Created attachment 289909 [details] Output of alsa-info.sh With kernel 5.4.49 the audio output through my Dell WD19 Docking station is now laggy and distorted (sounds like only bass frequencies are emitted). Previously, audio output through the dock worked just fine with all versions of 4.19.x and 5.4.x. After looking through the changelog for 5.4.49, I was able to locate the offending patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/patch/?id=aba41867dd66939d336fdf604e4d73b805d8039f Reverting this patch allows audio to play normally through my docking station.
Bring this to the attention of the person who wrote the offending commit: On Sun, Jun 28, 2020 at 03:19:06PM +0000, bugzilla-daemon@bugzilla.kernel.org wrote: > https://bugzilla.kernel.org/show_bug.cgi?id=208353 > > Bug ID: 208353 > Summary: Distorted Audio Output Through USB C Docking Station > Product: Drivers > Version: 2.5 > Kernel Version: 5.4.49 > Hardware: All > OS: Linux > Tree: Mainline > Status: NEW > Severity: normal > Priority: P1 > Component: USB > Assignee: drivers_usb@kernel-bugs.kernel.org > Reporter: pbui@kernel.bx612.space > Regression: No > > Created attachment 289909 [details] > --> https://bugzilla.kernel.org/attachment.cgi?id=289909&action=edit > Output of alsa-info.sh > > With kernel 5.4.49 the audio output through my Dell WD19 Docking station is > now > laggy and distorted (sounds like only bass frequencies are emitted). > Previously, audio output through the dock worked just fine with all versions > of > 4.19.x and 5.4.x. > > After looking through the changelog for 5.4.49, I was able to locate the > offending patch: > > > https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/patch/?id=aba41867dd66939d336fdf604e4d73b805d8039f This is upstream commit f0bd62b64016 ("ALSA: usb-audio: Improve frames size computation"). Alan Stern > Reverting this patch allows audio to play normally through my docking > station.
Please attach the output of "lsusb -v -d 0bda:402e" command.
Created attachment 289915 [details] Output of lsusb -v -d 0bda:402e Attached is the output "lsusb -v -d 0bda:402e" command.
Thanks! I think I've completely missed datainterval in my calculations. I'll prepare a patch for you to test. (Also CCing Takashi)
Created attachment 289917 [details] test.patch Here is a quick patch to test my theory. Please give it a try.
The patch works! After applying it to 5.4.49, I can play audio via my USB C docking station now. Thanks for the quick response and fix!
Alex, this is also https://bugzilla.kernel.org/show_bug.cgi?id=208317 with kernel 5.7.6. I applied your single line patch to my Arch 5.7.6 kernel and it does fix the issue. Thanks for the quick response.
I can confirm that the patch works for me on a ThinkPad X1 Carbon Gen 8 with a ThinkPad Ultra Docking Station running Linux 5.7.6 (as did reverting the original commit). Thank you!
Fedora is shipping 5.7.z kernels in our updates repo and it is raining bug-reports because of this issue. Can we please get a non test-patch fix for this ASAP? Or get a revert of the commit causing this regression queued up for all the stable series where this got added and is causing regressions (also ASAP)?
Also see: bug 208311, bug 208317 and: https://bugzilla.redhat.com/show_bug.cgi?id=1852564 https://bugzilla.redhat.com/show_bug.cgi?id=1853707 https://bugzilla.redhat.com/show_bug.cgi?id=1854166
I too can confirm the patch works, applied it to the Fedora 32 5.7.7 kernel and the problem is solved. Thanks for the fix!
I have similar issue with audio output from Dell WD15 USB-C Dock. I decoded notorious song "Never Gonna Give You Up" to raw audio and ran it throug aplay. Before the commit it plays normally, since the commit it plays like this: https://central.szn.cz/s/tr9otisNwiFEE2A The first official release with the issue is 5.7.6, it is still present in 5.8-rc4. Bisecting found the following commit d288dc74f8cf95cb7ae0aaf245b7128627a49bf3 (ALSA: usb-audio: Improve frames size computation), which is based on upstream commit f0bd62b64016508938df9babe47f65c2c727d25c. Release 5.7.7 with the patch from comment #5 seems okay.
The fix patch landed in Linus tree yesterday. Sorry that it took long, as I've been off. FWIW, Greg already reverted the original commit that caused a problem on stable trees.
*** Bug 208403 has been marked as a duplicate of this bug. ***
*** Bug 208317 has been marked as a duplicate of this bug. ***
*** Bug 208501 has been marked as a duplicate of this bug. ***