This bug is missing log files that will aid in diagnosing the problem.
While running an Ubuntu kernel (not a mainline or third-party kernel)
please enter the following command in a terminal window:
apport-collect 2016706
and then change the status of the bug to 'Confirmed'.
If, due to the nature of the issue you have encountered, you are unable
to run this command, please add a comment stating that fact and change
the bug status to 'Confirmed'.
This change has been made by an automated script, maintained by the
Ubuntu Kernel Team.
** Changed in: linux (Ubuntu)
Status: New => Incomplete
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/2016706
Title:
The 5.19 kernel backport is broken
Status in linux package in Ubuntu:
Incomplete
Bug description:
The latest 5.19 backport which migrated from 22.10 to 22.04 is broken on
arm64. The kernel team decided to include the following change:
5e0137612430 ("video/aperture: Disable and unregister sysfb devices via
aperture helpers")
it's the change 89314ff239e1933357419fa91b20190150f114a8 in the kinetic
kernel repo.
This change was part of a much larger series that was making sure that
fb devices were properly releasing the pci resources which is
necessary for proper drivers to load. Because the backport misses all
those other changes it breaks PCI resource release for efifb and
prevents specific drivers (e.g. vmwgfx) from being loaded. It doesn't
affect many other arm64 systems because most arm64 system don't have
dedicated PCI gpu's, like vmwgfx is, which is presumably why it was
missed.
Without the rest of the changes in that series efifb won't release BAR 2,
which results in :
vmwgfx 0000:00:0f.0 BAR 2: can't reserve [mem 0x7000000-0x77ffffff 64bit pref]
and a black screen on boot. Because the way drm drivers work is that first
they unload the legacy fb drivers, in this case efifb, and then they continue
their own initialization. In this case vmwgfx can't initialize because even
though efifb has been unloaded it never released the resource at BAR 2, so
vmwgfx can't claim it. Currently there's no mechanism in the kernel to reload
the original generic fb driver in case the specific gpu driver fails the
initialization. This results in black screen on booth with vmware hypervisor on
arm64 and Ubuntu's 5.19 kernel.
To fix it either the above mentioned change needs to be backed out or
the rest of the patches from that series needs to be backported as
well.
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2016706/+subscriptions
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : kernel-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
