Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Bug#1029373: kernel panic on Odroid U2 with default environment
12 views
Skip to first unread message
Joost van Zwieten
Jan 21, 2023, 5:30:04 PM1/21/23
to
Package: u-boot-exynos
Version: 2021.01+dfsg-5
Dear maintainers,
Booting a kernel with initrd using U-Boot from bullseye with default
environment results in a kernel panic, because U-Boot loads the initrd
in highmem [1]. The problem has been fixed upstream [2]. Please consider
including the fix in bullseye. I've tested the attached patch, which
applies to v2021.01 and which is a backport of the fix (variable
`CONFIG_SYS_BOOTMAPSZ` has been renamed to `CFG_SYS_BOOTMAPSZ` in
v2023.01).
Best, Joost
[1]: https://lists.denx.de/pipermail/u-boot/2023-January/504469.html
[2]: https://lists.denx.de/pipermail/u-boot/2023-January/505358.html
Version: 2021.01+dfsg-5
Dear maintainers,
Booting a kernel with initrd using U-Boot from bullseye with default
environment results in a kernel panic, because U-Boot loads the initrd
in highmem [1]. The problem has been fixed upstream [2]. Please consider
including the fix in bullseye. I've tested the attached patch, which
applies to v2021.01 and which is a backport of the fix (variable
`CONFIG_SYS_BOOTMAPSZ` has been renamed to `CFG_SYS_BOOTMAPSZ` in
v2023.01).
Best, Joost
[1]: https://lists.denx.de/pipermail/u-boot/2023-January/504469.html
[2]: https://lists.denx.de/pipermail/u-boot/2023-January/505358.html
Vagrant Cascadian
Jan 21, 2023, 9:00:04 PM1/21/23
to
On 2023-01-21, Joost van Zwieten wrote:
> Booting a kernel with initrd using U-Boot from bullseye with default
> environment results in a kernel panic, because U-Boot loads the initrd
> in highmem [1]. The problem has been fixed upstream [2]. Please consider
> including the fix in bullseye. I've tested the attached patch, which
> applies to v2021.01 and which is a backport of the fix (variable
> `CONFIG_SYS_BOOTMAPSZ` has been renamed to `CFG_SYS_BOOTMAPSZ` in
> v2023.01).
Thanks!
> Booting a kernel with initrd using U-Boot from bullseye with default
> environment results in a kernel panic, because U-Boot loads the initrd
> in highmem [1]. The problem has been fixed upstream [2]. Please consider
> including the fix in bullseye. I've tested the attached patch, which
> applies to v2021.01 and which is a backport of the fix (variable
> `CONFIG_SYS_BOOTMAPSZ` has been renamed to `CFG_SYS_BOOTMAPSZ` in
> v2023.01).
Could you please also test against the version in unstable? Generally we
need fixes to go to unstable before backporting them to stable... and
given that bookworm will likely release this year, I would like for the
version in bookworm (currently based on upstream v2023.01) to work as
well...
Looks like the patch got merged upstream:
https://source.denx.de/u-boot/u-boot/-/commit/ace75d642fa71034b248cde30709c79e06c52d61
Would be good to include a link to the upstream fix in the comments of
the patch, and also explain that it is backported with regards to
CFG_SYS_BOOTMAPSZ vs. CONFIG_SYS_BOOTMAPSZ in the patch itself... it was
a little confusing to see the text of patch use one and the patch
describe another...
live well,
vagrant
Joost van Zwieten
Jan 22, 2023, 5:40:03 PM1/22/23
to
On Sat, Jan 21, 2023 at 17:46, Vagrant Cascadian <vag...@debian.org>
wrote:
> On 2023-01-21, Joost van Zwieten wrote:
>> Booting a kernel with initrd using U-Boot from bullseye with default
>> environment results in a kernel panic, because U-Boot loads the
>> initrd
>> in highmem [1]. The problem has been fixed upstream [2]. Please
>> consider
>> including the fix in bullseye. I've tested the attached patch, which
>> applies to v2021.01 and which is a backport of the fix (variable
>> `CONFIG_SYS_BOOTMAPSZ` has been renamed to `CFG_SYS_BOOTMAPSZ` in
>> v2023.01).
>
> Thanks!
>
> Could you please also test against the version in unstable? Generally
> we
> need fixes to go to unstable before backporting them to stable... and
> given that bookworm will likely release this year, I would like for
> the
> version in bookworm (currently based on upstream v2023.01) to work as
> well...
Unfortunately I can't. In v2021.10 and v2023.01 two separate problems
>> Booting a kernel with initrd using U-Boot from bullseye with default
>> environment results in a kernel panic, because U-Boot loads the
>> initrd
>> in highmem [1]. The problem has been fixed upstream [2]. Please
>> consider
>> including the fix in bullseye. I've tested the attached patch, which
>> applies to v2021.01 and which is a backport of the fix (variable
>> `CONFIG_SYS_BOOTMAPSZ` has been renamed to `CFG_SYS_BOOTMAPSZ` in
>> v2023.01).
>
> Thanks!
>
> Could you please also test against the version in unstable? Generally
> we
> need fixes to go to unstable before backporting them to stable... and
> given that bookworm will likely release this year, I would like for
> the
> version in bookworm (currently based on upstream v2023.01) to work as
> well...
were introduced [1], [2]. A fix for one of them is pending. For the
other I don't have an idea how to fix this yet (reverting a commit works
for me, but might break other boards).
[1]: https://lists.denx.de/pipermail/u-boot/2023-January/504116.html
[2]: https://lists.denx.de/pipermail/u-boot/2023-January/504115.html
> Looks like the patch got merged upstream:
>
>
> https://source.denx.de/u-boot/u-boot/-/commit/ace75d642fa71034b248cde30709c79e06c52d61
>
> Would be good to include a link to the upstream fix in the comments of
> the patch, and also explain that it is backported with regards to
> CFG_SYS_BOOTMAPSZ vs. CONFIG_SYS_BOOTMAPSZ in the patch itself... it
> was
> a little confusing to see the text of patch use one and the patch
> describe another...
>
> live well,
> vagrant
0 new messages