On 08.04.22 04:50, Thorsten Leemhuis wrote: > First off: David, Filipe, many thx for your answers, that helped me a > lot to get a better picture of the situation! > > On 08.04.22 17:55, Filipe Manana wrote: >> On Fri, Apr 08, 2022 at 04:52:22PM +0200, David Sterba wrote: >>> On Fri, Apr 08, 2022 at 12:32:20PM +0200, Thorsten Leemhuis wrote: >>>> Hi, this is your Linux kernel regression tracker. Top-posting for once, >>>> to make this easily accessible to everyone. >>>> >>>> Btrfs maintainers, what's up here? Yes, this regression report was a bit >>>> confusing in the beginning, but Bruno worked on it. And apparently it's >>>> already fixed in 5.16, but still in 5.15. Is this caused by a change >>>> that is to big to backport or something? >>> >>> I haven't identified possible fixes in 5.16 so I can't tell how much >>> backport efforts it could be. As the report is related to performance on >>> package updates, my best guess is that the patches fixing it are those >>> from Filipe related to fsync/logging, and there are several of such >>> improvements in 5.16. Or something else that fixes it indirectly. >> >> So there's a lot of confusion in the thread, > > Yeah, definitely. That basically why I had hoped from a rough assessment > from the btrfs maintainers. > >> and the original openSUSE >> bugzilla [1] is also a bit confusing and large to follow. >> >> Let me try to make it clear: >> >> 1) For some reason, outside btrfs' control, inode eviction is triggered >> a lot on 5.15 kernels in Bruno's test machine when doing package >> installations/updates with zypper. > > So I assume there are no other reports like this? Great! Well, "inode eviction is triggered a lot on 5.15 kernels in Bruno's test machine when doing package installations/updates with zypper" is a little misleading IMHO. While this was true at the very beginning it became more than that. Now the high inode eviction is already know to be 100% reproducible on: - at least one real workload (opensuse package update). - 3 different bare metal machines with different hardware configuration. - 3 different cpus from 2 different cpu manufactures. - one synthetic worklod (the scripts I provided). - 4 different distributions on virtual machines. - all 5.15 kernels that I tried on. About the lack of reports, my only guess is that most users must choose the compress mount option instead of manually setting the compression property: - the compress mount option doesn't trigger the regression. - the compression property does trigger the regression. >> [...] > >> 6) In short, it is not known what causes the excessive evictions on 5.15 >> on his machine for that specific workload - we don't have a commit to >> point at and say it caused a regression. [...] > > Bruno, under these circumstances I'd say you need to bisect this to get > us closer to the root of the problem (and a fix for it). Sadly that how > it is sometimes, as briefly explained here: > > https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/tree/Documentation/admin-guide/reporting-regressions.rst#n140 Ok Thorsten. It's not sad at all: I had a great time researching this regression and gained a lot of knowledge while doing so. The problem is that I am just a simple user at its limits here and additional bisection is probably beyond my abilities. I can only hope some kernel developer feels motivated to further investigate this subject. Again, sorry for the confusion and thanks a lot for your patience and for the directions. >> This thread is also basically a revamp of an older thread [3]. >> >> [1] https://bugzilla.opensuse.org/show_bug.cgi?id=1193549 >> [2] https://lore.kernel.org/linux-btrfs/cover.1642676248.git.fdmanana@xxxxxxxx/ >> [3] https://lore.kernel.org/linux-fsdevel/MN2PR20MB251235DDB741CD46A9DD5FAAD24E9@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/ > > Yeah, but it was this thread that made me aware of the issue -- and just > like [3] it didn't get a single reply from a btrfs maintainer, so I had > to assume the report was ignored. A quick "we have no idea what my cause > this issue and it's the only report with such symptoms so far; could you > please bisect" would have made me happy already. :-D > > Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat) > > P.S.: As the Linux kernel's regression tracker I'm getting a lot of > reports on my table. I can only look briefly into most of them and lack > knowledge about most of the areas they concern. I thus unfortunately > will sometimes get things wrong or miss something important. I hope > that's not the case here; if you think it is, don't hesitate to tell me > in a public reply, it's in everyone's interest to set the public record > straight. > Grazie, Bruno.
