| === asac_ is now known as asac | ||
| TheMuso | GrueMaster: Could you please action bug 746023? A lot of these patches introduce new functionality, and IMO need an FFE. Thanks. | 07:26 |
|---|---|---|
| TheMuso | ogra_: ^^ if you could look into it as well, that would be ood, thanks. | 07:29 |
| TheMuso | good | 07:29 |
| GrueMaster | TheMuso: I'll ping ogra_ or rsalveti in the morning. | 08:18 |
| ppisati | what's the status of pm in omap4? | 10:29 |
| amitk | WIP | 10:32 |
| ppisati | there's a patch in maverick/omap4 related to pm, that makes the kernel panic whenever a cpu hotplugged | 10:33 |
| ppisati | but since i | 10:33 |
| ppisati | it's not done yet, i'll just revert it | 10:33 |
| amitk | you should ask ndec for an update. I do think we've got hotplug working on panda in the linaro kernel. | 10:34 |
| ppisati | yep, it works there, in fact you don't have that patch there :) | 10:35 |
| ppisati | it's just in maverick, i think it's a leftover | 10:35 |
| ppisati | https://bugs.launchpad.net/ubuntu/+bug/707003 | 10:55 |
| ppisati | here it is | 10:55 |
| ogra_ | ndec, our audio guys arent actually happy with the amount of changes to alsa-lib | 12:03 |
| diwic | ogra_, I'm here as well if I'm needed | 12:14 |
| ogra_ | diwic, ah, cool | 12:14 |
| diwic | ogra_, and, were there changes to alsa-lib? I thought it was just PA at this point. | 12:15 |
| diwic | PA = pulseaudio | 12:15 |
| ogra_ | well, the tarball doesnt seem to have PA patches at all | 12:16 |
| ogra_ | i only see alsa | 12:16 |
| diwic | ogra_, if you follow luke's link, you have three files, the top two I can't open, the third one has PA patches | 12:17 |
| ogra_ | the "commit message" one ? | 12:17 |
| diwic | yeah | 12:18 |
| ogra_ | ah, the name is misleading :P | 12:18 |
| ogra_ | havent opened that one at all | 12:18 |
| ogra_ | urgh ! | 12:18 |
| ogra_ | thats another 8 | 12:18 |
| diwic | can you open the top two? | 12:19 |
| ogra_ | hmm, they seem to be the same as in PA_UCM_patches.tar.gz | 12:20 |
| ogra_ | weird | 12:20 |
| diwic | aha, they're labeled .tar.gz but should be .tar.bz2 ...? | 12:22 |
| ogra_ | i dont think they are zipped at all | 12:22 |
| ogra_ | tar xvf should work | 12:23 |
| ogra_ | (for all these files) | 12:23 |
| diwic | so the "cofigfiles" (sp!) file, that's just config files, I'm not worried about those affecting anything else. It's the PA patches that worry me. | 12:26 |
| ogra_ | yep | 12:27 |
| ogra_ | same here | 12:27 |
| diwic | ogra_, is TheMuso clear on what to apply? Because I'm not and that makes it difficult to review. | 12:29 |
| ogra_ | diwic, i dont think he is, he didnt say anything but to add a FFE request to the bug | 12:29 |
| ogra_ | diwic, i dont have more info than you can read on the bug and in the backlog of this channel | 12:30 |
| ogra_ | though comment #3 somewhat indictaes he looked at all of them | 12:31 |
| ogra_ | and just wants release team approval | 12:31 |
| ogra_ | he doesnt question the quality there it seems | 12:32 |
| === stenzel_ is now known as stenzel | ||
| TheMuso | No, I have only quickly glanced at them. I just think that there is enough new functionality here that we need to get the release team's ok. | 12:34 |
| ogra_ | i dont have probs to get that OK, but kate will ask about the impact | 12:34 |
| TheMuso | Yeah I know | 12:35 |
| TheMuso | The tarballs are actually bzip2, but have the wrong extension. | 12:35 |
| diwic | TheMuso, figured that out but it took a while :-) | 12:37 |
| diwic | TheMuso, so are you clear on what to apply? | 12:37 |
| TheMuso | diwic: I think so, they are ordered after all. I just need to wait for the refactored patch on the pulse list that you commented on. | 12:39 |
| diwic | TheMuso, as for my comments on the PA mailing list, I just had a glance that the jack detection stuff as I'm interested in having that working for Oneiric. I didn't look much on the UCM stuff. | 12:40 |
