Bugzilla – Bug 446025
Headphones don't function properly on Dell Studio 15
Last modified: 2008-11-29 02:19:27 UTC
The anomalous headset behaviour is described as follows ... note the Studio 1535 has two headset output jacks one beside the other on the side of the laptop. The jack closer to the back of the laptop would play sound on the headset when plugged in, but the laptop's integrated speaker's volume would NOT be muted. The jack closer to the front of the laptop, when the headset was plugged in, would mute the laptop's integrated speakers, but there would be no sound in the headset. This makes the headset functionality not useful. It is important there is sound in the headset and that speaker volume be muted when the headset is plugged in. I documented my problem (which I believe is the same) here: http://forums.opensuse.org/1896927-post4.html [^] which is part of this thread of users with similar problems here: http://forums.opensuse.org/hardware/laptop/391698-dell-studio-15-sound-problems.html [^] As noted on that page, output of tsalsa script is here: http://nopaste.com/p/azPmEwv1T [^] openSUSE-11.1 beta5 comes with 2.6.27 kernel and 1.0.18 of alsa. The hardware audio codec of the Dell Studio 15 is IDT 92HD73C1X5 The Dell Studio 15 is also known as a Studio 1535. The above was observed on openSUSE-11.1 beta5 (KDE-4.1.2 liveCD) on my Dell Studio 15 (which is also known as a Dell Studio 1535). In my case I had to add "options snd-hda-intel model=dell-m6" to the /etc/modprobe.d/sound file to get any sound (restarting alsa with "rcalsasound restart" after making that edit). But as noted above, only speakers functioned properly and not the headsets. Reference the impact of the problem: Not having a functionally working headset is a big limitation for users who use a laptop for multimedia purposes, as one often likes to use the laptop in public places. But having blaring sound in public places is not desirable, and hence that aspect of the laptops usability is not present, due to this bug. If there are any diagnostic apps I can run to help sort this problem, please advise. I'll try to be better in monitoring my bug reports for this bug in particular (functional sound is very important to me). Finally, this could be an upstream alsa driver problem, as I annotated an alsa bug report by a Mandriva user (with what appears to be the same problem) here: https://bugtrack.alsa-project.org/alsa-bug/view.php?id=4198
First off, blame DELL that they never ever ship laptops with the correct BIOS... Basically you shouldn't need any model option as long as you have a correct BIOS setup. Please run /usr/sbin/alsa-info.sh with --no-upload option and attach the generated file. Don't use pastebin or whatever. They can be easily deleted at any time. So, use always attachments for bug reports. Then, try model=dell-eq option instead. The verb setup is slightly different from dell-m6. Also, try KOTD whether the problem still exists: ftp://ftp.suse.com/pub/projects/kernel/kotd/HEAD/ You can also try to build the latest ALSA driver modules from the snapshot tarball below: ftp://ftp.kernel.org/pub/linux/kernel/people/tiwai/snapshot/alsa-driver-snapshot.tar.gz In both cases, upload the alsa-info.sh output again.
Oh, and try *without* model option again with the latest KOTD before trying model=dell-eq option.
Thanks for the recommendations. I understand your two comments. It may take me a few days to do this. My previous investigations were based on the KDE-4.1 live CD of 11.1 beta5. So I will install 11.1 beta5 on my Dell Studio 15 and then follow your comments/recommendations to test. I'm currently sorting what the three separate Dell and Vista partitions on this laptop do, before I do my 11.1 beta5 install (which will replace the Vista MBR contents with the 11.1 Grub). Hopefully I'll have that sorted today or tomorrow, and then I'll give these sound investigations top priority.
I have exactly the same problems on a new DELL Studio 1537 notebook. A few additional data points: I installed the latest ALSA driver modules, but problems persist. In particular, with "model=dell-eq" I no longer get any sound, while I do get sound with "model=dell-m6". Also, routing of Audio CD data and internal microphone does not work. Internal microphone does seem to work without "model" module parameter (as observed by level display of krecord). Running diff on output of cat /proc/asound/card0/codec#0 with-out any model parameter, and with "model=dell-m6" shows interesting differences: --- codec.unknown 2008-11-18 13:39:04.000420000 +0000 +++ codec.dell-m6 2008-11-18 13:39:04.000586000 +0000 @@ -11,7 +11,7 @@ Default Amp-In caps: ofs=0x00, nsteps=0x03, stepsize=0x27, mute=0 Default Amp-Out caps: ofs=0x7f, nsteps=0x7f, stepsize=0x02, mute=1 GPIO: io=8, o=0, i=0, unsolicited=1, wake=1 - IO[0]: enable=0, dir=0, wake=0, sticky=0, data=0 + IO[0]: enable=1, dir=1, wake=0, sticky=0, data=1 IO[1]: enable=0, dir=0, wake=0, sticky=0, data=0 IO[2]: enable=0, dir=0, wake=0, sticky=0, data=0 IO[3]: enable=0, dir=0, wake=0, sticky=0, data=0 @@ -86,7 +86,7 @@ Pin Default 0x0321101f: [Jack] HP Out at Ext Left Conn = 1/8, Color = Black DefAssociation = 0x1, Sequence = 0xf - Pin-ctls: 0x00: + Pin-ctls: 0x40: OUT Unsolicited: tag=3f, enabled=1 Connection: 4 0x15* 0x16 0x17 0x1e @@ -127,7 +127,7 @@ Conn = Digital, Color = Unknown DefAssociation = 0x6, Sequence = 0x0 Misc = NO_PRESENCE - Pin-ctls: 0x20: IN + Pin-ctls: 0x00: Node 0x14 [Pin Complex] wcaps 0x400003: Stereo Amp-In Amp-In caps: N/A Amp-In vals: [0x00 0x00] @@ -135,16 +135,16 @@ Pin Default 0x4f0000f0: [N/A] Line Out at Ext UNKNOWN Conn = Unknown, Color = Unknown DefAssociation = 0xf, Sequence = 0x0 - Pin-ctls: 0x20: IN + Pin-ctls: 0x00: Node 0x15 [Audio Output] wcaps 0xd0c05: Stereo Amp-Out R/L Amp-Out caps: N/A - Amp-Out vals: [0x49 0x49] + Amp-Out vals: [0x64 0x64] Converter: stream=0, channel=0 Power: setting=D0, actual=D0 Delay: 13 samples Node 0x16 [Audio Output] wcaps 0xd0c05: Stereo Amp-Out R/L Amp-Out caps: N/A - Amp-Out vals: [0x2e 0x2e] + Amp-Out vals: [0x6d 0x6d] Converter: stream=0, channel=0 Power: setting=D0, actual=D0 Delay: 13 samples @@ -201,7 +201,7 @@ Amp-Out caps: ofs=0x00, nsteps=0x0f, stepsize=0x05, mute=1 Amp-Out vals: [0x0d 0x0d] Connection: 12 - 0x0a 0x0b 0x0c 0x0d 0x0e 0x0f 0x10 0x11 0x12 0x13 0x14 0x1d* + 0x0a 0x0b 0x0c 0x0d 0x0e 0x0f 0x10 0x11 0x12* 0x13 0x14 0x1d Node 0x21 [Audio Selector] wcaps 0x30090d: Stereo Amp-Out R/L Amp-Out caps: ofs=0x00, nsteps=0x0f, stepsize=0x05, mute=1 Amp-Out vals: [0x80 0x80]
