From: Eric Biggers <ebiggers@kernel.org>
To: linux-cifs@vger.kernel.org
Cc: linux-ext4@vger.kernel.org, syzkaller-bugs@googlegroups.com,
linux-kernel@vger.kernel.org
Subject: [PATCH v3] cifs: clear PF_MEMALLOC before exiting demultiplex thread
Date: Sun, 8 Mar 2020 22:58:20 -0700 [thread overview]
Message-ID: <20200309055820.272703-1-ebiggers@kernel.org> (raw)
In-Reply-To: <20200309055621.GA272474@sol.localdomain>
From: Eric Biggers <ebiggers@google.com>
Leaving PF_MEMALLOC set when exiting a kthread causes it to remain set
during do_exit(). That can confuse things. For example, if BSD process
accounting is enabled and the accounting file has FS_SYNC_FL set and is
located on an ext4 filesystem without a journal, then do_exit() can end
up calling ext4_write_inode(). That triggers the
WARN_ON_ONCE(current->flags & PF_MEMALLOC) there, as it assumes
(appropriately) that inodes aren't written when allocating memory.
This was originally reported for another kernel thread, xfsaild() [1].
cifs_demultiplex_thread() also exits with PF_MEMALLOC set, so it's
potentially subject to this same class of issue -- though I haven't been
able to reproduce the WARN_ON_ONCE() via CIFS, since unlike xfsaild(),
cifs_demultiplex_thread() is sent SIGKILL before exiting, and that
interrupts the write to the BSD process accounting file.
Either way, leaving PF_MEMALLOC set is potentially problematic. Let's
clean this up by properly saving and restoring PF_MEMALLOC.
[1] https://lore.kernel.org/r/0000000000000e7156059f751d7b@google.com
Signed-off-by: Eric Biggers <ebiggers@google.com>
---
v3: improved commit message
v2: added missing include
fs/cifs/connect.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 4804d1df8c1cf..97b8eb585cf9d 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -21,6 +21,7 @@
#include <linux/fs.h>
#include <linux/net.h>
#include <linux/string.h>
+#include <linux/sched/mm.h>
#include <linux/sched/signal.h>
#include <linux/list.h>
#include <linux/wait.h>
@@ -1164,8 +1165,9 @@ cifs_demultiplex_thread(void *p)
struct task_struct *task_to_wake = NULL;
struct mid_q_entry *mids[MAX_COMPOUND];
char *bufs[MAX_COMPOUND];
+ unsigned int noreclaim_flag;
- current->flags |= PF_MEMALLOC;
+ noreclaim_flag = memalloc_noreclaim_save();
cifs_dbg(FYI, "Demultiplex PID: %d\n", task_pid_nr(current));
length = atomic_inc_return(&tcpSesAllocCount);
@@ -1320,6 +1322,7 @@ cifs_demultiplex_thread(void *p)
set_current_state(TASK_RUNNING);
}
+ memalloc_noreclaim_restore(noreclaim_flag);
module_put_and_exit(0);
}
--
2.25.1
prev parent reply other threads:[~2020-03-09 5:59 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <0000000000000e7156059f751d7b@google.com>
2020-03-08 4:36 ` [PATCH] cifs: clear PF_MEMALLOC before exiting demultiplex thread Eric Biggers
2020-03-08 6:16 ` [PATCH v2] " Eric Biggers
2020-03-08 18:43 ` Steve French
2020-03-09 5:56 ` Eric Biggers
2020-03-09 5:58 ` Eric Biggers [this message]
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=20200309055820.272703-1-ebiggers@kernel.org \
--to=ebiggers@kernel.org \
--cc=linux-cifs@vger.kernel.org \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=syzkaller-bugs@googlegroups.com \
/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).