| diwic | TheMuso, would it be possible for you to make a feature branch or package (for the PA patches) which we can use for review and testing? | 12:42 |
| TheMuso | diwic: Will do first thing tomorrow. I am online doing work on other projects this evening. | 12:47 |
| TheMuso | ogra_: I'm sure you understand that I don't want to throw it in, without testing on non-Arm, non-TI, hardware. | 12:51 |
| TheMuso | To make sure things don't break elsewhere. | 12:51 |
| ogra_ | TheMuso, ok, i'll add that to the feature request then and will discuss with kate | 12:53 |
| TheMuso | ok thanks | 12:53 |
| TheMuso | But yeah, will get packages ready for testing first thing tomorrow morning. | 12:54 |
| diwic | TheMuso, thanks. | 12:57 |
| ndec | ogra_: diwic: TheMuso: the UCM patches we provided are PA patches only. the only requirement for alsa is to use 0.24 which has UCM support, and we already have it in natty | 13:11 |
| diwic | ndec, could you clarify on which set of patches is the one to apply? | 13:12 |
| ogra_ | ndec, well, the patchset in https://bugs.launchpad.net/ubuntu/+source/alsa-lib/+bug/746023 is definitely a lot more than just PA | 13:12 |
| ogra_ | ndec, if it would be possible to define the minimal amount of stuff to apply to make it work that would be great | 13:12 |
| ndec | ogra_: my understanding is that all patches in the tar ball are PA patches. | 13:13 |
| ogra_ | i dont think thats the case | 13:13 |
| ogra_ | and its a total of 19 patches in two tarballs | 13:14 |
| diwic | three tarballs | 13:14 |
| ogra_ | plus 7 config files | 13:14 |
| ogra_ | in another tarball | 13:14 |
| diwic | I think all 19 patches are for PA but some of them are overlapping | 13:14 |
| ogra_ | yeah, it seems the second tarball duplicates some of the patches from the first one | 13:15 |
| diwic | and then there's the set posted on the PA mailing-list | 13:15 |
| diwic | you can get confused for less ;-) | 13:15 |
| ogra_ | which seems to differ again :) | 13:15 |
| diwic | yeah, | 13:16 |
| diwic | . | 13:16 |
| ndec | ogra_: diwic: this a 6-patch series for PA which is sent upstream too (https://tango.0pointer.de/pipermail/pulseaudio-discuss/2011-April/009746.html) | 13:16 |
| ogra_ | ndec, different patrches from whats in the tarballs apparently | 13:16 |
| ogra_ | and as stated above the tarballs have more than 6 | 13:16 |
| ndec | ogra_: the second tar ball has all the required PA patches. and it replaces the first tarball. that's what alejandro said in his email. | 13:17 |
| ogra_ | hrm | 13:17 |
| ogra_ | (not on the bug though) | 13:17 |
| ogra_ | ndec, also the second set of patches is named UCM-patches-commit-message.tar.bz2 .... i didnt even bother to look for patches in it firstly | 13:18 |
| ndec | ogra_: http://afuera.cortijodelrio.net/~ddiaz/paucm/UCM-patches-commit-message.tar.bz2 has all patches , and it's in PA only, no alsa. and http://afuera.cortijodelrio.net/~ddiaz/paucm/PA_UCM_cofigFiles.tar.gz is for the config files | 13:18 |
| ogra_ | right, thats clear now | 13:18 |
| ndec | ogra_: sorry about the confusion. alejandro is in mexico. he will be online soon i guess. | 13:18 |
| hrw | ndec: is it known when TI will rebuild omap4-extras-multimedia deps to current natty level? | 13:18 |
| ogra_ | hrw, post release | 13:18 |
| ndec | hrw: hi! it's known that it won't be ready for 11.04 release day ;-) | 13:19 |
| hrw | thanks | 13:19 |
| diwic | ndec, ogra_, So what you want me to test and review is http://afuera.cortijodelrio.net/~ddiaz/paucm/UCM-patches-commit-message.tar.bz2 and *not* the PA mailinglist stuff and *not* the tarball named "PA_UCM_patches.tar.gz" | 13:19 |
| ndec | hrw: however the GFX package will be ready (and is already). so i guess that if you asked it's because you care about 1080p... | 13:19 |
| hrw | so will not have a need to switch monitor on ;D | 13:20 |
| ndec | diwic: the UCM-patches-commit-message tarball should contain the same series as sent to PA ML + a 2 other patches. that's what needs to be reviewed | 13:20 |
| hrw | ndec: just spotted omap4 extras and wanted to check status | 13:20 |
