block: move bd_set_size() above rescan_partitions() in __blkdev_get()
authorTejun Heo <tj@kernel.org>
Mon, 23 May 2011 11:26:07 +0000 (13:26 +0200)
committerJens Axboe <jaxboe@fusionio.com>
Mon, 23 May 2011 11:26:07 +0000 (13:26 +0200)
commit7e69723fef8771a9d57bd27d36281d756130b4b5
tree0b43d02f5c2eccb932564ed7c3999eb4e649b066
parent95cf3dd9dbe6883a0328724e2110e3fc6465630b
block: move bd_set_size() above rescan_partitions() in __blkdev_get()

02e352287a4 (block: rescan partitions on invalidated devices on
-ENOMEDIA too) relocated partition rescan above explicit bd_set_size()
to simplify condition check.  As rescan_partitions() does its own bdev
size setting, this doesn't break anything; however,
rescan_partitions() prints out the following messages when adjusting
bdev size, which can be confusing.

  sda: detected capacity change from 0 to 146815737856
  sdb: detected capacity change from 0 to 146815737856

This patch restores the original order and remove the warning
messages.

stable: Please apply together with 02e352287a4 (block: rescan
        partitions on invalidated devices on -ENOMEDIA too).

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Tony Luck <tony.luck@gmail.com>
Tested-by: Tony Luck <tony.luck@gmail.com>
Cc: stable@kernel.org

Stable note: 2.6.39 only.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
fs/block_dev.c