Skip to main content

Browse the archive

https://github.com/torvalds/linux
13 June 2022, 11:31:42 UTC
Revision a8c32a5c98943d370ea606a2e7dc04717eb92206 authored by Jens Axboe on 06 November 2012, 11:24:26 UTC, committed by Jens Axboe on 23 November 2012, 13:32:54 UTC 
dm: fix deadlock with request based dm and queue request_fn recursion
 
Request based dm attempts to re-run the request queue off the
request completion path. If used with a driver that potentially does
end_io from its request_fn, we could deadlock trying to recurse
back into request dispatch. Fix this by punting the request queue
run to kblockd.

Tested to fix a quickly reproducible deadlock in such a scenario.

Cc: stable@kernel.org
Acked-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
1 parent eac7cc5
Raw File
Tip revision: a8c32a5c98943d370ea606a2e7dc04717eb92206 authored by Jens Axboe on 06 November 2012, 11:24:26 UTC
dm: fix deadlock with request based dm and queue request_fn recursion
Tip revision: a8c32a5
nonet.c 
/*
 * net/nonet.c
 *
 * Dummy functions to allow us to configure network support entirely
 * out of the kernel.
 *
 * Distributed under the terms of the GNU GPL version 2.
 * Copyright (c) Matthew Wilcox 2003
 */

#include <linux/module.h>
#include <linux/errno.h>
#include <linux/fs.h>
#include <linux/init.h>
#include <linux/kernel.h>

static int sock_no_open(struct inode *irrelevant, struct file *dontcare)
{
	return -ENXIO;
}

const struct file_operations bad_sock_fops = {
	.owner = THIS_MODULE,
	.open = sock_no_open,
	.llseek = noop_llseek,
};
back to top