From 300b419486c1ca88e33938f182d5d5a8b90fb73f Mon Sep 17 00:00:00 2001 From: Rickard Green Date: Thu, 17 Jun 2010 10:23:50 +0200 Subject: Rewrite ethread library Large parts of the ethread library have been rewritten. The ethread library is an Erlang runtime system internal, portable thread library used by the runtime system itself. Most notable improvement is a reader optimized rwlock implementation which dramatically improve the performance of read-lock/read-unlock operations on multi processor systems by avoiding ping-ponging of the rwlock cache lines. The reader optimized rwlock implementation is used by miscellaneous rwlocks in the runtime system that are known to be read-locked frequently, and can be enabled on ETS tables by passing the `{read_concurrency, true}' option upon table creation. See the documentation of `ets:new/2' for more information. The ethread library can now also use the libatomic_ops library for atomic memory accesses. This makes it possible for the Erlang runtime system to utilize optimized atomic operations on more platforms than before. Use the `--with-libatomic_ops=PATH' configure command line argument when specifying where the libatomic_ops installation is located. The libatomic_ops library can be downloaded from: http://www.hpl.hp.com/research/linux/atomic_ops/ The changed API of the ethread library has also caused modifications in the Erlang runtime system. Preparations for the to come "delayed deallocation" feature has also been done since it depends on the ethread library. Note: When building for x86, the ethread library will now use instructions that first appeared on the pentium 4 processor. If you want the runtime system to be compatible with older processors (back to 486) you need to pass the `--enable-ethread-pre-pentium4-compatibility' configure command line argument when configuring the system. --- erts/include/internal/i386/atomic.h | 25 +++++++++++++++++++++++-- 1 file changed, 23 insertions(+), 2 deletions(-) (limited to 'erts/include/internal/i386/atomic.h') diff --git a/erts/include/internal/i386/atomic.h b/erts/include/internal/i386/atomic.h index 90b4c5f773..f28258059f 100644 --- a/erts/include/internal/i386/atomic.h +++ b/erts/include/internal/i386/atomic.h @@ -32,8 +32,11 @@ typedef struct { volatile long counter; } ethr_native_atomic_t; -#ifdef __x86_64__ +#if defined(__x86_64__) || !defined(ETHR_PRE_PENTIUM4_COMPAT) #define ETHR_MEMORY_BARRIER __asm__ __volatile__("mfence" : : : "memory") +#define ETHR_WRITE_MEMORY_BARRIER __asm__ __volatile__("sfence" : : : "memory") +#define ETHR_READ_MEMORY_BARRIER __asm__ __volatile__("lfence" : : : "memory") +#define ETHR_READ_DEPEND_MEMORY_BARRIER __asm__ __volatile__("" : : : "memory") #else #define ETHR_MEMORY_BARRIER \ do { \ @@ -42,7 +45,9 @@ do { \ } while (0) #endif -#ifdef ETHR_TRY_INLINE_FUNCS +#define ETHR_ATOMIC_HAVE_INC_DEC_INSTRUCTIONS 1 + +#if defined(ETHR_TRY_INLINE_FUNCS) || defined(ETHR_AUX_IMPL__) #ifdef __x86_64__ #define LONG_SUFFIX "q" @@ -158,6 +163,22 @@ ethr_native_atomic_xchg(ethr_native_atomic_t *var, long val) return tmp; } +/* + * Atomic ops with at least specified barriers. + */ + +#define ethr_native_atomic_read_acqb ethr_native_atomic_read +#define ethr_native_atomic_inc_return_acqb ethr_native_atomic_inc_return +#if defined(__x86_64__) || !defined(ETHR_PRE_PENTIUM4_COMPAT) +#define ethr_native_atomic_set_relb ethr_native_atomic_set +#else +#define ethr_native_atomic_set_relb ethr_native_atomic_xchg +#endif +#define ethr_native_atomic_dec_relb ethr_native_atomic_dec +#define ethr_native_atomic_dec_return_relb ethr_native_atomic_dec_return +#define ethr_native_atomic_cmpxchg_acqb ethr_native_atomic_cmpxchg +#define ethr_native_atomic_cmpxchg_relb ethr_native_atomic_cmpxchg + #undef LONG_SUFFIX #endif /* ETHR_TRY_INLINE_FUNCS */ -- cgit v1.2.3