Issues with suspend not actually suspending?
Hi all, did/does anyone else experience issues with suspending on Tumbleweed/MicroOS/Kalpa? Normally I can suspend my laptop by using the suspend buttons in Plasma6 or by issuing "systemctl suspend". Now it only works every third or fourth time. I works more reliably if I shut programs like Firefox and Thunderbird, but it still is not bullet-proof 100% working. It shows the look screen, freezes from some (20) seconds and then the login screen is shown again and the mouse responds again. I think it might have started Thursday a week ago. Two laptops, one Tumbleweed and one Kalpa immutable, both with Plasma6. Thinkpad T460s and Thinkpad T14 G2. Both updated almost daily to the latest snapshot. (This started before Plasma 6.1, so that should not be the reason.) Kind Regards, Johannes
On 26. 06. 24, 10:30, Johannes Kastl wrote:
Hi all,
did/does anyone else experience issues with suspending on Tumbleweed/MicroOS/Kalpa?
Normally I can suspend my laptop by using the suspend buttons in Plasma6 or by issuing "systemctl suspend".
Now it only works every third or fourth time. I works more reliably if I shut programs like Firefox and Thunderbird, but it still is not bullet-proof 100% working. It shows the look screen, freezes from some (20) seconds and then the login screen is shown again and the mouse responds again.
I think it might have started Thursday a week ago.
That might be an update to kernel 6.9.5. Could you test with 6.9.4 or 6.9.6? One is in history [1], one in Kernel:stable [2]. [1] http://download.opensuse.org/history/20240619/tumbleweed/repo/oss/x86_64/ker... [2] https://download.opensuse.org/repositories/Kernel:/stable/standard/x86_64/ke... thanks, -- js suse labs
Hi Jiri, On 26.06.24 10:51 Jiri Slaby wrote:
On 26. 06. 24, 10:30, Johannes Kastl wrote:
[ suspend not working ]
That might be an update to kernel 6.9.5.
Could you test with 6.9.4 or 6.9.6?
I have rebooted and used 6.9.4 now, let's see if this makes any difference. If not, I'll check 6.9.6. I'll report back later, need to suspend some times... :-) $ rpm -qa|grep kernel-default kernel-default-6.9.5-1.1.x86_64 kernel-default-6.9.4-1.1.x86_64 $ Kind Regards, Johannes
OP: Are you using nvidia drivers? I have had numerous issues with suspend while nvidia drivers were selected by the system, over my choice for nouveau . . . . Nouveau has less suspend issues in my experience, several machines, all with nvidia cards.
On 2024-06-26 19:18, Fritz Hudnut wrote:
OP:
Are you using nvidia drivers? I have had numerous issues with suspend while nvidia drivers were selected by the system, over my choice for nouveau . . . . Nouveau has less suspend issues in my experience, several machines, all with nvidia cards.
If your preference is Nouveau, something has added the NVidia repository on your back. There is a package that does that addition which you have to remove. Which one, I don't remember, but hopefully someone will remember with those hints. -- Cheers / Saludos, Carlos E. R. (from 15.5 x86_64 at Telcontar)
Carlos E. R. wrote:
If your preference is Nouveau, something has added the NVidia repository on your back. There is a package that does that addition which you have to remove. Which one, I don't remember, but hopefully someone will remember with those hints.
@Carlos: Yes, there is . . . but, not sure if this relates to the OP's problem or not, but in my case it was something like "G06 xxxxxx" that "blacklisted" nouveau and reset the video to nvidia, without my knowledge . . . . It hit a few of my OpenSUSE systems, and I locked it at G05??? It kept recurring. I have a couple of threads on the Help forum . . . . One of them: https://forums.opensuse.org/t/how-to-unblacklist-nouveau-in-leap-15-6/172170...
On 2024-06-26 21:38, Fritz Hudnut wrote:
Carlos E. R. wrote:
If your preference is Nouveau, something has added the NVidia repository on your back. There is a package that does that addition which you have to remove. Which one, I don't remember, but hopefully someone will remember with those hints.
@Carlos:
Yes, there is . . . but, not sure if this relates to the OP's problem or not, but in my case it was something like "G06 xxxxxx" that "blacklisted" nouveau and reset the video to nvidia, without my knowledge . . . .
Yes, I know. There is an unrelated package that will add the NVidia repository behind your back. You have first to uninstall and block for ever that package. Then you have to remove the NVidia repository, and then the NVidia packages. That is what I remember of the issue. I don't remember the name of that package. [...] Found it. Archived-At: <https://lists.opensuse.org/archives/list/users@lists.opensuse.org/message/OOSV4CIJYRPTRY2G45KGTJBOYA5B3IRV/> From: Andrei Borzenkov <arvidjaar@gmail.com> Date: Fri, 20 Oct 2023 09:42:09 +0300 Message-ID: <CAA91j0WqnPXcgKT8MC7hG7cZYBFuzJr-sjfaUYK9qTGX9xK5hg@mail.gmail.com> Subject: Re: Backup files deleted by: "Repository 'repo-oss (15.5)' is managed by service" To: oS-EN <users@lists.opensuse.org> -- Cheers / Saludos, Carlos E. R. (from 15.5 x86_64 at Telcontar)
On 26.06.24 19:18 Fritz Hudnut wrote:
OP:
Are you using nvidia drivers? I have had numerous issues with suspend while nvidia drivers were selected by the system, over my choice for nouveau . . . . Nouveau has less suspend issues in my experience, several machines, all with nvidia cards.
Nope, Thinkpad T460 with Intel graphics. And the T14 I think also uses Intel graphics. Kind Regards, Johannes
On 26.06.24 10:51 Jiri Slaby wrote:
Could you test with 6.9.4 or 6.9.6?
I rebooted and started with 6.9.4: Linux baghira 6.9.4-1-default #1 SMP PREEMPT_DYNAMIC Wed Jun 12 10:44:05 UTC 2024 (ad8659c) x86_64 GNU/Linux The first attempt was not successfull, I needed several tries from within Plasma6 to suspend the laptop. I just rebooted and after login in fresh, just starting the autostart applications, the suspend was working immediately. But I am going to test with 6.9.6 next. Not sure if this has to do with me using U2F devices for unlocking the plasma lock screen, but also without the device present it does not reliably suspend. Is there any special place where I could look for logs? On a short glance I found nothing in the journal, but there is a lot going on, I might have missed it... Kind Regards, Johannes
On 26. 06. 24, 21:40, Johannes Kastl wrote:
On 26.06.24 10:51 Jiri Slaby wrote:
Could you test with 6.9.4 or 6.9.6?
I rebooted and started with 6.9.4:
Linux baghira 6.9.4-1-default #1 SMP PREEMPT_DYNAMIC Wed Jun 12 10:44:05 UTC 2024 (ad8659c) x86_64 GNU/Linux
The first attempt was not successfull, I needed several tries from within Plasma6 to suspend the laptop.
I just rebooted and after login in fresh, just starting the autostart applications, the suspend was working immediately.
But I am going to test with 6.9.6 next.
Not sure if this has to do with me using U2F devices for unlocking the plasma lock screen, but also without the device present it does not reliably suspend.
Is there any special place where I could look for logs? On a short glance I found nothing in the journal, but there is a lot going on, I might have missed it...
I would start with dmesg around "Suspending": dmesg | grep -C 30 Suspending regards, -- js suse labs
On 26.06.24 21:40 Johannes Kastl wrote:
But I am going to test with 6.9.6 next.
Now running with 6.9.6 and I had three successful suspends over the day. Sometimes the Plasma lockscreen is shown before the screen shuts off, sometimes it goes to black directly. But in all three attempts the suspend worked! Hooray! Linux baghira 6.9.6-3.gd344d3a-default #1 SMP PREEMPT_DYNAMIC Tue Jun 25 07:48:26 UTC 2024 (d344d3a) x86_64 GNU/Linux Kind Regards, Johannes
Hi all, On 26.06.24 10:30 Johannes Kastl wrote:
did/does anyone else experience issues with suspending on Tumbleweed/MicroOS/Kalpa?