Ok, I installed openSUSE-11.1 beta5 on my Dell Studio 15. A correction to my above post, ... after reading Hartwig Linux's post, I checked the documentation with my Dell, and noted it is also referred to as a Studio 1537 and not a 1535 as I incorrectly noted above. I did my testing from KDE-3.5.10. I used an alsa-conf built /etc/modprobe.d/sound file as my basis: alias snd-card-0 snd-hda-intel alias sound-slot-0 snd-hda-intel I tested it as above (no sound at all). Then I added the following line (and restarted alsa) "options snd-hda-intel model=dell-m6" and sound worked with same anamalous behaviour as before with headsets. I then replaced "dell-m6" with "dell-eq" in that file and restarted alsa. Contrary to what "Hartwig Linux" observed, I obtained the exact same behaviour with "dell-eq" and "dell-m6". This was with the "baseline" openSUSE-11.1 beta5 rpms and with the 2.6.27.5-2-default kernel. I'll try a kotd-head kernel next, and see if it makes any difference.
Created attachment 253206 [details] alsa-info script output with no model entry in /etc/modprobe.d file, run with HEAD kernel OK, using rpm -ivh I installed the 2.6.27.6-HEAD_20081118175713-default kernel (kernel-default, kernel-default-extra and kernel-default-base). I then rebooted to that kernel. I ran the alsa-info.sh script (attached) and I tested sound with no model option in the /etc/modprobe.d/sound file, with a model=dell-m6 option in the /etc/modprobe.d/sound file, and with a model=dell-eq option. In all cases with the 2.6.27.6-HEAD_20081118175713 kernel the behaviour was identical ... ie with this kernel, one need not specify an option to get speaker sound (unlike when booting to the 2.6.27.5-2 kernel). However the anomalous headset behaviour was also the same for those cases. I'll attempt a compile of my own alsa against the tarball you referenced either tomorrow or day after (I've done this before, so hopefully I have not forgotten how :) ). One additional note, while sound is playing, with the headphone plugged in the rear-most side headphone jack, if in kmix one mutes the "Headphone as line out" selection, then both speaker sound and the rear-most side headphone are muted at the same time. If the headphone is not plugged in the headphone jack, and in kmix one mutes the "Headphone as line out" selection, there is no affect - sound still plays on the speakers. Hence based on that observation it appears (to me) that there is some sort of autosensing there, but we have not guessed the Dell implementation yet.
OK, then model=dell-m6 must be the right model for your laptops. Let's forget about dell-eq model now. Could you write up the problem of your headset briefly in a few lines? Pointing a forum isn't helpful for debugging. A bug report should be self-consistent. Please always try to write a concise and brief summary. And, I still need the whole alsa-info.sh output. Load the driver with model=dell-m6, and run alsa-info.sh with and without headphones. Also -- don't forget to reset NEEDINFO flag after you provide the requested information. Thanks.
Created attachment 253563 [details] opensuse-11.1 beta 5 kde-3.5.10 kmix application SUMMARY: Studio 1537 Laptop has a feature with two headphone output jacks on the side. With the nominal openSUSE kernels, sound does not work in the headphones nor in the speakers. Integrated speaker Sound can be made to work in the speakers in the openSUSE-11.1 beta5 kernel (on KDE-3.5.10) with the option "model=dell-m6". The rear-most of the two side headphone operates simultaneous with the speakers. The front-most of the two side headphones does not work at all (ie sound never comes out of the front-most side headphone jack). DETAILS: With sound playing, when a headphone is plugged in the rear of the two jacks, the headphone works, but the integrated speaker sound on the laptop is not muted. But it should be muted. When the headphone is moved to the more front of the two jacks, that headphone does not work in the front jack, and also the integrated speaker sound is then muted (and hence one has NO sound). KMIX BEHAVIOUR: Reference kmix in KDE-3.5.10, I noted these unusual effects: While sound is playing, when the headphone is plugged in the rear-most side headphone jack, if in kmix one mutes the "Headphone as line out" selection, then both integraed speaker sound and the rear-most side headphone jacks' sound are both muted at the same time. ie there is NO sound. However if the headphone is NOT plugged in any headphone jack, and if in kmix one then mutes the "Headphone as line out" selection, there is no affect in this case (ie the sound still plays on the integrated speakers). I've attached a screen print of what the mixer looks like, which hopefully helps explain this. alsa-info script outputs to follow. ... I still need to compile alsa, so while I have reset the "provide info" flag, I still need to provide you more info
Created attachment 253564 [details] alsa-info script output with no headset connected.
Created attachment 253565 [details] alsa-info script output with headset plugged in rear-most headphone jack. Headphone sound works but mute does not function.
Created attachment 253566 [details] headphone plugged in front most headphone jack. All sound is muted. (ie NO sound).
I tried to compile the alsa in the tarball you referenced ( ftp://ftp.kernel.org/pub/linux/kernel/people/tiwai/snapshot/alsa-driver-snapshot.tar.gz ) against the 2.6.27.6-HEAD_20081118175713 kernel, but I failed. ./configure worked, but the "make" aborted early with this error: ========================================= In file included from /home/lee/tars/alsa-driver/acore/hrtimer.c:1: /home/lee/tars/alsa-driver/include/adriver.h:1598:1: warning: "page_to_pfn" redefined In file included from include2/asm/page.h:196, from include2/asm/paravirt.h:7, from include2/asm/irqflags.h:55, from /usr/src/linux-2.6.27.6-HEAD_20081118175713/include/linux/irqflags.h:57, from include2/asm/system.h:11, from include2/asm/processor.h:17, from /usr/src/linux-2.6.27.6-HEAD_20081118175713/include/linux/prefetch.h:14, from /usr/src/linux-2.6.27.6-HEAD_20081118175713/include/linux/list.h:6, from /usr/src/linux-2.6.27.6-HEAD_20081118175713/include/linux/module.h:9, from /home/lee/tars/alsa-driver/include/adriver.h:57, from /home/lee/tars/alsa-driver/acore/hrtimer.c:1: /usr/src/linux-2.6.27.6-HEAD_20081118175713/include/asm-generic/memory_model.h:78:1: warning: this is the location of the previous def inition make[4]: *** [/home/lee/tars/alsa-driver/acore/hrtimer.o] Error 1 make[3]: *** [/home/lee/tars/alsa-driver/acore] Error 2 make[2]: *** [_module_/home/lee/tars/alsa-driver] Error 2 make[1]: *** [sub-make] Error 2 make[1]: Leaving directory `/usr/src/linux-2.6.27.6-HEAD_20081118175713' make: *** [compile] Error 2 ========================================= I'm hoping my being able to sucessfully compile this kernel is not critical to the solution of this Dell Studio 15 headset problem. :)
Created attachment 253640 [details] codec#0.svg is output from codec graph when pc boot to 2.6.27.5-2 default kernel with model=dell-m6 in etc/modprobe.d/sound file The text file attached is an .svg file (can be displayed with gimp). If this does not attach well, I can convert it to a .pdf file. This is the output of codecgraph from this web site: http://helllabs.org/codecgraph/ I don't know if this will be of any help, but I ran it anyway when booted to the 11.1 beta5 default 2.6.27.5-2 kernel. I can also run it from the 2.6.27.6-HEAD_20081118175713 kernel if that is deemed to be more useful.
Thanks for information. "Headphone as Line Out" switch is the option to make an extra headphone (i.e. rear one) behaving as (surround) outputs. In your case, this should be off. Then both front and rear headphones are handled as headphones, i.e. mutes the speaker output when plugged. In short, the only remaining problem is that the headphone output doesn't work, no matter which one. I guess this is a GPIO setup problem and is already fixed in the latest ALSA driver. Thus it'd be really helpful if you can build the driver manually and test it. About the compilation of alsa-driver package: Did you install the corresponding kernel source? It's safer to back to the original kernel-pae and kernel-source packages from the latest beta. HEAD packages are updated too frequently.
Yes, I did install the corresponding kernel source (ie I had the 2.6.27.6-HEAD_20081118175713 kernel-source) . However I can not recall if I installed kernel-syms nor linux-kernel-headers. I think I had those installed for the 2.6.27.5-2 kernel, but not for the linux-2.6.27.6-HEAD_20081118175713 kernel. I have to run to work. But I'll restore kernel-source back to the 2.6.27.5-2 kernel and I'll try to compile alsa against the 2.6.27.5-2 kernel tonight after work. I'll leave the "NEEDINFO" status flagged on this thread, until I know the results of tonight's compilation efforts.
Note that you don't need kernel-syms.rpm usually unless you want to build the modules for all kernel flavors. Building a module only for your running kernel, kernel-source.rpm alone should suffice. linux-kernel-headers.rpm is for user-space programs. This isn't used for compiling a kernel, but it must be harmless (and required by glibc-devel).
I have the same experience with my Dell Studio 1535 laptop. Referring to Comment #2 I happily blame Dell for the wrong BIOS, but it it does not solve the problem, and on the other hand the sound with Ubuntu 8.04.1 work perfectly out-of-box. It means that the problem is solvable from the Linux side somehow. Unfortunately I replaced Ubuntu with openSUSE 11.1 beta 5 since then, so I can't send Ubuntu configs. I may reinstall Ubuntu again and send any information that can help.
