pm / runtime: force memory allocation with no I/O during Runtime PM callbcack

Apply the introduced memalloc_noio_save() and memalloc_noio_restore() to
force memory allocation with no I/O during runtime_resume/runtime_suspend
callback on device with the flag of 'memalloc_noio' set.

Signed-off-by: Ming Lei <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Eric Dumazet <[email protected]>
Cc: David Decotigny <[email protected]>
Cc: Tom Herbert <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Jens Axboe <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Alan Stern <[email protected]>
Cc: Oliver Neukum <[email protected]>
Cc: Jiri Kosina <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: KAMEZAWA Hiroyuki <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: "Rafael J. Wysocki" <[email protected]>
Cc: Greg KH <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
This commit is contained in:
Ming Lei 2013-02-22 16:34:19 -08:00 • committed by Linus Torvalds
commit db88175f41

View file

@ -348,7 +348,24 @@ static int rpm_callback(int (*cb)(struct device *), struct device *dev)
if (!cb)
return -ENOSYS;
retval = __rpm_callback(cb, dev);
if (dev->power.memalloc_noio) {
unsigned int noio_flag;
/*
* Deadlock might be caused if memory allocation with
* GFP_KERNEL happens inside runtime_suspend and
* runtime_resume callbacks of one block device's
* ancestor or the block device itself. Network
* device might be thought as part of iSCSI block
* device, so network device and its ancestor should
* be marked as memalloc_noio too.
*/
noio_flag = memalloc_noio_save();
retval = __rpm_callback(cb, dev);
memalloc_noio_restore(noio_flag);
} else {
retval = __rpm_callback(cb, dev);
}
dev->power.runtime_error = retval;
return retval != -EACCES ? retval : -EIO;