=== asac_ is now known as asac === Jack87|Away is now known as Jack87 === fairuz-away is now known as fairuz === doko_ is now known as doko === jkridner_ is now known as jkridner === Jack87 is now known as Jack87|Away [09:29] anyone with a beagle xm around? [09:29] i want to rule out a hw issue === amitk_ is now known as amitk [13:51] ogra_, what's the output of "alsaucm listcards"? [13:52] ogra@panda:~$ alsaucm listcards [13:52] Im setting defaults [13:52] 0: SDP4430 [13:52] ogra_, that "Im settings defaults" is annoying. I'm sure that if we just correct "Im" to "I'm", sound will start to work ;-) [13:53] what's the output of "alsaucm list SDP4430"? [13:53] lol [13:53] ogra@panda:~$ alsaucm list SDP4430 [13:53] Im setting defaults [13:53] alsaucm: error failed to get list SDP4430: No such file or directory [13:56] ogra_, how about "alsaucm list _verbs" or "alsaucm list _enadevs"? [13:57] http://paste.ubuntu.com/596013/ [13:57] ogra@panda:~$ alsaucm list _enadevs [13:57] Im setting defaults [13:57] alsaucm: error failed to get list _enadevs: Invalid argument [14:01] alsaucm set _verb HiFi [14:01] ogra@panda:~$ alsaucm set _verb HiFi [14:01] Im setting defaults [14:01] pasuspender -- speaker-test -c 2 -t sine -D plughw:SDP4430 [14:02] any sound? [14:02] hmm, that needs me to be near the board and listen, right ? [14:02] right [14:02] * ogra_ is remotely logged into the board, i need to move over to the office, one moment [14:03] i see it doing stuff on the cmdline though [14:07] * ogra hugs diwic [14:07] works ! [14:07] ogra, great :-) [14:07] so its the alsaucm set _verb HiFi we need ? [14:07] let me reboot the board and see if it persists [14:09] ogra, likely needs to be set on every boot - might persist a reboot but not a complete power-off with capacitors unloaded etc [14:09] hmm, i will test that too [14:10] it wouldnt be a prob to have a one time call of it in our first boot config tool [14:10] putting it in place for every boot is more effort [14:11] ogra, the point is that the PA UCM stuff should make it possible to set these verbs from PA. [14:11] aha, after reboot the UI only has HDMI again [14:12] hmm, no [14:14] ogra, does the other verbs in the list make sense to you? [14:14] voice low power etc [14:15] yep. omap4/panda is a mobile development platform [14:16] all the verbs make sense there [14:16] i'm trying a reboot with "alsaucm set _verb HiFi" in rc.local, lets see what comes out there [14:16] ogra, so question is if "HiFi" is the correct one to set at boot (if we should set anything at all at boot) [14:16] seems pulse isnt picking it up without a restart [14:17] we want to default to headphone on boot, so hifi is fine [14:17] the prob is that people wanting to use HDMI sound will have to hack it up [14:17] else we reset their sound every boot [14:18] is HDMI going through the SDP4430 as well, or is it something separate? [14:19] not sure, i think its going through the same chip [14:20] aplay -l [14:21] http://paste.ubuntu.com/596025/ [14:23] ogra, looks like HDMI is on a separate sound card, probably not using UCM at all, and should not be affected by the "_verb HiFi" stuff [14:23] k === arune is now known as arunemec [14:25] ok. adding the line to rc.local (and actually writing hifi with a capital F) gets me working sound OOTB [14:26] diwic, so do you think its safe to pull in the pulse package with the patches ? i can care for addint an upstart script or some such that sets the default (or a udev rule) [14:27] ogra, I'm not following, pull to where? The main ubuntu archive? === arunemec is now known as arune [14:29] diwic, well, the pulse packages with the TI patches werent uploaded yet [14:29] and we only have two days left to do so [14:30] ... getting sound to work this time is kind of important since we already failed with that last release [14:30] ogra, a udev rule sounds like a good solution