Date: Tue, 25 Mar 2025 10:44:32 -0700 From: David Wolfskill <david@catwhisker.org> To: Olivier Certner <olce@freebsd.org> Cc: freebsd-current@freebsd.org, src-committers@freebsd.org, Gleb Smirnoff <glebius@freebsd.org> Subject: Re: March 2025 stabilization week Message-ID: <Z-LrgOuNgTi8dp5Y@albert.catwhisker.org> In-Reply-To: <6881633.Qb2xPn7ZBO@ravel> References: <Z-ERC0xDcivdQ7tD@cell.glebi.us> <6881633.Qb2xPn7ZBO@ravel>
index | next in thread | previous in thread | raw e-mail
[-- Attachment #1 --]
On Tue, Mar 25, 2025 at 10:39:56AM +0100, Olivier Certner wrote:
> ...
> I'd like to recommend cherry picking commit 718d1928f874 on top of the stabweek tag for people using DRM kmod's amdgpu driver. It fixes longstanding bug PR 277476 (GUI freezes after some time, due to memory fragmentation) for DRM kmod 5.15 or higher (5.10 is unaffected).
>
> ...
>
> If you test it, please report about your experience (even if it's just to say "it works", or "I don't see any difference") and your GPU.
>
> Thanks and regards.
>
> --
> Olivier Certner
TL;DR: No issues seen here.
OK; well... not sure if this will help, but:
* I track both stable/14 and head (and update installed ports under
stable/14) on 3 machines daily: one headless build machine & 2 laptops.
* The machines are set up so the same /usr/local is in place whether
they are running stable/14 or head.
* /etc/src.conf is set up so that ports-resident kmods are rebuilt
(under the running OS) each time the kernel is rebuilt.
* I had had some issues in head as of 15 March; "git bisect" had
identified as main-n275968-19df0c5abcb9.
* That identification was a "red herring" -- my laptop "head" kernels
also have "options DEBUG_REDZONE" -- and it took "a while" for
cleverer folks than I to see what the problem really was. Once markj
provided a patch, I tried it, and that seemed to resolve the issue.
* Sunday last, I updated head to main-n276051-29eb4de61a4c-dirty
("-dirty" because of the patch from markj).
* Mark committed (& pushed) main-n276054-74361d693aec on Sunday; I
managed to remember to "git restore sys/kern/kern_malloc.c".
* Monday, I had updated head to main-n276073-67c1c4dfd1cc without issue.
* Today (Tuesday, in my time zone as I write this), I updated head to
main-n276079-718d1928f874 (which is the commit cite above); no issues.
Caveat: I don't tend to run head much longer than it takes to build the
next head & reboot into X11 (in the case of the laptops).
I am using graphics/drm-61-kmod on the laptops.
Peace,
david
--
David H. Wolfskill david@catwhisker.org
"Any unauthorized release of classified information is a violation of
the law and will be treated as such.” Hmmm....
See https://www.catwhisker.org/~david/publickey.gpg for my public key.
[-- Attachment #2 --]
-----BEGIN PGP SIGNATURE-----
iNUEARYKAH0WIQSTLzOSbomIK53fjFliipiWhXYx5QUCZ+LrgF8UgAAAAAAuAChp
c3N1ZXItZnByQG5vdGF0aW9ucy5vcGVucGdwLmZpZnRoaG9yc2VtYW4ubmV0OTMy
RjMzOTI2RTg5ODgyQjlEREY4QzU5NjI4QTk4OTY4NTc2MzFFNQAKCRBiipiWhXYx
5TFgAP4sW6/sOGZbSx9gm3MkzBu4DCVGjCl1/ft888TLsKMajAEA18oQeHgYjQBK
fD2R5m+3mFM44o9a/HgV3pGU8fJGCQY=
=fAwZ
-----END PGP SIGNATURE-----
help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Z-LrgOuNgTi8dp5Y>