Linux regressions report for mainline [2023-06-25]
From: Regzbot (on behalf of Thorsten Leemhuis)
Date: Sun Jun 25 2023 - 09:51:43 EST
Hi Linus, find below regzbot's report about the remaining regressions
that were introduced this cycle. Not much has changed since my "quick &
small additional regressions report pre -rc8/final", so I'm quoting it
here:
> * kernel test robot noticed a -20.0% regression of
> vm-scalability.throughput post -rc7;
BTW, that was caused by 9425c591e06 ("page cache: fix
page_cache_next/prev_miss off by one") [v6.4-rc7]
> Mike posted reverts akpm applied
> that landed in next recently:
>
> https://lore.kernel.org/lkml/202306211346.1e9ff03e-oliver.sang@xxxxxxxxx/
> https://lore.kernel.org/lkml/20230621212403.174710-1-mike.kravetz@xxxxxxxxxx/
Akpm moved to those reverts from the mm-unstable to the mm-stable branch
(not totally sure, but I think those are branch with changes for the
next merge window):
https://lore.kernel.org/all/20230624000036.57829C433CB@xxxxxxxxxxxxxxx/
https://lore.kernel.org/all/20230624000035.3C79BC433C0@xxxxxxxxxxxxxxx/
> * Swapnil Sapkal from AMD found another performance regression; from
> skimming the thread it seems a partial fix is in -next for some time
> already; part of the problem is also caused by df323337e50 "apparmor:
> Use a memory pool instead per-CPU caches" [v5.5-rc1]
>
> https://lore.kernel.org/lkml/09e0f469-a3f7-62ef-75a1-e64cec2dcfc5@xxxxxxx/
> https://lore.kernel.org/lkml/20230515143536.114960-1-mathieu.desnoyers@xxxxxxxxxxxx/
> https://lore.kernel.org/lkml/6c693e3b-b941-9acf-6821-179e7a7fe2b8@xxxxxxxxxxxx/
>
> [this one is not tracked by regzbot, something went sideways, need to
> look closer]
>
>
> * Mario came up with a fix for the ppc/tpm kexec hang recently reported:
>
> https://lore.kernel.org/lkml/99B81401-DB46-49B9-B321-CF832B50CAC3@xxxxxxxxxxxxx/
> https://lore.kernel.org/all/20230623030427.908-1-mario.limonciello@xxxxxxx/
>
>
> * still no solution in sight for the "Dragonboard 845c broken due to
> asynchronous probe" regression
>
> https://lore.kernel.org/lkml/CAMi1Hd1avQDcDQf137m2auz2znov4XL8YGrLZsw5edb-NtRJRw@xxxxxxxxxxxxxx/
> https://lore.kernel.org/lkml/20230515145323.1693044-1-amit.pundir@xxxxxxxxxx/
> https://lore.kernel.org/lkml/20230602161246.1855448-1-amit.pundir@xxxxxxxxxx/
>
>
> * no progress wrt to the MPTCP tests regressions due to 9p changes
>
> https://lore.kernel.org/regressions/855a232a-76d3-7e7b-b2b5-2ebc41bcadd6@xxxxxxxxxxxx/
>
>
> * brand new: transmit queue timeouts with r8169 due to 2ab19de62d6
> ("r8169: remove ASPM restrictions now that ASPM is disabled during NAPI
> poll") [v6.4-rc1]
>
> https://lore.kernel.org/netdev/c3465166-f04d-fcf5-d284-57357abb3f99@xxxxxxxxxx/
>
>
> * just noticed: patches for a "WiFi / LTE won't work at all on trogdor
> devices" regression might still be heading your way:
>
> https://lore.kernel.org/lkml/20230616081440.v2.1.Ie79b5f0ed45739695c9970df121e11d724909157@changeid/
Ciao, Thorsten
---
Hi, this is regzbot, the Linux kernel regression tracking bot.
Currently I'm aware of 6 regressions in linux-mainline. Find the
current status below and the latest on the web:
https://linux-regtracking.leemhuis.info/regzbot/mainline/
Bye bye, hope to see you soon for the next report.
Regzbot (on behalf of Thorsten Leemhuis)
======================================================
current cycle (v6.3.. aka v6.4-rc), culprit identified
======================================================
[ *NEW* ] transmit queue timeout on r8169
-----------------------------------------
https://linux-regtracking.leemhuis.info/regzbot/regression/c3465166-f04d-fcf5-d284-57357abb3f99@xxxxxxxxxx/
https://lore.kernel.org/netdev/c3465166-f04d-fcf5-d284-57357abb3f99@xxxxxxxxxx/
By Tobias Klausmann; 3 days ago; 4 activities, latest 1 days ago.
Introduced in 2ab19de62d67 (v6.4-rc1)
Recent activities from: Tobias Klausmann (2), Thorsten Leemhuis (1),
Bagas Sanjaya (1)
regulator: qcom-rpmh: Dragonboard 845c broken due to asynchronous probe
-----------------------------------------------------------------------
https://linux-regtracking.leemhuis.info/regzbot/regression/CAMi1Hd1avQDcDQf137m2auz2znov4XL8YGrLZsw5edb-NtRJRw@xxxxxxxxxxxxxx/
https://lore.kernel.org/lkml/CAMi1Hd1avQDcDQf137m2auz2znov4XL8YGrLZsw5edb-NtRJRw@xxxxxxxxxxxxxx/
By Amit Pundir; 42 days ago; 52 activities, latest 1 days ago.
