http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c1 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |josua.m@t-online.de, | |tiwai@suse.com Flags| |needinfo?(josua.m@t-online. | |de) --- Comment #1 from Takashi Iwai --- The attachment missing? Just to be sure, could you test with the kernel in OBS Kernel:openSUSE-15.1 repo? The kernel bundled in the image is still old. I don't believe it'll fix much, but better to confirm at first. If the problem persists with the latest Leap 15.1 kernel, please check the recent upstream kernel in OBS Kernel:stable repo. Or you can try even 5.0 kernel in OBS Kernel:HEAD repo. If this works better, there is still some hope to get it fixed. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c3 Josua Mayer changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(josua.m@t-online. | |de) | --- Comment #3 from Josua Mayer --- Created attachment 797640 --> http://bugzilla.opensuse.org/attachment.cgi?id=797640&action=edit dmesg.log with 4.12.14-lp151.6.g360bd3f-default from Kernel:openSUSE-15.1/standard oops, you are right :facepalm: So here attached is me booting the apparently latest kernel from Kernel:openSUSE-15.1/standard. Observations: - no cpu lockup so far - DRI_PRIME=1 glxinfo doesn't lock up - DRI_PRIME=1 glxinfo only reports intel gpu as renderer - dmesg still has call traces from nouveau -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c5 --- Comment #5 from Takashi Iwai --- OK, that's good to hear. Yes, nouveau spews too much scaring errors. I'm going to suppress the kernel stack trace in that timeout error, which is simply annoying. The problem seems like that nouveau driver can't deal with the runtime PM properly for the hybrid GPU (maybe only some models). In the earlier code, it tries to get from the loop via timeout reading the GPU timer register, but it didn't work because the GPU itself stalled so the GPU timer wasn't updated. That said, the stall should have been worked around. But it's still more or less broken. I'm still interested in whether this problem appears with the latest upstream. I guess yes. Could you give it a try? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c7 --- Comment #7 from Josua Mayer --- Created attachment 797922 --> http://bugzilla.opensuse.org/attachment.cgi?id=797922&action=edit dmes.log with 5.0.0-rc8-1.g4ddf057-vanilla - from Kernel:HEAD/standard Also tried the last update 5.0.0-rc8-1.g4ddf057-vanilla Sadly this version has an additional breakage over 5.0-rc7 :( DRI_PRIME=1 glxinfo <-- hangs -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c30 --- Comment #30 from Josua Mayer --- I haven't seen the issue of this report with 4.12.13-lp151.25-default yet. However I have a new problem: Waking up from suspend to RAM :( - choose suspend in menu - display turns off, power led turns off, and flashes in slow interval - wake up by pressing a key - power led is on again - display never turns on Any way to get logs here? Has this been observed in another report? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c32 --- Comment #32 from Takashi Iwai --- Please check with the upstream 5.1 whether the problem persists. IIRC, your machine got a crash with 5.0.x, so if 5.1 works, it means that there is a fix between them. OTOH, if it still crashes with 5.1, we need to track rather in the upstream bugzilla. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c34 --- Comment #34 from Josua Mayer --- Created attachment 804524 --> http://bugzilla.opensuse.org/attachment.cgi?id=804524&action=edit dmesg.log with kernel-default 5.1.0-1.1.ga974d8b -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c36 --- Comment #36 from Takashi Iwai --- It implies that some recent upstream "fix" broke your case. This really needs to be tracked in the upstream. Can we move to the upstream bugzilla? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c38 --- Comment #38 from Josua Mayer --- G'day once more. I can now confirm that things worked as expected with: kernel-default-4.12.14-lp151.25.3.x86_64 I verified this by making a tarball of the kernel files from the snapshot before updating. So works in .3 breaks in kernel-default-4.12.14-lp151.25.4.x86_64 Can we get a diff of these for the upstream report? Or maybe just for my own exepriments? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c40 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(josua.m@t-online. | |de) --- Comment #40 from Takashi Iwai --- Could you get the GIT ID of both the working and the broken kernels? e.g. check via % rpm -qi kernel-default-4.12.14-lp151.25.3 | grep '^GIT' and % rpm -qi kernel-default-4.12.14-lp151.25.4 | grep '^GIT' I guess both are same. And if that's true, it means that the problem isn't about the kernel code but something else. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c42 --- Comment #42 from Josua Mayer --- Supplementary: This version information confused me so much that I had to retry! The thing that I tarred up as 4.12.14-lp151.25.4 reports in uname: Linux linux-aflp 4.12.14-lp151.25-default #1 SMP Sat Apr 27 12:09:14 UTC 2019 (bb82b7a) x86_64 x86_64 x86_64 GNU/Linux And, it has the [ 24.176889] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 6013d4 [ TIMEOUT ] problem. Now the one I tared up from the snapshot as 4.12.14-lp151.25.3 reports in uname: Linux linux-aflp 4.12.14-lp151.25-default #1 SMP Sat Apr 27 12:09:14 UTC 2019 (bb82b7a) x86_64 x86_64 x86_64 GNU/Linux And it does *not* have the problem! What is going on here? This is the same? Really? I did make sure to rerun dracut -f --kver 4.12.14-lp151.25-default after unpacking the respecive tarball, and before rebooting. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c44 Josua Mayer changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(josua.m@t-online. | |de) | --- Comment #44 from Josua Mayer --- Sorted out my version issue. Final report: 4.12.14-lp151.25.3 works fine 4.12.14-lp151.25.4 breaks nouveau 4.12.14-lp151.25.4 works fine WITH MODULES FROM 4.12.14-lp151.25.3 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c46 --- Comment #46 from Josua Mayer --- I was making a tarball from the output of rpm -q --list - passing the output to xrags, then tar. I think it is time for a massive facepalm, because: the tarball is missing lotso f modules, including nouveau.ko. After the leaves had settled I did a cp -r /.snapshots/218/snapshot/lib/modules/4.12.14-lp151.25-default /lib/modules/ cp /.snapshots/218/snapshot/boot/vmlinuz-4.12.14-lp151.25-default /boot/ dracut -f --kver 4.12.14-lp151.25-default Rebooted, and all good. rpm -r /.snapshots/218/snapshot -qf /boot/vmlinuz-4.12.14-lp151.25-default kernel-default-4.12.14-lp151.25.4.x86_64 Meanwhile on the latest 15.1 version in the repos: kernel-default-4.12.14-lp151.26.4.x86_64 the glxinfo test succeeds; and even Minecraft is playable with DRI_PRIME=1 and the nouveau driver! That means we were hunting a ghost between 25.3 and 25.4 (snapshots 217 and 218) :( I should take a break now till I remember the proper use of tar. Also, I don't know where the ghost went. Cold versus warm boot? Do I need to power it off for a while? Who knows. I will not touch the current system, and just turn it back on tomorrow to see if it still works. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 Andres Nogueiras changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |anogueiras@yahoo.es -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c51 --- Comment #51 from Andres Nogueiras --- Created attachment 818774 --> http://bugzilla.opensuse.org/attachment.cgi?id=818774&action=edit X log 0 from actual session -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c53 --- Comment #53 from Andres Nogueiras --- Created attachment 818776 --> http://bugzilla.opensuse.org/attachment.cgi?id=818776&action=edit Xorg log 1 from crashed session -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1126480 http://bugzilla.opensuse.org/show_bug.cgi?id=1126480#c55 Andres Nogueiras changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P4 - Low --- Comment #55 from Andres Nogueiras --- I just run through similar issues with my laptop during the last month, as I moved from 15.0 to 15.1 The first symptom is that at boot it takes longer than the previous distribution, due to nouveau driver trowing messages. Then, once the X session starts, if wanted to change to a console, it takes around 15 or 20 seconds. Going back to X is fast. And when leaving the X session (closing) it hangs and do not let me switch to tty. I have to remotely log in, and the only thing I can do is poweroff the laptop. For this takes around 5 to 10 minutes. -- You are receiving this mail because: You are on the CC list for the bug.