kernel/linux.git/fs/cifs/dfs_cache.c, branch linux-6.0.y

cifs: avoid use of global locks for high contention data

2022-08-01T06:34:45+00:00

During analysis of multichannel perf, it was seen that the global locks cifs_tcp_ses_lock and GlobalMid_Lock, which were shared between various data structures were causing a lot of contention points. With this change, we're breaking down the use of these locks by introducing new locks at more granular levels. i.e. server->srv_lock, ses->ses_lock and tcon->tc_lock to protect the unprotected fields of server, session and tcon structs; and server->mid_lock to protect mid related lists and entries at server level. Signed-off-by: Shyam Prasad N Signed-off-by: Steve French

cifs: fix uninitialized pointer in error case in dfs_cache_get_tgt_share

2022-06-04T18:33:42+00:00

Set default value of ppath to null. Reported-by: kernel test robot Reviewed-by: Paulo Alcantara (SUSE) Signed-off-by: Steve French

cifs: skip trailing separators of prefix paths

2022-06-03T19:14:58+00:00

During DFS failover, prefix paths may change, so make sure to not leave trailing separators when parsing thew in dfs_cache_get_tgt_share(). The separators of prefix paths are already handled by build_path_from_dentry_optional_prefix(). Consider the following DFS link: //dom/dfs/link: [\srv1\share\dir1, \srv2\share\dir1] Before commit: mount.cifs //dom/dfs/link tree connect to \\srv1\share; prefix_path=dir1 disconnect srv1; failover to srv2 tree connect to \\srv2\share; prefix_path=dir1\ mv foo bar ... SMB2 430 Create Request File: dir1\\foo;GetInfo Request FILE_INFO/SMB2_FILE_ALL_INFO;Close Request SMB2 582 Create Response File: dir1\\foo;GetInfo Response;Close Response SMB2 430 Create Request File: dir1\\bar;GetInfo Request FILE_INFO/SMB2_FILE_ALL_INFO;Close Request SMB2 286 Create Response, Error: STATUS_OBJECT_NAME_NOT_FOUND;GetInfo Response, Error: STATUS_OBJECT_NAME_NOT_FOUND;Close Response, Error: STATUS_OBJECT_NAME_NOT_FOUND SMB2 462 Create Request File: dir1\\foo;SetInfo Request FILE_INFO/SMB2_FILE_RENAME_INFO NewName:dir1\\bar;Close Request SMB2 478 Create Response File: dir1\\foo;SetInfo Response, Error: STATUS_OBJECT_NAME_INVALID;Close Response After commit: mount.cifs //dom/dfs/link tree connect to \\srv1\share; prefix_path=dir1 disconnect srv1; failover to srv2 tree connect to \\srv2\share; prefix_path=dir1 mv foo bar ... SMB2 430 Create Request File: dir1\foo;GetInfo Request FILE_INFO/SMB2_FILE_ALL_INFO;Close Request SMB2 582 Create Response File: dir1\foo;GetInfo Response;Close Response SMB2 430 Create Request File: dir1\bar;GetInfo Request FILE_INFO/SMB2_FILE_ALL_INFO;Close Request SMB2 286 Create Response, Error: STATUS_OBJECT_NAME_NOT_FOUND;GetInfo Response, Error: STATUS_OBJECT_NAME_NOT_FOUND;Close Response, Error: STATUS_OBJECT_NAME_NOT_FOUND SMB2 462 Create Request File: dir1\foo;SetInfo Request FILE_INFO/SMB2_FILE_RENAME_INFO NewName:dir1\bar;Close Request SMB2 478 Create Response File: dir1\foo;SetInfo Response;Close Response Signed-off-by: Paulo Alcantara (SUSE) Signed-off-by: Steve French

