[Bug 72283] New: [<ffffffffa01b7ecb>] dce6_afmt_write_speaker_allocation+0xdb/0x140 [radeon]
https://bugs.freedesktop.org/show_bug.cgi?id=72283
Priority: medium Bug ID: 72283 Assignee: dri-devel@lists.freedesktop.org Summary: [<ffffffffa01b7ecb>] dce6_afmt_write_speaker_allocation+0xdb/0x140 [radeon] Severity: critical Classification: Unclassified OS: Linux (All) Reporter: bgunteriv@gmail.com Hardware: x86-64 (AMD64) Status: NEW Version: git Component: Drivers/Gallium/r600 Product: Mesa
Not sure if this has been known.
After trying to update to kernel 3.13rc2, I get problems booting my mini.iso system. Once the system loads, it loses the HDMI signal, and I get no picture. I get errors talking about the HAWAII drivers not loaded: update-initramfs: Generating /boot/initrd.img-3.13.0-031300rc2-generic W: Possible missing firmware /lib/firmware/radeon/HAWAII_smc.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_sdma.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_rlc.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_mc.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_mec.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_ce.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_me.bin for module radeon W: Possible missing firmware /lib/firmware/radeon/HAWAII_pfp.bin for module radeon
here are my log files:xbmc@xbmc:~$ dmesg | pastebinit http://paste.ubuntu.com/6514897/ xbmc@xbmc:~$ cat ~/.xbmc/temp/xbmc.log | pastebinit http://paste.ubuntu.com/6514899/ xbmc@xbmc:~$ cat /var/log/Xorg.0.log | pastebinit http://paste.ubuntu.com/6514901/ xbmc@xbmc:~$ DISPLAY=:0 vdpauinfo | pastebinit [1]+ Stopped DISPLAY=:0 vdpauinfo | pastebinit xbmc@xbmc:~$ dpkg -l | grep mesa | pastebinit http://paste.ubuntu.com/6514909/
as you can see, when I try to run DISPLAY=:0 vdpauinfo, my system hangs, and I have to force kill the command.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #1 from Alex Deucher agd5f@yahoo.com --- Created attachment 90196 --> https://bugs.freedesktop.org/attachment.cgi?id=90196&action=edit possible fix
Does the attached patch fix the issue? Also, please attach files rather than pointing to pastebin since pastebin entries can disappear.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
Michel Dänzer michel@daenzer.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Product|Mesa |DRI Version|git |unspecified Component|Drivers/Gallium/r600 |DRM/Radeon
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #2 from bgunteriv@gmail.com --- this patch did not fix the issue.
should I attach newer logs?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #3 from Alex Deucher agd5f@yahoo.com --- (In reply to comment #2)
this patch did not fix the issue.
should I attach newer logs?
Does the patch fix the segfault in dce6_afmt_write_speaker_allocation?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #4 from Alex Deucher agd5f@yahoo.com --- (In reply to comment #2)
this patch did not fix the issue.
should I attach newer logs?
Yes, please attach the logs with the patch applied.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #5 from bgunteriv@gmail.com --- now it's saying: perf samples too long (* > *), lowering kernel.per_event_max_sample_rate to ***
this repeats for about 5 lines where I have asteriks(*), are random numbers that keep changing every time a new line shows up.
this is the last line: perf samples too long (463153 > 250000), lowering kernel.per_event_max_sample_rate to 500
seems to have stopped here.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #6 from bgunteriv@gmail.com --- okay,
i was able to get past that last error. I apologize, it was a bad kernel build (i'm pretty sure).
now I'm able to get the kernel to load a little more, and I'm now getting this
on my screen i get: *ERROR* UVD not responding, trying to reset the VCPU!!!
this repeats until it gives up then it says: *ERROR* radeon: failed initializing UVD (-1)
i've had other people report this similar experience with this patch
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #7 from Alex Deucher agd5f@yahoo.com --- (In reply to comment #6)
okay,
i was able to get past that last error. I apologize, it was a bad kernel build (i'm pretty sure).
now I'm able to get the kernel to load a little more, and I'm now getting this
on my screen i get: *ERROR* UVD not responding, trying to reset the VCPU!!!
this repeats until it gives up then it says: *ERROR* radeon: failed initializing UVD (-1)
Make sure you have installed the latest rlc ucode for your chip and make sure you've updated the ucode in your initrd if you are using one.
i've had other people report this similar experience with this patch
Which patch?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #8 from bgunteriv@gmail.com ---
Make sure you have installed the latest rlc ucode for your chip and make sure you've updated the ucode in your initrd if you are using one.
