* [PATCH 1/2] ARM: locks: use early clobber in arch_spin_trylock
@ 2013-07-30 17:57 Will Deacon
2013-07-30 17:57 ` [PATCH 2/2] ARM: rwlocks: retry trylock operation if strex fails on free lock Will Deacon
0 siblings, 1 reply; 2+ messages in thread
From: Will Deacon @ 2013-07-30 17:57 UTC (permalink / raw)
To: linux-arm-kernel
The res variable is written before we've finished with the input
operands (namely the lock address), so ensure that we mark it as `early
clobber' to avoid unintended register sharing.
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/arm/include/asm/spinlock.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/arm/include/asm/spinlock.h b/arch/arm/include/asm/spinlock.h
index f8b8965..7ed43f68 100644
--- a/arch/arm/include/asm/spinlock.h
+++ b/arch/arm/include/asm/spinlock.h
@@ -107,7 +107,7 @@ static inline int arch_spin_trylock(arch_spinlock_t *lock)
" subs %1, %0, %0, ror #16\n"
" addeq %0, %0, %4\n"
" strexeq %2, %0, [%3]"
- : "=&r" (slock), "=&r" (contended), "=r" (res)
+ : "=&r" (slock), "=&r" (contended), "=&r" (res)
: "r" (&lock->slock), "I" (1 << TICKET_SHIFT)
: "cc");
} while (res);
--
1.8.2.2
^ permalink raw reply related [flat|nested] 2+ messages in thread
* [PATCH 2/2] ARM: rwlocks: retry trylock operation if strex fails on free lock
2013-07-30 17:57 [PATCH 1/2] ARM: locks: use early clobber in arch_spin_trylock Will Deacon
@ 2013-07-30 17:57 ` Will Deacon
0 siblings, 0 replies; 2+ messages in thread
From: Will Deacon @ 2013-07-30 17:57 UTC (permalink / raw)
To: linux-arm-kernel
Commit 15e7e5c1ebf5 ("ARM: 7749/1: spinlock: retry trylock operation if
strex fails on free lock") modifying our arch_spin_trylock to retry the
acquisition if the lock appeared uncontended, but the strex failed.
This patch does the same for rwlocks, which were missed by the original
patch.
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/arm/include/asm/spinlock.h | 49 +++++++++++++++++++++++++----------------
1 file changed, 30 insertions(+), 19 deletions(-)
diff --git a/arch/arm/include/asm/spinlock.h b/arch/arm/include/asm/spinlock.h
index 7ed43f68..b07c09e 100644
--- a/arch/arm/include/asm/spinlock.h
+++ b/arch/arm/include/asm/spinlock.h
@@ -168,17 +168,20 @@ static inline void arch_write_lock(arch_rwlock_t *rw)
static inline int arch_write_trylock(arch_rwlock_t *rw)
{
- unsigned long tmp;
+ unsigned long contended, res;
- __asm__ __volatile__(
-" ldrex %0, [%1]\n"
-" teq %0, #0\n"
-" strexeq %0, %2, [%1]"
- : "=&r" (tmp)
- : "r" (&rw->lock), "r" (0x80000000)
- : "cc");
+ do {
+ __asm__ __volatile__(
+ " ldrex %0, [%2]\n"
+ " mov %1, #0\n"
+ " teq %0, #0\n"
+ " strexeq %1, %3, [%2]"
+ : "=&r" (contended), "=&r" (res)
+ : "r" (&rw->lock), "r" (0x80000000)
+ : "cc");
+ } while (res);
- if (tmp == 0) {
+ if (!contended) {
smp_mb();
return 1;
} else {
@@ -254,18 +257,26 @@ static inline void arch_read_unlock(arch_rwlock_t *rw)
static inline int arch_read_trylock(arch_rwlock_t *rw)
{
- unsigned long tmp, tmp2 = 1;
+ unsigned long contended, res;
- __asm__ __volatile__(
-" ldrex %0, [%2]\n"
-" adds %0, %0, #1\n"
-" strexpl %1, %0, [%2]\n"
- : "=&r" (tmp), "+r" (tmp2)
- : "r" (&rw->lock)
- : "cc");
+ do {
+ __asm__ __volatile__(
+ " ldrex %0, [%2]\n"
+ " mov %1, #0\n"
+ " adds %0, %0, #1\n"
+ " strexpl %1, %0, [%2]"
+ : "=&r" (contended), "=&r" (res)
+ : "r" (&rw->lock)
+ : "cc");
+ } while (res);
- smp_mb();
- return tmp2 == 0;
+ /* If the lock is negative, then it is already held for write. */
+ if (contended < 0x80000000) {
+ smp_mb();
+ return 1;
+ } else {
+ return 0;
+ }
}
/* read_can_lock - would read_trylock() succeed? */
--
1.8.2.2
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2013-07-30 17:57 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-07-30 17:57 [PATCH 1/2] ARM: locks: use early clobber in arch_spin_trylock Will Deacon
2013-07-30 17:57 ` [PATCH 2/2] ARM: rwlocks: retry trylock operation if strex fails on free lock Will Deacon
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).