		: "+r"(tmp), "=m"(sem->count)
-		: "r"(sem), "m"(sem->count)
-		: "memory");
+LOCK_PREFIX	"xadd %0,%1"
+		: "+r" (tmp), "+m" (sem->count)
+		: : "memory");
 
 	return tmp+delta;
 }

include/asm-i386/semaphore.h

···
 		"call __down_failed\n\t"
 		"jmp 1b\n"
 		LOCK_SECTION_END
-		:"=m" (sem->count)
+		:"+m" (sem->count)
 		:
 		:"memory","ax");
 }
···
 		"call __down_failed_interruptible\n\t"
 		"jmp 1b\n"
 		LOCK_SECTION_END
-		:"=a" (result), "=m" (sem->count)
+		:"=a" (result), "+m" (sem->count)
 		:
 		:"memory");
 	return result;
···
 		"call __down_failed_trylock\n\t"
 		"jmp 1b\n"
 		LOCK_SECTION_END
-		:"=a" (result), "=m" (sem->count)
+		:"=a" (result), "+m" (sem->count)
 		:
 		:"memory");
 	return result;
···
 		"jmp 1b\n"
 		LOCK_SECTION_END
 		".subsection 0\n"
-		:"=m" (sem->count)
+		:"+m" (sem->count)
 		:
 		:"memory","ax");
 }

include/asm-i386/spinlock.h

···
 	alternative_smp(
 		__raw_spin_lock_string,
 		__raw_spin_lock_string_up,
-		"=m" (lock->slock) : : "memory");
+		"+m" (lock->slock) : : "memory");
 }
 
 /*
···
 	alternative_smp(
 		__raw_spin_lock_string_flags,
 		__raw_spin_lock_string_up,
-		"=m" (lock->slock) : "r" (flags) : "memory");
+		"+m" (lock->slock) : "r" (flags) : "memory");
 }
 #endif
 
···
 	char oldval;
 	__asm__ __volatile__(
 		"xchgb %b0,%1"
-		:"=q" (oldval), "=m" (lock->slock)
+		:"=q" (oldval), "+m" (lock->slock)
 		:"0" (0) : "memory");
 	return oldval > 0;
 }
···
 
 #define __raw_spin_unlock_string \
 	"movb $1,%0" \
-		:"=m" (lock->slock) : : "memory"
+		:"+m" (lock->slock) : : "memory"
 
 
 static inline void __raw_spin_unlock(raw_spinlock_t *lock)
···
 
 #define __raw_spin_unlock_string \
 	"xchgb %b0, %1" \
-		:"=q" (oldval), "=m" (lock->slock) \
+		:"=q" (oldval), "+m" (lock->slock) \
 		:"0" (oldval) : "memory"
 
 static inline void __raw_spin_unlock(raw_spinlock_t *lock)
···
 
 static inline void __raw_read_unlock(raw_rwlock_t *rw)
 {
-	asm volatile(LOCK_PREFIX "incl %0" :"=m" (rw->lock) : : "memory");
+	asm volatile(LOCK_PREFIX "incl %0" :"+m" (rw->lock) : : "memory");
 }
 
 static inline void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	asm volatile(LOCK_PREFIX "addl $" RW_LOCK_BIAS_STR ", %0"
-				 : "=m" (rw->lock) : : "memory");
+				 : "+m" (rw->lock) : : "memory");
 }
 
 #endif /* __ASM_SPINLOCK_H */