Some other weird things: - There are two sound cards: 0. 82801H (ICH8 Family) HD Audio Controller 1. RV620 Audio device [Radeon HD 34xx Series] Earlier when I logged in into KDE there were no sound but an error message: sound 0. card (HDA Intel in the KMix) could not be initialized so sound is switched to the other card (HDMI, which has no sound, anyway). But at the next sound event the sound 0 card worked again. What is more weird: from this morning there is no error message and I can hear the login sound of KDE.
The KDE problem is likely irrelevant. Please avoid to mix things up in a single bug report. As mentioned, the problem could be already fixed in the latest ALSA driver code, but I didn't put it to 2.6.28 kernel material (note that SUSE11.1 kernel has backport patches from 2.6.28-rc) because it wasn't obvious. So, please try to build the latest alsa-driver snapshot, and check whether it works. Don't forget to turn off "Headphone as Line Out" switch, BTW.
I donloaded, compiled and installed the latest alsa driver (1.0.18a from Nov. 12.) but the effect is the same. Let me mention again that the sound stuff works perfectly out-of-box with Ubuntu 8.04.1, which probably contains a way older alsa driver (it might be 1.0.12 or even 1.0.8) and a way older kernel (maybe 2.6.24).
1.0.18a is way too old. Try the *snapshot* tarball in comment#1. I know that it worked on Ubuntu (and that they don't give any feedback to the upstream). But, I, as the upstream kernel maintainer, am interested in fixing the upstream bugs first. And, this should have been fixed there. That's why I've been asking to try the latest snapshot version. Once after we confirm it's working, then we can backport the patch to SUSE kernel.
Sorry, but I can not get the alsa-driver-snapshot to compile on openSUSE-11.1 beta5. Out of curiousity, I downloaded it to my openSUSE-10.3 PC and it compiled successfully. So there is some quirk wrt the openSUSE-11.1 factory development applications, that I don't understand. Possibly related to the kernel, as I successfully compiled codec-graph on the 11.1 beta5 PC with no problem. But the alsa-driver-snapshot keeps giving me the same error during the "make" (that I noted above).
OK, this seems to be a problem of configure script in alsa-driver. I checked only x86-64 recently with SUSE kernels. Now I fixed it, and this should work finally even with i386. Give it a try again.
BTW, Mészáros, could you also attach your alsa-info.sh output? It might be that you have a slightly different PCI SSID. Thanks.
Created attachment 254393 [details] The requested output of alsa-info.sh Sir, yes, Sir!
BTW, Hungarians do everything the opposite way, so Mészáros (Butcher) is my family name, Gyula (thought to be in connection with Julius, but it is not) is my given name. :-)
Thanks, Gyula (interestingly, Japanese usually writes the name in the order of "surname firstname" in the native language, but writes as "firstname surname" in English or in other latin letters...) Actually, your device is a different model from Lee's, and you don't have to pass model=dell-m6 option. And, you reported that it's also Dell Studio 1535, right?
In the meantime, I updated to the 2.6.27.5-3 kernel ... maybe I should not have as it creates a moving configuration. I can go back to the older 2.6.27.5-2 kernel if it will help. I downloaded the update alsa-driver-snapshot and I was able to ./configure and compile with make. However instead of running "make install" I ran "checkinstall" to create a custom rpm (easy for tracking on my own PC). But that rpm would not install, as I obtained massive dependency errors with kernel (you can see the errors here) default-base: http://pastebin.be/14983 [post good for 1 month] I'll try again (this time with make install) and see if that makes any difference. I'll keep the flag "NEEDINFO" set until I finish trying "make install" with the 2.6.27.5-3 kernel. If that fails, I'll go back to the older 2.6.27.5-2 kernel and try.
"make install" (with the 2.6.27.5-3 kernel) fails with: linux-brlh:/home/lee/tars/alsa-driver # make install if [ -L /include/sound ]; then \ rm -f /include/sound; \ ln -sf /home/lee/tars/alsa-driver/include/sound /include/sound; \ else \ rm -rf /include/sound; \ install -d -m 755 -g root -o root /include/sound; \ for f in include/sound/*.h; do \ install -m 644 -g root -o root $f /include/sound; \ done \ fi install: cannot stat `include/sound/*.h': No such file or directory make: *** [install-headers] Error 1 ========= I'll go back to the older kernel and try.
Cannot reproduce on my system. You don't have to install headers. Just run "make install-modules" to install only modules.
(In reply to comment #27 from Takashi Iwai) > Actually, your device is a different model from Lee's, and you don't have to > pass model=dell-m6 option. And, you reported that it's also Dell Studio 1535, > right? > Yes, Iwai San, it is Dell Studio 1535, according the sticker on the bottom. Lee has written that his model is really 1537, see Comment #5.
Created attachment 254436 [details] alsa-info script output with alsa-driver-snapshot installed via checkinstall created rpm Ok, I successfully compiled alsa-driver-snapshot against the 2.6.27.5-2 kernel. Again I used checkinstall. I installed it, rebooted, and lost all sound :) ... Maybe I should not have used checkinstall. I tried configuring with alsaconf, but still no sound. Speaker test gives errors: lee@linux-brlh:~> speaker-test -c2 -l10 -twav speaker-test 1.0.18 Playback device is default Stream parameters are 48000Hz, S16_LE, 2 channels WAV file(s) ALSA lib confmisc.c:768:(parse_card) cannot find card '0' ALSA lib conf.c:3513:(_snd_config_evaluate) function snd_func_card_driver returned error: No such file or directory ALSA lib confmisc.c:392:(snd_func_concat) error evaluating strings ALSA lib conf.c:3513:(_snd_config_evaluate) function snd_func_concat returned error: No such file or directory ALSA lib confmisc.c:1251:(snd_func_refer) error evaluating name ALSA lib conf.c:3513:(_snd_config_evaluate) function snd_func_refer returned error: No such file or directory ALSA lib conf.c:3985:(snd_config_expand) Evaluate error: No such file or directory ALSA lib pcm.c:2202:(snd_pcm_open_noupdate) Unknown PCM default ============ So I tried configuring with YaST > Hardware > Sound and obtained classic error: "An error occured during the installation of 82801I (ICH9 Family) HD Audio Controller The kernel module snd-hda-intel for sound support could not be loaded. This can be caused by incorrect module parameters, including invalid IO or IRQ parameters." I tried removing /etc/modprobe.d/sound file and restart YaST , but same error. I noted "rcalsasound" was missing. I'm beginning to think compilation and build, although it finished, was incomplete (with checkinstall). The script alsa-info.sh is attached.
