October 2024 - openSUSE Kernel Bugs - openSUSE Mailing Lists

[Bug 1189456] Add ashmem and binder drivers
by bugzilla_noreply＠suse.com 18 Oct '24

18 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1189456 https://bugzilla.suse.com/show_bug.cgi?id=1189456#c14 --- Comment #14 from Takashi Iwai <tiwai(a)suse.com> --- Not yet. Rust for Linux requires CONFIG_MODVERSION to be disabled, but it's a mandatory feature for (open)SUSE kernels, hence R4L can't be enabled. Once after it's addressed in the upstream, we might consider enabling Rust for Linux for Tumbleweed. -- You are receiving this mail because: You are on the CC list for the bug. You are the assignee for the bug.

1 0

[Bug 1189456] Add ashmem and binder drivers
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1189456 https://bugzilla.suse.com/show_bug.cgi?id=1189456#c13 llyyr <llyyr.public(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |llyyr.public(a)gmail.com --- Comment #13 from llyyr <llyyr.public(a)gmail.com> --- ashmem is not required anymore, since memfd exists now. And there's ongoing effort on a rust_binder driver to replace the old insecure binder driver. Will TW enable the rust_binder driver in the default kernel? -- You are receiving this mail because: You are the assignee for the bug. You are on the CC list for the bug.

