Created attachment 90786 [details] Kernel log with monitor attached to display port I own an ASUS Nvidia GTX 660 card. If I connect a monitor to the HDMI connector, sound works. If I connect a monitor to the display port connector, sound does not work. In the latter case, all /proc/asound/NVidia/eld* show the output monitor_present 0 eld_valid 0 If I connect a monitor to the HDMI port, "monitor_present" shows "1". Kernel: 3.12.5 (Gentoo) Dmesg for the display port case (failure case): See attachment
Probably missing some initialization. Can you boot with 'nouveau.debug=debug' (which turns on all nv_debug messages) -- there's some stuff in DP init that might make sense to someone. [Unfortunately I know very little about DP.] If you'd like to push this further yourself, you could make an mmiotrace of the blob of a unplug/plug sequence, that might give someone some hints. See https://wiki.ubuntu.com/X/MMIOTracing for some hints as to how to start.
It's not expected to work actually. Nobody as of yet has cared at all, and I totally forgot about it, personally.
There appear to be some patches for it headed for linux-3.18. You can give http://cgit.freedesktop.org/nouveau/linux-2.6/log/?h=linux-3.18 a whirl and see if that makes it work.
How does patch reviewing work for things like this? It seems kind of like eldritch incantations.... Are there docs that explain what these mean or does this come from replaying register writes from the blob? ~ Patrick On Mon, Sep 15, 2014 at 12:23 PM, <bugzilla-daemon@freedesktop.org> wrote: > *Comment # 3 <https://bugs.freedesktop.org/show_bug.cgi?id=72717#c3> on > bug 72717 <https://bugs.freedesktop.org/show_bug.cgi?id=72717> from Ilia > Mirkin <imirkin@alum.mit.edu> * > > There appear to be some patches for it headed for linux-3.18. You can give > http://cgit.freedesktop.org/nouveau/linux-2.6/log/?h=linux-3.18 > > a whirl and see if that makes it work. > > ------------------------------ > You are receiving this mail because: > > - You are the assignee for the bug. > > > _______________________________________________ > Nouveau mailing list > Nouveau@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/nouveau > >
(In reply to comment #4) > Are there docs that explain what these mean or Of course (I hope!). But they're internal to NVIDIA and they have no real interest in sharing. Sometimes things can be found out from occasoinal code drops, e.g. the more recent gk20a android driver. (Of course that HW has no display component, so no use for something like this.) > does this come from replaying register writes from the blob? Seeing what the blob does + manual poking based on observed patterns.
On Mon, Sep 15, 2014 at 2:17 PM, <bugzilla-daemon@freedesktop.org> wrote: > *Comment # 5 <https://bugs.freedesktop.org/show_bug.cgi?id=72717#c5> on > bug 72717 <https://bugs.freedesktop.org/show_bug.cgi?id=72717> from Ilia > Mirkin <imirkin@alum.mit.edu> * > > (In reply to comment #4 <https://bugs.freedesktop.org/show_bug.cgi?id=72717#c4>)> Are there docs that explain what these mean or > > Of course (I hope!). But they're internal to NVIDIA and they have no real > interest in sharing. > > Hah. In other words, no. :| Oh well, keep up the awesome work guys!
DP audio should work starting with kernel 3.18. Feel free to reopen if it still doesn't.
Thanks, it works on kernel 4.0.5-gentoo. :-D
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.