| ogra_ | hrw, the GLES drivers should be ready | 13:21 |
| ndec | hrw: understand. | 13:21 |
| ogra_ | the rest will be missing until post release | 13:21 |
| diwic | ndec, they differ. | 13:21 |
| ndec | ogra_: hrw: wlan will be ready for 11.04 release since we switch to mac802.11 stack, and since our firmware for panda is in linux-firmware package with no license restriction. | 13:22 |
| ndec | diwic: a lot? | 13:22 |
| ndec | diwic: i didn't check anything myself.. but alejandro from my team worked on this. | 13:22 |
| diwic | ndec, e g, in patch 5/6, the function jack_insert_new_hook_callback differs, in the tarball, it's a stub, in the PA mailinglist, it seems to contain test code. Neither looks like good enough to release. | 13:24 |
| ndec | diwic: ok. i would guess this part could be skipped maybe | 13:24 |
| ogra_ | will it still work with only 1-4 applied ? | 13:26 |
| diwic | unlikely | 13:27 |
| ogra_ | well, i'm taking about general sound output ;) | 13:27 |
| ogra_ | vs no sound at all | 13:27 |
| ogra_ | jack detection seems to me like something i should be able to select in the audio settings still | 13:28 |
| ogra_ | even if no autodetection is there | 13:28 |
| diwic | 5/6 is the big patch of them | 13:28 |
| ndec | diwic: ok sorry.. 5 is required. what I meant is jack detect support is not needed to get sound. | 13:28 |
| diwic | let's wait for Mexico to wake up to clarify this a bit further | 13:30 |
| ogra_ | yes | 13:30 |
| diwic | or Margarita if she's available | 13:30 |
| ogra_ | ndec, whats her IRC nick, do you know ? | 13:30 |
| ndec | ogra_: i forgot... but i will send an email to her and alejandro | 13:32 |
| ogra_ | great, thanks | 13:32 |
| diwic | ogra_, fwiw, if basic output is what's needed you can just call alsaucm to set the right ucm profile. No PA patches necessary | 13:36 |
| XorA|gone | hey ndec hows life in Nice? | 13:39 |
| ppisati | don't we have a natty's omap4 image somewhere? | 13:43 |
| ogra_ | ppisati, we have two, what do you need, headless/serial or netbook ? | 13:45 |
| ppisati | uhm... both? :) | 13:45 |
| ogra_ | diwic, hmm, so only the config files would be needed and pulse would magically work ? | 13:46 |
| ogra_ | ppisati, http://cdimage.ubuntu.com/ubuntu-netbook/releases/natty/beta-1/ and http://cdimage.ubuntu.com/ubuntu-headless/releases/natty/beta-1/ | 13:46 |
| ogra_ | ppisati, note the latter has no graphical output at all, the former has no serial | 13:47 |
| ogra_ | ppisati, and two directory levels higher there are also daily builds | 13:47 |
| ppisati | ogra_: k, thanks | 13:47 |
| ppisati | i want to tackle the "X turn off" bug now | 13:48 |
| ogra_ | then you should use netbook | 13:48 |
| ogra_ | headless will need you to install a ton of stuff first | 13:49 |
| ppisati | with maverick i can't get it, and the upgrade process was PAINFULL (18 hrs!!!) | 13:49 |
| ppisati | yep | 13:49 |
| ppisati | i'll add serial to netbook | 13:49 |
| ogra_ | yeah, the SD speed is far from optimal | 13:49 |
| ppisati | yep | 13:49 |
| ppisati | and from time to time it seems it goes in pm mode | 13:49 |
| ogra_ | the SD ? | 13:49 |
| ppisati | if i leave it idle for a bit, and then issue an "ls" it takes many seconds to spit the output | 13:50 |
| ogra_ | weird, anyting in dmesg ? | 13:50 |
| ppisati | dunno if the SD or anything else, but it's noticeable | 13:50 |
| ppisati | uhm | 13:50 |
| ppisati | no | 13:50 |
| ppisati | actually yes | 13:51 |
| ppisati | the video/hdmi goes in pm mode | 13:51 |
| ogra_ | i know there were issues in the past if read and write occured at the same time | 13:51 |
| ogra_ | wrt SD cards | 13:51 |
| ppisati | uhm no | 13:51 |
| ppisati | nothing regarding i/o, sd, etcetc | 13:51 |
| ogra_ | the video shouldnt impact the SD IO | 13:51 |
| ogra_ | the sync lost bug is an old one from maverick, given we use a forward ported graphics driver its no wonder if it exposes the same bug | 13:52 |
| ppisati | let's see | 13:53 |