Normally I can suspend my laptop by using the suspend buttons in Plasma6 or by issuing "systemctl suspend".
Now it only works every third or fourth time. I works more reliably if I shut programs like Firefox and Thunderbird, but it still is not bullet-proof 100% working. It shows the look screen, freezes from some (20) seconds and then the login screen is shown again and the mouse responds again.
I need to warm up that old thread again, as the issue still persist and is getting more and more annoying. Thinkpad T460s, Intel Graphics, Tumbleweed, KDE Plasma, Kernel 6.11.8-1-default I see the following in the logs
[ 180.055854] [ T5267] PM: suspend entry (deep) [ 180.386532] [ T5267] Filesystems sync: 0.330 seconds [ 180.441680] [ T5267] Freezing user space processes [ 180.443615] [ T5267] Freezing user space processes completed (elapsed 0.001 seconds) [ 180.443622] [ T5267] OOM killer disabled. [ 180.443623] [ T5267] Freezing remaining freezable tasks [ 200.450388] [ T5267] Freezing remaining freezable tasks failed after 20.007 seconds (0 tasks refusing to freeze, wq_busy=1): [ 200.450417] [ T5267] Showing freezable workqueues that are still busy: [ 200.450426] [ T5267] workqueue pm: flags=0x4 [ 200.450435] [ T5267] pwq 2: cpus=0 node=0 flags=0x0 nice=0 active=0 refcnt=2 [ 200.450454] [ T5267] inactive: pm_runtime_work [ 200.450474] [ T5267] workqueue btrfs-endio-meta: flags=0xe [ 200.450490] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=2 [ 200.450503] [ T5267] inactive: btrfs_end_bio_work [btrfs] [ 200.450946] [ T5267] workqueue btrfs-endio-write: flags=0xe [ 200.450961] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=19 [ 200.450974] [ T5267] inactive: 18*btrfs_work_helper [btrfs] [ 200.451418] [ T5267] workqueue btrfs-qgroup-rescan: flags=0x2000e [ 200.451432] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=1 refcnt=6 [ 200.451444] [ T5267] in-flight: 42:btrfs_work_helper [btrfs] [ 200.452059] [ T5267] Restarting kernel threads ... done. [ 200.453253] [ T5267] OOM killer enabled. [ 200.453263] [ T5267] Restarting tasks ... done. [ 200.458456] [ T5267] random: crng reseeded on system resumption [ 200.514712] [ T5267] PM: suspend exit [ 200.514794] [ T5267] PM: suspend entry (s2idle) [ 200.599113] [ T5267] Filesystems sync: 0.084 seconds [ 200.599349] [ T5267] Freezing user space processes [ 200.601418] [ T5267] Freezing user space processes completed (elapsed 0.002 seconds) [ 200.601425] [ T5267] OOM killer disabled. [ 200.601427] [ T5267] Freezing remaining freezable tasks [ 220.604678] [ T5267] Freezing remaining freezable tasks failed after 20.003 seconds (0 tasks refusing to freeze, wq_busy=1): [ 220.604708] [ T5267] Showing freezable workqueues that are still busy: [ 220.604721] [ T5267] workqueue btrfs-endio-meta: flags=0xe [ 220.604730] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=2 [ 220.604749] [ T5267] inactive: btrfs_end_bio_work [btrfs] [ 220.605189] [ T5267] workqueue btrfs-endio-write: flags=0xe [ 220.605206] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=55 [ 220.605220] [ T5267] inactive: 54*btrfs_work_helper [btrfs] [ 220.605675] [ T5267] workqueue btrfs-qgroup-rescan: flags=0x2000e [ 220.605690] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=1 refcnt=6 [ 220.605702] [ T5267] in-flight: 42:btrfs_work_helper [btrfs] [ 220.606301] [ T5267] Restarting kernel threads ... done. [ 220.606895] [ T5267] OOM killer enabled. [ 220.606908] [ T5267] Restarting tasks ... done. [ 220.615591] [ T5267] random: crng reseeded on system resumption [ 220.669956] [ T5267] PM: suspend exit
Anyone else having these issues? I'll search bugzilla (again) and open a bugreport, if there is none. Kind Regards, Johannes
On Wed, Nov 20, 2024 at 10:41 AM Johannes Kastl <mail@johannes-kastl.de> wrote:
Hi all,
On 26.06.24 10:30 Johannes Kastl wrote:
did/does anyone else experience issues with suspending on Tumbleweed/MicroOS/Kalpa?