Created attachment 254453 [details] alsa-info.txt output with alsa-driver-snapshot installed (correction to post#32) Ignore attachment in post #32. I attached wrong file. Apologies. Correct alsa-info output attached, run with alsa-driver-snapshot installed.
Re: comment#30 > Cannot reproduce on my system. > You don't have to install headers. Just run "make install-modules" > to install only modules. I've restored my PC back to the original beta5 alsa-2.5.27.5-2, and also restored the original alsa rpms that came with that. I removed (I think) all traces of the previous alsa-driver-snapshot attempts. I'll try compiling alsa-driver-snapshot one last time, this time with "make install-modules". I should be able to post the results within an hour or so.
Created attachment 254465 [details] alsa-info.txt output with 2.6.27.5-2 kernel with alsa-driver-snapshot installed. No sound from any headset I installed the alsa-driver-snapshot with ./configure, make, make install-modules. I then ensured model=dell-m6 was in /etc/modprobe.d/sound file, rebooted and tested audio. I have only speaker sound. No audio at all on any headphone jack. DETAILS Before installing alsa-driver-snapshot I had audio on the rear-most side headphone jack. Now there is NO audio from that headphone jack. Also, before when plugging the headphone into the front-most side headphone jack, the speaker sound was muted, and there was no sound from the headphone hack. Now when plugging in the headphone into that jack, there is no affect on the speakers (sound keeps playing) and there is still no sound from that jack. I've attached the latest alsa-info.sh output with this configuration.
One piece of strange behaviour I noted with post #35. This is reliably reproduceable. One needs to follow a special routine to get sound functioning. Immediate after every reboot, there is no sound on headset-nor-speakers with the configuration with alsa-driver-snapshot installed via the "make install-modules" (above) with the model=dell-m6 in the /etc/modprobe.d/sound file. Checking the mixer indicated all settings are ok. Plugging the headset into the side-rear most headphone hack has no affect, still not sound on headset nor speaker. But plugging the headset into the side-front-most headset jack, gives no sound initally, but as soon as the headphone is removed from the jack, speaker sound starts (but still no headset sound). This is the only way to "kickstart" the speaker sound in this new "alsa-driver-snapshot" configuration.
A new development after playing with the KDE mixer in KDE-3.5.10 on openSUSE-11.1 beta5 with the alsa-driver-snapshot and the 2.6.27.5-2 kernel. When playing sound (through the speakers) if I plug in the headset to the front-most side headphone jack, the speaker sound keeps playing and there is no sound (as noted above). However if in the kmix I select mute on the "headphones as line out" the speaker sound stops, and the headphone sound starts. If I then "unmute" the "headphones as line out" the speaker remains quiet and the headphone continues to play. So this bizarre workaround is a BIG step forward. :) With the alsa-driver-snapshot tarball installed I can no longer get sound out of the rear-most slide headphone jack. It does appear the mapping ( ? ) of the headphone controls needs work, and I don't know where to start, nor how to do this.
Created attachment 254499 [details] alsa-info.txt with alsa-driver-snapshot, 2.6.27.5-2 kernel, and front-most side headphone working As noted in above post, the attached alsa-info.txt is with alsa-driver-snapshot, 2.6.27.5-2 kernel, and front-most side headphone working and speaker muted. rear-most headphone is not working. Toggling the kmix "headphone as line out" seemed to have initiated this headphone functionality (with speakers muted). I hope this file is of some help.
Lee, read comment#14. "Headphone as Line-out" should be OFF, muted. You just turned a wrong switch on, like a famous self-destruction button. I seriously think to remove this switch for dell model...
I note your comment #14 and your comment #39. I acknowledge "Headphone as Line-out" should be muted. I did not express my observation well. I think what I have observed is important, and I am not sure I epxressed this well enough for you to understand what I observed. With the alsa-driver-snapshot applied and dell-m6 in the /etc/modprobe.d/sound file, there is no sound upon boot. None. Adjusting mixer settings has no effect (I left "Headphone as Line-out" muted OFF at this stage and did not touch it). There is no speaker sound. If one plugs a headphone in/out of a headset jack from one to a few times, sound then starts to work on the speakers. There is still no headset sound, no matter what mixer setting is applied (leaving the "Headphone as Line-out" untouched). Now the behaviour I observed, in this state where ONLY speaker sound functions, is if one changes the software switch state (and note I say change the state and not mute nor unmute) of the "Headphone as Line-out" switch in the mixer WHEN THE PHYSICAL STATE of the front-most side-headphone-jack is changed, then sound will switch from the Speaker to the headphone, or visa versa. This is only true for the frontmost side headphone jack. The rear most headphone jack has no affect on this, and from now on when I state headphone, I am only referring to the front headphone, as the rear has no affect. For example (1) Sound coming out of speaker, no headphone attached. Of course, only sound from speaker. Plug in headphone jack and there is no sound in headphone. Sound in speaker still. Unplug headphone. Still sound in speaker. Replug in headphone. No sound in headphone but still sound in speakers. Now physical state (speaker plugged in) is different as headphone is plugged in while sound is coming from speaker. Change state of "Headphone as Line-out" switch (ie reverse what it was) and sound changes from speaker to headphone. Now only sound comes from headphone. If one unplugs headphone. No sound at all. That leads to ... (2) Sound coming out of head phone. No speaker sound. Remove headphone. No speaker sound. Plug in headphone, have headphone sound and still no speaker sound. Remove headphone. Still no speaker sound. Now there is a physical state change (as no headphone is attached) but we know sound works with the headphone. Change state of "Headphone as Line-out" switch (ie reverse whatever its current state might be) and sound now comes out of speaker. Plug in headphone, and no headphone sound and speaker still has sound. One can repeat this over and over. Its highly repeatable. Note ONLY this switch "Headphone as Line-out" has any effect in changing sound from headphone to speaker. NOTHING else works. So to change sound from headphone to speaker and visa versa, one has to change the state of the switch one is not supposed to touch ( "Headphone as Line-out" ) in order to change the sound from Speaker to Headphone, and visa versa. I understand that switch should NOT be used to change sound from speaker to headset, but currently it is the ONLY switch that works with the alsa-driver-snapshot. I hope I succeeded better in the above in explaining my observations. I wish there was more that I could do to help you debug this, but my knowledge in applying special settings to the config files is very limited here, and I have no knowledge of the inside of the alsa driver.
I don't know if it will help, but I found on the web this detailed page on the hardware audio codec ( IDT 92HD73C1 ) that is on the Dell Studio 15. http://www.idt.com/?genID=92HD73C1 with datasheet, design documents, application notes, white paper ... etc ...
Reference comment #27 and #31, I requested from Dell a Dell Studio-15 laptop. The precise 1535 or 1537 model number was not important to me when ordering. (It is important for other reasons). The sales and order receipts state that I ordered and state that Dell delivered a Dell 1537. When I enter the Dell Service Tag (glued on the bottom of this laptop) on the Dell web site, it states that I have a Dell 1535. ... < go figure > :) ... Since I ordered a Dell laptop with English language keyboard/Visa/MS-Works/documentation in Germany, I think that may explain the Dell Germany confusion. I'm going to check the BIOS to see if it gives any superior indication as to my precise model. I note my BIOS is A03 and there is an update to A04 for the Dell 1535 and an update to A05 for the Dell 1537. Before updating to a more recent BIOS, I need to sort exactly which model I have.
Reference Comment#42, the BIOS on my Dell Studio 15 clearly states: Product Name: Studio 1537. Assuming that the BIOS is correct, that is rather shoddy database population on the part of Dell web site people. I could have installed the wrong updated BIOS.
OK, then do the following to understand what's happening. 1. Clean reboot once to runlevel 3 as root: don't login to graphical desktop. This is to avoid any stupid sound-subsystem mess. Don't touch any mixer at this stage. 2. Try to play via "aplay -vv YOURFILE.wav". Check speaker, front hp and rear hp outputs. 3. Make sure that "headphone as line-out" off. Get alsa-info.sh without any headphones. 4. Plug the front HP, get alsa-info.sh, make a diff to the file of #3. (At best, run "diff -up old-alsa-info new-alsa-info") 5. Plug the rear HP, get alsa-info.sh, make a diff to the file of #3. 6. At this stage, if you have the silent problems, try the procedure you described above to get the sound back. Once if you get the sound, get alsa-info.sh to make a diff. Basically, it's to see the difference of HD-audio codec registers between different states. It's the only way to see what's really difference. BTW, the latest alsa-driver snapshot won't create "headphone as line-out" switch for laptops. Thus try the above with the older snapshot version first.
Created attachment 254919 [details] Difference between no sound with no HP plugged in, and speaker sound with front HP plugged in I rebooted the PC to run level-3. I have the alsa-driver-snap shot from 21-Nov-08 still installed. I executed aplay -vv test.wav with no headphones (HP) connected. There is no sound from the speaker. I let the playing of the .wav complete (still no sound). I ran alsa-info.sh to establish a baseline (and renamed the file to 02-speakers-nosound.txt). I then plugged in the headphone (HP) into the front HP jack. I ran aplay -vv test.wav. I immediately had sound from the speaker and no sound from the plugged in headphone. When the .wav playing completed, I ran alsa-info.sh (and then renamed the file to 03-front-hp-with-speaker-sound.txt. I then plugged the HP into the rear HP jack. I ran aplay -vv test.wav. I still had sound from only the speaker and no sound from the plugged in headphone. When the .wav playing completed, I ran alsa-info.sh (and then renamed the file to 04-rear-hp-with-speaker-sound.txt. I ran amixer > amixer.txt to record the mixer state. I note this setting: Simple mixer control 'Headphone as Line Out',0 Capabilities: pswitch pswitch-joined Playback channels: Mono Mono: Playback [off] I then ran “diff -up 02-speakers-nosound.txt 03-front-hp-with-speaker-sound.txt > difference02-to-03.txt I then ran “diff -up 02-speakers-nosound.txt 04-rear-hp-with-speaker-sound.txt > difference02-to-04.txt difference02-to-03.txt is attached to this post. I will attach difference02-to-04.txt in a follow on post. Note, at no time did I have sound coming from the headset.
