Re: [PATCH 0/4] block: Fixes for bdi handling

[Arthur Marsh]

Jan Kara wrote on 09/03/17 03:18:

Hi!

patches in this series fix the most urgent bugs that were introduced by commit
165a5e22fafb "block: Move bdi_unregister() to del_gendisk()" and by
0dba1314d4f8 "scsi, block: fix duplicate bdi name registration crashes".
In fact before these commits we had a different set of problems in the
code but they were less visible :).

I'm still waiting for test confirmation from Omar and Arthur Marsh who reported
issues but I'm not able to hit any problem anymore in my testing.  I think it
would be nice to get the patches to rc2 so to speed up things I'm posting the
patches now so that review can happen in parallel with the testing.

Other BDI handling fixes I have in my queue can wait a bit more since they are
either theoretical or long-standing issues. So I'll repost them once these four
are sorted out.

Honza



Sorry for the delay in replying, I had to leave the kernel with all 4 
patches applied rebuilding while I was at work and just booted it.


I've only done a kexec reboot so far but there were no problems - no 
errors in dmesg, all disks were recognised and all attempted mounts worked.


Thanks very much for the quick fix!

Arthur.

Re: [PATCH 0/4] block: Fixes for bdi handling

[Omar Sandoval]

On Wed, Mar 08, 2017 at 08:50:51AM -0800, Omar Sandoval wrote:
> On Wed, Mar 08, 2017 at 05:48:30PM +0100, Jan Kara wrote:
> > Hi!
> > 
> > patches in this series fix the most urgent bugs that were introduced by 
> > commit
> > 165a5e22fafb "block: Move bdi_unregister() to del_gendisk()" and by
> > 0dba1314d4f8 "scsi, block: fix duplicate bdi name registration crashes".
> > In fact before these commits we had a different set of problems in the
> > code but they were less visible :).
> > 
> > I'm still waiting for test confirmation from Omar and Arthur Marsh who 
> > reported
> > issues but I'm not able to hit any problem anymore in my testing.  I think 
> > it
> > would be nice to get the patches to rc2 so to speed up things I'm posting 
> > the
> > patches now so that review can happen in parallel with the testing.
> > 
> > Other BDI handling fixes I have in my queue can wait a bit more since they 
> > are
> > either theoretical or long-standing issues. So I'll repost them once these 
> > four
> > are sorted out.
> > 
> > Honza
> 
> Applying patches 1 and 2 indeed fixed my virtio-scsi problem. I'll apply
> the whole series and test that my sd/sr test cases still work.

Yup, everything is working here. For the series:

Tested-by: Omar Sandoval 

Re: [PATCH 0/4] block: Fixes for bdi handling

[Jens Axboe]

On 03/08/2017 09:48 AM, Jan Kara wrote:
> Hi!
> 
> patches in this series fix the most urgent bugs that were introduced by commit
> 165a5e22fafb "block: Move bdi_unregister() to del_gendisk()" and by
> 0dba1314d4f8 "scsi, block: fix duplicate bdi name registration crashes".
> In fact before these commits we had a different set of problems in the
> code but they were less visible :).
> 
> I'm still waiting for test confirmation from Omar and Arthur Marsh who 
> reported
> issues but I'm not able to hit any problem anymore in my testing.  I think it
> would be nice to get the patches to rc2 so to speed up things I'm posting the
> patches now so that review can happen in parallel with the testing.
> 
> Other BDI handling fixes I have in my queue can wait a bit more since they are
> either theoretical or long-standing issues. So I'll repost them once these 
> four
> are sorted out.

Agree, we'll need to get this sorted for -rc2, thanks a lot for looking
into it. I'll queue these up for testing as well, and hopefully ship
them off tomorrow (or Friday, at the latest).

-- 
Jens Axboe

Re: [PATCH 0/4] block: Fixes for bdi handling

[Omar Sandoval]

On Wed, Mar 08, 2017 at 05:48:30PM +0100, Jan Kara wrote:
> Hi!
> 
> patches in this series fix the most urgent bugs that were introduced by commit
> 165a5e22fafb "block: Move bdi_unregister() to del_gendisk()" and by
> 0dba1314d4f8 "scsi, block: fix duplicate bdi name registration crashes".
> In fact before these commits we had a different set of problems in the
> code but they were less visible :).
> 
> I'm still waiting for test confirmation from Omar and Arthur Marsh who 
> reported
> issues but I'm not able to hit any problem anymore in my testing.  I think it
> would be nice to get the patches to rc2 so to speed up things I'm posting the
> patches now so that review can happen in parallel with the testing.
> 
> Other BDI handling fixes I have in my queue can wait a bit more since they are
> either theoretical or long-standing issues. So I'll repost them once these 
> four
> are sorted out.
> 
>   Honza

Applying patches 1 and 2 indeed fixed my virtio-scsi problem. I'll apply
the whole series and test that my sd/sr test cases still work.