Normally I can suspend my laptop by using the suspend buttons in Plasma6 or by issuing "systemctl suspend".
Now it only works every third or fourth time. I works more reliably if I shut programs like Firefox and Thunderbird, but it still is not bullet-proof 100% working. It shows the look screen, freezes from some (20) seconds and then the login screen is shown again and the mouse responds again.
I need to warm up that old thread again, as the issue still persist and is getting more and more annoying.
Thinkpad T460s, Intel Graphics, Tumbleweed, KDE Plasma, Kernel 6.11.8-1-default
I see the following in the logs
[ 180.055854] [ T5267] PM: suspend entry (deep) [ 180.386532] [ T5267] Filesystems sync: 0.330 seconds [ 180.441680] [ T5267] Freezing user space processes [ 180.443615] [ T5267] Freezing user space processes completed (elapsed 0.001 seconds) [ 180.443622] [ T5267] OOM killer disabled. [ 180.443623] [ T5267] Freezing remaining freezable tasks [ 200.450388] [ T5267] Freezing remaining freezable tasks failed after 20.007 seconds (0 tasks refusing to freeze, wq_busy=1): [ 200.450417] [ T5267] Showing freezable workqueues that are still busy: [ 200.450426] [ T5267] workqueue pm: flags=0x4 [ 200.450435] [ T5267] pwq 2: cpus=0 node=0 flags=0x0 nice=0 active=0 refcnt=2 [ 200.450454] [ T5267] inactive: pm_runtime_work [ 200.450474] [ T5267] workqueue btrfs-endio-meta: flags=0xe [ 200.450490] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=2 [ 200.450503] [ T5267] inactive: btrfs_end_bio_work [btrfs] [ 200.450946] [ T5267] workqueue btrfs-endio-write: flags=0xe [ 200.450961] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=19 [ 200.450974] [ T5267] inactive: 18*btrfs_work_helper [btrfs] [ 200.451418] [ T5267] workqueue btrfs-qgroup-rescan: flags=0x2000e [ 200.451432] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=1 refcnt=6 [ 200.451444] [ T5267] in-flight: 42:btrfs_work_helper [btrfs]
Did you check with "btrfs quota rescan --status" if any scan is in progress? Does it work if you disable quota on btrfs filesytem(s)?
[ 200.452059] [ T5267] Restarting kernel threads ... done. [ 200.453253] [ T5267] OOM killer enabled. [ 200.453263] [ T5267] Restarting tasks ... done. [ 200.458456] [ T5267] random: crng reseeded on system resumption [ 200.514712] [ T5267] PM: suspend exit [ 200.514794] [ T5267] PM: suspend entry (s2idle) [ 200.599113] [ T5267] Filesystems sync: 0.084 seconds [ 200.599349] [ T5267] Freezing user space processes [ 200.601418] [ T5267] Freezing user space processes completed (elapsed 0.002 seconds) [ 200.601425] [ T5267] OOM killer disabled. [ 200.601427] [ T5267] Freezing remaining freezable tasks [ 220.604678] [ T5267] Freezing remaining freezable tasks failed after 20.003 seconds (0 tasks refusing to freeze, wq_busy=1): [ 220.604708] [ T5267] Showing freezable workqueues that are still busy: [ 220.604721] [ T5267] workqueue btrfs-endio-meta: flags=0xe [ 220.604730] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=2 [ 220.604749] [ T5267] inactive: btrfs_end_bio_work [btrfs] [ 220.605189] [ T5267] workqueue btrfs-endio-write: flags=0xe [ 220.605206] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=55 [ 220.605220] [ T5267] inactive: 54*btrfs_work_helper [btrfs] [ 220.605675] [ T5267] workqueue btrfs-qgroup-rescan: flags=0x2000e [ 220.605690] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=1 refcnt=6 [ 220.605702] [ T5267] in-flight: 42:btrfs_work_helper [btrfs] [ 220.606301] [ T5267] Restarting kernel threads ... done. [ 220.606895] [ T5267] OOM killer enabled. [ 220.606908] [ T5267] Restarting tasks ... done. [ 220.615591] [ T5267] random: crng reseeded on system resumption [ 220.669956] [ T5267] PM: suspend exit
Anyone else having these issues?
