Linux RAID subsystem development
 help / color / mirror / Atom feed
* re: md: raid1,10: Handle REQ_WRITE_SAME flag in write bios
@ 2013-01-08 13:44 Dan Carpenter
  2013-01-09  4:11 ` Joe Lawrence
  0 siblings, 1 reply; 3+ messages in thread
From: Dan Carpenter @ 2013-01-08 13:44 UTC (permalink / raw)
  To: Joe.Lawrence; +Cc: linux-raid

Hello Joe Lawrence,

This is a semi-automatic email about new static checker warnings.

The patch aa992f57bd76: "md: raid1,10: Handle REQ_WRITE_SAME flag in 
write bios" from Dec 14, 2012, leads to the following Smatch 
complaint:

drivers/md/raid10.c:3630 run()
	 error: we previously assumed 'mddev->queue' could be null (see line 3619)

drivers/md/raid10.c
  3618		chunk_size = mddev->chunk_sectors << 9;
  3619		if (mddev->queue) {
                    ^^^^^^^^^^^^
Old check.

  3620			blk_queue_max_discard_sectors(mddev->queue,
  3621						      mddev->chunk_sectors);
  3622			blk_queue_io_min(mddev->queue, chunk_size);
  3623			if (conf->geo.raid_disks % conf->geo.near_copies)
  3624				blk_queue_io_opt(mddev->queue, chunk_size * conf->geo.raid_disks);
  3625			else
  3626				blk_queue_io_opt(mddev->queue, chunk_size *
  3627						 (conf->geo.raid_disks / conf->geo.near_copies));
  3628		}
  3629	
  3630		blk_queue_max_write_same_sectors(mddev->queue, mddev->chunk_sectors);
                                                 ^^^^^^^^^^^^
New dereference.

  3631		rdev_for_each(rdev, mddev) {
  3632			long long diff;

regards,
dan carpenter

^ permalink raw reply	[flat|nested] 3+ messages in thread
* re: md: raid1,10: Handle REQ_WRITE_SAME flag in write bios
@ 2013-01-08 13:23 Dan Carpenter
  0 siblings, 0 replies; 3+ messages in thread
From: Dan Carpenter @ 2013-01-08 13:23 UTC (permalink / raw)
  To: Joe.Lawrence; +Cc: linux-raid

Hello Joe Lawrence,

This is a semi-automatic email about new static checker warnings.

The patch aa992f57bd76: "md: raid1,10: Handle REQ_WRITE_SAME flag in
write bios" from Dec 14, 2012, leads to the following Smatch
complaint:

drivers/md/raid1.c:2861 run()
	 warn: variable dereferenced before check 'mddev->queue' (see line 2823)

drivers/md/raid1.c
  2822	
  2823		blk_queue_max_write_same_sectors(mddev->queue, mddev->chunk_sectors);
                                                 ^^^^^^^^^^^^
New dereference.

  2824		rdev_for_each(rdev, mddev) {
  2825			if (!mddev->gendisk)
  2826				continue;
  2827			disk_stack_limits(mddev->gendisk, rdev->bdev,
  2828					  rdev->data_offset << 9);
  2829			if (blk_queue_discard(bdev_get_queue(rdev->bdev)))
  2830				discard_supported = true;
  2831		}
  2832	
  2833		mddev->degraded = 0;
  2834		for (i=0; i < conf->raid_disks; i++)
  2835			if (conf->mirrors[i].rdev == NULL ||
  2836			    !test_bit(In_sync, &conf->mirrors[i].rdev->flags) ||
  2837			    test_bit(Faulty, &conf->mirrors[i].rdev->flags))
  2838				mddev->degraded++;
  2839	
  2840		if (conf->raid_disks - mddev->degraded == 1)
  2841			mddev->recovery_cp = MaxSector;
  2842	
  2843		if (mddev->recovery_cp != MaxSector)
  2844			printk(KERN_NOTICE "md/raid1:%s: not clean"
  2845			       " -- starting background reconstruction\n",
  2846			       mdname(mddev));
  2847		printk(KERN_INFO 
  2848			"md/raid1:%s: active with %d out of %d mirrors\n",
  2849			mdname(mddev), mddev->raid_disks - mddev->degraded, 
  2850			mddev->raid_disks);
  2851	
  2852		/*
  2853		 * Ok, everything is just fine now
  2854		 */
  2855		mddev->thread = conf->thread;
  2856		conf->thread = NULL;
  2857		mddev->private = conf;
  2858	
  2859		md_set_array_sectors(mddev, raid1_size(mddev, 0, 0));
  2860	
  2861		if (mddev->queue) {
                    ^^^^^^^^^^^^
Old check.

  2862			mddev->queue->backing_dev_info.congested_fn = raid1_congested;
  2863			mddev->queue->backing_dev_info.congested_data = mddev;

regards,
dan carpenter

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2013-01-09  4:11 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-01-08 13:44 md: raid1,10: Handle REQ_WRITE_SAME flag in write bios Dan Carpenter
2013-01-09  4:11 ` Joe Lawrence
  -- strict thread matches above, loose matches on Subject: below --
2013-01-08 13:23 Dan Carpenter

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox