[PATCH v2] dm-zoned: Avoid triggering reclaim from inside dmz_map()

This patch avoids that lockdep reports the following:

======================================================
WARNING: possible circular locking dependency detected
4.18.0-rc1 #62 Not tainted
------------------------------------------------------
kswapd0/84 is trying to acquire lock:
00000000c313516d (&xfs_nondir_ilock_class){++++}, at: xfs_free_eofblocks+0xa2/0x1e0

but task is already holding lock:
00000000591c83ae (fs_reclaim){+.+.}, at: __fs_reclaim_acquire+0x5/0x30

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 (fs_reclaim){+.+.}:
  kmem_cache_alloc+0x2c/0x2b0
  radix_tree_node_alloc.constprop.19+0x3d/0xc0
  __radix_tree_create+0x161/0x1c0
  __radix_tree_insert+0x45/0x210
  dmz_map+0x245/0x2d0 [dm_zoned]
  __map_bio+0x40/0x260
  __split_and_process_non_flush+0x116/0x220
  __split_and_process_bio+0x81/0x180
  __dm_make_request.isra.32+0x5a/0x100
  generic_make_request+0x36e/0x690
  submit_bio+0x6c/0x140
  mpage_readpages+0x19e/0x1f0
  read_pages+0x6d/0x1b0
  __do_page_cache_readahead+0x21b/0x2d0
  force_page_cache_readahead+0xc4/0x100
  generic_file_read_iter+0x7c6/0xd20
  __vfs_read+0x102/0x180
  vfs_read+0x9b/0x140
  ksys_read+0x55/0xc0
  do_syscall_64+0x5a/0x1f0
  entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #1 (&dmz->chunk_lock){+.+.}:
  dmz_map+0x133/0x2d0 [dm_zoned]
  __map_bio+0x40/0x260
  __split_and_process_non_flush+0x116/0x220
  __split_and_process_bio+0x81/0x180
  __dm_make_request.isra.32+0x5a/0x100
  generic_make_request+0x36e/0x690
  submit_bio+0x6c/0x140
  _xfs_buf_ioapply+0x31c/0x590
  xfs_buf_submit_wait+0x73/0x520
  xfs_buf_read_map+0x134/0x2f0
  xfs_trans_read_buf_map+0xc3/0x580
  xfs_read_agf+0xa5/0x1e0
  xfs_alloc_read_agf+0x59/0x2b0
  xfs_alloc_pagf_init+0x27/0x60
  xfs_bmap_longest_free_extent+0x43/0xb0
  xfs_bmap_btalloc_nullfb+0x7f/0xf0
  xfs_bmap_btalloc+0x428/0x7c0
  xfs_bmapi_write+0x598/0xcc0
  xfs_iomap_write_allocate+0x15a/0x330
  xfs_map_blocks+0x1cf/0x3f0
  xfs_do_writepage+0x15f/0x7b0
  write_cache_pages+0x1ca/0x540
  xfs_vm_writepages+0x65/0xa0
  do_writepages+0x48/0xf0
  __writeback_single_inode+0x58/0x730
  writeback_sb_inodes+0x249/0x5c0
  wb_writeback+0x11e/0x550
  wb_workfn+0xa3/0x670
  process_one_work+0x228/0x670
  worker_thread+0x3c/0x390
  kthread+0x11c/0x140
  ret_from_fork+0x3a/0x50

-> #0 (&xfs_nondir_ilock_class){++++}:
  down_read_nested+0x43/0x70
  xfs_free_eofblocks+0xa2/0x1e0
  xfs_fs_destroy_inode+0xac/0x270
  dispose_list+0x51/0x80
  prune_icache_sb+0x52/0x70
  super_cache_scan+0x127/0x1a0
  shrink_slab.part.47+0x1bd/0x590
  shrink_node+0x3b5/0x470
  balance_pgdat+0x158/0x3b0
  kswapd+0x1ba/0x600
  kthread+0x11c/0x140
  ret_from_fork+0x3a/0x50

other info that might help us debug this:

