Message-ID: <1fdfce01346136012f5f92b8845d272f8de58652.camel@scientia.org>
Date: Thu, 06 Mar 2025 22:58:47 +0100
From: Christoph Anton Mitterer <calestyo@...entia.org>
To: linux-wireless@...r.kernel.org, netdev@...r.kernel.org
Cc: Chandrashekar Devegowda <chandrashekar.devegowda@...el.com>
Subject: hibernate suddenly broken (by )

Hey.

[Sorry for the cross posting, but I guess the cause for the issue
described below is in iwlwifi and/or mtk_t7xx, so adding the lists
mentioned in MAINTAINERS for those, as well as the mtk_t7xx maintainer.
:-) ]


Since a few days, hibernation is broken (i.e. it aborts after trying to
hibernate).


Things that may be worth to mention are:
- running on Debian unstable
- full disk encryption with dm-crypt, booting from USB
- hibernation goes into btrfs swapfile (within the same dm-crypt LUKS
  volume)

That setup has worked now since quite a while, at least since btrfs-
progs added the btrfs inspect-internal map-swapfile to find out the
proper resume offset (which was in 6.1... so around end of 2022).


For your information, but probably unrelated to this issue is, that S3-
suspend (i.e. "deep") has been broken ever since I got that device:
https://bugzilla.kernel.org/show_bug.cgi?id=216998
I recently (see below) had another conversation with a Fujitsu support
guy, who then claimed S3 would have been deprecated and thus not been
implemented by that device.


Normally, when I go into hibernate, Cinnamon locks the desktop, I see
how the clock (respectively the processes) freeze, the screen goes
completely off (with backlight), then goes on again (still frozen) and
then it shuts off.
With the bug, after the screen went on again, it unfroze (and then
stayed running).


Detailed kernel logs attached, with the following done:


0. In a few tries for which I haven't attached the logs hibernate
   always seemed to succeed once, and subsequent tries would fail
   (until I freshly booted without resume)


1. hibernate.1.log
   Kernel 2.12.17
   Booted with WiFi disabled (via NetworkManager), where
   enabling/disabling seems to simply set the soft block on the device,
   i.e.:
   disabled
      # rfkill list 
      1: phy0: Wireless LAN
      	Soft blocked: yes
      	Hard blocked: no
   enabled:
      # rfkill list 
      1: phy0: Wireless LAN
      	Soft blocked: no
      	Hard blocked: no
      
   Description what I did, within the log.
   In short, as long as I've kept WiFi disabled (though I did only a
   few tries, so this may be coincidental) it seemed to have kept
   working.
   After enabling WiFi, it already failed on the next try.
   Disabling it again, didn't help.
   In some of the #0 tries I even tried to remove iwlwifi modules (but
   not mtk_t7xx), but that didn't help either.


2. Same as #1, just with WiFi enabled from the fresh boot on.


3. There were only two recent package upgrade which could have been
   related.
   One was upgrade from 6.12.16 to 6.12.17, though IIRC I had already
   run a few days with 6.12.17, and it worked.
   Still downgraded and tried with that.

   But also failed with 6.12.16.


4. The only other upgrade was NetworkManager (which went from 1.50.3-1
   -> 1.52.0-2).
   So tried with that downgraded. Failed again eventually.


5. In my 5th test series I've disabled the WWAN device in the BIOS
   (and surprise what mtk_t7xx is doing ;-) ):
   I don't use it anyway, but I'm pretty sure that I've had that
   (accidentally enabled already like half a year or so ago), because
   I kept seeing mtk_t7xx errors during previous boots since around
   6.10 or so maybe also a bit earlier, which apparently caused however
   never problems with hibernate.

   With the WWAN device disabled in BIOS mtk_t7xx isn't even loaded
   anymore.

   After 5 hibernation/resume cycles, which all worked, I've stopped
   testing.


The only other "change" was that I've recently tried to flash a EC
firmware upgrade from Fujitsu, but the flasher (which I had to run from
some Windows USB stick), fails already before actually doing anything,
an the BIOS/UEFI still shows the EC firmware on the old version.

So I'd strongly guess that no flashing actually took place, and that
this is unrelated.


So I guess the problem are these:

Mar 06 21:07:40 heisenberg kernel: mtk_t7xx 0000:58:00.0: [PM] SAP suspend error: -110
Mar 06 21:07:40 heisenberg kernel: mtk_t7xx 0000:58:00.0: can't suspend (t7xx_pci_pm_runtime_suspend [mtk_t7xx] returned -110)

respectively

Mar 06 21:08:08 heisenberg kernel: mtk_t7xx 0000:58:00.0: [PM] SAP suspend error: -110
Mar 06 21:08:08 heisenberg kernel: mtk_t7xx 0000:58:00.0: PM: pci_pm_freeze(): t7xx_pci_pm_suspend [mtk_t7xx] returns -110
Mar 06 21:08:08 heisenberg kernel: mtk_t7xx 0000:58:00.0: PM: dpm_run_callback(): pci_pm_freeze returns -110
Mar 06 21:08:08 heisenberg kernel: mtk_t7xx 0000:58:00.0: PM: failed to freeze async: error -110

and/or these:

Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3: CSR_RESET = 0x10
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3: Host monitor block 0x0 vector 0x0
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 0]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 1]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 2]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 3]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 4]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 5]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 6]: 0x3f1d0430
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 7]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 8]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 9]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 10]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 11]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 12]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 13]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 14]: 0x3f1d0030
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3: Host monitor block 0x0 vector 0x1
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 0]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 1]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 2]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 3]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 4]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 5]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 6]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 7]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 8]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 9]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 10]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 11]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 12]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 13]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 14]: 0x040066c2
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3: Host monitor block 0x0 vector 0x6
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 0]: 0xa00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 1]: 0xe00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 2]: 0xf00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 3]: 0xf00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 4]: 0xb00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 5]: 0xb00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 6]: 0xb00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 7]: 0xf00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 8]: 0xf00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 9]: 0xf00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 10]: 0xb00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 11]: 0xa00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 12]: 0xa00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 13]: 0xe00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 14]: 0xe00c7c75
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3: Host monitor block 0x22 vector 0x0
Mar 06 21:47:34 heisenberg kernel: iwlwifi 0000:00:14.3:     value [iter 0]: 0x61cec019



No idea which change caused the sudden breakage of hibernate... and the
workaround (disabling the device) is okay for me,... still, would be of
course good if this could be fixed.


If I can help anything or you need further information, don't hesitate
to ask. :-)


Cheers,
Chris.

Download attachment "logs.tar.xz" of type "application/x-xz-compressed-tar" (39084 bytes)

Powered by blists - more mailing lists