Bug#1120831: Acknowledgement (eslinux-image-6.17.8+deb14-amd64: failed command: READ FPDMA QUEUED after boot)
Hi Wolf,
On Sat, Dec 20, 2025 at 08:46:43AM +0000, Wolf wrote:
>
> On Thursday, 18 December 2025 at 21:18, Salvatore Bonaccorso <carnil@debian.org> wrote:
>
> >
>
> >
>
> > Hi Wolf,
> >
>
> > On Wed, Dec 17, 2025 at 10:57:21AM +0000, Wolf wrote:
> >
>
> > > On Wednesday, 17 December 2025 at 11:57, Salvatore Bonaccorso carnil@debian.org wrote:
> > >
>
> > > > Control: tags -1 + moreinfo
> > >
>
> > > > Hi Wolf,
> > >
>
> > > > On Wed, Dec 10, 2025 at 05:12:57PM +0000, Wolf wrote:
> > >
>
> > > > > The error still appears with nvidia blacklisted.
> > >
>
> > > > Ok that is actually "great" now. Given you can reproduce the issue,
> > > > can you do the bisect work between the known good kernel and first bad
> > > > one?
> > >
>
> > > > Do you need instructions on how to do it?
> > >
>
> > > > Regards,
> > > > Salvatore
> > >
>
> > > Hi, Salvatore,
> > >
>
> > > Last good kernel I know is 6.16.12-1, which I'm using now (installed at 2025-10-13).
> > >
>
> > > All others are bad, starting with 6.17.6-1 (from 2025-11-02).
> > >
>
> > > I reported the bug only after I found 6.17.6-1, 6.17.7-1 and 6.17.8-1 bugged.
> >
>
> >
>
> > Is the issue present as well in 6.17.2-1~exp1 which was in
> > experimental? If yes then I suggest to do the following as next steps:
> >
>
> > Check upstream v6.16 directly and v6.17. The procedure can be as
> > follows:
> >
>
> > git clone https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
> > cd linux
> > git checkout v6.16
> > cp /boot/config-$(uname -r) .config
> > yes '' | make localmodconfig
> > make savedefconfig
> > mv defconfig arch/x86/configs/my_defconfig
> >
>
> > # test 6.16 to ensure this is "good"
> > make my_defconfig
> > make -j $(nproc) bindeb-pkg
> > ... install the resulting .deb package and confirm it successfully boots / problem does not exist
> >
>
> > # test 6.17 to ensure this is "bad"
> > git checkout v6.17
> > make my_defconfig
> > make -j $(nproc) bindeb-pkg
> > ... install the resulting .deb package and confirm it fails to boot / problem exists
> >
>
> > With that confirmed, the bisection can start:
> >
>
> > git bisect start
> > git bisect good v6.16
> > git bisect bad v6.17
> >
>
> > In each bisection step git checks out a state between the oldest
> > known-bad and the newest known-good commit. In each step test using:
> >
>
> > make my_defconfig
> > make -j $(nproc) bindeb-pkg
> > ... install, try to boot / verify if problem exists
> >
>
> > and if the problem is hit run:
> >
>
> > git bisect bad
> >
>
> > and if the problem doesn't trigger run:
> >
>
> > git bisect good
> >
>
> > . Please pay attention to always select the just built kernel for
> > booting, it won't always be the default kernel picked up by grub.
> >
>
> > Iterate until git announces to have identified the first bad commit.
> >
>
> > Then provide the output of
> >
>
> > git bisect log
> >
>
> > In the course of the bisection you might have to uninstall previous
> > kernels again to not exhaust the disk space in /boot. Also in the end
> > uninstall all self-built kernels again.
> >
>
> > This prodecure will lead us to a single commit ideally where the
> > problem starts. This would be needed to properly report the issue
> > upstream and together with upstream people understand what the problem
> > is.
> >
>
> > Regards,
> > Salvatore
>
> Hi, Salvatore,
>
> 6.17.2-1~exp1 is not yet available in SID.
No, it was a version in experimental, but those versions are
superseeded already, so the version need to be fetched from the
snapshot.debian.org service:
It can be fetched from https://snapshot.debian.org/package/linux-signed-amd64/6.17.2%2B1~exp1/
But once we have the more close range of versions the next step is the
bisect. I realize this is asking involving a couple of kernel
versions, but the above proceure should make it efficient enough.
Regards,
Salvatore
Reply to: