* Re: Scarlett 16i16 4th Gen: I/O error on kernel 6.19-rc6 (works on 6.18) [not found] <CAF7epbtx7YDwa2NBDZY+9NuxCEirY90jNonY58KgWfx8pq6QyQ@mail.gmail.com> @ 2026-01-24 16:24 ` Geoffrey D. Bennett 2026-01-25 8:44 ` Takashi Iwai 0 siblings, 1 reply; 4+ messages in thread From: Geoffrey D. Bennett @ 2026-01-24 16:24 UTC (permalink / raw) To: Ruslan Garynin; +Cc: Jaroslav Kysela, Takashi Iwai, linux-sound Hi Ruslan, Audio playback for Focusrite USB interfaces is handled by the generic USB audio driver (snd-usb-audio); the FCP driver I maintain only handles the control protocol (mixer, routing, etc.) over a separate USB endpoint — it's not involved in the actual audio data transfer. I've CC'd the ALSA maintainers and linux-sound list to continue the discussion there. Regards, Geoffrey. On Sat, Jan 24, 2026 at 08:56:38PM +0500, Ruslan Garynin wrote: > Hi Geoffrey, > > I found a regression affecting Scarlett 16i16 4th Gen playback on kernel > 6.19-rc6. > > Summary: > - Working: 6.18.6-zen1-1-zen > - Broken: 6.19.0-rc6-1-mainline > - Issue: I/O error on USB audio playback > - HDA/HDMI audio works fine on 6.19-rc6, so it's USB audio specific > > Test: > systemctl --user stop pipewire pipewire-pulse wireplumber > speaker-test -D hw:3 -c 18 -t sine -F S32_LE > > On 6.18 (works): > 0 - Front Left > 1 - Front Right > ... > 15 - Channel 16 > (sound plays) > > On 6.19-rc6 (fails): > 0 - Front Left > Write error: -5,Input/output error > > Device info: > USB ID: 1235:821b > FCP driver loads OK: "Focusrite Control Protocol Driver ready" > Implicit Feedback Mode: Yes > No errors in dmesg > > /proc/asound/card3/stream0: > Format: S32_LE > Channels: 18 > Endpoint: 0x01 (1 OUT) (ASYNC) > Rates: 44100, 48000 > Sync Endpoint: 0x81 (1 IN) > Implicit Feedback Mode: Yes > > Note: This might get fixed on its own when mainline is updated further > before > the stable 6.19 release. I'm just reporting this in case it doesn't, so > you're > aware of a potential issue that could affect users when 6.19 changes > eventually > land in stable kernels. > > Let me know if you need any additional info or testing. > > Thanks for maintaining the Scarlett driver! > > Best regards, > Ruslan Garynin ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Scarlett 16i16 4th Gen: I/O error on kernel 6.19-rc6 (works on 6.18) 2026-01-24 16:24 ` Scarlett 16i16 4th Gen: I/O error on kernel 6.19-rc6 (works on 6.18) Geoffrey D. Bennett @ 2026-01-25 8:44 ` Takashi Iwai [not found] ` <CAF7epbs1+q9YVtW9TorTTNj-hWc4EDrrHss09q1Y5jhP4ifqZA@mail.gmail.com> 0 siblings, 1 reply; 4+ messages in thread From: Takashi Iwai @ 2026-01-25 8:44 UTC (permalink / raw) To: Ruslan Garynin Cc: Geoffrey D. Bennett, Jaroslav Kysela, Takashi Iwai, linux-sound On Sat, 24 Jan 2026 17:24:07 +0100, Geoffrey D. Bennett wrote: > > Hi Ruslan, > > Audio playback for Focusrite USB interfaces is handled by the generic > USB audio driver (snd-usb-audio); the FCP driver I maintain only > handles the control protocol (mixer, routing, etc.) over a separate > USB endpoint — it's not involved in the actual audio data transfer. > > I've CC'd the ALSA maintainers and linux-sound list to continue the > discussion there. > > Regards, > Geoffrey. > > On Sat, Jan 24, 2026 at 08:56:38PM +0500, Ruslan Garynin wrote: > > Hi Geoffrey, > > > > I found