[dm-devel] [PATCH] dm: fix request-based dm's use of dm_wait_for_completion

2018-12-11 Thread Mike Snitzer
The md->wait waitqueue is used by both bio-based and request-based DM. Commit dbd3bbd291 ("dm rq: leverage blk_mq_queue_busy() to check for outstanding IO") lost sight of the requirement that dm_wait_for_completion() must work with all types of DM devices. Fix md_in_flight() to call the blk-mq or

Re: [dm-devel] [PATCH] dm: fix request-based dm's use of dm_wait_for_completion

2018-12-11 Thread Jens Axboe
On 12/11/18 7:10 AM, Mike Snitzer wrote: > The md->wait waitqueue is used by both bio-based and request-based DM. > Commit dbd3bbd291 ("dm rq: leverage blk_mq_queue_busy() to check for > outstanding IO") lost sight of the requirement that > dm_wait_for_completion() must work with all types of DM de

Re: [dm-devel] [PATCH v2 23/24] domap(): never return DOMAP_RETRY in daemon mode

2018-12-11 Thread Benjamin Marzinski
On Sun, Dec 09, 2018 at 10:06:05PM +0100, Martin Wilck wrote: > On Mon, 2018-12-03 at 17:45 -0600, Benjamin Marzinski wrote: > > TL;DR: I'm 99.7% sure we don't need lock_multipath() any more. > > The historic reason is 4d7a270: > > 'Multiple multipath(8) execs can race with udev storm. >

Re: [dm-devel] [PATCH v3 00/24] multipath-tools: improve logging at -v3

2018-12-11 Thread Benjamin Marzinski
On Mon, Dec 10, 2018 at 10:49:35AM +0100, Martin Wilck wrote: > Changes in v3: > > Resent full series on Christophe's request. All patches except 22/24, 23/24 > are > the same as before. Added Ben's "Reviewed-by:" where appropriate. Belated ACKs on 23/24 and v4 of 22/24 -Ben > > - 22/24: add

Re: [dm-devel] [PATCH] libmultipath: Increase SERIAL_SIZE to 256 bytes

2018-12-11 Thread Martin Wilck
Kyle, please resubmit with SERIAL_SIZE=128. That should be sufficient for the time being. Martin On Tue, 2018-08-28 at 16:34 -0500, KyleMahlkuch wrote: > Certain IBM FlashSystem LUNs can return up to 85 bytes of serial > number in the Unit Serial Number VPD page, which is larger than > the curre

[dm-devel] [PATCH 0/1] update to v2 Fix false removes

2018-12-11 Thread Benjamin Marzinski
I accidentally didn't include Christophe in my [PATCH v2 0/1] Fix false removes patchset, so v1 of "libmultipath: fix false removes in dmevents polling code" went in. This commit just patches the code to my v2 version. Sorry for the confusion. I've included Martin's Reviewed-by as the end res

[dm-devel] [PATCH 1/1] libmultipath: dm_is_mpath cleanup

2018-12-11 Thread Benjamin Marzinski
Add condlog() message in dm_is_mpath() fails and change the dm_is_mpath() call in watch_dmevents() to check the return value with the same syntax as all the other callers. Fixes: 9050cd5a "libmultipath: fix false removes in dmevents polling code" Reviewed-by: Martin Wilck Signed-off-by: Benjamin

Re: [dm-devel] for-next hangs on test srp/012

2018-12-11 Thread Mike Snitzer
On Tue, Dec 11 2018 at 5:58pm -0500, Bart Van Assche wrote: > Hi Jens, > > If I run the following subset of blktests: > > while :; do ./check -q srp && ./check -q nvmeof-mp; done > > against today's for-next branch (commit dd2bf2df85a7) then after some > time the following hang is reported:

Re: [dm-devel] for-next hangs on test srp/012

2018-12-11 Thread Bart Van Assche
On Tue, 2018-12-11 at 18:05 -0500, Mike Snitzer wrote: > On Tue, Dec 11 2018 at 5:58pm -0500, > Bart Van Assche wrote: > > > Hi Jens, > > > > If I run the following subset of blktests: > > > > while :; do ./check -q srp && ./check -q nvmeof-mp; done > > > > against today's for-next branch (

Re: [dm-devel] [PATCH 0/1] update to v2 Fix false removes

2018-12-11 Thread Benjamin Marzinski
On Tue, Dec 11, 2018 at 03:23:03PM -0600, Benjamin Marzinski wrote: > I accidentally didn't include Christophe in my > > [PATCH v2 0/1] Fix false removes Clearly, something is wrong with my git send-mail setup, since you got ignored again. Sorry. I'll figure out what's going on. -Ben > patchs

Re: [dm-devel] [PATCH 0/1] update to v2 Fix false removes

2018-12-11 Thread Christophe Varoqui
FYI, I did receive the v2 patch and wrapping message. Thanks Le mer. 12 déc. 2018 à 00:47, Benjamin Marzinski a écrit : > On Tue, Dec 11, 2018 at 03:23:03PM -0600, Benjamin Marzinski wrote: > > I accidentally didn't include Christophe in my > > > > [PATCH v2 0/1] Fix false removes > > Clearly,

Re: [dm-devel] for-next hangs on test srp/012

2018-12-11 Thread Mike Snitzer
how things go. > > Hi Mike, > > Thank you for your reply. I should have mentioned that I used Jens' for-next > branch as starting point and that that commit is present in the branch I used > as a starting point for my tests: > > $ git branch --contains c4576aed8d85d808

[dm-devel] xfstests generic/347 was never correct [was: Re: dm: fix inflight IO check]

2018-12-11 Thread Mike Snitzer
On Mon, Dec 10 2018 at 7:32pm -0500, Mike Snitzer wrote: > On Mon, Dec 10 2018 at 5:45pm -0500, > Jens Axboe wrote: > > > After switching to percpu inflight counters, the inflight check > > is totally buggy. It's perfectly valid for some counters to be > > non-zero while having a total inflig

Re: [dm-devel] xfstests generic/347 was never correct [was: Re: dm: fix inflight IO check]

2018-12-11 Thread Eric Sandeen
On 12/11/18 8:34 PM, Mike Snitzer wrote: > On Mon, Dec 10 2018 at 7:32pm -0500, > Mike Snitzer wrote: > >> On Mon, Dec 10 2018 at 5:45pm -0500, >> Jens Axboe wrote: >> >>> After switching to percpu inflight counters, the inflight check >>> is totally buggy. It's perfectly valid for some counte