Quick regression status pre 7.2 final/-rc8

[Thorsten Leemhuis]

Hi Linus! Just FYI, here is a quick regression status from my side about
regressions introduced this cycle:

Regressions with available fixes that might hit more than a user or two
-----------------------------------------------------------------------

* As it happens the Touchpad in my Thinkpad recently stopped working.
The patch author and Mario concluded that a revert is the best idea for
7.2-final:
https://lore.kernel.org/all/444caa3c-01ae-4516-b63a-576e095fef94@xxxxxxx/

  The revert was recently submitted as "Revert "i2c: designware: defer
probe if child GpioInt controllers are not bound" here:
https://lore.kernel.org/all/20260814100719.9548-1-hardikprakash.official@xxxxxxxxx/

  Andi, the i2c maintainers, picked it up Friday Night, but afaics
hasn't sent it to you or asked you to pull it directly, so I assume he
thinks it's not worth the risk.

* There is also a mt76 know since a few weeks that causes self-deadlocks
with 6 GHz wifi APs. A fix was posted a while ago in
https://lore.kernel.org/all/20260629083543.153564-1-jb.tsai@xxxxxxxxxxxx/
and is in next as 1f83a8378eb26c ("wifi: mt76: mt7921: refactor regd
update to fix recursive mutex deadlock") since next-20260806. But it's
on track to be merged for 7.3-rc1, despite one user and myself asking
the wifi driver and net maintainers to considered it for 7.2 final:
https://lore.kernel.org/all/20260804185118.19705-1-lucid_duck@xxxxxxxxxxxxx/
https://lore.kernel.org/all/7e66ca09-fa70-4a3c-948a-0c46b268655c@xxxxxxxxxxxxx/

  On Saturday another user ran into this and asked for the fix to make
7.2, too [after submitting a second fix for it]:
https://lore.kernel.org/all/20260816135909.52136-1-mikhail.v.gavrilov@xxxxxxxxx/

  And on Friday there was a report that sounded like it might be the
same problem:
https://lore.kernel.org/all/066b30cc-a9e6-4aeb-964d-71551e8ea3ef@xxxxxxxxx/


Regressions that cause big problems for some that have available fixes
----------------------------------------------------------------------

* PCI: imx6: PCI link creation fails on Apalis iMX6, which breaks booting:
  Report:
https://lore.kernel.org/all/20260810211209.GB4821@francesco-nb/
  Bjorn had considered a huge revert on Wednesday, but that wasn't
submitted afaics:
https://lore.kernel.org/lkml/20260811214900.GA763236@bhelgaas/
  Proposed fix from Thursday:
https://lore.kernel.org/lkml/20260813095003.356062-1-hongxing.zhu@xxxxxxxxxxx/


* powerpc/entry: indefinitely warnings and system never completes boot
  Report:
https://lore.kernel.org/lkml/6f9bfb0f-b14c-468e-bb9f-c157d120d0dc@xxxxxxxxxxxxx/
  Proposed fix:
https://lore.kernel.org/lkml/20260812152035.1661781-1-mkchauras@xxxxxxxxx/


Regressions that cause smaller problems
---------------------------------------

* PCI: booting takes a few seconds longer

  Reports:
https://lore.kernel.org/lkml/20260801092152.5643-1-andiwild@xxxxxxxxx/
https://bugzilla.kernel.org/show_bug.cgi?id=221801
  At least a partial fix:
https://lore.kernel.org/lkml/20260801201244.4421-1-andiwild@xxxxxxxxx/


HTH, Ciao, Thorsten