a regression affecting Scarlett 16i16 4th Gen playback on kernel > > 6.19-rc6. > > > > Summary: > > - Working: 6.18.6-zen1-1-zen > > - Broken: 6.19.0-rc6-1-mainline > > - Issue: I/O error on USB audio playback > > - HDA/HDMI audio works fine on 6.19-rc6, so it's USB audio specific > > > > Test: > > systemctl --user stop pipewire pipewire-pulse wireplumber > > speaker-test -D hw:3 -c 18 -t sine -F S32_LE > > > > On 6.18 (works): > > 0 - Front Left > > 1 - Front Right > > ... > > 15 - Channel 16 > > (sound plays) > > > > On 6.19-rc6 (fails): > > 0 - Front Left > > Write error: -5,Input/output error > > > > Device info: > > USB ID: 1235:821b > > FCP driver loads OK: "Focusrite Control Protocol Driver ready" > > Implicit Feedback Mode: Yes > > No errors in dmesg > > > > /proc/asound/card3/stream0: > > Format: S32_LE > > Channels: 18 > > Endpoint: 0x01 (1 OUT) (ASYNC) > > Rates: 44100, 48000 > > Sync Endpoint: 0x81 (1 IN) > > Implicit Feedback Mode: Yes > > > > Note: This might get fixed on its own when mainline is updated further > > before > > the stable 6.19 release. I'm just reporting this in case it doesn't, so > > you're > > aware of a potential issue that could affect users when 6.19 changes > > eventually > > land in stable kernels. > > > > Let me know if you need any additional info or testing. > > > > Thanks for maintaining the Scarlett driver! > > > > Best regards, > > Ruslan Garynin Could you check whether the earlier 6.19-rc shows the same problem? Also check the very latest Linus tree, too. Takashi ^ permalink raw reply [flat|nested] 4+ messages in thread
[parent not found: <CAF7epbs1+q9YVtW9TorTTNj-hWc4EDrrHss09q1Y5jhP4ifqZA@mail.gmail.com>]
* Re: Scarlett 16i16 4th Gen: I/O error on kernel 6.19-rc6 (works on 6.18) [not found] ` <CAF7epbs1+q9YVtW9TorTTNj-hWc4EDrrHss09q1Y5jhP4ifqZA@mail.gmail.com> @ 2026-01-25 9:07 ` Takashi Iwai 2026-01-25 10:36 ` Ruslan Garynin 0 siblings, 1 reply; 4+ messages in thread From: Takashi Iwai @ 2026-01-25 9:07 UTC (permalink / raw) To: Ruslan Garynin Cc: Takashi Iwai, Geoffrey D. Bennett, Jaroslav Kysela, Takashi Iwai, linux-sound On Sun, 25 Jan 2026 09:59:00 +0100, Ruslan Garynin wrote: > > > Hi Takashi, > > I checked my pacman logs and found that: > > - Working version: 6.19-rc3 > - Broken version: 6.19-rc6 > > The upgrade from rc3 to rc6 happened on 2026-01-23. > > I will try to rebuild and test rc4 and rc5 to narrow down > exactly which version introduced the regression. > > Will report back with the results. If it worked, it might be the recent fix that landed in rc6, and this was corrected in Linus tree for the upcoming rc7 via commit 70b4db7d258118a7464f039112a74ddb49a95b06 ALSA: usb-audio: Use the right limit for PCM OOB check So please check the Linus tree or cherry-pick the commit above. thanks, Takashi > > Regards, > Ruslan Garynin > > On Sun, Jan 25, 2026 at 1:44 PM Takashi Iwai <tiwai@suse.de> wrote: > > On Sat, 24 Jan 2026 17:24:07 +0100, > Geoffrey D. Bennett wrote: > > > > Hi Ruslan, > > > > Audio playback for Focusrite USB interfaces is handled by the generic > > USB audio driver (snd-usb-audio); the FCP driver I maintain only > > handles the control protocol (mixer, routing, etc.) over a separate > > USB endpoint — it's not involved in the actual audio data transfer. > > > > I've CC'd the