Introduced in ad44ac082fd (v6.4-rc1)
Recent activities from: Amit Pundir (2), Thorsten Leemhuis (1), Bjorn
Andersson (1)
One patch associated with this regression:
* Re: [PATCH] regulator: qcom-rpmh: Revert "regulator: qcom-rpmh: Use PROBE_FORCE_SYNCHRONOUS"
https://lore.kernel.org/lkml/552345c5-b1e9-41f6-f275-b6eeeb51df25@xxxxxxxxxx/
42 days ago, by Caleb Connolly
Noteworthy links:
* [PATCH] Revert "regulator: qcom-rpmh: Revert "regulator: qcom-rpmh: Use PROBE_FORCE_SYNCHRONOUS""
https://lore.kernel.org/lkml/20230515145323.1693044-1-amit.pundir@xxxxxxxxxx/
40 days ago, by Amit Pundir; thread monitored.
* [PATCH] arm64: dts: qcom: sdm845-db845c: Move LVS regulator nodes up
https://lore.kernel.org/lkml/20230602161246.1855448-1-amit.pundir@xxxxxxxxxx/
22 days ago, by Amit Pundir; thread monitored.
9p: MPTCP tests regressions due to new 9p features in v6.4
----------------------------------------------------------
https://linux-regtracking.leemhuis.info/regzbot/regression/855a232a-76d3-7e7b-b2b5-2ebc41bcadd6@xxxxxxxxxxxx/
https://lore.kernel.org/regressions/855a232a-76d3-7e7b-b2b5-2ebc41bcadd6@xxxxxxxxxxxx/
By Matthieu Baerts; 18 days ago; 6 activities, latest 1 days ago.
Introduced in d9bc0d11e33b (v6.4-rc1)
Recent activities from: Thorsten Leemhuis (1)
tpm/ppc: crash during a kexec
-----------------------------
https://linux-regtracking.leemhuis.info/regzbot/regression/99B81401-DB46-49B9-B321-CF832B50CAC3@xxxxxxxxxxxxx/
https://lore.kernel.org/lkml/99B81401-DB46-49B9-B321-CF832B50CAC3@xxxxxxxxxxxxx/
By Sachin Sant; 10 days ago; 10 activities, latest 1 days ago.
Introduced in bd8621ca1510 (v6.4-rc1)
Recent activities from: Limonciello, Mario (2), Thorsten Leemhuis (1),
Greg KH (1), Mario Limonciello (1), Sachin Sant (1), Michael
Ellerman (1)
2 patch postings are associated with this regression, the latest is this:
* Re: [6.4-rc6] Crash during a kexec operation (tpm_amd_is_rng_defective)
https://lore.kernel.org/lkml/3f5a37f1-70e0-3dcf-3bd3-acc8a04e53ee@xxxxxxx/
2 days ago, by Limonciello, Mario
Noteworthy links:
* [PATCH] tpm: Ensure that tpm chip has ops to check if it's defective
https://lore.kernel.org/lkml/20230623030427.908-1-mario.limonciello@xxxxxxx/
2 days ago, by Mario Limonciello; thread monitored.
[ *NEW* ] mm/page cache: performance regression (2 reverts pending)
-------------------------------------------------------------------
https://linux-regtracking.leemhuis.info/regzbot/regression/202306211346.1e9ff03e-oliver.sang@xxxxxxxxx/
https://lore.kernel.org/lkml/202306211346.1e9ff03e-oliver.sang@xxxxxxxxx/
By kernel test robot; 4 days ago; 14 activities, latest 1 days ago.
Introduced in 9425c591e06a (v6.4-rc7)
Fix incoming:
* Revert "page cache: fix page_cache_next/prev_miss off by one"
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=master&id=555d5a30e7b3020e6c4137ce7da8cd028c67caea
=======================================================
on back burner, but with activity since the last report
=======================================================
mm: performance regression on shell-heavy workloads
---------------------------------------------------
https://linux-regtracking.leemhuis.info/regzbot/regression/20230608111408.s2minsenlcjow7q3@quack3/
https://lore.kernel.org/lkml/20230608111408.s2minsenlcjow7q3@quack3/
By Jan Kara; 17 days ago; 7 activities, latest 6 days ago.
Introduced in f1a7941243c (v6.2-rc1)
Recent activities from: Michal Hocko (1)
=============
End of report
=============
All regressions marked '[ *NEW* ]' were added since the previous report,
which can be found here:
https://lore.kernel.org/r/168709975910.455445.13798099813754443067@xxxxxxxxxxxxx
Thanks for your attention, have a nice day!
Regzbot, your hard working Linux kernel regression tracking robot
P.S.: Wanna know more about regzbot or how to use it to track regressions
for your subsystem? Then check out the getting started guide or the
reference documentation:
https://gitlab.com/knurd42/regzbot/-/blob/main/docs/getting_started.md
https://gitlab.com/knurd42/regzbot/-/blob/main/docs/reference.md
The short version: if you see a regression report you want to see
tracked, just send a reply to the report where you Cc
regressions@xxxxxxxxxxxxxxx with a line like this:
#regzbot introduced: v5.13..v5.14-rc1
If you want to fix a tracked regression, just do what is expected
anyway: add a 'Link:' tag with the url to the report, e.g.:
Link: https://lore.kernel.org/all/30th.anniversary.repost@xxxxxxxxxxxxxxxxxx/