summaryrefslogtreecommitdiff
path: root/lib/libc/stdlib/jemalloc.c
AgeCommit message (Expand)Author
2008-06-23Split choose_arena() back out into inline and non-inline portions.ad
2008-06-05Check to see if MALLOC_OPTIONS is actually set in the environment beforead
2008-03-08Redo the CPU-based arena selection slightly differently. With the latestad
2007-12-04- use a non localized version of strerror_r.christos
2007-12-01Back out the per-cpu arena changes. With this, ld.so magically stopsad
2007-11-30Fix the other half of the 'store int as a thread specific pointer'/dsl
2007-11-30converting a pointer to unsigned is bad; use uintptr_t.christos
2007-11-29Make the allocation arenas per-CPU in the most optimistic case, butad
2007-11-19Enable posix_memalign().ad
2007-10-22Revert the CSE parts of the previous change. "run" is calculated withsimonb
2007-10-19- do some CSE.christos
2007-10-16use mremap for huge -> huge realloc.yamt
2007-10-15make lint happy.yamt
2007-10-15malloc_print_stats: print huge_allocated correctly. it's already in bytes.yamt
2007-10-15use MAP_ALIGNED.yamt
2007-10-09Add defs for hppa. From he@.ad
2007-10-07On sparc64, both __sparc__ and __sparc64__ is defined. Avoid redefinitionhe
2007-10-05Port to NetBSD. Note: posix_memalign() is here, but it's not exportedad
2007-10-05Pull in jemalloc from FreeBSD:ad