ALSA maintainers and linux-sound list to continue the > > discussion there. > > > > Regards, > > Geoffrey. > > > > On Sat, Jan 24, 2026 at 08:56:38PM +0500, Ruslan Garynin wrote: > > > Hi Geoffrey, > > > > > > I found a regression affecting Scarlett 16i16 4th Gen playback on > kernel > > > 6.19-rc6. > > > > > > Summary: > > > - Working: 6.18.6-zen1-1-zen > > > - Broken: 6.19.0-rc6-1-mainline > > > - Issue: I/O error on USB audio playback > > > - HDA/HDMI audio works fine on 6.19-rc6, so it's USB audio specific > > > > > > Test: > > > systemctl --user stop pipewire pipewire-pulse wireplumber > > > speaker-test -D hw:3 -c 18 -t sine -F S32_LE > > > > > > On 6.18 (works): > > > 0 - Front Left > > > 1 - Front Right > > > ... > > > 15 - Channel 16 > > > (sound plays) > > > > > > On 6.19-rc6 (fails): > > > 0 - Front Left > > > Write error: -5,Input/output error > > > > > > Device info: > > > USB ID: 1235:821b > > > FCP driver loads OK: "Focusrite Control Protocol Driver ready" > > > Implicit Feedback Mode: Yes > > > No errors in dmesg > > > > > > /proc/asound/card3/stream0: > > > Format: S32_LE > > > Channels: 18 > > > Endpoint: 0x01 (1 OUT) (ASYNC) > > > Rates: 44100, 48000 > > > Sync Endpoint: 0x81 (1 IN) > > > Implicit Feedback Mode: Yes > > > > > > Note: This might get fixed on its own when mainline is updated > further > > > before > > > the stable 6.19 release. I'm just reporting this in case it > doesn't, so > > > you're > > > aware of a potential issue that could affect users when 6.19 > changes > > > eventually > > > land in stable kernels. > > > > > > Let me know if you need any additional info or testing. > > > > > > Thanks for maintaining the Scarlett driver! > > > > > > Best regards, > > > Ruslan Garynin > > Could you check whether the earlier 6.19-rc shows the same problem? > Also check the very latest Linus tree, too. > > Takashi > ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Scarlett 16i16 4th Gen: I/O error on kernel 6.19-rc6 (works on 6.18) 2026-01-25 9:07 ` Takashi Iwai @ 2026-01-25 10:36 ` Ruslan Garynin 0 siblings, 0 replies; 4+ messages in thread From: Ruslan Garynin @ 2026-01-25 10:36 UTC (permalink / raw) To: Takashi Iwai Cc: Geoffrey D. Bennett, Jaroslav Kysela, Takashi Iwai, linux-sound Hi Takashi, I can confirm the fix works! I applied commit 70b4db7d258118a7464f039112a74ddb49a95b06 ("ALSA: usb-audio: Use the right limit for PCM OOB check") on top of 6.19-rc6 and rebuilt the kernel. Test result: - Before patch: "Write error: -5, Input/output error" on first channel - After patch: All 18 channels play successfully Thank you for the quick fix! Tested-by: Ruslan Garynin <ufurifkf@gmail.com> Regards, Ruslan Garynin Regards, Ruslan. On Sun, Jan 25, 2026 at 2:07 PM Takashi Iwai <tiwai@suse.de> wrote: > > On Sun, 25 Jan 2026 09:59:00 +0100, > Ruslan Garynin wrote: > > > > > > Hi Takashi, > > > > I checked my pacman logs and found that: > > > > - Working version: 6.19-rc3 > > - Broken version: 6.19-rc6 > > > > The upgrade from rc3 to rc6 happened on 2026-01-23. > > > > I will try to rebuild and test rc4 and rc5 to narrow down > > exactly which version introduced the regression. > > > > Will report back with the results. > > If it worked, it might be the recent fix that landed in rc6, and this > was corrected in Linus tree for