I'll search bugzilla (again) and open a bugreport, if there is none.
Kind Regards, Johannes
Hi, thanks for the fast reply, Andrei! On 20.11.24 08:53 Andrei Borzenkov wrote:
On Wed, Nov 20, 2024 at 10:41 AM Johannes Kastl <mail@johannes-kastl.de> wrote:
[ 200.450388] [ T5267] Freezing remaining freezable tasks failed after 20.007 seconds (0 tasks refusing to freeze, wq_busy=1): [ 200.450417] [ T5267] Showing freezable workqueues that are still busy: [ 200.450426] [ T5267] workqueue pm: flags=0x4 [ 200.450435] [ T5267] pwq 2: cpus=0 node=0 flags=0x0 nice=0 active=0 refcnt=2 [ 200.450454] [ T5267] inactive: pm_runtime_work [ 200.450474] [ T5267] workqueue btrfs-endio-meta: flags=0xe [ 200.450490] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=2 [ 200.450503] [ T5267] inactive: btrfs_end_bio_work [btrfs] [ 200.450946] [ T5267] workqueue btrfs-endio-write: flags=0xe [ 200.450961] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=0 refcnt=19 [ 200.450974] [ T5267] inactive: 18*btrfs_work_helper [btrfs] [ 200.451418] [ T5267] workqueue btrfs-qgroup-rescan: flags=0x2000e [ 200.451432] [ T5267] pwq 16: cpus=0-3 flags=0x4 nice=0 active=1 refcnt=6 [ 200.451444] [ T5267] in-flight: 42:btrfs_work_helper [btrfs]
Did you check with "btrfs quota rescan --status" if any scan is in progress? Does it work if you disable quota on btrfs filesytem(s)?
I have not enabled quota intentionally, if this is active it has been since the Tumbleweed installation. I'll see if I can find out more. BTW, that command you gave does not work, I'll check the manpage to see what I need to do. Kind Regards, Johannes
On 20.11.24 08:53 Andrei Borzenkov wrote:
Did you check with "btrfs quota rescan --status" if any scan is in progress? Does it work if you disable quota on btrfs filesytem(s)?
Disabling quota on / apparently makes this a lot more reliable. I suspended multiple times and not had one failing attempt. I'll keep an eye on it. The question is: Was this working with enabled quota before? Or did the quota get enabled in the last couple of month for whatever reason? As I wrote, I did not enable the quota manually. Kind Regards, Johannes
On Wed, Nov 20, 2024 at 11:54 AM Johannes Kastl <mail@johannes-kastl.de> wrote:
On 20.11.24 08:53 Andrei Borzenkov wrote:
Did you check with "btrfs quota rescan --status" if any scan is in progress? Does it work if you disable quota on btrfs filesytem(s)?
Disabling quota on / apparently makes this a lot more reliable. I suspended multiple times and not had one failing attempt. I'll keep an eye on it.
The question is: Was this working with enabled quota before? Or did the quota get enabled in the last couple of month for whatever reason? As I wrote, I did not enable the quota manually.
Quota has been enabled by default in openSUSE for years, it is needed to get space based snapshot cleanup. Depending on the number of snapshots and filesystem size quota scan may take quite a lot of time. Whether it is normal that it also blocks suspend I do not know. It also could be an indication of some filesystem issues. That is something you better ask on btrfs list.
participants (5)
-
Andrei Borzenkov -
Carlos E. R. -
Fritz Hudnut -
Jiri Slaby -
Johannes Kastl