$ sudo journalctl --since "20 min ago" > journal_last_20_min_hibernate.txt
20 minute journaling:
(You can preview it, you don’t need to download it.)
sudo journalctl --boot=0 --no-hostname --since "20 min ago" > journal_last_20_min.txt
sudo journalctl -b --since "20 min ago" > journal_last_20_min.txt
I would leave this to people with more experience with hibernation and/or nvidia issues, but after quickly skimming these logs, I see at first glance that there is an nvidia problem indicated: it seems your kernel gets tainted and has errors with relations to nvidia-sleep.sh.
I add the tag nvidia to maybe have some people with nvidia experience check this out. Maybe someone has seen that before.
1 Like
Added nvidia
Thank you Sir
1 Like
nvidia kernel modules always taint the kernel. It does not indicate there is an error. Any kernel module that is not provided by the kernel* packages will taint the kernel, just to indicate that a foreign kernel module has been loaded.
I didn’t mean the “standard” taint that is generally caused by nvidia, which is not an error taint.
Here is an extract of what I meant:
...
Kas 12 15:46:16 fedora kernel: WARNING: CPU: 0 PID: 9348 at include/linux/rwsem.h:80 follow_pte+0x1f0/0x220
Kas 12 15:46:16 fedora kernel: Modules linked in: uinput snd_seq_dummy snd_hrtimer vboxnetadp(OE) vboxnetflt(OE) vboxdrv(OE) rfkill nf_conntrack_netbios_ns nf_conntrack_broadcast nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nft_chain_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_set qrtr nvidia_drm(POE) nvidia_modeset(POE) nvidia_uvm(POE) nf_tables joydev nvidia(POE) sunrpc binfmt_misc snd_soc_avs intel_rapl_msr snd_soc_hda_codec intel_rapl_common ee1004 snd_hda_ext_core intel_tcc_cooling snd_hda_codec_realtek x86_pkg_temp_thermal snd_soc_core intel_powerclamp snd_hda_codec_hdmi snd_hda_codec_generic snd_hda_scodec_component coretemp ppdev snd_compress ac97_bus snd_pcm_dmaengine iTCO_wdt intel_pmc_bxt mei_hdcp iTCO_vendor_support kvm_intel snd_hda_intel mei_pxp snd_intel_dspcfg snd_intel_sdw_acpi snd_hda_codec snd_hda_core kvm snd_hwdep snd_seq snd_seq_device rapl snd_pcm intel_cstate snd_timer r8169 intel_uncore intel_wmi_thunderbolt pcspkr i2c_i801 realtek
Kas 12 15:46:16 fedora kernel: i2c_smbus vfat snd fat drm_ttm_helper mei_me ttm soundcore mei parport_pc parport intel_pmc_core intel_vsec pmt_telemetry pmt_class acpi_pad loop nfnetlink zram hid_logitech_hidpp crct10dif_pclmul crc32_pclmul crc32c_intel polyval_clmulni polyval_generic ghash_clmulni_intel sha512_ssse3 sha256_ssse3 sha1_ssse3 video wmi hid_logitech_dj ip6_tables ip_tables fuse i2c_dev
Kas 12 15:46:16 fedora kernel: CPU: 0 UID: 0 PID: 9348 Comm: nvidia-sleep.sh Tainted: P W OE 6.11.6-300.fc41.x86_64 #1
Kas 12 15:46:16 fedora kernel: Tainted: [P]=PROPRIETARY_MODULE, [W]=WARN, [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
Kas 12 15:46:16 fedora kernel: Hardware name: Gigabyte Technology Co., Ltd. Z170-HD3/Z170-HD3-CF, BIOS F22f 03/09/2018
Kas 12 15:46:16 fedora kernel: RIP: 0010:follow_pte+0x1f0/0x220
...
This seems to occur (along with a long list of related entries) some seconds after the hibernation is attempted. More details from the kernel are in the logs.
I click on the suspend button.
Sometimes it falls asleep. It works fine.
Sometimes the monitor disconnects, but the pc keeps working. After 10-15 minutes it goes to sleep.
Sometimes the monitor disconnects and the pc works for hours. It never falls asleep. I have to shut it down by long pressing the pc power button. Here are the logs I have experienced this:
sudo journalctl -b --since "14 min ago" > journal_boot_last_14_min_suspend.txt
I also had this problem with fedora 40.
After upgrading to fedora 41, I started having it more often.
I think it might be nvidia, but I’m not a developer, not a maintainer, I don’t understand this log data.
@gosling203 I moved your other post/topic (the one that starts with “I click on the suspend button”) to his one: you have best chances to resolve the issue if you consolidate all knowledge in one topic. However, it was a good idea of you to also add the suspend-resume tag, but I do it manually for this topic, in order to have all related tags assigned within this one topic (nvidia, suspend-resume, etc.).
1 Like
Added suspend-resume
Sorry, It wasn’t clear which log file to look at. In the first log file “journal_last_20_min_hibernate.txt” you actually got 21 kernel crash events (aka oops).
The next step would be to eliminate the foreign modules, that is VirtualBox and Nvidia modules, from the kernel and then try again to see if it works.
How do I do it?
Like this:
sudo dnf remove virtualbox*
sudo dnf remove akmods-nvidia nvidia*
sudo dnf remove VirtualBox '*nvidia*' --exclude nvidia-gpu-firmware should remove both with a single command. Be certain to reboot after that completes.
I deleted it. Now I have no nvidia driver.
I put it in hibernate mode, it woke up as soon as I fell asleep and the account login screen opened when it woke up.
It didn’t sleep, it didn’t go into hibernate mode. It’s weird.
I also tested “Suspend then hibernate” and “Hybrid sleep” modes.
It slept very fast. I don’t think it slept, it just shut down. Because when I pressed the pc power button to wake it up, the pc started but it wouldn’t turn on, it couldn’t even connect to the monitor. I waited. It wouldn’t turn on. I pressed the power button, he pretended to sleep again, but he didn’t sleep, the pc shut down. I pressed the power button again, it didn’t wake up.
I had to turn it off by long pressing the power button.
At this rate the SSD will be scrapped.
Hibernate requires a physical swap partition or file of at least large enough to contain everything in RAM.
Suspend/sleep does not.
Now I’ve also tested the “Suspend” mode.
He fell asleep very fast. I think it didn’t sleep, it just shut down. Because when I pressed the pc power button to wake it up, the pc started but it wouldn’t turn on, it wouldn’t even connect to the monitor. I waited. It wouldn’t turn on. I pressed the power button, he pretended to sleep again, but he didn’t sleep, the pc shut down. I pressed the power button again, it didn’t wake up.
I had to turn it off by long pressing the power button.
The same thing happened to me in suspend mode. I think suspend is now completely broken.