Bugzilla – Bug 1208069
Leap 15.5 & Gecko rolling plasma does not revive video from suspend?
Last modified: 2023-04-01 16:54:46 UTC
Having a recurring issue with Leap 15.5 not reviving from suspend on '12 Mac Pro, and the issue was also happening in another installed system, Gecko rolling Plasma. Leap 15.5 will suspend the machine and clicking on shift key will again start the "engine" and the display light turns blue, but the screen stays black. In the case of the Gecko install clicking suspend the machine makes some funny noise like it is "trying" to suspend, but the machine keeps running, the screen goes black. In one of the installs I got the display to come back up, but mousing around and clicking on the GUI does nothing . . . I think that was the Leap 15.5. I was switching systems to see if it was a "hardware" issue, but when I got to my Manjaro install it's fine; I can suspend and wake from suspend. I'm typing this in Manjaro. A forum request has been started on opensuse, so far the only suggestion in it has been to "file a bug report on bugzilla." The bugzilla server showed me another similar bug report from 2017 on a Plasma system; didn't have time today to read through it . . . just making the report here. https://bugzilla.opensuse.org/show_bug.cgi?id=1047729
A few more of the gory details are posted in the forum thread: https://forums.opensuse.org/t/leap-15-5-not-reviving-from-suspend-no-packman-in-the-repos/164258/2
Could you retest with the kernel in OBS Kernel:SLE15-SP5 repo? If the bug persists, please give the hwinfo output.
(In reply to Takashi Iwai from comment #2) > Could you retest with the kernel in OBS Kernel:SLE15-SP5 repo? If the bug > persists, please give the hwinfo output. @Takashi: Happy to do that, but jammed today. The other problem is not knowing how to do what you are suggesting . . . ??? Add "SLE15-SP5" to the repos?? Search "OBS kernel:SLE15-SP5" in YaST? and then install it? For the most part I have a bunch of linux installs that I maintain, but don't "get under the hood" with them, except the fix them . . . which lately TW and Leap 15.5 have been requesting more of my attention that I have to spare . . . . The "third" confirmation is that I had the same problem with another openSUSE based system yesterday, a Gecko rolling MATE iteration had the same problem as the other two, failure to revive from suspending, display stays black, etc. Historically, when these "failure to suspend or revive" issues showed up it was an "nvidia" issue . . . so years ago I stopped using proprietary drivers. But recently in TW there was a "G06" package that somehow insinuated itself into the system, wiping out video in older cards like my GTX780. After that issue I scanned the upgrade packages looking for "G06" but didn't see any in the Leap or Gecko zypper listings . . . but possibly some aspects of "nvidia-ness" have been adapted by the system??
Go to the download URL for the OBS repo, and download kernel-default.rpm, kernel-default-extra.rpm and kernel-default-optional.rpm for the architecture (x86_64): http://download.opensuse.org/repositories/Kernel:/SLE15-SP5/pool/ Then install them via zypper install (as root): % zypper in kernel-default-*.rpm But, it'd be safer to increase the number of installable kernels beforehand. Edit /etc/zypp/zypp.conf (as root), and modify the line defining "multiversion.kernels" to add more entries, e.g. multiversion.kernels = latest,latest-1,latest-2,latest-3,running This will keep up to 4 kernels without purge.
@Takashi: So I managed to get the suggested kernel packages installed and suspended the machine. 10 mins later I attempted to revive from suspend, the display did power up and show the GUI, but as before, GUI was "frozen" and I could not use the mouse. So, in that sense the newer kernel is "worse" than the older one, where I could at least mouse around, but clicking on stuff failed to respond--here I couldn't mouse. I rebooted and used grub menu to select the previous kernel and it booted to TTY "emergency" mode with error, "unable to open display."
I just found a bug report I filed on Leap 15.5 GUI freezing back in Nov '22, doesn't seem like it was "solved" via the bug report, but it might show "line of inquiry" . . . other similarity is the "XFCE" component. https://bugzilla.suse.com/show_bug.cgi?id=1205452
Could you give hwinfo output, and the dmesg output before the suspend? Also, check journal outputs if it got some kernel Oops or crash. Last but not least, please check with the kernel from OBS Kernel:stable:Backport repo (only kernel-default.rpm) for the latest 6.1.x kernel, and verify whether the problem persists with it or not.
(In reply to Takashi Iwai from comment #7) > Could you give hwinfo output, and the dmesg output before the suspend? > Also, check journal outputs if it got some kernel Oops or crash. > > Last but not least, please check with the kernel from OBS > Kernel:stable:Backport repo (only kernel-default.rpm) for the latest 6.1.x > kernel, and verify whether the problem persists with it or not. @Takashi: Sorry, I missed that request for the "hwinfo" output from the previous, "if the problem persists" . . . . It'll be a day or so before I rotate back to the Leap 15.5 system. By the "latest kernel" I'm assuming you mean from the same site as before, which was for "xxxxx.80.rpm" so you would want something newer than that, like "xxxx.81"??? Or, this is a completely other site for stable kernels??? Lastly, I can report that in running a Gecko Plasma zypper upgrade yesterday, that it seemed like zypper worked it's way past the problem and I could suspend and revive the system . . . wasn't super-zippy, but it did return to full function of the GUI.
(In reply to Takashi Iwai from comment #7) > Could you give hwinfo output, and the dmesg output before the suspend? > Also, check journal outputs if it got some kernel Oops or crash. > > Last but not least, please check with the kernel from OBS > Kernel:stable:Backport repo (only kernel-default.rpm) for the latest 6.1.x > kernel, and verify whether the problem persists with it or not. @Takashi Had a few spare moments to try to find this "OBS Kernel" data, but it seemed to lead to a list of links, none of which said "6.1" . . . so I would need a direct link to whatever site you are suggesting that I check into to get a newer kernel going in Leap 15.5 . . . !!???
(In reply to Fritz Hudnut from comment #9) > (In reply to Takashi Iwai from comment #7) > > Could you give hwinfo output, and the dmesg output before the suspend? > > Also, check journal outputs if it got some kernel Oops or crash. > > > > Last but not least, please check with the kernel from OBS > > Kernel:stable:Backport repo (only kernel-default.rpm) for the latest 6.1.x > > kernel, and verify whether the problem persists with it or not. > > @Takashi > > Had a few spare moments to try to find this "OBS Kernel" data, but it seemed > to lead to a list of links, none of which said "6.1" . . . so I would need a > direct link to whatever site you are suggesting that I check into to get a > newer kernel going in Leap 15.5 . . . !!??? It's available at http://download.opensuse.org/repositories/Kernel:/stable:/Backport/standard/
(In reply to Takashi Iwai from comment #10) > (In reply to Fritz Hudnut from comment #9) > > (In reply to Takashi Iwai from comment #7) > > > Could you give hwinfo output, and the dmesg output before the suspend? > > > Also, check journal outputs if it got some kernel Oops or crash. > > > > > > Last but not least, please check with the kernel from OBS > > > Kernel:stable:Backport repo (only kernel-default.rpm) for the latest 6.1.x > > > kernel, and verify whether the problem persists with it or not. > > > > @Takashi > > > > Had a few spare moments to try to find this "OBS Kernel" data, but it seemed > > to lead to a list of links, none of which said "6.1" . . . so I would need a > > direct link to whatever site you are suggesting that I check into to get a > > newer kernel going in Leap 15.5 . . . !!??? > > It's available at > > http://download.opensuse.org/repositories/Kernel:/stable:/Backport/standard/ @Takashi OK, great, precerate that. I was looking at some site that was titled "OBS" which did not seem to get past 5.14 kernels . . . didn't make sense. So you mean something like "kernel-default-6.1.12-lp154.2.1.g373f017.x86_64.rpm" would be what I would be looking for? And I would need now to remove one of the other kernels first, or in theory I can add one more??
(In reply to Takashi Iwai from comment #10) > (In reply to Fritz Hudnut from comment #9) > > (In reply to Takashi Iwai from comment #7) > > > Could you give hwinfo output, and the dmesg output before the suspend? > > > Also, check journal outputs if it got some kernel Oops or crash. > > > > > > Last but not least, please check with the kernel from OBS > > > Kernel:stable:Backport repo (only kernel-default.rpm) for the latest 6.1.x > > > kernel, and verify whether the problem persists with it or not. > > > > @Takashi > > > > Had a few spare moments to try to find this "OBS Kernel" data, but it seemed > > to lead to a list of links, none of which said "6.1" . . . so I would need a > > direct link to whatever site you are suggesting that I check into to get a > > newer kernel going in Leap 15.5 . . . !!??? > > It's available at > > http://download.opensuse.org/repositories/Kernel:/stable:/Backport/standard/ @Takashi: Alrighty, got the 6.1.12 xxx kernel installed and used SuperGrub2 to boot that selection, and that does seem to have revived from suspend several times and the GUI is working (as I'm typing this) . . . so that would be the preferred condition . . . . After I booted the new kernel Package Updater showed "84 packages+-" to update and when I ran the zypper dup -l it showed "2 problems" relating to, "do I want to move the 5.14 kernels from OBS to openSUSE??" Couldn't make the decision right at that moment, like, I don't need that kernel because it didn't fix the problem, so it isn't too critical . . . but, then it might be one of the back ups??? Anyway, I'll monitor the operation of Leap 15.5 for a bit, but it does seem like the 6.1.12 Backports kernel addresses the problem. Thanks kindly for the help on it.
@Takashi: So after I installed the 6.1.12 backports kernel I ran update-bootloader in the TW system that controls grub and on reboot I checked uname -r and it shows: "uname -r 5.14.21-150500.39-default" So that isn't the "39.2" that was shown when I ran the zypper dup -l while I was in Leap, to see if that would support the new kernel, the 6.1.12 item that seemed to fix the problem. Is that just a "temporary" condition to check to see if it would fix the issue, but now something else would have to be done to make it permanent?? Or now we have too many kernels installed to let the 6.1 option bubble up?? Forgot that I didn't shut down after I ran the update-bootloader, which might reset grub to where it was . . . don't believe that this 5.14.21 kernel will work to revive the system.
Yes. I can confirm that the 5.14.21-150500.39-default kernel that the internal grub is loading does not work to revive from suspend. The question is, I ran the "update-bootloader" command after the 6.1.12 kernel was installed, but grub doesn't seem to recognize it, nor does it seem to recognize the previous 5.14.21 xxxxx .80 kernel either. I have to use the ext Super Grub2 drive option to select the 6.1.12 option. So far it is only the newest 6.1.12 kernel that revives the machine from suspend, so the question remains, how to get that kernel to be selected as "running"??? Do I have to remove all of the other kernels?? I thought grub would select the newest kernel installed to be the "running" kernel?
I've been still waiting for the output of hwinfo, preferably from both working and non-working systems. This is the minimal requirement for further looking at the problem. Without such logs, one can't know what system you have and what's going on.
@Takashi: Sorry . . . perhaps because I didn't hear back, that I "moved on" . . . . I think the Leap side did get "fixed" . . . but I recall a few days back that the Gecko install continues to "not revive from suspend." I'm jammed right now, but when I get back to the Leap and Gecko partitions I'll try to check out your request and post it here. But, system is '12 Mac Pro with Quad-core i7, 16 GB RAM . . . . Reviving from suspend works fine on the machine running TW, Manjaro, and Debian systems.
Spent a few minutes, while in Debian Sid to try to run "hwinfo" but Sid didn't understand the request. Did a quick google of "hwinfo" . . . possibly it is iinstalled on Leap, but not on Sid? In looking at a video on "hwinfo" it looks like possibly a large amount of information is provided? What specific commands would you like to see the output of in the hwinfo output and/or in dmesg??
Well, the request was the hwinfo program output and the dmesg program output from openSUSE installations.
(In reply to Takashi Iwai from comment #18) > Well, the request was the hwinfo program output and the dmesg program output > from openSUSE installations. Sure. I'm just booted in Sid today, so I was checking the command out there, which didn't seem "applicable" or installed. So I'm going to run #hwinfo or something like #cat hwinfo >hwinfo.log.txt ???? Or, something more "granular" in the hwinfo output?? I don't mess around "under the hood" so I'm looking for specific commands to run to provide the info you are looking for?
Just run "hwinfo > hwinfo.txt" as root or such would be enough. Ditto for dmesg. This is the very starting point to see what hardware components are included. The debug of suspend/resume is often really tough, unfortunately, as it can't be done remotely and it can't leave useful debug info in many cases (because the device itself is suspended when something happens, hence you can't record it). If we are lucky, it's a crash and leaves some kernel Oops, etc.
Created attachment 866010 [details] hwinfo file from Leap 15.5
Created attachment 866011 [details] dmesg file from Leap 15.5
I'm in Leap 15.5 this morning, which is now suspending and reviving OK. Tomorrow might be the Gecko rolling day, that system does not revive from suspend.
Created attachment 866048 [details] hwinfo file from Gecko rolling, same machine
Created attachment 866049 [details] dmesg file from Gecko rolling on same machine
So, today booted in Gecko MATE rolling, and this is the OS that is not reviving from suspend, tested that out again today. And it appears that I did also upgrade the kernel to 6.2 lp xxxx and that did not seem to help out with the issue. Ran a zypper and it hit a dependency on nvidia G06 xxxxx . . . which previously created problems in TW, and there I locked it out of the rotation. Possibly missed it over here in Gecko?? [CODE] uname -r 6.2.4-lp154.2.g0532a55-default [/CODE]