Puzzled by total downgrade of all kernel-firmware

I went to do a zypper dup as normal this morning and the entire set of kernel-firmware packages will be downgraded if I run it. Just puzzled and is this correct?


The following 33 packages are going to be downgraded:
  kernel-firmware-all kernel-firmware-amdgpu kernel-firmware-ath10k kernel-firmware-atheros kernel-firmware-bluetooth kernel-firmware-bnx2 kernel-firmware-brcm kernel-firmware-chelsio
  kernel-firmware-dpaa2 kernel-firmware-i915 kernel-firmware-intel kernel-firmware-iwlwifi kernel-firmware-liquidio kernel-firmware-marvell kernel-firmware-media kernel-firmware-mediatek
  kernel-firmware-mellanox kernel-firmware-mwifiex kernel-firmware-network kernel-firmware-nfp kernel-firmware-nvidia kernel-firmware-platform kernel-firmware-prestera kernel-firmware-qlogic
  kernel-firmware-radeon kernel-firmware-realtek kernel-firmware-serial kernel-firmware-sound kernel-firmware-ti kernel-firmware-ueagle kernel-firmware-usb-network ucode-amd yast2-network

They are going from 20201023-2.1 to 20201023-1.1.

Stuart

Same here. Haven’t read anything on the factory mailing list that indicated there may have been problems… so whether it was intentional or not, I don’t know…

I did the downgrade and rebooted. Suspend still works, but I don’t have a swap partition, so I can’t test suspend to RAM. So far, I don’t see any difference in overall behavior.

Gene

Because I updated at the weekend there is not much else to change so far so I think I’ll leave it for 24 hours and see if anything comes up as to why this is being done.

Stuart

Two more dowgraded package today ucode-amd and yast2-network. So I have locked all the downgraded packages for now until I find out more. The rest updated fine with no issues (some 22 pakages).

Stuart

Not updated to 20201108 yet.

Maybe you should ask on the factory mailing list, likely someone there will know the reason…

More fun:

**3400G:~ #** grep 'install|kernel-firmware-all' /var/log/zypp/history | cut -d '|' -f 1-4              
2019-11-02 19:57:52|install|kernel-firmware-all|20191023-1.1 
2019-12-19 14:42:42|install|kernel-firmware-all|20191118-1.1 
2020-03-06 15:30:38|install|kernel-firmware-all|20200207-1.1 
2020-03-14 07:06:07|install|kernel-firmware-all|20200302-1.1 
2020-06-16 13:47:55|install|kernel-firmware-all|20200519-2.1 
2020-06-23 04:21:33|install|kernel-firmware-all|20200610-1.1 
2020-10-06 15:19:49|install|kernel-firmware-all|20200916-1.1 
2020-10-10 05:39:48|install|kernel-firmware-all|20201005-1.1 
2020-10-11 15:22:45|install|kernel-firmware-all|20201005-1.1 
2020-10-12 11:55:18|install|kernel-firmware-all|20201005-1.1 
2020-10-15 17:29:38|install|kernel-firmware-all|20201005-3.1 
2020-10-17 20:15:04|install|kernel-firmware-all|20201005-36.1 
2020-10-18 09:04:44|install|kernel-firmware-all|20201005-3.1 
2020-10-18 17:47:37|install|kernel-firmware-all|20201005-36.1 
2020-10-21 17:39:05|install|kernel-firmware-all|20201005-3.1 
2020-10-30 06:50:05|install|kernel-firmware-all|20201005-3.1 
2020-11-02 19:10:48|install|kernel-firmware-all|20201023-2.1 
2020-11-09 11:17:33|install|kernel-firmware-all|20201023-1.1 
**3400G:~ #**


See also: 1177428 – AMDGPU resume fail

Looking at that bug I can see the KVM disabled messages and the platform integrity messages on my Journal. I don’t seem to have any apparent issues and I don’t use suspend or hibernate because my system now boots in seconds with the SSD.

I think I’ll leave those packages locked for now until I have an issue because of the lock or there is another update for them.

Stuart

It might only be a rebuild of the firmware packages that didn’t change much at all.

If I remember correctly, all these packages had been in the Tumpleweed Update Repo and now have been pushed back into the main repo.

