Skip to main content

Browse the archive

Revision 774a1221e862b343388347bac9b318767336b20b authored by Tejun Heo on 16 January 2013, 02:52:51 UTC, committed by Linus Torvalds on 16 January 2013, 17:05:33 UTC 
module, async: async_synchronize_full() on module init iff async is used
 
If the default iosched is built as module, the kernel may deadlock
while trying to load the iosched module on device probe if the probing
was running off async.  This is because async_synchronize_full() at
the end of module init ends up waiting for the async job which
initiated the module loading.

 async A				modprobe

 1. finds a device
 2. registers the block device
 3. request_module(default iosched)
					4. modprobe in userland
					5. load and init module
					6. async_synchronize_full()

Async A waits for modprobe to finish in request_module() and modprobe
waits for async A to finish in async_synchronize_full().

Because there's no easy to track dependency once control goes out to
userland, implementing properly nested flushing is difficult.  For
now, make module init perform async_synchronize_full() iff module init
has queued async jobs as suggested by Linus.

This avoids the described deadlock because iosched module doesn't use
async and thus wouldn't invoke async_synchronize_full().  This is
hacky and incomplete.  It will deadlock if async module loading nests;
however, this works around the known problem case and seems to be the
best of bad options.

For more details, please refer to the following thread.

  http://thread.gmane.org/gmane.linux.kernel/1420814

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Alex Riesen <raa.lkml@gmail.com>
Tested-by: Ming Lei <ming.lei@canonical.com>
Tested-by: Alex Riesen <raa.lkml@gmail.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent 406089d
Raw File
netdevices.c 
/* AFS network device helpers
 *
 * Copyright (c) 2007 Patrick McHardy <kaber@trash.net>
 */

#include <linux/string.h>
#include <linux/rtnetlink.h>
#include <linux/inetdevice.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <net/net_namespace.h>
#include "internal.h"

/*
 * get a MAC address from a random ethernet interface that has a real one
 * - the buffer will normally be 6 bytes in size
 */
int afs_get_MAC_address(u8 *mac, size_t maclen)
{
	struct net_device *dev;
	int ret = -ENODEV;

	BUG_ON(maclen != ETH_ALEN);

	rtnl_lock();
	dev = __dev_getfirstbyhwtype(&init_net, ARPHRD_ETHER);
	if (dev) {
		memcpy(mac, dev->dev_addr, maclen);
		ret = 0;
	}
	rtnl_unlock();
	return ret;
}

/*
 * get a list of this system's interface IPv4 addresses, netmasks and MTUs
 * - maxbufs must be at least 1
 * - returns the number of interface records in the buffer
 */
int afs_get_ipv4_interfaces(struct afs_interface *bufs, size_t maxbufs,
			    bool wantloopback)
{
	struct net_device *dev;
	struct in_device *idev;
	int n = 0;

	ASSERT(maxbufs > 0);

	rtnl_lock();
	for_each_netdev(&init_net, dev) {
		if (dev->type == ARPHRD_LOOPBACK && !wantloopback)
			continue;
		idev = __in_dev_get_rtnl(dev);
		if (!idev)
			continue;
		for_primary_ifa(idev) {
			bufs[n].address.s_addr = ifa->ifa_address;
			bufs[n].netmask.s_addr = ifa->ifa_mask;
			bufs[n].mtu = dev->mtu;
			n++;
			if (n >= maxbufs)
				goto out;
		} endfor_ifa(idev);
	}
out:
	rtnl_unlock();
	return n;
}
back to top