cifs: fix potential deadlock in direct reclaim

2022-06-01T05:03:18+00:00

The srv_mutex is used during writeback so cifs should ensure that allocations done when that mutex is held are done with GFP_NOFS, to avoid having direct reclaim ending up waiting for the same mutex and causing a deadlock. This is detected by lockdep with the splat below: ====================================================== WARNING: possible circular locking dependency detected 5.18.0 #70 Not tainted ------------------------------------------------------ kswapd0/49 is trying to acquire lock: ffff8880195782e0 (&tcp_ses->srv_mutex){+.+.}-{3:3}, at: compound_send_recv but task is already holding lock: ffffffffa98e66c0 (fs_reclaim){+.+.}-{0:0}, at: balance_pgdat which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (fs_reclaim){+.+.}-{0:0}: fs_reclaim_acquire kmem_cache_alloc_trace __request_module crypto_alg_mod_lookup crypto_alloc_tfm_node crypto_alloc_shash cifs_alloc_hash smb311_crypto_shash_allocate smb311_update_preauth_hash compound_send_recv cifs_send_recv SMB2_negotiate smb2_negotiate cifs_negotiate_protocol cifs_get_smb_ses cifs_mount cifs_smb3_do_mount smb3_get_tree vfs_get_tree path_mount __x64_sys_mount do_syscall_64 entry_SYSCALL_64_after_hwframe -> #0 (&tcp_ses->srv_mutex){+.+.}-{3:3}: __lock_acquire lock_acquire __mutex_lock mutex_lock_nested compound_send_recv cifs_send_recv SMB2_write smb2_sync_write cifs_write cifs_writepage_locked cifs_writepage shrink_page_list shrink_lruvec shrink_node balance_pgdat kswapd kthread ret_from_fork other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(fs_reclaim); lock(&tcp_ses->srv_mutex); lock(fs_reclaim); lock(&tcp_ses->srv_mutex); *** DEADLOCK *** 1 lock held by kswapd0/49: #0: ffffffffa98e66c0 (fs_reclaim){+.+.}-{0:0}, at: balance_pgdat stack backtrace: CPU: 2 PID: 49 Comm: kswapd0 Not tainted 5.18.0 #70 Call Trace: dump_stack_lvl dump_stack print_circular_bug.cold check_noncircular __lock_acquire lock_acquire __mutex_lock mutex_lock_nested compound_send_recv cifs_send_recv SMB2_write smb2_sync_write cifs_write cifs_writepage_locked cifs_writepage shrink_page_list shrink_lruvec shrink_node balance_pgdat kswapd kthread ret_from_fork Fix this by using the memalloc_nofs_save/restore APIs around the places where the srv_mutex is held. Do this in a wrapper function for the lock/unlock of the srv_mutex, and rename the srv_mutex to avoid missing call sites in the conversion. Note that there is another lockdep warning involving internal crypto locks, which was masked by this problem and is visible after this fix, see the discussion in this thread: https://lore.kernel.org/all/20220523123755.GA13668@axis.com/ Link: https://lore.kernel.org/r/CANT5p=rqcYfYMVHirqvdnnca4Mo+JQSw5Qu12v=kPfpk5yhhmg@mail.gmail.com/ Reported-by: Shyam Prasad N Suggested-by: Lars Persson Reviewed-by: Ronnie Sahlberg Reviewed-by: Enzo Matsumiya Signed-off-by: Vincent Whitchurch Signed-off-by: Steve French

cifs: return ENOENT for DFS lookup_cache_entry()

2022-05-20T22:44:34+00:00

EEXIST didn't make sense to use when dfs_cache_find() couldn't find a cache entry nor retrieve a referral target. It also doesn't make sense cifs_dfs_query_info_nonascii_quirk() to emulate ENOENT anymore. Signed-off-by: Enzo Matsumiya Reviewed-by: Paulo Alcantara (SUSE) Signed-off-by: Steve French

cifs: fix NULL ptr dereference in refresh_mounts()