for setting the default verb [14:31] k [14:31] ogra, well, before we pull them in, we should get them working - do you think they are working? [14:32] i got sound, thats enough imh ;) [14:32] *imho [14:32] with the rc.loacl entry at least [14:32] ogra, yeah, but that does not need the PA patches [14:33] alsaucm is at the alsa-lib level [14:33] are you sure ? [14:33] * ogra reverts to the archive pulse version, gimme a few min [14:33] hmm, good question :-) [14:34] gar [14:34] now i rteinstalled pulse on my laptop :P [14:35] grmbl [14:35] * ogra whacks apt over the head [14:36] oh, trying to downgrade PulseAudio? [14:36] well, it has a ~ppa version [14:36] apt should respect that [14:37] sigh [14:37] ogra, see your email, got it from Conor who had managed to downgrade pulseaudio after several rounds in the padded cell [14:39] ogra, that's what he had to do - you should probably skip the modules you do not use [14:39] grmbl, it complains about packages i dont even have installed [14:39] bah, sigh, a version for every single package ? [14:40] ogra, guess you could make a shell variable [14:40] for the version number [14:40] yeah [14:42] k, that seems to work [14:42] will take a while ... slow SD card [14:43] diwic, in any case i think the pulse patches are important for UI exposure, you wont be able to select if you want HiFi or HDMI output without them [14:44] but we'll see [14:46] * ogra scratches head ... why the heck does pulse trigger update-initramfs ? [14:54] diwic, ok, seems it works without the patches too [14:54] ogra, how does it look in pulse now? One soundcard or two? [14:54] the default selected output in the UI is HDMI though [14:54] two cards [14:54] great [14:54] sound playing from headphone ... [14:55] but if i open the UI it shows HDMI as the selected output device [14:55] how are you playing the sound? [14:55] if i switch back and forth and leave HDMI the default, sound is dead [14:55] if i switch back to SDP its fine then [14:56] i'm hitting back in the terminal to hear the kbd bell [14:56] and i use the test speakers button in the UI [14:56] ok [14:56] i'll try banshee too later, but i dont think we'll have issues there [14:57] a udev rule should live in alsa or pulse though [14:57] (i can indeed add one as a hack but i would prefer not to) [14:59] ogra, I wasn't exactly following what the "sound is dead" thing was [14:59] by default the UI shows HDMI as selected output device [15:00] while its using SDP [15:00] if i switch the radio button to SDP and back to HDMI, sound on the headphone is dead [15:00] hmm, that's strange, I'm assuming you haven't messed around with pavucontrol? [15:00] nope [15:01] its just showing a wronmg default on first start is what i meant to say [15:01] banshee works fine [15:01] ok, that would be my conclusion as well [15:01] but I don't recognize that problem from anywhere else, seems strange [15:02] well, a bug we can surely release with [15:02] you might want to make sure recording is working as well [15:03] that might need an additional verb...? [15:03] hmm, the UI doesnt offer me *any* input devices [15:04] i have mono and stereo outputs for SDP and a stereo output for HDMI [15:04] could you tell me what physical inputs and outputs there are on this card? [15:05] only a mic jack ... [15:05] i'm not sure there are additional inputs on the expansion header though [15:05] but getting the mic jack to work should be enough [15:05] there seems to be a recod verb [15:06] aha [15:07] alsaucm set _verb Record [15:07] that gets me duplex support in SDP [15:07] ogra, question is if they are exclusive, so "HiFi" turns off record and vice versa [15:07] they arent it seems i constantly listen to radio from banshee [15:07] cool [15:07] let me try a reboot and add it to rc.local [15:08] * ogra reboots [15:10] * ogra hears login sound [15:12] hmm, record doesnt seem to use the mic [15:15] how are you checking the mic? [15:16] i have my headset plugged in ... checking the meter in sound prefs and testing sound recorder [15:16] i have set it from record to Voice_Call now, lets see [15:16] instead of Record [15:17] still have a login sound it seems [15:18] hmmk [15:18] doesnt get my any input device at all [15:19] It's interesting that the verb affects what devices show up in PA. [15:21] well, no matter which one i set, i dont seem to be able to get input [15:21] i would be willing to live without though, if we dont find a solution [15:21] only Record actually gets me duplex support [15:22] but it seems to not be routed to the mic jack [15:22] or its muted on alsa level [15:22] alsaucm should unmute the relevant things...but everything can be buggy of course [15:22] yeah [15:23] well, having playback at release is a massive improvement already [15:23] is there any reason why alsaucm requires these ugly underscores ? [15:24] ogra, No idea - I just figured them out by reading the source code... [15:25] could you give me the output of arecord -l ? [15:27] http://paste.ubuntu.com/596061/ [15:30] ogra, hmm, from what you know about this board, would you suspect the mic jack is at "9 DMIC capture" or "11 Analog capture"? I'm suspecting 11 as the "D" in "DMIC" might suggest a digital mic is needed [15:30] right, i would say that too [15:31] was recording working with the init scripts we had in maverick? [15:31] no [15:32] ok [15:32] and we never really had the full implementation in maverick [15:32] only parts of it [15:34] what is the output of "amixer -D hw:SDP4430"? [15:36] http://paste.ubuntu.com/596064/ [15:39] ogra, hrm, you'll almost need the datasheet of the SDP4430 for that to make sense...is that available? [15:39] ndec, sebjan ^^^^ ? [15:39] do we have one for the sound chip ? [15:40] i know there are several papers but i'm not sure the sound is covered === prpplague^2 is now known as prpplague [15:42] ogra, found something on google, claiming it's connected to a TWL6040 chip, is that correct? [15:42] i think there was 6040 in one of the pastes [15:43] http://paste.ubuntu.com/596025/ yep [15:43] right that's on channel 11 as well [15:45] the twl6040 has one "Analoc mic" and one "Headset mic". Can you verify which one you have connected your mic to? [15:45] heh, not really [15:45] there is nothing printed on the jack [15:45] on the PCB it just says headphone [15:46] which might or might not be the headset mic [15:48] it seems to be the headset mic [15:48] from looking at the schematics ;-) [15:50] :) [15:51] assuming you have the "record" verb fixed, could you try this command [15:51] pasuspender -- arecord -D plughw:SDP4430 -f cd -d 10 [15:51] sorry [15:52] prints garbage [15:52] pasuspender -- arecord -D plughw:SDP4430 -f cd -d 10 /tmp/test.wav [15:52] and then speak into the mic for 10 seconds [15:53] papplay only revels silence [15:54] *reveals [15:54] ok [15:54] and i have overrun!!! (at least 16.473 ms long) [15:54] twice [15:54] hrm [15:55] I think you should talk to TI about what verb you should use for recording from mic and then try my above command to record [15:55] k [15:55] My time is about up [15:55] But at least we got some playback :-) [15:55] well, can we add the udev rule to at least get HiFi in ? [15:55] i guess that should go into alsa or pulse [15:56] if so, alsa - not pulse [15:56] right [15:56] if you get the mic working on alsa level, we can consider the pulse stuff, but if it ain't working on the alsa level, it won't work on the additional pulse layer either. [15:57] right [15:57] good you called upon me for this, I thought it was all okay. [15:57] i thought you had the bug on your radar [15:57] but i somehow assumed the same for the TI devs [15:58] i'm babbling on the bug since a while with my unsuccessfull findings [15:58] but if you ask me, I'm happy for having a udev