aboutsummaryrefslogtreecommitdiffstats
path: root/erts/include/internal/ppc32/rwlock.h
diff options
context:
space:
mode:
authorRickard Green <[email protected]>2011-01-02 10:03:54 +0100
committerRickard Green <[email protected]>2011-06-14 11:40:19 +0200
commit7f19af0423934f85c74ccb75546e5e3a6b6d10e8 (patch)
tree612d1010f37517f813a94d8a5f38cfd0126ce3f8 /erts/include/internal/ppc32/rwlock.h
parent4a5a75811e2cd590b5c94f71864a5245fd511ccf (diff)
downloadotp-7f19af0423934f85c74ccb75546e5e3a6b6d10e8.tar.gz
otp-7f19af0423934f85c74ccb75546e5e3a6b6d10e8.tar.bz2
otp-7f19af0423934f85c74ccb75546e5e3a6b6d10e8.zip
Improve ethread atomics
The ethread atomics API now also provide double word size atomics. Double word size atomics are implemented using native atomic instructions on x86 (when the cmpxchg8b instruction is available) and on x86_64 (when the cmpxchg16b instruction is available). On other hardware where 32-bit atomics or word size atomics are available, an optimized fallback is used; otherwise, a spinlock, or a mutex based fallback is used. The ethread library now performs runtime tests for presence of hardware features, such as for example SSE2 instructions, instead of requiring this to be determined at compile time. There are now functions implementing each atomic operation with the following implied memory barrier semantics: none, read, write, acquire, release, and full. Some of the operation-barrier combinations aren't especially useful. But instead of filtering useful ones out, and potentially miss a useful one, we implement them all. A much smaller set of functionality for native atomics are required to be implemented than before. More or less only cmpxchg and a membar macro are required to be implemented for each atomic size. Other functions will automatically be constructed from these. It is, of course, often wise to implement more that this if possible from a performance perspective.
Diffstat (limited to 'erts/include/internal/ppc32/rwlock.h')
-rw-r--r--erts/include/internal/ppc32/rwlock.h15
1 files changed, 9 insertions, 6 deletions
diff --git a/erts/include/internal/ppc32/rwlock.h b/erts/include/internal/ppc32/rwlock.h
index 19ec26ab68..311f000b69 100644
--- a/erts/include/internal/ppc32/rwlock.h
+++ b/erts/include/internal/ppc32/rwlock.h
@@ -1,7 +1,7 @@
/*
* %CopyrightBegin%
*
- * Copyright Ericsson AB 2005-2010. All Rights Reserved.
+ * Copyright Ericsson AB 2005-2011. All Rights Reserved.
*
* The contents of this file are subject to the Erlang Public License,
* Version 1.1, (the "License"); you may not use this file except in
@@ -23,12 +23,14 @@
*
* Based on the examples in Appendix E of Motorola's
* "Programming Environments Manual For 32-Bit Implementations
- * of the PowerPC Architecture". Uses eieio instead of sync
- * in the unlock sequence, as suggested in the manual.
+ * of the PowerPC Architecture".
*/
#ifndef ETHREAD_PPC_RWLOCK_H
#define ETHREAD_PPC_RWLOCK_H
+#define ETHR_HAVE_NATIVE_RWSPINLOCKS 1
+#define ETHR_NATIVE_RWSPINLOCK_IMPL "ethread"
+
/* Unlocked if zero, read-locked if negative, write-locked if +1. */
typedef struct {
volatile int lock;
@@ -47,9 +49,10 @@ ethr_native_read_unlock(ethr_native_rwlock_t *lock)
{
int tmp;
- /* this is eieio + ethr_native_atomic_inc() - isync */
+ ETHR_MEMBAR(ETHR_LoadStore|ETHR_StoreStore);
+
+ /* this is ethr_native_atomic_inc() - isync */
__asm__ __volatile__(
- "eieio\n\t"
"1:\t"
"lwarx %0,0,%1\n\t"
"addic %0,%0,1\n\t"
@@ -105,7 +108,7 @@ ethr_native_read_lock(ethr_native_rwlock_t *lock)
static ETHR_INLINE void
ethr_native_write_unlock(ethr_native_rwlock_t *lock)
{
- __asm__ __volatile__("eieio" : : : "memory");
+ ETHR_MEMBAR(ETHR_LoadStore|ETHR_StoreStore);
lock->lock = 0;
}