http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 Hillwood Yang changed: What |Removed |Added ---------------------------------------------------------------------------- Summary|The SB live 5.1 card lost |Leap 15.0 lost Analog |Analog output in Leap 15.0 |output for SB live 5.1 | |sound card -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c2 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |hillwoodroc@gmail.com Flags| |needinfo?(hillwoodroc@gmail | |.com) --- Comment #2 from Takashi Iwai --- (In reply to Hillwood Yang from comment #0)

Created attachment 769977 [details] output list in Leap 15.0

This sound card works fine in Leap 42.3 and Tumbleweed. But there is not any Analog output option in output list for SB live 5.1 (EMU10K1).

I see "Analog Output" in this screenshot. The screenshot in comment 1 misses it, though. Did you attach the pictures wrongly? There should be no big difference in the kernel, so at most it's likely a PulseAudio problem. Try to upgrade pulseaudio related packages from TW or OBS multimedia:libs project. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 Hillwood Yang changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #769978|output list in Tumbleweed |output list in Leap 15 description|20180502 | -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 Hillwood Yang changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #769978|output list in Leap 15 |output list in Leap 15 with description| |the bug -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c4 --- Comment #4 from Takashi Iwai --- Hrm, OK, then please try to install TW kernel from OBS Kernel:stable repo, and test with it. If this makes the difference, I should be able to fix it easily. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c6 --- Comment #6 from Takashi Iwai --- Aha, interesting. Could you take alsa-info.sh outputs on both 4.16 and 4.12 kernels, and attach to bugzilla? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c8 --- Comment #8 from Hillwood Yang --- Created attachment 770115 --> http://bugzilla.opensuse.org/attachment.cgi?id=770115&action=edit alsa-info output with kernel 4.16.8 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c10 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(hillwoodroc@gmail | |.com) --- Comment #10 from Takashi Iwai --- A test kernel is being built in OBS home:tiwai:bnc1093027 repo. It'll take some time (an hour or so) until the build finishes. Once after that, please test the kernel. If it still doesn't work, please upload alsa-info.sh output from the kernel again. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c12 --- Comment #12 from Takashi Iwai --- Hm, then the patches are irrelevant with the allocation issues. Could you give the whole kernel messages after the boot? Basically the issue isn't new, and it's interesting that it worked before. (Do I understand correctly that you didn't have the problem with Leap 42.3 kernel, right?) There were a big rewrite of swiotlb in the recent kernels, and this might make things working again... -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c14 --- Comment #14 from Takashi Iwai --- Thanks. This is an AMD IOMMU, and I thought this should have been a long-standing issue with it. Didn't it happen in the past? Interesting. In anyway, I'll try to hack and add a temporary workaround. A new kernel is being built in OBS home:tiwai:bnc1093027-2 repo. Please try it later. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c16 --- Comment #16 from Hillwood Yang --- Created attachment 770441 --> http://bugzilla.opensuse.org/attachment.cgi?id=770441&action=edit dmesg-4.12.14.g7c82cb4.log Sadly, this building still can not fix this bug.... -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c18 --- Comment #18 from Takashi Iwai --- Thanks, so it's not the failure at the allocation time but it's a bad page fault happening in iommu, as the AMD-Vi error messages indicate. Could you try to boot with iommu=soft option and see whether it suppresses the issue? It should be effective to genuine Leap kernels, too. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c20 --- Comment #20 from Hillwood Yang --- Created attachment 770455 --> http://bugzilla.opensuse.org/attachment.cgi?id=770455&action=edit dmesg-4.12.14-lp150.11-default-with-iommu=soft.txt -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c22 --- Comment #22 from Hillwood Yang --- Created attachment 770457 --> http://bugzilla.opensuse.org/attachment.cgi?id=770457&action=edit alsainfo-kernel-4_12_14-softiommu.txt -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c24 --- Comment #24 from Takashi Iwai --- Err, I meant the hardware side :) Which AMD chip it is? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c26 --- Comment #26 from Takashi Iwai --- Did you update BIOS to the recent version? I'm asking this because there were some issues on some Ryzen boards that showed similar IOMMU page fault errors, and got fixed by BIOS update. So just to be sure... And one more favor: could you try kernel-vanilla-4.12.* package from Leap 15.0, and check whether you get the same issue? It's to confirm whether the issue is brought by our own patches or not. Thanks. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c28 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(hillwoodroc@gmail | |.com) | --- Comment #28 from Takashi Iwai --- OK, thanks for confirmation. So the conclusion is that it's a generic problem of AMD IOMMU stuff on 4.12 kernel, which has been eventually fixed in the upstream. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1093027 http://bugzilla.opensuse.org/show_bug.cgi?id=1093027#c30 --- Comment #30 from Takashi Iwai --- BTW, although the bug itself could be worked around with the option, I merged a few emu10k1 fixes from upstream, as relevant with iommu, too. -- You are receiving this mail because: You are on the CC list for the bug.