Re: Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-13 Thread Mike Christie
On 11/12/15, 3:58 PM, Chris Leech wrote: On Fri, Nov 06, 2015 at 09:52:57AM -0600, Michael Christie wrote: Or and Shlomo, I am going to just revert the patch for now until I can do some performance testing with a patch to add back correct locking and also re-rerview the patch. Maybe instead of

Re: Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-12 Thread Chris Leech
On Fri, Nov 06, 2015 at 09:52:57AM -0600, Michael Christie wrote: > Or and Shlomo, I am going to just revert the patch for now until I can > do some performance testing with a patch to add back correct locking > and also re-rerview the patch. Maybe instead of locks I can just use > some kfifos

Re: Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-11 Thread Mike Christie
On 11/09/2015 07:40 AM, Guilherme G. Piccoli wrote: > On 11/06/2015 01:52 PM, Michael Christie wrote: > >>> On Nov 5, 2015, at 6:56 PM, Chris Leech wrote: >>> >>> On Thu, Nov 05, 2015 at 02:33:44PM -0200, Guilherme G. Piccoli wrote: >>> >>> Interesting. From the trace, the

Re: Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-09 Thread Guilherme G. Piccoli
On 11/06/2015 01:52 PM, Michael Christie wrote: On Nov 5, 2015, at 6:56 PM, Chris Leech wrote: On Thu, Nov 05, 2015 at 02:33:44PM -0200, Guilherme G. Piccoli wrote: Interesting. From the trace, the list debugging code is detecting corruption when removing a task from some

Re: Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-06 Thread Michael Christie
> On Nov 5, 2015, at 6:56 PM, Chris Leech wrote: > > On Thu, Nov 05, 2015 at 02:33:44PM -0200, Guilherme G. Piccoli wrote: >> Hello Shlomo and Or, >> >> I'm Guilherme Piccoli from LTC/IBM - firstly, sorry to bother you. >> >> >> We are running some tests with iSCSI and we

Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-05 Thread Guilherme G. Piccoli
Hello Shlomo and Or, I'm Guilherme Piccoli from LTC/IBM - firstly, sorry to bother you. We are running some tests with iSCSI and we found an issue caused possibly by commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path"). After some time (+/- 1 hour) of testing with a

Antw: Re: Possible issue in commit 659743b02c41 ("libiscsi: Reduce locking contention in fast path")

2015-11-05 Thread Ulrich Windl
>>> Chris Leech schrieb am 06.11.2015 um 01:56 in Nachricht <20151106005608.ga18...@straylight.hirudinean.org>: [...] > Am I missing something, or is splitting a linked list across two locks a > major failing of this change? [...] Could you explain your question again for