Re: ext4 filesystem corruption with 4.10-rc2 on ppc64le

On Wednesday, January 04, 2017 10:28:37 AM Theodore Ts'o wrote:

On Wed, Jan 04, 2017 at 11:32:42AM +0530, Chandan Rajendra wrote:

quoted

On Wednesday, January 04, 2017 04:18:08 PM Anton Blanchard wrote:

quoted

I'm consistently seeing ext4 filesystem corruption using a mainline
kernel. It doesn't take much to trigger it - download a ppc64le Ubuntu
cloud image, boot it in KVM and run:

sudo apt-get update
sudo apt-get dist-upgrade
sudo reboot

And it never makes it back up, dying with rather severe filesystem
corruption.

The patch at https://patchwork.kernel.org/patch/9488235/ should fix the
bug.

It looks like this patch is already queued up on the "for-linus"
branch on the linux-block.git tree.

Chandra, thanks for pointing this out!  I had missed your e-mail from
Christmas day, and it was on my todo list to figure out why I was
seeing lots of 1k block regressions on gce-xfstests post-merge window
that wasn't showing up on the ext4.git tree before I sent my pull
request to Linus.

Jens, could you expedite a pull request to Linus?  This is affecting
ext4 on 1k block file systems on x86/x86_64, so this is not a ppc-only
regression.  

Anton or Chandan, could you do me a favor and verify whether or not
64k block sizes are working for you on ppcle on ext4 by running
xfstests?  Light duty testing works for me but when I stress ext4 with
pagesize==blocksize on ppcle64 via xfstests, it blows up.  I suspect
(but am not sure) it's due to (non-upstream) device driver issues, and
a verification that you can run xfstests on your ppcle64 systems using
standard upstream device drivers would be very helpful, since I don't
have easy console access on the machines I have access to at $WORK.  :-(

Hi Ted,

I found one regression w.r.t 64k blocksize. I posted a patch
(http://marc.info/?l=linux-block&m=148388687722745&w=2) to fix the issue. 

-- 
chandan