RE: [PATCH 1/2] btrfs: reada: limit max works count

-----Original Message-----
From: Chris Mason [mailto:clm@fb.com]
Sent: Thursday, January 21, 2016 1:48 AM
To: Zhao Lei <redacted>; linux-btrfs@vger.kernel.org
Subject: Re: [PATCH 1/2] btrfs: reada: limit max works count

On Wed, Jan 20, 2016 at 10:16:27AM -0500, Chris Mason wrote:

quoted

On Tue, Jan 12, 2016 at 03:46:26PM +0800, Zhao Lei wrote:

quoted

reada create 2 works for each level of tree in recursion.

In case of a tree having many levels, the number of created works is
2^level_of_tree.
Actually we don't need so many works in parallel, this patch limit
max works to BTRFS_MAX_MIRRORS * 2.

Hi,

I don't think you end up calling atomic_dec() for every time that
reada_start_machine() is called.  Also, I'd rather not have a global
static variable to limit the parallel workers, when we have more than
one FS mounted it'll end up limiting things too much.

With this patch applied, I'm seeing deadlocks during btrfs/066.    You
have to run the scrub tests as well, basically we're just getting
fsstress run alongside scrub.

I'll run a few more times with it reverted to make sure, but I think
it's the root cause.

I spoke too soon, it ended up deadlocking a few tests later.  Sorry for now I'm
pulling all the reada patches.  We'll sort out bug fixes vs cleanups in later rcs.

With all of the reada patches removed, the deadlocks are gone.

Sorry for hear it.

Actually I run xfstests with all patch applied, and see no regression in my env:

FSTYP         -- btrfs
PLATFORM      -- Linux/x86_64 lenovo 4.4.0-rc6_HEAD_8e16378041f7f3531c256fd3e17a36a4fca92d29_+
MKFS_OPTIONS  -- /dev/sdb6
MOUNT_OPTIONS -- /dev/sdb6 /var/ltf/tester/scratch_mnt

btrfs/066 151s ... 164s
Ran: btrfs/066
Passed all 1 tests

I'll investigate the root reason.

Thanks
Zhaolei

-chris