Hey Folks,
As most of you might know, Kernel 6.7 Test Week is currently happening[0] I have figured out that, I can't login to FAS from the Kernel Test Result site[1]. If you are also someone who is stuck, please follow the infra issue filed here[2]. And please submit the results here[3]
If you have still not participated and have some free time, we urge you to take part in testing!
We are also sorry for the inconvenience caused due to login issues. Thanks for testing :)
[0] http://fedoraproject.org/wiki/Test_Day:2024-01-21_Kernel_6.7_Test_Week [1] https://apps.fedoraproject.org/kerneltest/ [2] https://pagure.io/fedora-infrastructure/issue/11744 [3] https://testdays.fedoraproject.org/events/173
Kernel-6.7.3 still has debugging enabled which will break the nvidia driver
https://bugzilla.rpmfusion.org/show_bug.cgi?id=6859
https://forums.developer.nvidia.com/t/linux-6-7-beta-550-40-07-error-modpost...
On Thu, Feb 1, 2024 at 1:45 AM Leigh Scott leigh123linux@gmail.com wrote:
Kernel-6.7.3 still has debugging enabled which will break the nvidia driver
This is not accurate. The 6.7.3 (and 6.6.15) updates did break the nvidia driver. It has nothing to do with debugging being enabled. A proper and valid bugfix:
5ec8e8ea8b7783fab150cf86404fc38cb4db8800 mm/sparsemem: fix race in accessing memory_section->usage
added calls to rcu_read_lock() and rcu_read_unlock() which use symbols that are exported GPL only. This was not some attempt to stab at nvidia, but their non-GPL module is failing as a result because that bug fix now pulls those calls into their driver build. Some people have suggested removing the EXPORT_SYMBOL_GPL on these symbols and leaving them as regular exports. That is not a valid fix for Fedora to carry for many reasons. I will leave 6.7.3 in updates testing for a while to give nvidia some time to come up with a work around, but at some point things have to move forward, and I will not hold non nvidia users bugfixes hostage because of an out of tree driver, this has always been Fedora's policy.
Justin
https://bugzilla.rpmfusion.org/show_bug.cgi?id=6859
https://forums.developer.nvidia.com/t/linux-6-7-beta-550-40-07-error-modpost...
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Thu, Feb 1, 2024 at 1:45 AM Leigh Scott <leigh123linux(a)gmail.com> wrote:
This is not accurate. The 6.7.3 (and 6.6.15) updates did break the nvidia driver. It has nothing to do with debugging being enabled. A proper and valid bugfix:
5ec8e8ea8b7783fab150cf86404fc38cb4db8800 mm/sparsemem: fix race in accessing memory_section->usage
added calls to rcu_read_lock() and rcu_read_unlock() which use symbols that are exported GPL only. This was not some attempt to stab at nvidia, but their non-GPL module is failing as a result because that bug fix now pulls those calls into their driver build. Some people have suggested removing the EXPORT_SYMBOL_GPL on these symbols and leaving them as regular exports. That is not a valid fix for Fedora to carry for many reasons. I will leave 6.7.3 in updates testing for a while to give nvidia some time to come up with a work around, but at some point things have to move forward, and I will not hold non nvidia users bugfixes hostage because of an out of tree driver, this has always been Fedora's policy.
Justin
https://forums.developer.nvidia.com/t/linux-6-7-beta-550-40-07-error-modp...
-- _______________________________________________ devel mailing list -- devel(a)lists.fedoraproject.org To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org
Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Thanks for holding the push to testing, I have managed to patch nvidia 545.xx and 550.xx so
On 2/2/24 10:48, Leigh Scott wrote:
Thanks for holding the push to testing, I have managed to patch nvidia 545.xx and 550.xx so
470.xx is also patched.
I think it is ok to push the new kernel to testing, the legacy nvidia drivers shouldn't hold up the new kernel.
Thank you for doing this fix so quickly, it is greatly appreciated.
-- Ian Laurie FAS: nixuser | IRC: nixuser TZ: Australia/Sydney
On Thu, Feb 1, 2024 at 5:49 PM Leigh Scott leigh123linux@gmail.com wrote:
Thanks for holding the push to testing, I have managed to patch nvidia 545.xx and 550.xx so
470.xx is also patched.
I think it is ok to push the new kernel to testing, the legacy nvidia drivers shouldn't hold up the new kernel.
Much appreciated. It has already been noted that the updated packages work.
-- _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue