lustre-devel-lustre.org archive mirror
 help / color / mirror / Atom feed
From: James Simmons <jsimmons@infradead.org>
To: Andreas Dilger <adilger@whamcloud.com>,
	Oleg Drokin <green@whamcloud.com>, NeilBrown <neilb@suse.de>
Cc: Lustre Development List <lustre-devel@lists.lustre.org>
Subject: [lustre-devel] [PATCH 17/19] lustre: llite: mend the trunc_sem_up_write()
Date: Sun, 28 Nov 2021 18:27:52 -0500	[thread overview]
Message-ID: <1638142074-5945-18-git-send-email-jsimmons@infradead.org> (raw)
In-Reply-To: <1638142074-5945-1-git-send-email-jsimmons@infradead.org>

From: Bobi Jam <bobijam@whamcloud.com>

The original lli_trunc_sem replace change (commit ae9e437745) fixed a
lock scenario:

t1 (page fault)          t2 (dio read)              t3 (truncate)
|- vm_mmap_pgoff()       |- vvp_io_read_start()     |- vvp_io_setattr
|- down_write(mmap_sem)  |- down_read(trunc_sem)            _start()
|- do_map()              |- ll_direct_IO_impl()
|- vvp_io_fault_start    |- ll_get_user_pages()

                                                    |- down_write(
                         |- down_read(mmap_sem)        trunc_sem)
|- down_read(trunc_sem)

t1 waits for read semaphore of trunc_sem which is hindered by t3,
since t3 is waiting for the write semaphore while t2 take its read
semaphore,and t2 is waiting for mmap_sem which has been taken by t1,
and a deadlock ensues.

commit ae9e437745 changes the down_read(trunc_sem) to
trunc_sem_down_read_nowait() in page fault path, to make it ignore
that there is a down_write(trunc_sem) waiting, just takes the read
semaphore if no writer has taken the semaphore, and breaks the
deadlock.

But there is a delicacy in using wake_up_var(), wake_up_var()->
__wake_up_bit()->waitqueue_active() locklessly test for waiters on the
queue, and if it's called without explicit smp_mb() it's possible for
the waitqueue_active() to ge hoisted before the condition store such
that we'll observe an empty wait list and the waiter might not
observe the condition, and the waiter won't get woke up whereafter.

Fixes: ae9e437745 ("lustre: llite: replace lli_trunc_sem")
WC-bug-id: https://jira.whamcloud.com/browse/LU-14713
Lustre-commit: 39745c8b5493159bb ("LU-14713 llite: mend the trunc_sem_up_write()")
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/43844
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
 fs/lustre/llite/llite_internal.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/lustre/llite/llite_internal.h b/fs/lustre/llite/llite_internal.h
index 7768c99..ce7431f 100644
--- a/fs/lustre/llite/llite_internal.h
+++ b/fs/lustre/llite/llite_internal.h
@@ -365,6 +365,8 @@ static inline void trunc_sem_down_write(struct ll_trunc_sem *sem)
 static inline void trunc_sem_up_write(struct ll_trunc_sem *sem)
 {
 	atomic_set(&sem->ll_trunc_readers, 0);
+	/* match the smp_mb() in wait_var_event()->prepare_to_wait() */
+	smp_mb();
 	wake_up_var(&sem->ll_trunc_readers);
 }
 
-- 
1.8.3.1

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

  parent reply	other threads:[~2021-11-28 23:28 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-28 23:27 [lustre-devel] [PATCH 00/19] lustre: update to OpenSFS tree Nov 28, 2021 James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 01/19] lnet: fix delay rule crash James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 02/19] lnet: change tp_nid to 16byte in lnet_test_peer James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 03/19] lnet: extend preferred nids in struct lnet_peer_ni James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 04/19] lnet: switch to large lnet_processid for matching James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 05/19] lnet: libcfs: add timeout to cfs_race() to fix race James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 06/19] lustre: llite: tighten condition for fault not drop mmap_sem James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 07/19] lnet: o2iblnd: map_on_demand not needed for frag interop James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 08/19] lnet: o2iblnd: Fix logic for unaligned transfer James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 09/19] lnet: Reset ni_ping_count only on receive James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 10/19] lustre: ptlrpc: fix timeout after spurious wakeup James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 11/19] lnet: Fail peer add for existing gw peer James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 12/19] lustre: ptlrpc: remove bogus LASSERT James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 13/19] lustre: quota: optimize capability check for root squash James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 14/19] lustre: llite: skip request slot for lmv_revalidate_slaves() James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 15/19] lnet: set eth routes needed for multi rail James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 16/19] lustre: llite: Do not count tiny write twice James Simmons
2021-11-28 23:27 ` James Simmons [this message]
2021-11-28 23:27 ` [lustre-devel] [PATCH 18/19] lnet: Netlink improvements James Simmons
2021-11-28 23:27 ` [lustre-devel] [PATCH 19/19] lnet: libcfs: separate daemon_list from cfs_trace_data James Simmons

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1638142074-5945-18-git-send-email-jsimmons@infradead.org \
    --to=jsimmons@infradead.org \
    --cc=adilger@whamcloud.com \
    --cc=green@whamcloud.com \
    --cc=lustre-devel@lists.lustre.org \
    --cc=neilb@suse.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).