I'm not sure what you mean. can you point me in the right direction? thanks in advance.
i've had other people report this similar experience with this patch
Which patch?
the patch you provided above.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #9 from bgunteriv@gmail.com --- i copied over the HAWAII drivers from here -> http://people.freedesktop.org/~agd5f/radeon_ucode/
so when i run: update-initramfs -k all -u
it no longer complains about the HAWAII firmware not installed. was this the proper way to do this?
when i pull the git version of linux-firmware, these files are not in the repo.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #10 from bgunteriv@gmail.com --- (In reply to comment #8)
Make sure you have installed the latest rlc ucode for your chip and make sure you've updated the ucode in your initrd if you are using one.
my system is just not very stable.
sometimes i get the UVD errors sometimes i get the reference to [<ffffffffa01b7ecb>] dce6_afmt_write_speaker_allocation+0xdb/0x140 [radeon]
and a couple of times it will get a little bit further, and then freeze. but i can't get any logs, cause it never gets to loading my network.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #11 from bgunteriv@gmail.com --- not sure if these reports will help you. http://forum.xbmc.org/showthread.php?tid=174854&page=123
i see that you and fritsch are working together on this.
thanks for all your help and support. let us know if there is any information that we can give you to help out.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #12 from Alex Deucher agd5f@yahoo.com --- (In reply to comment #9)
i copied over the HAWAII drivers from here -> http://people.freedesktop.org/~agd5f/radeon_ucode/
so when i run: update-initramfs -k all -u
it no longer complains about the HAWAII firmware not installed. was this the proper way to do this?
Yes. You can ignore the hawaii errors though since you aren't using a hawaii board.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #13 from bgunteriv@gmail.com --- Created attachment 90534 --> https://bugs.freedesktop.org/attachment.cgi?id=90534&action=edit dmesg
Here are new logs with the latest drm-intel kernel (http://kernel.ubuntu.com/~kernel-ppa/mainline/drm-intel-nightly/current/)
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #14 from bgunteriv@gmail.com --- Created attachment 90535 --> https://bugs.freedesktop.org/attachment.cgi?id=90535&action=edit X.org.0.log
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #15 from bgunteriv@gmail.com --- still no signal with the latest kernel from today Dec. 10.
do you need more logs?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #16 from Alex Deucher agd5f@yahoo.com --- (In reply to comment #15)
still no signal with the latest kernel from today Dec. 10.
Can you bisect?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #17 from bgunteriv@gmail.com --- i can try.
i can tell you that this started happening as soon as 3.13 started rolling out. so, i will try to start there.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #18 from bgunteriv@gmail.com --- which git repo should i use? https://www.kernel.org/
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #19 from Alex Deucher agd5f@yahoo.com --- Yes, please use Linus' git tree: http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #20 from bgunteriv@gmail.com --- 3017 revisions between 3.12 - 3.13rc1 wow!!! this is gonna take a while.
is there a faster way???