Chain exists of:
  &xfs_nondir_ilock_class --> &dmz->chunk_lock --> fs_reclaim

Possible unsafe locking scenario:

     CPU0                    CPU1
     ----                    ----
lock(fs_reclaim);
                             lock(&dmz->chunk_lock);
                             lock(fs_reclaim);
lock(&xfs_nondir_ilock_class);

*** DEADLOCK ***

3 locks held by kswapd0/84:
 #0: 00000000591c83ae (fs_reclaim){+.+.}, at: __fs_reclaim_acquire+0x5/0x30
 #1: 000000000f8208f5 (shrinker_rwsem){++++}, at: shrink_slab.part.47+0x3f/0x590
 #2: 00000000cacefa54 (&type->s_umount_key#43){.+.+}, at: trylock_super+0x16/0x50

stack backtrace:
CPU: 7 PID: 84 Comm: kswapd0 Not tainted 4.18.0-rc1 #62
Hardware name: Supermicro Super Server/X10SRL-F, BIOS 2.0 12/17/2015
Call Trace:
 dump_stack+0x85/0xcb
 print_circular_bug.isra.36+0x1ce/0x1db
 __lock_acquire+0x124e/0x1310
 lock_acquire+0x9f/0x1f0
 down_read_nested+0x43/0x70
 xfs_free_eofblocks+0xa2/0x1e0
 xfs_fs_destroy_inode+0xac/0x270
 dispose_list+0x51/0x80
 prune_icache_sb+0x52/0x70
 super_cache_scan+0x127/0x1a0
 shrink_slab.part.47+0x1bd/0x590
 shrink_node+0x3b5/0x470
 balance_pgdat+0x158/0x3b0
 kswapd+0x1ba/0x600
 kthread+0x11c/0x140
 ret_from_fork+0x3a/0x50

Reported-by: Masato Suzuki <redacted>
Fixes: 4218a9554653 ("dm zoned: use GFP_NOIO in I/O path")
Signed-off-by: Bart Van Assche <redacted>
Cc: Damien Le Moal <redacted>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: <redacted>
---

Changes compared to v1: added "Cc: stable"

 drivers/md/dm-zoned-target.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/md/dm-zoned-target.c b/drivers/md/dm-zoned-target.c
index 3c0e45f4dcf5..a44183ff4be0 100644
--- a/drivers/md/dm-zoned-target.c
+++ b/drivers/md/dm-zoned-target.c

@@ -787,7 +787,7 @@ static int dmz_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
 	/* Chunk BIO work */
 	mutex_init(&dmz->chunk_lock);
-	INIT_RADIX_TREE(&dmz->chunk_rxtree, GFP_KERNEL);
+	INIT_RADIX_TREE(&dmz->chunk_rxtree, GFP_NOIO);
 	dmz->chunk_wq = alloc_workqueue("dmz_cwq_%s", WQ_MEM_RECLAIM | WQ_UNBOUND,
 					0, dev->name);
 	if (!dmz->chunk_wq) {

-- 
2.17.1

OK - but I think a proper fix would be to preallocate the chunks and the 
radix tree when the device is created.

If the system is highly stressed, it may be possible that the GFP_NOIO 
allocation would wait for some data being written back - and the write 
back may be directed back to the dm-zoned device, waiting for the GFP_NOIO 
allocation to succeed.

Mikulas



On Fri, 22 Jun 2018, Bart Van Assche wrote:

On Wed, 2018-06-27 at 11:14 -0400, Mikulas Patocka wrote:

OK - but I think a proper fix would be to preallocate the chunks and the 
radix tree when the device is created.

If the system is highly stressed, it may be possible that the GFP_NOIO 
allocation would wait for some data being written back - and the write 
back may be directed back to the dm-zoned device, waiting for the GFP_NOIO 
allocation to succeed.

Hello Mikulas,

Thank you for the feedback. We will look further into this.

Bart.