From 483cbbeddd5fe2c80fd4141ff0748fa06c4ff146 Mon Sep 17 00:00:00 2001 From: Alexei Naberezhnov Date: Tue, 27 Mar 2018 16:54:16 -0700 Subject: md/raid5: fix 'out of memory' during raid cache recovery This fixes the case when md array assembly fails because of raid cache recovery unable to allocate a stripe, despite attempts to replay stripes and increase cache size. This happens because stripes released by r5c_recovery_replay_stripes and raid5_set_cache_size don't become available for allocation immediately. Released stripes first are placed on conf->released_stripes list and require md thread to merge them on conf->inactive_list before they can be allocated. Patch allows final allocation attempt during cache recovery to wait for new stripes to become availabe for allocation. Cc: linux-raid@vger.kernel.org Cc: Shaohua Li Cc: linux-stable # 4.10+ Fixes: b4c625c67362 ("md/r5cache: r5cache recovery: part 1") Signed-off-by: Alexei Naberezhnov Signed-off-by: Song Liu --- drivers/md/raid5.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'drivers/md/raid5.c') diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 4990f0319f6c..cecea901ab8c 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -6369,6 +6369,7 @@ raid5_show_stripe_cache_size(struct mddev *mddev, char *page) int raid5_set_cache_size(struct mddev *mddev, int size) { + int result = 0; struct r5conf *conf = mddev->private; if (size <= 16 || size > 32768) @@ -6385,11 +6386,14 @@ raid5_set_cache_size(struct mddev *mddev, int size) mutex_lock(&conf->cache_size_mutex); while (size > conf->max_nr_stripes) - if (!grow_one_stripe(conf, GFP_KERNEL)) + if (!grow_one_stripe(conf, GFP_KERNEL)) { + conf->min_nr_stripes = conf->max_nr_stripes; + result = -ENOMEM; break; + } mutex_unlock(&conf->cache_size_mutex); - return 0; + return result; } EXPORT_SYMBOL(raid5_set_cache_size); -- cgit v1.2.3-59-g8ed1b