Hi Edward & Yang-Rong, Thanks for the link - I gave it a try today.
On Thu, 2008-07-31 at 09:22 +0800, Edward Shu wrote: > Tim, > Have you tried the latest version in the following link? > > -------- ???????????? -------- > > ??????: > > Re: [Fwd: ] > > ??????: > > Thu, 31 Jul 2008 08:53:23 +0800 > > ?????????: > > Yang-Rong Zhou <Yangrong.Zhou at Sun.COM> > > > > > > > > > > > > > > Hi Garrett, > > > > The hiss exists in the old version, but hopefully it is reduced in the > > latest one. > > Have you updated your audiohd driver to the latest one? You can get the > > latest one in the following path: > > > > http://blogs.sun.com/sming56/entry/audiohd_driver_trial This is a bit more successful. The debug version of the 32-bit driver still doesn't load at all, the same missing symbols that I reported last time. Trying the non-debug 32-bit driver (the eeepc 701 isn't 64-bit, so I didn't test that driver) it loaded and attached correctly. Audio output works, both through the speaker and headphone socket. The monitor and gain controls also work fine using the buildin mic -- changing levels for monitor and gain, and making a noise near the mic show that we're definitely picking up audio as expected. Testing gnome-sound-recorder is less successful, it seems to detect either "Monitor" or "Gain" as input devices (which doesn't seem quite right) and is unable to record audio from the builtin mic. I tested suspend/resume, but the system was unable to suspend: Jul 31 09:38:08 beag genunix: [ID 535284 kern.notice] System is being suspended Jul 31 09:38:11 beag acpica: [ID 530733 kern.notice] NOTICE: acpica_ddi_setwake: could not evaluate _PRW Jul 31 09:38:11 beag genunix: [ID 122848 kern.warning] WARNING: Unable to suspend device pci1043,82a1 at 1b. Jul 31 09:38:11 beag genunix: [ID 537702 kern.warning] WARNING: Device is busy or does not support suspend/resume. Jul 31 09:38:13 beag unix: [ID 954099 kern.info] NOTICE: IRQ16 is being shared by drivers with different interrupt levels. Jul 31 09:38:13 beag This may result in reduced system performance. Jul 31 09:38:15 beag genunix: [ID 583038 kern.notice] System has been resumed. After the system was resumed, audio failed to function - audioplay just hangs Truss on "audioplay" showed: munmap(0xCE880000, 2407) = 0 close(4) = 0 ioctl(5, AUDIO_DRAIN, 0x00000000) (sleeping...) In mdb, I see: > ::ptree ! grep audioplay dfb06270 audioplay > dfb06270::walk thread | ::findstack -v stack pointer for thread d7388200: d63edb88 d63edbbc swtch+0x17d() d63edbe0 cv_wait_sig+0x12f(d955a3f4, d955a3ec) d63edc0c str_cv_wait+0x92(d955a3f4, d955a3ec, fffffc18, 0) d63edc74 strdoioctl+0x312(d955a3a0, d63edce0, 100002, 1, dbb00150, d63edf78) d63ede68 strioctl+0x3663(ed4ae800, 20004103, 0, 100002, 1, dbb00150) d63ede90 spec_ioctl+0x4b(ed4ae800, 20004103, 0, 100002, dbb00150, d63edf78) d63eded0 fop_ioctl+0x49(ed4ae800, 20004103, 0, 100002, dbb00150, d63edf78) d63edf84 ioctl+0x155() d63edfac sys_call+0x10c() I tried to modunload the audiohd driver to see if the system would suspend without audiohd, but it always reported that the device was busy. Is there a way to forcibly unload the driver even when using GNOME? [normally when using the OSS driver, the system suspends ok, just that audio doesn't function when the system resumes] Finally, I've included the message logs when booting the system with this version of the driver. Hope it's useful - would be great to get audio recording and suspend/resume working! cheers, tim Jul 31 09:22:32 beag audiohd: [ID 294501 kern.notice] NOTICE: Intel HD audio PCI 0x78 is: 0x0 Jul 31 09:22:32 beag audiohd: [ID 123783 kern.notice] NOTICE: Cannot no-snoop, ok Jul 31 09:22:32 beag audiohd: [ID 550065 kern.notice] NOTICE: power=0x0, is ok Jul 31 09:22:32 beag audiohd: [ID 829518 kern.notice] NOTICE: BAR0 pci reg 0x10 = 0xf7eb8004 Jul 31 09:22:32 beag audiohd: [ID 833715 kern.notice] NOTICE: bar1 pci reg 0x14 = 0x0 Jul 31 09:22:32 beag audiohd: [ID 901379 kern.notice] NOTICE: pci reg 0x40 = 0x3 Jul 31 09:22:32 beag audiohd: [ID 578956 kern.notice] NOTICE: pci reg 0x50 = 0xc8426001 Jul 31 09:22:32 beag audiohd: [ID 449975 kern.notice] NOTICE: pci reg 0x54 = 0x0 Jul 31 09:22:32 beag audiohd: [ID 194031 kern.notice] NOTICE: GCAP = 0x4401 Jul 31 09:22:32 beag audiohd: [ID 723211 kern.notice] NOTICE: in reset_controller: mask is 0x1 Jul 31 09:22:32 beag audiosup: [ID 107980 kern.info] NOTICE: audiohd0: codec 0: vid = 0x10ec0662, sid = 0x00000000, rev = 0x00100101 Jul 31 09:22:32 beag audiosup: [ID 199890 kern.notice] NOTICE: audiohd0: (DAC:6 Jul 31 09:22:32 beag audiosup: [ID 401376 kern.notice] NOTICE: audiohd0: 30:4) Jul 31 09:22:32 beag audiosup: [ID 199886 kern.notice] NOTICE: audiohd0: (DAC:2 Jul 31 09:22:32 beag audiosup: [ID 895635 kern.notice] NOTICE: audiohd0: 12:2 Jul 31 09:22:32 beag audiosup: [ID 397280 kern.notice] NOTICE: audiohd0: 20:4) Jul 31 09:22:32 beag audiosup: [ID 199888 kern.notice] NOTICE: audiohd0: (DAC:4 Jul 31 09:22:32 beag audiosup: [ID 284433 kern.notice] NOTICE: audiohd0: 14:2 Jul 31 09:22:32 beag audiosup: [ID 414803 kern.notice] NOTICE: audiohd0: 27:4) Jul 31 09:22:32 beag audiosup: [ID 199887 kern.notice] NOTICE: audiohd0: (DAC:3 Jul 31 09:22:32 beag audiosup: [ID 590034 kern.notice] NOTICE: audiohd0: 13:2 Jul 31 09:22:32 beag audiosup: [ID 528352 kern.notice] NOTICE: audiohd0: 21:4) Jul 31 09:22:32 beag audiosup: [ID 659424 kern.notice] NOTICE: audiohd0: 22:4) Jul 31 09:22:32 beag audiosup: [ID 283731 kern.notice] NOTICE: audiohd0: 26:4) Jul 31 09:22:32 beag audiosup: [ID 518718 kern.notice] NOTICE: audiohd0: The pin: 29 has no output path Jul 31 09:22:32 beag audiosup: [ID 518716 kern.notice] NOTICE: audiohd0: The pin: 28 has no output path Jul 31 09:22:32 beag audiosup: [ID 470141 kern.notice] NOTICE: audiohd0: Pin in:24 Jul 31 09:22:32 beag audiosup: [ID 470142 kern.notice] NOTICE: audiohd0: Pin in:25 Jul 31 09:22:32 beag audiosup: [ID 540096 kern.notice] NOTICE: audiohd0: Monitor Mic found! Jul 31 09:22:32 beag last message repeated 1 time Jul 31 09:22:32 beag audiosup: [ID 789852 kern.notice] NOTICE: audiohd0: Exclusive Moniter found success Jul 31 09:22:32 beag audiosup: [ID 743564 kern.notice] NOTICE: audiohd0: Share Moniter found success! Jul 31 09:22:32 beag last message repeated 3 times Jul 31 09:22:32 beag unix: [ID 954099 kern.info] NOTICE: IRQ16 is being shared by drivers with different interrupt levels. Jul 31 09:22:32 beag This may result in reduced system performance. Jul 31 09:22:32 beag npe: [ID 236367 kern.info] PCI Express-device: pci1043,82a1 at 1b, audiohd0 Jul 31 09:22:32 beag genunix: [ID 936769 kern.info] audiohd0 is /pci at 0,0/pci1043,82a1 at 1b Jul 31 09:22:32 beag audiosup: [ID 190399 kern.notice] NOTICE: audiohd0: Audiohd Version 1.3 attach finished successfully! > -------- ???????????? -------- > ??????: > Re: [Fwd: ] > ??????: > Thu, 31 Jul 2008 08:53:23 +0800 > ?????????: > Yang-Rong Zhou <Yangrong.Zhou at Sun.COM> > ?????????: > Garrett.Damore at Sun.COM > > > > > Hi Garrett, > > The hiss exists in the old version, but hopefully it is reduced in the latest > one. > Have you updated your audiohd driver to the latest one? You can get the > latest one in the following path: > /net/moonshot.prc/build2/yz224750/audiohd_ws/audiohddrv/ > or > http://blogs.sun.com/sming56/entry/audiohd_driver_trial > > > Regards, > Yangrong > > Garrett D'Amore ??????: > > > > > > > > ____________________________________________________________________ > > > > ??????: > > ?????????: Tim Foster <Tim.Foster at Sun.COM> > > ??????: Wed, 30 Jul 2008 15:09:13 +0100 > > ?????????: Garrett D'Amore <Garrett.Damore at Sun.COM> > > ?????????: Garrett D'Amore <Garrett.Damore at Sun.COM> > > ??????: eeepc-discuss at opensolaris.org > > > > Hi Garret > > > > ( & hi eeepc-interest at opensolaris.org! Bcc:d a sun.com alias about > > eeepcs - you guys should join the external alias too ) > > > > I tested out your new audiohd driver[1] on my Eee PC 701 laptop, which I > > have running 2008.11, the nv_93 update, which is bfu'd to onnv_95 (to > > get S3 suspend/resume, yes I'm impatient :-) > > > > Here's the audio device we have: > > > > pci bus 0x0000 cardnum 0x1b function 0x00: vendor 0x8086 device 0x2668 > > Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) High Definition > > Audio Controller > > > > Trying to boot with the debug version of the driver results in: > > > > Jul 30 10:22:01 beag genunix: [ID 819705 kern.notice] /kernel/drv/audiohd: > > undefined symbol > > Jul 30 10:22:01 beag genunix: [ID 826211 kern.notice] 'audio_tb_pos' > > Jul 30 10:22:01 beag genunix: [ID 819705 kern.notice] /kernel/drv/audiohd: > > undefined symbol > > Jul 30 10:22:01 beag genunix: [ID 826211 kern.notice] 'audio_tb_seq' > > Jul 30 10:22:01 beag genunix: [ID 819705 kern.notice] /kernel/drv/audiohd: > > undefined symbol > > Jul 30 10:22:01 beag genunix: [ID 826211 kern.notice] 'audio_tb_siz' > > Jul 30 10:22:01 beag genunix: [ID 819705 kern.notice] /kernel/drv/audiohd: > > undefined symbol > > Jul 30 10:22:01 beag genunix: [ID 826211 kern.notice] 'audio_tb_lock' > > Jul 30 10:22:01 beag genunix: [ID 819705 kern.notice] /kernel/drv/audiohd: > > undefined symbol > > Jul 30 10:22:01 beag genunix: [ID 826211 kern.notice] 'audio_trace_buffer' > > Jul 30 10:22:01 beag genunix: [ID 472681 kern.notice] WARNING: mod_load: > > cannot load module 'audiohd' > > > > Booting the non-debug driver is more interesting. On boot, we get a loud > > hissing being emitted from the system speakers (think air being let out > > of a bicycle tyre). > > > > According to gnome-volume-control, the sound level looks like it's about > > 75%. > > > > Doing an audioplay of /usr/demo/SOUND/sounds/bark.au plays the sample > > correctly, at the expected level, but with the same hissing in the > > background unfortunately, so it's hard to hear, but the sample does get > > played. > > > > Opening up gnome-volume-control, we see that built-in speaker and > > line-out outputs are selected. Changing the volume doesn't have any > > effect on the loudness of the hissing, but muting the audio turns it off > > completely. > > > > Turning off the line-out output also turns off the hissing completely, > > but while that's turned off, plugging a set of headphones into the > > headphone socket shows that the hissing is redirected to the headphones. > > > > Going to the "Recording" tab, I expected perhaps that we'd have monitor > > & gain turned up to max (hence the hissing), but this wasn't the case - > > so I'm not sure that's related. The gui shows one mic input. > > > > > > Finally, trying a suspend/resume with this loaded results in us being > > unable to suspend - messages are in the attached archive. > > > > > > I've attached some output that was produced during boot along with > > prtconf -v output for this machine. > > > > Hope this is useful? For now, I'll go back to the hdaudio OSS driver I'm > > using (which while it "works" ok, we don't get any audio > > post-s3-suspend, just several messages along the lines of: > > > > Jul 30 15:10:25 beag hdaudio: [ID 545374 kern.warning] WARNING: RIRB > > timeout (cad=0, nid=2, verb=f0d, parm=0) > > > > ) > > > > cheers, > > tim > > > > [1] > > http://gdamore.blogspot.com/2008/07/new-experimental-audiohd-driver.html > > > > > -- > Best Regards, > Ming. > > ------------------------------------------ > -Edward Shu > -Solaris x86 Engineering, Sun Microsystems > -Blog: http://blogs.sun.com/sming56/ > -tele: +86-10-62673100 > __________________________________________ >