Created attachment 254920 [details] Difference between no sound with no HP plugged in, and speaker sound with rear HP plugged in Second file associated with post#45 (rear Head Phones comparison this time).
Created attachment 254926 [details] This file is a summary of differences from alsa-info.sh script run in Run Level 5 with different headphone states and a mixer change This is a test you did NOT ask for in Run Level 5 where I obtained both speaker sound and then headphone sound (by changing state of a mixer parameter that you do no like me to touch). You may wish to ignore this test, but I thought the comparison of speaker vs headset sound might be of interest. Again with 21-Nov-08 alsa-driver-snapshot installed. I booted to run level 5 (KDE-3.5.10) and ran in a konsole aplay -vv test.wav with no Headphone (HP) connected. I had no sound from the speakers. I then ran alsa-info.sh and renamed the file to 11-rl5-no-speaker-sound.txt. I waited for test.wav to complete. I then plugged the HP in the front HP jack and executed aplay -vv test.wav. There was speaker sound only. (No HP sound). I then ran alsa-info.sh and renamed the file to 12-rl5-hp-front-with-speaker-sound.txt. I then unplugged the HP. I executed the aplay -vv test.wav with no HP connected. There was still speaker sound. I then ran alsa-info.sh and renamed the file to 13-rl5-no-hp-but-still-speaker-sound.txt Then with HP unplugged, in kmix I changed the state of the “Headphone as Line Out”. Note the HP was still not connected. I then executed the aplay -vv test.wav. There was no speaker sound. I then plugged the HP in the front HP jack and executed the aplay -vv test.wav. This time there was HP sound and still no speaker sound. I then ran alsa-info.sh and renamed the file to 14-rl5-front-hp-sound-but-no-speaker-sound.txt I then ran diff -up 11-rl5-no-speaker-sound.txt 12-rl5-hp-front-with-speaker-sound.txt > difference11-to-12.txt I then ran diff -up 11-rl5-no-speaker-sound.txt 13-rl5-no-hp-but-still-speaker-sound.txt > difference11-to-13.txt I then ran diff -up 11-rl5-no-speaker-sound.txt 14-rl5-front-hp-sound-but-no-speaker-sound.txt > difference11-to-14.txt I then ran diff -up 12-rl5-hp-front-with-speaker-sound.txt 14-rl5-front-hp-sound-but-no-speaker-sound.txt > difference12-to-14.txt I combined those 4 diff outputs into one file called run-level-5-differences.txt. I thought the last diff in particular (functional speaker sound vs functional headset sound) might be of interest.
Thanks, this problem might be deeper than I originally thought. The codec diff shows that the whole sections didn't appear in the first shot. It implies some missing power-management bits. The patch below is against the latest alsa-driver tree. Grab the latest alsa-driver-snapshot tarball, and apply it on alsa-driver/alsa-kernel directory with -p2 option: % cd alsa-driver/alsa-kernel % patch -p2 < /somewhere/stac-more-fixes.diff then run make, and "make install-modules" as usual. And check again the procedure above. At this time, there will be no "Headphone as line-out" switch. If the first alsa-info.sh output (or /proc/asound/card0/codec#* file) doesn't show any valid codec contents (e.g. GPIO bits are shown as 255), then something is still wrong.
Created attachment 254995 [details] More fixes of patch_sigmatel.c
Not sure how helpful it is to jump in at the current time, but my problem seems identical to Lee's. At least his observations match mine, as far as I can tell. There might be subtle differences in the hardware of our notebook, however. Have built the latest snapshot with the patch patch_sigmatel.c, and run the test-cases. Summary: 1. without HP plugged in: no sound from speakers 2. with HP plugged into both front/rear HP Jack, I do get sound from the HPs This is both times without specifying any model option. Another interesting observation. With a snapshot from around Nov.18, I did get sound from speakers(without HP), which was muted if I plugged a HP in front Jack. In this case, I also got sound on the HP, so this was close to intended behavior. However, neither MIC Jack, Audio CD nor built-in MIC worked. Attached are the output from alsa-info.sh for the 3 cases: 1. alsa-info.without-hp.txt (no headphones plugged in) 1. alsa-info.hp-front.txt (headphones plugged into front HP Jack) 1. alsa-info.hp-middle.txt (headphones plugged into rear HP Jack) Hope this helps, and thanks for the efforts, Hartwig
Created attachment 255077 [details] alsa-info.sh output without HP plugged in
Created attachment 255078 [details] alsa-info.sh output with HP plugged into front jack
Created attachment 255079 [details] alsa-info.sh output with HP plugged into rear HP jack (middle jack)
(In reply to comment #50 from Hartwig Linux) > Another interesting observation. With a snapshot from around Nov.18, I did get > sound from speakers(without HP), which was muted if I plugged a HP in front > Jack. In this case, I also got sound on the HP, so this was close to intended > behavior. However, neither MIC Jack, Audio CD nor built-in MIC worked. To be more specific, this was with specifying "model=dell-m6 option". Without that option, krecord's level indicator that the internal MIC recorded sound (but none of the outputs worked).
The model option won't be needed with the latest kernel, but still needed for beta5. Anyway, more testers are welcome. Can any of you test the patch in comment#49 ? Also, kernel packages with some fix patches are available at: ftp://ftp.suse.com/pub/people/tiwai/11.1-kernel-test/ Please test it if possible.
(In reply to comment #55 from Takashi Iwai) > Can any of you test the patch in comment#49 ? Please note that the alsa-info.sh output above was generated with that patch applied. > Also, kernel packages with some fix patches are available at: > ftp://ftp.suse.com/pub/people/tiwai/11.1-kernel-test/ > Please test it if possible. I'll try todo this tonight/tommorow, no time at the moment, unfortunately.