My laptop is doing the same downgrade atm

I’ve just experienced - before I started the downgrade - some total failures of audio eg playing a youtube video in firefox - sound not syncing nicely, and then trying to play an iplayer video (BBC) - no sound at all, just some “blip, weeble, blip, squeegle” sounds

Tried using xine and vlc to play an mp4 that would have worked. yeah, sound doesn’t work…

have rebooted, checked this forum, and am not doing the kernel downgrade - or sudo zypper dup

Earlier today, I have chromium-browser bin out when I used pavucontrol to check what microphone levels were set (and which mic was selected) during a (previously stable) jitsi video conference call.

I’d previously been playing music using Clementine, but I’d halted that before starting up chromium

I did a reboot, and sound worked, and didn’t bin out

(PS for anyone who also might think about Jitsi for video conferencing, chrome based browsers are the only ones that work, now)

And I also now have no sound in a (Windows) Virtual Box VM that I was using to watch (UK) Channel 4 yesterday evening

Anyway, my system is almost ready for it’s reboot after downgrading about 45 packages

We’ll see whether I can hear sound in an mp4 after the reboot . . .

:wink:

I have synced sound and video using xine and an mp4 file after the reboot I’ve just played a youtube vid (briefly) in firefox and the sound is synced, and then gone back and restarted the mp4 video in xine, and sound works fine in that

And just loggedin (firefox agaain) and played a video (briefly), then played a bit more of a youtube clip, then the mp4 again and all of the problems I had earlier have disappeared.

NB Although I am on a laptop, I do shut down properly, I don’t just close the lid.
So it has been a full reboot every time, not a hibernate / suspend or anything like that So my audio problems are fixed by this downgrade . . .

Sorry. I spoke too soon

Started playing a mp4 using xine, paused it to go to the kitchen, came back and the sound is “weeble, squiggle, wweee, pib” again

Tried the same mp4 in vlc, and it’s broken, there, too

Reboot time (again)

Sorry, some of the above is rubbish

Just found I had a faulty (or intermittently faulty) inline volume control extension between my headphone socket and my headphones, and it was occasionally garbling one (or both) channels

So, I did have earlier (at work, with different headphones and a different extension) a condition where changing volumes / mic in pavucontrol did make chromium-browser crash as I was trying to set mic volume levels in a jitsi video conference call

All the other audio problems are down to a faulty cable

Sorry

Is there an emoticon for “hanging head in shame” ?

If so apply it here

Finally decided to remove the freezes and update everything, all went well on both TW machines.

Stuart

The mess has been cleaned up:

**3400G:~ #** rpm -q --changelog kernel-firmware-all|head -33 
* Thu Oct 29 2020 tiwai@suse.com 
- Update to version 20201023 (git commit dae4b4cd0841): 
  * cypress: add Cypress firmware and clm_blob files 
  * rtl_bt: Update RTL8821C BT FW to 0xAA6C_A99E 
  * ath10k: add SDIO firmware for QCA9377 WiFi 
  * ice: update package file to 1.3.16.0 
  * mediatek: separate venc service thread 
  * QCA : Updated firmware file for WCN3991 
  * iwlwifi: update and add new FWs from core56-54 release 
  * iwlwifi: update 3168, 7265D, 8000C and 8265 firmwares 
  * i915: Add DG1 DMC v2.02 
  * qcom : updated venus firmware files for v5.4 
- Add _constraints to fix the build error (bsc#1178242) 

* Mon Oct 12 2020 tiwai@suse.de 
- Use fdupes -s (bsc#1177428); 
  this seems fixing the mess up of hard-linked files by the package 
  update 

* Sun Oct 11 2020 tiwai@suse.de 
- Fix symlink creations for split packages (bsc#1177554); 
  also show the copy/link status verbosely at packge build 

* Mon Oct 05 2020 tiwai@suse.de 
- Placeholder for SLE15-SP3 package sync: 
  bsc#1143331, bsc#1160204, bsc#1154395 bsc#1155307, jsc#SLE-8379 

* Mon Oct 05 2020 tiwai@suse.com 
- Update to version 20201005 (git commit 58d41d0facca): 
  * ice: Add comms package file for Intel E800 series driver 
  * copy-firmware: Always write Link: entries 
  * amdgpu: update vega20 firmware for 20.40 
  * amdgpu: update vega12 firmware for 20.40 
**3400G:~ #**