Hi Adrian, On 22.05.23 11:15, John Paul Adrian Glaubitz wrote:
On Mon, 2023-05-22 at 11:06 +0200, Frank Scheiner wrote:And it's definitely better to catch the regression before the release as this way you will get the fix landed for 6.4 instead of 6.5.I know I did that in the past for sparc64 kernels, but I seem to be unable to redo that for ia64 at the moment. I always get:You can just download a complete cross-toolchain for ia64 from here:https://mirrors.edge.kernel.org/pub/tools/crosstool/Just download the ia64 toolchain, extract it somewhere and add the bin sub-directory to your PATH environment variable. That's all you need.
Very useful indeed! Thanks for the pointer.
An update from my side:
In the meantime Linux 6.4-rc3 is out and I gave it some testing on my
rx2620 both with modules compiled in and not compiled in.
As already mentioned off-list, 6.4-rc kernels with modules compiled in
work on the rx2620, interestingly w/o kernel oopses (6.4-rc2 still had
them but otherwise worked). This lead me to assume the problem was gone,
but checking with the modular kernel showed the problem still in effect
on the rx2620.
Checking 6.4-rc1 also already shows the problem, from there I did
quickly scan through the merge commits between 6.3 and 6.4-rc1 and noticed:
```
commit b6a7828502dc769e1a5329027bc5048222fa210a
Merge: d06f5a3f7140 8660484ed1cf
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date: Thu Apr 27 16:36:55 2023 -0700
Merge tag 'modules-6.4-rc1' of
git://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux
Pull module updates from Luis Chamberlain:
[...]
```
I then tested with the next older merge commit
(cec24b8b6bb841a19b5c5555b600a511a8988100) and the suspected one
(b6a7828502dc769e1a5329027bc5048222fa210a) and indeed
cec24b8b6bb841a19b5c5555b600a511a8988100 is good and
b6a7828502dc769e1a5329027bc5048222fa210a is bad.
I'll bisect between those tomorrow to determine the exact breaking commit.
Cheers,
Frank