1 0

[Bug 1229999] High idle power usage on Lenovo Slim 7 16" AMD
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1229999 https://bugzilla.suse.com/show_bug.cgi?id=1229999#c11 --- Comment #11 from varval761 <variable_valuables761(a)simplelogin.com> --- I tested a few other distros, including Kubuntu and previous versions of Fedora and Tumbleweed, and they all had similar minimum idle power draw as the latest Tumbleweed/Krypton (25W+). So it seems so far that only recent Fedora versions have lower power draw (~14W). Does this maybe have any relevant info? https://fedoraproject.org/wiki/Changes/TunedAsTheDefaultPowerProfileManagem… (Fedora had less power draw even when replacing tuned with p-p-d or tlp though, in my previously mentioned tests) -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1231465] Kernel 6.11.2-1 only a fraction of ram is available
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1231465 https://bugzilla.suse.com/show_bug.cgi?id=1231465#c33 --- Comment #33 from Jiri Slaby <jslaby(a)suse.com> --- (In reply to Jiri Slaby from comment #32) > (In reply to Felix Niederwanger from comment #31) > > I forgot the KASAN sequence: > > > > > (poweroff) > > > 6.11.3-5-KASAN: error > > > 6.11.3-5-KASAN: error > > > 6.11.3-5-KASAN: error > > > > This kernel fails to boot from cold. FWIW I enabled kasan in my kernel, so that it contains both the workaround and KASAN: home:jirislaby:kernel-bsc1231465 It might be worth trying too. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1231465] Kernel 6.11.2-1 only a fraction of ram is available
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1231465 https://bugzilla.suse.com/show_bug.cgi?id=1231465#c32 Jiri Slaby <jslaby(a)suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(felix.niederwange | |r(a)suse.com) --- Comment #32 from Jiri Slaby <jslaby(a)suse.com> --- (In reply to Felix Niederwanger from comment #31) > I forgot the KASAN sequence: > > > (poweroff) > > 6.11.3-5-KASAN: error > > 6.11.3-5-KASAN: error > > 6.11.3-5-KASAN: error > > This kernel fails to boot from cold. It might die well due to the bug. Or not. Did you boot with "quiet" in cmdline? Can you see any messages if not? Or any chance it managed to store the kernel messages into the journal? Or would it be possible to boot with init=/bin/bash and look into the dmesg directly (I am not sure if it's not too late -- perhaps boot with rd.break, but that might be too late too)? But you at least confirmed 6.11.0 is ok and 6.11.2 is not. What about 6.12? And I've just pushed 6.11.4 -- still the same? Sorry, but I am not clever from this bug at all. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1225352] [Build 13.199] openQA test fails in prepare_firstboot: RPi3 not booting?
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1225352 https://bugzilla.suse.com/show_bug.cgi?id=1225352#c48 Fabian Vogt <fvogt(a)suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|FIXED |--- --- Comment #48 from Fabian Vogt <fvogt(a)suse.com> --- (In reply to Guillaume GARDET from comment #47) > (In reply to Patrik Jakobsson from comment #46) > > Why is the test still failing? Shouldn't QA have the fix by now? > > There is a regression, since the test passed 9 days ago: > https://openqa.opensuse.org/tests/4543603 Also affecting 15.5 and 15.6 again. Can you check the serial log? Maybe it's the same cause again, maybe something else. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1231465] Kernel 6.11.2-1 only a fraction of ram is available
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1231465 https://bugzilla.suse.com/show_bug.cgi?id=1231465#c29 --- Comment #29 from Vlastimil Babka <vbabka(a)suse.com> --- Some parts of the following comment don't match the current theory for me. Is it because of imprecise wording or is something else happening? (In reply to Felix Niederwanger from comment #12) > Another observation: kernel 6.12.0-rc2-3.g4f5e31e-default only works on the > first boot. This part is ok. > When the machine is powered down, then also with this kernel the > machine only detects 1.3 GiB of memory. "powered down" would mean another cold boot? That wouldn't match the theory. If there was "rebooted", it would. > I powered the machine on this morning with the same kernel as Friday > (6.12.0-rc2-3.g4f5e31e) and run into the OOM issues. Running into the OOM issue immediately on cold boot doesn't match the theory? > I rolled back to the > 6.11.0-1 kernel and everything was fine. Then I re-installed the current > kernel:HEAD (now 6.12.0-rc3-1gefb99), rebooted and the system was running > fine with 32 GiB. This means here 6.11.0 would not corrupt EFI memory if a reboot from that to 6.12.0 was fine? > Then I powered off the laptop and the same kernel then > detected only 1.3 GiB of memory. I rebooted 3 times with this kernel, always > 1.3 GiB. Again if it's "powered off" the first boot should have been ok? > I used then an older snapshot with the 6.11.0-1 kernel and everything was > fine again (32 GiB memory). From there rebooting into the 6.12.0-rc3 kernel > only shows again 1.3 GiB of memory. But here it seems 6.11.0 is corrupting the EFI memory, so it's a different result than above? Maybe the corruption is not completely deterministic? -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1231465] Kernel 6.11.2-1 only a fraction of ram is available
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1231465 https://bugzilla.suse.com/show_bug.cgi?id=1231465#c28 --- Comment #28 from Vlastimil Babka <vbabka(a)suse.com> --- (In reply to Felix Niederwanger from comment #12) > I used then an older snapshot with the 6.11.0-1 kernel and everything was > fine again (32 GiB memory). From there rebooting into the 6.12.0-rc3 kernel > only shows again 1.3 GiB of memory. This suggests 6.11.0 itself already was corrupting the EFI memory? -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1231465] Kernel 6.11.2-1 only a fraction of ram is available
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1231465 https://bugzilla.suse.com/show_bug.cgi?id=1231465#c27 --- Comment #27 from Jiri Slaby <jslaby(a)suse.com> --- (In reply to Jiri Slaby from comment #26) > (In reply to Felix Niederwanger from comment #20) > > To cross-check I needed 2 reboot attempts with the 6.11.2 kernel to end up > > with only 1.3 GiB. > > Side note: you have to boot each tested kernel twice to actually *see* the > behavior. The first boot corrupts the EFI memory, the second one loads the > memory. And cold-boot/shutdown in between to reset to the original BIOS/EFI state. > Having that in mind, you still confirm 6.12 is fine, 6.11 is fine, but > 6.11.2 is not? -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1231465] Kernel 6.11.2-1 only a fraction of ram is available
by bugzilla_noreply＠suse.com 17 Oct '24

17 Oct '24

https://bugzilla.suse.com/show_bug.cgi?id=1231465 https://bugzilla.suse.com/show_bug.cgi?id=1231465#c26 --- Comment #26 from Jiri Slaby <jslaby(a)suse.com> --- (In reply to Felix Niederwanger from comment #20) > To cross-check I needed 2 reboot attempts with the 6.11.2 kernel to end up > with only 1.3 GiB. Side note: you have to boot each tested kernel twice to actually *see* the behavior. The first boot corrupts the EFI memory, the second one loads the memory. Having that in mind, you still confirm 6.12 is fine, 6.11 is fine, but 6.11.2 is not? -- You are receiving this mail because: You are the assignee for the bug.

1 0