rule added [15:58] such a rule should be trivial and non-invasive [15:59] k, i'll try to come up with a one liner until tomorrow morning [15:59] that should be enough to make the release [15:59] given that hiFi is enough for now [15:59] you can probably match against ATTRS{id}=="SDP4430" [15:59] yeah [16:00] yeah and see if you can get the mic working with the help of the TI folks [16:00] ATTRS{id}=="SDP4430" IMPORT{program}="alsaucm set _verb HiFi" [16:00] should that work ? [16:00] ppisati: I saw your call earlier for a beaglexm, what's up? [16:02] ogra, I would have tried this, but I'm not a udev expert: ATTRS{id}=="SDP4430", RUN+="alsaucm set _verb HiFi" ? [16:02] ah, right, RUN [16:03] IMPORT will hand over the return value to the caller [16:03] ok, let's talk more tomorrow, gotta go. [16:03] diwic, i think we'll keep recording for a possible SRU if i dont find a way to make it work tonight, should i attach the one liner to the bug ? [16:04] ogra, yes and preferrably as a debdiff against the relevant package [16:04] ok, will prepare everything [16:04] ogra, and maybe poke TheMuso about it if he's awake when you are [16:05] GrueMaster, something to test for you, add: ATTRS{id}=="SDP4430", RUN+="alsaucm set _verb HiFi" to /lib/udev/rules.d/60-persistent-alsa.rules before persistent_alsa_end and reboot, see (hear) if you have sound then on panda [16:05] diwic, will, do, many many thanks for the help !!! [16:05] ogra: Any other changes/packages to install first? [16:06] GrueMaster, no ! thats the awesome bit :) [16:06] we already have working sound since ages, UCM is just not initialized [16:07] well, your panda should be up to date indeed :) === diwic is now known as diwic_afk [16:21] Trying now on yesterday's image. [16:22] did you try headless since beta btw ? [16:22] * ogra_ hasnt and has a suspicion that the ppa stuff breaks jasper on headless [16:23] ppa stuff? I haven't had a change as my pandas were doing Maverick -> Natty upgrade tests (not recommended on SD BTW). [16:24] yeah, there was a fix to the ppa handling in jasper added after beta [16:25] i think it might fail if the target dir doesnt exists (which it likely doesnt without any desktop packages) [16:25] dont worry, i'll test myself tomorrow and fix if my suspicion is true [16:28] No sound here. [16:28] Panda 2.0 [16:28] really ?!? [16:28] Lower jack, right? [16:28] yes [16:29] do you see the SDP4430 in the sound prefs ? [16:29] no [16:29] sigh [16:29] try calling: alsaucm set _verb HiFi [16:30] from a terminal (close the sound prefs first) [16:30] and killal pulseaudio [16:30] then start the sound prefs again [16:30] see if that changes it, probably udev initializes it to late [16:32] if i add that line to rc.local here, it makes output work fine [16:35] rebooting. Typo in udev rule. (oops). [16:35] phew [16:39] pulseaudio is crashing now. Need to look into it. [16:39] hrm [16:41] does alsaucm list _verbs get you a list of use cases ? [16:42] (it should spit out 8 use cases) [16:43] also alsaucm listcards [16:43] whould list the SDP4430 [16:43] *should [16:46] * ogra_ doesnt get why it works flawless for him [16:49] GrueMaster, oh, there was a kernel upgrade i havent installed yet, i wonder if something went wrong with the tree merge wrt alsa/ucm [16:49] Do you have the pulseaudio from the ppa installed or the base image stuff? [16:49] all archive [16:52] * ogra_ fires off a dist-upgrade [16:52] pulseaudio is crashing here with main.c: Daemon startup failed. Going to dive in and see why. [16:52] What is your image based on? Beta2? [16:52] well, what do you get from the alsaucm commands ? [16:52] yes, its my peta2 test install [16:52] alsaucm is fne. [16:52] fine [16:53] so you get the list and also get the soundcard listed ? [16:53] (caffeine is building up) [16:53] yes [16:53] weird [16:53] pulse shouldnt crash [16:53] But pulseaudio is