From b0140891a8cea36469f58d23859e599b1122bd37 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 10 May 2011 17:49:01 +1000 Subject: md: Fix race when creating a new md device. There is a race when creating an md device by opening /dev/mdXX. If two processes do this at much the same time they will follow the call path __blkdev_get -> get_gendisk -> kobj_lookup The first will call -> md_probe -> md_alloc -> add_disk -> blk_register_region and the race happens when the second gets to kobj_lookup after add_disk has called blk_register_region but before it returns to md_alloc. In the case the second will not call md_probe (as the probe is already done) but will get a handle on the gendisk, return to __blkdev_get which will then call md_open (via the ->open) pointer. As mddev->gendisk hasn't been set yet, md_open will think something is wrong an return with ERESTARTSYS. This can loop endlessly while the first thread makes no progress through add_disk. Nothing is blocking it, but due to scheduler behaviour it doesn't get a turn. So this is essentially a live-lock. We fix this by simply moving the assignment to mddev->gendisk before the call the add_disk() so md_open doesn't get confused. Also move blk_queue_flush earlier because add_disk should be as late as possible. To make sure that md_open doesn't complete until md_alloc has done all that is needed, we take mddev->open_mutex during the last part of md_alloc. md_open will wait for this. This can cause a lock-up on boot so Cc:ing for stable. For 2.6.36 and earlier a different patch will be needed as the 'blk_queue_flush' call isn't there. Signed-off-by: NeilBrown Reported-by: Thomas Jarosch Tested-by: Thomas Jarosch Cc: stable@kernel.org --- drivers/md/md.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'drivers/md/md.c') diff --git a/drivers/md/md.c b/drivers/md/md.c index 7d6f7f18a920..4a4c0f80bdeb 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -4347,13 +4347,19 @@ static int md_alloc(dev_t dev, char *name) disk->fops = &md_fops; disk->private_data = mddev; disk->queue = mddev->queue; + blk_queue_flush(mddev->queue, REQ_FLUSH | REQ_FUA); /* Allow extended partitions. This makes the * 'mdp' device redundant, but we can't really * remove it now. */ disk->flags |= GENHD_FL_EXT_DEVT; - add_disk(disk); mddev->gendisk = disk; + /* As soon as we call add_disk(), another thread could get + * through to md_open, so make sure it doesn't get too far + */ + mutex_lock(&mddev->open_mutex); + add_disk(disk); + error = kobject_init_and_add(&mddev->kobj, &md_ktype, &disk_to_dev(disk)->kobj, "%s", "md"); if (error) { @@ -4367,8 +4373,7 @@ static int md_alloc(dev_t dev, char *name) if (mddev->kobj.sd && sysfs_create_group(&mddev->kobj, &md_bitmap_group)) printk(KERN_DEBUG "pointless warning\n"); - - blk_queue_flush(mddev->queue, REQ_FLUSH | REQ_FUA); + mutex_unlock(&mddev->open_mutex); abort: mutex_unlock(&disks_mutex); if (!error && mddev->kobj.sd) { -- cgit v1.2.1