ftrace can fail to allocate per-CPU ring buffer on systems with a large number of CPUs coupled while large amounts of cache happening in the page cache. Currently the ring buffer allocation doesn't retry in the VM implementation even if direct-reclaim made some progress but still wasn't able to find a free page. On retrying I see that the allocations almost always succeed. The retry doesn't happen because __GFP_NORETRY is used in the tracer to prevent the case where we might OOM, however if we drop __GFP_NORETRY, we risk destabilizing the system if OOM killer is triggered. To prevent this situation, use the __GFP_DONTOOM flag introduced in earlier patches while droppping __GFP_NORETRY. With this the following succeed without destabilizing a system with 8 CPU cores and 4GB of memory: echo 100000 > /sys/kernel/debug/tracing/buffer_size_kb On an 8-core system, that would allocate ~800MB. Cc: Alexander Duyck <alexander.h.duyck@xxxxxxxxx> Cc: Mel Gorman <mgorman@xxxxxxx> Cc: Hao Lee <haolee.swjtu@xxxxxxxxx> Cc: Vladimir Davydov <vdavydov.dev@xxxxxxxxx> Cc: Johannes Weiner <hannes@xxxxxxxxxxx> Cc: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx> Cc: Steven Rostedt <rostedt@xxxxxxxxxxx> Cc: linux-mm@xxxxxxxxx Cc: linux-kernel@xxxxxxxxxxxxxxx Signed-off-by: Joel Fernandes <joelaf@xxxxxxxxxx> --- kernel/trace/ring_buffer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 4ae268e687fe..b1cdcac6ca89 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -1141,7 +1141,7 @@ static int __rb_allocate_pages(long nr_pages, struct list_head *pages, int cpu) * not destabilized. */ bpage = kzalloc_node(ALIGN(sizeof(*bpage), cache_line_size()), - GFP_KERNEL | __GFP_NORETRY, + GFP_KERNEL | __GFP_DONTOOM, cpu_to_node(cpu)); if (!bpage) goto free_pages; @@ -1149,7 +1149,7 @@ static int __rb_allocate_pages(long nr_pages, struct list_head *pages, int cpu) list_add(&bpage->list, pages); page = alloc_pages_node(cpu_to_node(cpu), - GFP_KERNEL | __GFP_NORETRY, 0); + GFP_KERNEL | __GFP_DONTOOM, 0); if (!page) goto free_pages; bpage->page = page_address(page); -- 2.13.2.725.g09c95d1e9-goog -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
