diff options
author | Andi Kleen <ak@linux.intel.com> | 2011-10-29 01:01:54 +0000 |
---|---|---|
committer | Andi Kleen <ak@gcc.gnu.org> | 2011-10-29 01:01:54 +0000 |
commit | d33ef9a52bb4a561ec8fb14ad4f45c68dc57bc72 (patch) | |
tree | 2071cd9a9389dac958320f1bd9ae2bb9c110cc2b /gcc | |
parent | bf72b0094aa097ec23fdac68b33d2f86274bfd1d (diff) | |
download | gcc-d33ef9a52bb4a561ec8fb14ad4f45c68dc57bc72.zip gcc-d33ef9a52bb4a561ec8fb14ad4f45c68dc57bc72.tar.gz gcc-d33ef9a52bb4a561ec8fb14ad4f45c68dc57bc72.tar.bz2 |
Free large chunks in ggc v2
This implements the freeing back of large chunks in the ggc madvise path
Richard Guenther asked for. This way on systems with limited
address space malloc() and other allocators still have
a chance to get back at some of the memory ggc freed. The
fragmented pages are still just given back, but the address space
stays allocated.
I tried freeing only aligned 2MB areas to optimize for 2MB huge
pages, but the hit rate was quite low, so I switched to 1MB+
unaligned areas.
v2: Hardcode free unit size instead of param
gcc/:
2011-10-18 Andi Kleen <ak@linux.intel.com>
* ggc-page (release_pages): First free large continuous
chunks in the madvise path.
From-SVN: r180648
Diffstat (limited to 'gcc')
-rw-r--r-- | gcc/ChangeLog | 5 | ||||
-rw-r--r-- | gcc/ggc-page.c | 48 |
2 files changed, 53 insertions, 0 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 65df15b..3b24734 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,5 +1,10 @@ 2011-10-18 Andi Kleen <ak@linux.intel.com> + * ggc-page (release_pages): First free large continuous + chunks in the madvise path. + +2011-10-18 Andi Kleen <ak@linux.intel.com> + * ggc-page.c (alloc_pages): Always round up entry_size. 2011-10-19 Andi Kleen <ak@linux.intel.com> diff --git a/gcc/ggc-page.c b/gcc/ggc-page.c index 077bc8e..7bef4c0 100644 --- a/gcc/ggc-page.c +++ b/gcc/ggc-page.c @@ -973,6 +973,54 @@ release_pages (void) page_entry *p, *start_p; char *start; size_t len; + size_t mapped_len; + page_entry *next, *prev, *newprev; + size_t free_unit = (GGC_QUIRE_SIZE/2) * G.pagesize; + + /* First free larger continuous areas to the OS. + This allows other allocators to grab these areas if needed. + This is only done on larger chunks to avoid fragmentation. + This does not always work because the free_pages list is only + approximately sorted. */ + + p = G.free_pages; + prev = NULL; + while (p) + { + start = p->page; + start_p = p; + len = 0; + mapped_len = 0; + newprev = prev; + while (p && p->page == start + len) + { + len += p->bytes; + if (!p->discarded) + mapped_len += p->bytes; + newprev = p; + p = p->next; + } + if (len >= free_unit) + { + while (start_p != p) + { + next = start_p->next; + free (start_p); + start_p = next; + } + munmap (start, len); + if (prev) + prev->next = p; + else + G.free_pages = p; + G.bytes_mapped -= mapped_len; + continue; + } + prev = newprev; + } + + /* Now give back the fragmented pages to the OS, but keep the address + space to reuse it next time. */ for (p = G.free_pages; p; ) { |