segfaulting. [16:53] check syslog [16:53] * ogra_ goes to get coffee before the call [16:55] which call? [16:55] Linaro? [16:58] yep [16:59] I'm skipping out. QA CoP meeting in :30 [16:59] k [17:12] ogra, might want to try pasuspender -- arecord -D plughw:SDP4430,11 -f cd -d 10 /tmp/test.wav [17:12] as an option to the previous command [17:13] will do (my mic is occupied by a meeting atm :) ) [17:18] ogra: there are problems on panda with the input mic. prpplague might know more than me on that. [17:20] ndec: the audio input is actually configured as a mono line-in currently [17:20] ndec, ah, ok, should the Record usecase be the right one then ? [17:20] you should be able to record using it as a line in [17:26] ogra: Not seeing an issue with headless on omap4. [17:26] k, thanks [18:12] ogra: On my panda audio, I am seeing good indicators for asoc early in the dmesg log, but later on I am getting asoc: machine hw_params failed followed by a series of specific asoc failures. [18:13] weird, i dont have such issues here [18:13] the alsaucm call should definitely set up everything for you [18:14] I even removed that line from udev. Still getting it. Not sure why. See http://members.dsl-only.net/~tdavis/syslog.gz [18:15] your syslog looks fine [18:15] I may have a theory. Let me check something. [18:16] asoc: no valid backend routes for PCM: SDP4430 Media should actually go away if alsaucm was initialized [18:16] i have that as well here until the card is properly initialized [18:17] the others are not errors [18:20] Do you still have /usr/share/alsa/init/omap4? [18:21] no [18:21] that was dropped from alsa [18:21] do you ? [18:21] Ok, because you said you were on the beta 2 image. [18:21] it shouldnt be there [18:21] It is in beta 2, not yesterday's image. [18:21] upgarded to whatever was recent on monday [18:22] and just upgrading again [18:22] Check if the file is still on your system please. Upgrading doesn't delete it necessarily. [18:22] it has to, else that would be a massive dpkg bug :) [18:22] but for your convenience ... [18:22] ogra@panda:~$ ls /usr/share/alsa/init/omap4 [18:22] ls: cannot access /usr/share/alsa/init/omap4: No such file or directory [18:26] lool, ping [18:27] GrueMaster, did you recently upgrade the kernel on your panda ? [18:28] I just did. This is yesterday's (20110418) image now rebooting with kernel & other updates. [18:28] (forgot to run flash-kernel after updating). [18:28] ?? [18:28] the postinst runs that [18:29] Yea, sure it does. [18:29] (not on mine). [18:29] UGH ! [18:29] do we have a bug for that ? [18:30] Kernel was released last night. I just upgraded and (5 minutes ago) saw that the new kernel was installed but not in the u-boot partition. Simple answer - NO. Give me some time. [18:30] but flash-kernel ran fine when you executed it manually ? [18:30] Ok, pulse is happy now with the new kernel. [18:31] Yes it did. [18:32] weird, i didnt have the new kernel here but it worked [18:33] I was running 38-1207 when I was seeing the error, after manually running flash-kernel I am now on 38-1208 kernel. [18:33] And audio is happy (so far - testing actual output next). [18:34] works now. [18:35] wohoo [18:36] I'm starting over with yesterday's image to see if I can pinpoint the issues, especially with the kernel update. [18:36] yeah [18:36] (the joys of having two systems). [18:37] i have seen an umount error in flash-kernel during my tests, thats why i brought it up [18:37] doesnt happen anymore though [18:38] i wonder if we should change umount to umount -l in the script [18:38] not sure. [18:38] Script is getting an overhaul though. May not be necessary. [18:39] GrueMaster, btw, you are aware that we have an --update-bootloader function on flash-kernel ? [18:39] * ogra_ remembers you saying something about bootloader updates [18:39] That was either Karmic or Lucid on babbage. [18:40] I could look