diff options
| author | Keir Fraser <keir.fraser@citrix.com> | 2010-02-15 17:54:04 +0000 |
|---|---|---|
| committer | Keir Fraser <keir.fraser@citrix.com> | 2010-02-15 17:54:04 +0000 |
| commit | 82dc4c9fd8b278972528b70fdfdf3d316e413642 (patch) | |
| tree | 1fac1dbf15b7769e534f92af7c3f529a1b4196d3 | |
| parent | 1b79179e637294d940636b438d292e943904f628 (diff) | |
| download | xen-82dc4c9fd8b278972528b70fdfdf3d316e413642.tar.gz xen-82dc4c9fd8b278972528b70fdfdf3d316e413642.tar.bz2 xen-82dc4c9fd8b278972528b70fdfdf3d316e413642.zip | |
When tmem is enabled, reserve a fraction of memory
for allocations of 0<order<9 to avoid fragmentation
issues.
Signed-off by: Dan Magenheimer <dan.magenheimer@oracle.com>
| -rw-r--r-- | xen/common/page_alloc.c | 19 |
1 files changed, 18 insertions, 1 deletions
diff --git a/xen/common/page_alloc.c b/xen/common/page_alloc.c index 735c207e80..54920ff4db 100644 --- a/xen/common/page_alloc.c +++ b/xen/common/page_alloc.c @@ -224,6 +224,10 @@ static heap_by_zone_and_order_t *_heap[MAX_NUMNODES]; static unsigned long *avail[MAX_NUMNODES]; static long total_avail_pages; +/* TMEM: Reserve a fraction of memory for mid-size (0<order<9) allocations.*/ +static long midsize_alloc_zone_pages; +#define MIDSIZE_ALLOC_FRAC 128 + static DEFINE_SPINLOCK(heap_lock); static unsigned long init_node_heap(int node, unsigned long mfn, @@ -304,6 +308,14 @@ static struct page_info *alloc_heap_pages( spin_lock(&heap_lock); /* + * TMEM: When available memory is scarce, allow only mid-size allocations + * to avoid worst of fragmentation issues. + */ + if ( opt_tmem && ((order == 0) || (order >= 9)) && + (total_avail_pages <= midsize_alloc_zone_pages) ) + goto fail; + + /* * Start with requested node, but exhaust all node memory in requested * zone before failing, only calc new node value if we fail to find memory * in target node, this avoids needless computation on fast-path. @@ -336,6 +348,7 @@ static struct page_info *alloc_heap_pages( return pg; } + fail: /* No suitable memory blocks. Fail the request. */ spin_unlock(&heap_lock); return NULL; @@ -504,6 +517,10 @@ static void free_heap_pages( avail[node][zone] += 1 << order; total_avail_pages += 1 << order; + if ( opt_tmem ) + midsize_alloc_zone_pages = max( + midsize_alloc_zone_pages, total_avail_pages / MIDSIZE_ALLOC_FRAC); + /* Merge chunks as far as possible. */ while ( order < MAX_ORDER ) { @@ -842,7 +859,7 @@ static unsigned long avail_heap_pages( unsigned long total_free_pages(void) { - return total_avail_pages; + return total_avail_pages - midsize_alloc_zone_pages; } void __init end_boot_allocator(void) |