2022-04-21T03:54:17+00:00

Either mount(2) or automount might not have server->origin_fullpath set yet while refresh_cache_worker() is attempting to refresh DFS referrals. Add missing NULL check and locking around it. This fixes bellow crash: [ 1070.276835] general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN NOPTI [ 1070.277676] KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] [ 1070.278219] CPU: 1 PID: 8506 Comm: kworker/u8:1 Not tainted 5.18.0-rc3 #10 [ 1070.278701] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.15.0-0-g2dd4b9b-rebuilt.opensuse.org 04/01/2014 [ 1070.279495] Workqueue: cifs-dfscache refresh_cache_worker [cifs] [ 1070.280044] RIP: 0010:strcasecmp+0x34/0x150 [ 1070.280359] Code: 00 00 00 fc ff df 41 54 55 48 89 fd 53 48 83 ec 10 eb 03 4c 89 fe 48 89 ef 48 83 c5 01 48 89 f8 48 89 fa 48 c1 e8 03 83 e2 07 <42> 0f b6 04 28 38 d0 7f 08 84 c0 0f 85 bc 00 00 00 0f b6 45 ff 44 [ 1070.281729] RSP: 0018:ffffc90008367958 EFLAGS: 00010246 [ 1070.282114] RAX: 0000000000000000 RBX: dffffc0000000000 RCX: 0000000000000000 [ 1070.282691] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000 [ 1070.283273] RBP: 0000000000000001 R08: 0000000000000000 R09: ffffffff873eda27 [ 1070.283857] R10: ffffc900083679a0 R11: 0000000000000001 R12: ffff88812624c000 [ 1070.284436] R13: dffffc0000000000 R14: ffff88810e6e9a88 R15: ffff888119bb9000 [ 1070.284990] FS: 0000000000000000(0000) GS:ffff888151200000(0000) knlGS:0000000000000000 [ 1070.285625] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1070.286100] CR2: 0000561a4d922418 CR3: 000000010aecc000 CR4: 0000000000350ee0 [ 1070.286683] Call Trace: [ 1070.286890] [ 1070.287070] refresh_cache_worker+0x895/0xd20 [cifs] [ 1070.287475] ? __refresh_tcon.isra.0+0xfb0/0xfb0 [cifs] [ 1070.287905] ? __lock_acquire+0xcd1/0x6960 [ 1070.288247] ? is_dynamic_key+0x1a0/0x1a0 [ 1070.288591] ? lockdep_hardirqs_on_prepare+0x410/0x410 [ 1070.289012] ? lock_downgrade+0x6f0/0x6f0 [ 1070.289318] process_one_work+0x7bd/0x12d0 [ 1070.289637] ? worker_thread+0x160/0xec0 [ 1070.289970] ? pwq_dec_nr_in_flight+0x230/0x230 [ 1070.290318] ? _raw_spin_lock_irq+0x5e/0x90 [ 1070.290619] worker_thread+0x5ac/0xec0 [ 1070.290891] ? process_one_work+0x12d0/0x12d0 [ 1070.291199] kthread+0x2a5/0x350 [ 1070.291430] ? kthread_complete_and_exit+0x20/0x20 [ 1070.291770] ret_from_fork+0x22/0x30 [ 1070.292050] [ 1070.292223] Modules linked in: bpfilter cifs cifs_arc4 cifs_md4 [ 1070.292765] ---[ end trace 0000000000000000 ]--- [ 1070.293108] RIP: 0010:strcasecmp+0x34/0x150 [ 1070.293471] Code: 00 00 00 fc ff df 41 54 55 48 89 fd 53 48 83 ec 10 eb 03 4c 89 fe 48 89 ef 48 83 c5 01 48 89 f8 48 89 fa 48 c1 e8 03 83 e2 07 <42> 0f b6 04 28 38 d0 7f 08 84 c0 0f 85 bc 00 00 00 0f b6 45 ff 44 [ 1070.297718] RSP: 0018:ffffc90008367958 EFLAGS: 00010246 [ 1070.298622] RAX: 0000000000000000 RBX: dffffc0000000000 RCX: 0000000000000000 [ 1070.299428] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000 [ 1070.300296] RBP: 0000000000000001 R08: 0000000000000000 R09: ffffffff873eda27 [ 1070.301204] R10: ffffc900083679a0 R11: 0000000000000001 R12: ffff88812624c000 [ 1070.301932] R13: dffffc0000000000 R14: ffff88810e6e9a88 R15: ffff888119bb9000 [ 1070.302645] FS: 0000000000000000(0000) GS:ffff888151200000(0000) knlGS:0000000000000000 [ 1070.303462] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1070.304131] CR2: 0000561a4d922418 CR3: 000000010aecc000 CR4: 0000000000350ee0 [ 1070.305004] Kernel panic - not syncing: Fatal exception [ 1070.305711] Kernel Offset: disabled [ 1070.305971] ---[ end Kernel panic - not syncing: Fatal exception ]--- Signed-off-by: Paulo Alcantara (SUSE) Cc: stable@vger.kernel.org Reviewed-by: Ronnie Sahlberg Signed-off-by: Steve French

cifs: use a different reconnect helper for non-cifsd threads

2022-03-19T04:12:03+00:00

The cifs_demultiplexer_thread should only call cifs_reconnect. If any other thread wants to trigger a reconnect, they can do so by updating the server tcpStatus to CifsNeedReconnect. The last patch attempted to use the same helper function for both types of threads, but that causes other issues with lock dependencies. This patch creates a new helper for non-cifsd threads, that will indicate to cifsd that the server needs reconnect. Fixes: 2a05137a0575 ("cifs: mark sessions for reconnection in helper function") Signed-off-by: Shyam Prasad N Signed-off-by: Steve French

cifs: call helper functions for marking channels for reconnect

2022-02-09T04:13:48+00:00

cifs_mark_tcp_ses_conns_for_reconnect helper function is now meant to be used by any of the threads to mark a channel (or all the channels) for reconnect. Replace all such manual changes to tcpStatus to use this helper function, which takes care that the right channels, smb sessions and tcons are marked for reconnect. Also includes one line minor change Reported-by: kernel test robot Signed-off-by: Shyam Prasad N Signed-off-by: Steve French

cifs: cifs_ses_mark_for_reconnect should also update reconnect bits

2022-01-19T17:10:58+00:00

Recent restructuring of cifs_reconnect introduced a helper func named cifs_ses_mark_for_reconnect, which updates the state of tcp session for all the channels of a session for reconnect. However, this does not update the session state and chans_need_reconnect bitmask. This change fixes that. Also, cifs_mark_tcp_sess_for_reconnect should mark set the bitmask for all channels when the whole session is marked for reconnect. Fixed that here too. Signed-off-by: Shyam Prasad N Signed-off-by: Steve French

cifs: introduce cifs_ses_mark_for_reconnect() helper

2021-11-16T16:57:08+00:00

Use new cifs_ses_mark_for_reconnect() helper to mark all session channels for reconnect instead of duplicating it in different places. Signed-off-by: Paulo Alcantara (SUSE) Signed-off-by: Steve French