up the bug if you like. [18:40] ah, no, dont care :P [18:40] Bug 365053 [18:40] Launchpad bug 365053 in flash-kernel "On armel (Babbage platform), kernel image upgrading breaks if Ubiquity is instructed not to install a bootloader" [High,Triaged] https://launchpad.net/bugs/365053 [18:40] (I had the list open) [18:40] ah, that one [18:40] ancient crap [18:41] must have been karmic [18:42] no babbge in lucid iirc [18:42] yes there was. [18:42] hrm [18:42] right, we dropped it in maverick [18:42] Babbage & dove. omap was an after-sprint bringup. [18:43] yeah, "tech preview image" and totally unsupported :) [18:43] Which reminds me, I had a notice in my inbox that ppsati was working on a Lucid omap bug. Can you tell hime to not care? [18:44] Bug 588243 [18:44] ppisati, stop caring ! [18:44] Launchpad bug 588243 in linux-ti-omap "kernel BUG at /build/buildd/linux-ti-omap-2.6.33/drivers/video/omap2/dss/core.c:323! " [High,Confirmed] https://launchpad.net/bugs/588243 [18:44] :) [18:44] i saw the bugmail too [18:49] I'm marking it as won't fix. Lucid was a demo release on omap anyways. I don't think we are on the hook for it, as it only supported Beagle C4. [18:50] yeah [18:50] good move [19:05] ogra: on which package should we report bugs on the installer? [19:06] depends, when does your issue occur ? [19:07] everything after first reboot is ubiquity ... everything before first reboot is jasper [19:08] ogra: ok. that answers it... i am just answering to someone on pandaboard ML,and wanted to ask to report bugs on LP... [19:09] the jasper package is actually jasper-initramfs [19:13] Yea for undecipherable ubiquity failures. My oem-config crashed hard for some reason. Probable a fs error, but no real indication. [19:14] Ah there it is, two pages up in syslog: mmcblk0: error -110 sending read/write command, response 0x900, card status 0xe00 [19:15] grmbl. [19:15] ogra: just fyi. http://groups.google.com/group/pandaboard/browse_thread/thread/aa46ddf46df2d946 [19:17] ndec: That looks normal, but I have been seeing some odd SD failures lately. Not reproducable, just seemingly random. [19:18] Usually flashing SD over a different image will cause this (i.e. if I had kubuntu on there previously). [19:19] reflashing seems to clear it up. [19:54] ogra_: "pong" [19:56] jcrigby: ogra_: tested the u-boot fix on B5 and it's now working fine with all beagle revisions [19:56] jcrigby: so I believe we should fix this asap, if still possible for natty [19:57] ogra_: do you know if we can we still upload package fixes? [20:07] rsalveti, if ogra_ could do the upload that would be great [20:07] or we can ask some other person with upload rights [20:08] ogra_: janimo: NCommander: anyone? ^ [20:09] need someone to sponsor an u-boot update, that has a fix for beagle b5 [20:12] rsalveti, link? [20:13] I'll get around to it in the next few hours unless someone else grabs it [20:13] bug 760350 [20:13] Launchpad bug 760350 in u-boot-linaro "2011.03 doesn't boot anymore on Beagle B5" [High,In progress] https://launchpad.net/bugs/760350 [20:13] code https://code.launchpad.net/~jcrigby/ubuntu/natty/u-boot-linaro/new- [20:13] proposed-2011.04.2 [20:13] jcrigby: is this the one to merge? [20:14] so basically needs a merge to main u-boot tree in LP? [20:14] then upload a package? [20:14] https://code.launchpad.net/~jcrigby/ubuntu/natty/u-boot-linaro/new-proposed-2011.04.2 [20:14] yes [20:14] were there deps tested as well? [20:15] jcrigby, debs I meant [20:15] Gahhh. oem-config has crashed 5 times now on 3 different SD cards (different brands, sizes, & speeds). This is getting very frustrating. [20:15] And it is only the netbook image that crashes. [20:15] GrueMaster, core dumped? [20:15] And not always. [20:15] No [20:15] rsalveti, its new-proposed, the old proposed was deleted [20:15] janimo: yes, I tested the produced debs with my beagles and panda [20:19]