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #21 from bgunteriv@gmail.com --- i think i have found the culprit.
it was a merge that happened right before rc2.
i'm still bisecting, i have about 6 more steps to go.
so, hopefully by the end of the week, i will have a report.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #22 from bgunteriv@gmail.com --- i'm still testing.
this is my log so far: root@xbmc:~/linux# git bisect log # bad: [dc1ccc48159d63eca5089e507c82c7d22ef60839] Linux 3.13-rc2 # good: [5e01dc7b26d9f24f39abace5da98ccbd6a5ceb52] Linux 3.12 git bisect start 'v3.13-rc2' 'v3.12' # good: [5cbb3d216e2041700231bcfc383ee5f8b7fc8b74] Merge branch 'akpm' (patches from Andrew Morton) git bisect good 5cbb3d216e2041700231bcfc383ee5f8b7fc8b74 # good: [d8fe4acc88da8fbbe360b6592c9d0abbb85117dc] Merge branch 'akpm' (patch-bomb from Andrew Morton) git bisect good d8fe4acc88da8fbbe360b6592c9d0abbb85117dc # good: [4914d7b458e35a7db2f9c7dc6eb014620254bbbf] alpha: Use qemu+cserve provided high-res clock and alarm. git bisect good 4914d7b458e35a7db2f9c7dc6eb014620254bbbf # bad: [82023bb7f75b0052f40d3e74169d191c3e4e6286] Merge tag 'pm+acpi-2-3.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm git bisect bad 82023bb7f75b0052f40d3e74169d191c3e4e6286 # bad: [1ea406c0e08c717241275064046d29b5bac1b1db] Merge tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband git bisect bad 1ea406c0e08c717241275064046d29b5bac1b1db
---------------
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #23 from Lee Donaghy lee295012@gmail.com --- Created attachment 90699 --> https://bugs.freedesktop.org/attachment.cgi?id=90699&action=edit dmesg
Same problem.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #24 from Lee Donaghy lee295012@gmail.com --- Created attachment 90700 --> https://bugs.freedesktop.org/attachment.cgi?id=90700&action=edit Xorg
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #25 from Alex Deucher agd5f@yahoo.com --- Created attachment 90723 --> https://bugs.freedesktop.org/attachment.cgi?id=90723&action=edit possible fix
Does this patch fix the issue?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #26 from Lee Donaghy lee295012@gmail.com --- Created attachment 90731 --> https://bugs.freedesktop.org/attachment.cgi?id=90731&action=edit dmesg
Still broken.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #27 from Lee Donaghy lee295012@gmail.com --- Created attachment 90732 --> https://bugs.freedesktop.org/attachment.cgi?id=90732&action=edit Xorg
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #28 from bgunteriv@gmail.com --- Created attachment 90735 --> https://bugs.freedesktop.org/attachment.cgi?id=90735&action=edit first bad commit
so, it says that this is the first bad commit for me.
doesn't seem quite right, and this seems to deal with Samba...
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #29 from bgunteriv@gmail.com --- Created attachment 90736 --> https://bugs.freedesktop.org/attachment.cgi?id=90736&action=edit git bisect log
Here's my full git bisect log.
still not sure if I did this correctly. I never saw a 3.13 kernel build...?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
Alex Deucher agd5f@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #90723|0 |1 is obsolete| |
--- Comment #30 from Alex Deucher agd5f@yahoo.com --- Created attachment 90737 --> https://bugs.freedesktop.org/attachment.cgi?id=90737&action=edit possible fix
Does this patch work any better?
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #31 from Lee Donaghy lee295012@gmail.com --- Created attachment 90739 --> https://bugs.freedesktop.org/attachment.cgi?id=90739&action=edit dmesg
i now have a working screen.
i get this error in dmesg: [drm:dce4_afmt_write_speaker_allocation] *ERROR* Couldn't read Speaker Allocation Data Block: 0 but video and audio is working for me.
many thanks.
https://bugs.freedesktop.org/show_bug.cgi?id=72283
--- Comment #32 from bgunteriv@gmail.com --- yup... this patch works!!!
now we just gotta figure out the artifacts :-)
https://bugs.freedesktop.org/show_bug.cgi?id=72283
Martin Peres martin.peres@free.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |MOVED
--- Comment #33 from Martin Peres martin.peres@free.fr --- -- GitLab Migration Automatic Message --
This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.
You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/414.
dri-devel@lists.freedesktop.org
-
bugzilla-daemon＠freedesktop.org