On Wednesday February 27, ross@willow.seitz.com wrote:
> Neil,
>
> I read through the md code this afternoon looking for uses of
> mddev->disks, and protected each use with reconfig_sem. These three are
> the only three I saw.
Thanks.
However I think that in each of these cases the semaphore is already
held.
mddev->reconfig_sem is normally claimed by calling lock_mddev(mddev),
which jusdt does down(&mddev->reconfig_sem).
The first chunk of yor patch is in bind_rdev_to_array.
This is called from
autorun_devices which has just allocated the mddev so the
semaphore is still locked (it is allocated locked, and
then unlocked).
add_new_disk and hot_add_disk which are only called from
md_ioctl *after* the mddev has been locked with
lock_mddev_interruptible(mddev)
So in that case the extra locking is not needed and would deadlock.
I think you will find that the same is true for the other chunks.
NeilBrown
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