OK, then with the latest snapshot + patch, you don't get the speaker no matter what you do, right?
(In reply to comment #57 from Takashi Iwai) > OK, then with the latest snapshot + patch, you don't get the speaker no matter > what you do, right? Yes, no speaker output whatsoever. Did not try any model options, though.
Just to be sure, turn "Surround Playback Switch" on. If the patch doesn't work as expected, we need to dig with a lowlevel too. Get hda-verb tarball from ftp://ftp.kernel.org/pub/linux/kernel/people/tiwai/misc/hda-verb-0.3.tar.gz Extract, and run make. Now, run the program as root like: # ./hda-verb /dev/snd/hwC0D0 0x01 0xfec 0 and show the output. Then, run # ./hda-verb /dev/snd/hwC0D0 0x01 0x7ec 0 and try aplay.
I guess I found a part of the problem. Now I updated the patch, merged into the git tree, and updated the snapshot tarball again. Please try the latest alsa-driver-snapshot tarball (without patch), and let me know if it works or not.
(In reply to comment #60 from Takashi Iwai) > Please try the latest alsa-driver-snapshot tarball (without patch), and let me > know if it works or not. Success with that snapshot! At least from a quick test. Tested 3 scenarios: 1. without headphones: sound through speaker 2. with headphones in front jack: speaker muted, sound in HP 2. with headphones in rear jack: speaker muted, sound in HP I'll test MIC/Audio CD input etc later today or tomorrow. Thanks for your efforts!
Good to hear finally something positive :) The audio CD won't work as analog input, likely, as most laptops don't have a cable from the DVD-ROM drive. The mic input is more interesting topic. In the very latest alsa-driver snapshot, now there are three different variants of dell-m6 models (check ALSA-Configuration.txt in the tarball): dell-m6-amic -- with analog mics dell-m6-dmic -- with digital mics dell-m6 -- with both analog and digital mics I guessed Studio 1537 as dell as -dmic variant. If it doesn't work, let me know, and try other model options.
(In reply to comment #62 from Takashi Iwai) > The audio CD won't work as analog input, likely, as most laptops don't have a > cable from the DVD-ROM drive. In windows, it does work, so there should be a cable. Or does windows extract the audio data? No Audio CD around to test at the moment, though, and that's definitely a minor point. > The mic input is more interesting topic. ... > I guessed Studio 1537 as dell as -dmic variant. If it doesn't work, let me > know, and try other model options. It does work, without any options. So the guess was right, at least for my notebook. Once more thank you for your efforts!, Greetings, Hartwig
I just got back from work, so I skipped the post #48/49 test, and went direct to the post#60 fix, running ./configure, make, and make install-modules. Then rebooted. I'm running 11.1 beta5 with 2.6.27.5-2 kernel on a 32-bit install on kde-3.5.10 on a Studio 1537. After a reboot, I obtained the same successful behaviour as noted by Hartwig Linux in his post#61. Takashi Iwai well done and thankyou very much. Hartwig Linux, thankyou for your input as well. I can't possibly match your superb testing turn around times. Needless to say I am quite happy about seeing the patch and I'm looking forward to the updated alsa, when this makes it into the upstream. I agree with the RESOLVED for this bug. One minor note (and this could be a kde-3.5.10 kmix problem) when successfully playing sound in the speakers, the kmix speaker mute switch has no effect. It mutes nothing. On the other hand, the analog loop back, which is normally muted, if "unmuted" will immediately mute speaker sound, or either headset sound (depending on which one's are playing). The kmix analog loopback1 and analog loopback2 have not effect. Please note I don't care about this kmix behaviour (I'm just reporting it for completeness). I'm very happy about this fix. How long do you think it will take before we see this in a "configuration controlled" release of an alsa update?
I can confirm the external mic input jack works for an external microphone. I just tested it a few minutes ago. My /etc/modprobe.d/sound file has no special options: alias snd-card-0 snd-hda-intel alias sound-slot-0 snd-hda-intel Again, the external input mic works on a Dell Studio 15 (1537 model) with the fix (alsa-driver-snapshot) noted in post #60 on openSUSE-11.1 beta5 with the 2.6.25.5.2-default kernel (32-bit) under KDE-3.5.10. I used kmix to setup the mic. My record test was: arecord -d 10 myrecording.wav There are two internal mics on this laptop, one on either side of the webcam. Since I have not setup the webcam, I have not tried to test those mics. Since sound is now working, I am going to change my configuration, and update my 11.1 beta5 with some newer releases (like the newer kernel). Thanks again for you work on this, and I hope to see this fix in an upstream alsa version some time soon (hopefully before Christmas :) ).
Thanks for checking. So far it sounds good. The fixes are already in the upstream git tree, so they'll be certainly to the next release. But I'm not sure when the next release will happen :) Note that the fix patches slipped from 11.1-rc1 deadline, thus rc1 kernel is still buggy for this device. But the patches shall go into rc2. KOTD package already contains the fix patches. If you have time, give it a try instead of the manual build of alsa-driver. If it (or upcoming rc2 kernel) doesn't work, please reopen the bug.
I've just installed the latest (11.27.) alsa build from Build Service and the headphones work well now from both jacks of my Dell Studio 1535. Thanks for the hard work!
I installed openSUSE-11.1 RC1 with KDE-3.5.10. I confess I did not even bother to test the sound with the "as packaged" alsa. Apologies as it was late and I was tired. I simply downloaded and installed the latest git alsa rpms and rebooted. Sound / headphones worked properly from both headphone jacks. Many thanks again.