the upcoming rc7 via commit > 70b4db7d258118a7464f039112a74ddb49a95b06 > ALSA: usb-audio: Use the right limit for PCM OOB check > > So please check the Linus tree or cherry-pick the commit above. > > > thanks, > > Takashi > > > > > Regards, > > Ruslan Garynin > > > > On Sun, Jan 25, 2026 at 1:44 PM Takashi Iwai <tiwai@suse.de> wrote: > > > > On Sat, 24 Jan 2026 17:24:07 +0100, > > Geoffrey D. Bennett wrote: > > > > > > Hi Ruslan, > > > > > > Audio playback for Focusrite USB interfaces is handled by the generic > > > USB audio driver (snd-usb-audio); the FCP driver I maintain only > > > handles the control protocol (mixer, routing, etc.) over a separate > > > USB endpoint — it's not involved in the actual audio data transfer. > > > > > > I've CC'd the ALSA maintainers and linux-sound list to continue the > > > discussion there. > > > > > > Regards, > > > Geoffrey. > > > > > > On Sat, Jan 24, 2026 at 08:56:38PM +0500, Ruslan Garynin wrote: > > > > Hi Geoffrey, > > > > > > > > I found a regression affecting Scarlett 16i16 4th Gen playback on > > kernel > > > > 6.19-rc6. > > > > > > > > Summary: > > > > - Working: 6.18.6-zen1-1-zen > > > > - Broken: 6.19.0-rc6-1-mainline > > > > - Issue: I/O error on USB audio playback > > > > - HDA/HDMI audio works fine on 6.19-rc6, so it's USB audio specific > > > > > > > > Test: > > > > systemctl --user stop pipewire pipewire-pulse wireplumber > > > > speaker-test -D hw:3 -c 18 -t sine -F S32_LE > > > > > > > > On 6.18 (works): > > > > 0 - Front Left > > > > 1 - Front Right > > > > ... > > > > 15 - Channel 16 > > > > (sound plays) > > > > > > > > On 6.19-rc6 (fails): > > > > 0 - Front Left > > > > Write error: -5,Input/output error > > > > > > > > Device info: > > > > USB ID: 1235:821b > > > > FCP driver loads OK: "Focusrite Control Protocol Driver ready" > > > > Implicit Feedback Mode: Yes > > > > No errors in dmesg > > > > > > > > /proc/asound/card3/stream0: > > > > Format: S32_LE > > > > Channels: 18 > > > > Endpoint: 0x01 (1 OUT) (ASYNC) > > > > Rates: 44100, 48000 > > > > Sync Endpoint: 0x81 (1 IN) > > > > Implicit Feedback Mode: Yes > > > > > > > > Note: This might get fixed on its own when mainline is updated > > further > > > > before > > > > the stable 6.19 release. I'm just reporting this in case it > > doesn't, so > > > > you're > > > > aware of a potential issue that could affect users when 6.19 > > changes > > > > eventually > > > > land in stable kernels. > > > > > > > > Let me know if you need any additional info or testing. > > > > > > > > Thanks for maintaining the Scarlett driver! > > > > > > > > Best regards, > > > > Ruslan Garynin > > > > Could you check whether the earlier 6.19-rc shows the same problem? > > Also check the very latest Linus tree, too. > > > > Takashi > > ^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2026-01-25 10:36 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <CAF7epbtx7YDwa2NBDZY+9NuxCEirY90jNonY58KgWfx8pq6QyQ@mail.gmail.com>
2026-01-24 16:24 ` Scarlett 16i16 4th Gen: I/O error on kernel 6.19-rc6 (works on 6.18) Geoffrey D. Bennett
2026-01-25 8:44 ` Takashi Iwai
[not found] ` <CAF7epbs1+q9YVtW9TorTTNj-hWc4EDrrHss09q1Y5jhP4ifqZA@mail.gmail.com>
2026-01-25 9:07 ` Takashi Iwai
2026-01-25 10:36 ` Ruslan Garynin
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.