| ppisati | the other one "high memory instabilities", well, i can't reproduce it in any way | 13:53 |
| ppisati | i left my panda compile for more than 5+ hrs, but no problem | 13:54 |
| ogra_ | with a default unmodified ubuntu image ? | 13:54 |
| ppisati | yep, maverick image with latest natty kernel | 13:56 |
| ppisati | lp633227 | 13:56 |
| ogra_ | strange, but i heard its hard to reproduce | 13:56 |
| ppisati | yep | 13:56 |
| ogra_ | i know the bug :) | 13:56 |
| ppisati | the real problem is that we can't get a: | 13:56 |
| ppisati | a) quick way to reproduce it | 13:56 |
| ppisati | b) a reliable way to reproduce it | 13:56 |
| ppisati | it's like chasing ghost | 13:56 |
| ogra_ | well, in the past we didnt have issues with a) | 13:57 |
| ogra_ | it died pretty early iirc | 13:57 |
| ogra_ | seems that changed with natty | 13:57 |
| ogra_ | but it was never reliable | 13:57 |
| XorA|gone | what good is a reliable bug :-) we would all be out of jobs | 13:57 |
| ogra_ | haha | 13:58 |
| ppisati | well, but when you can reproduce it, then real fun starts :) | 13:58 |
| XorA|gone | also hard to repeat bugs are good for hang over days | 13:59 |
| ppisati | i hate hang over days :) | 13:59 |
| ogra_ | depends on the evenig really | 13:59 |
| ogra_ | some are worth a good hangover ;) | 13:59 |
| XorA|gone | its always good to sleep at desk then tell your boss your trying to repeat really hard to repeat bug X :-D | 14:00 |
| * ogra_ woud be surprised to see his boss at his desk though ;) | 14:01 | |
| diwic | ogra_, so assume we didn't take the patches at all, I guess you could get output would be to write a PulseAudio profile instead. | 14:01 |
| diwic | that would not be invasive and would guarantee to not affect anything else | 14:01 |
| ogra_ | diwic, automatically with only calling one command ? | 14:01 |
| ogra_ | i could do that chrooted from jasper-iniramfs (our first boot tool) | 14:02 |
| diwic | ogra_, I don't know UCM syntax so I'm a little unsure if PA profiles support everything that UCM does | 14:06 |
| ogra_ | well, the question is if PA will ick it up at all without patches | 14:06 |
| ogra_ | calling alsaucm is surely no prob | 14:07 |
| * ogra_ goes to set up his panda for testing that | 14:09 | |
| ogra_ | diwic, hmm, intresting, i cnt even install alsa-utils on my panda, seems there is an issue with the postinst | 14:21 |
| ogra_ | Setting up alsa-utils (1.0.24.2-0ubuntu3) ... | 14:21 |
| ogra_ | dpkg: error: --compare-versions takes three arguments: <version> <relation> <version> | 14:21 |
| ogra_ | hmpf | 14:22 |
| ogra_ | comparing against a former version if there was none ... | 14:22 |
| diwic | ogra_, hmm, haven't seen that one before | 14:23 |
| ogra_ | well, i use a headless image as a base, that doesnt have anything more than ubuntu-minimal installed | 14:24 |
| hrw | ogra_: same here. upgrading of alsa-utils works fine, install fails | 14:24 |
| ogra_ | yeah | 14:24 |
| ogra_ | if dpkg --compare-versions $2 lt 1.0.24.2-0ubuntu1 ; then | 14:24 |
| ogra_ | seems $2 is empty on fresh installations | 14:24 |
| ogra_ | (thats from the postinst) | 14:25 |
| ogra_ | diwic, anyway, thats not what i wanted to test... see below: | 14:25 |
| ogra_ | ogra@panda:~$ alsaucm | 14:25 |
| ogra_ | No card found | 14:25 |
| ogra_ | ogra@panda:~$ cat /proc/asound/cards | 14:25 |
| ogra_ | 0 [SDP4430 ]: SDP4430 - SDP4430 | 14:25 |
| ogra_ | TI OMAP4 SDP4430 Board | 14:25 |
| ogra_ | so not working OOTB | 14:25 |
| hrw | confirmed | 14:26 |
| diwic | ogra_, try installing the config files if they're not there | 14:26 |
| ogra_ | same goes for the listcards command | 14:26 |
| alejandro-gm | Hi guys, hows the PA UCM working for you? | 14:27 |
| ogra_ | ogra@panda:/$ ls usr/share/alsa/ucm/SDP4430/ | 14:27 |
| ogra_ | FMAnalog hifi hifiLP record SDP4430.conf voice voiceCall | 14:27 |
| ogra_ | ogra@panda:/$ alsaucm | 14:27 |
| ogra_ | Im setting defaults | 14:27 |
| ogra_ | ogra@panda:/$ alsaucm listcards | 14:27 |
| ogra_ | Im setting defaults | 14:27 |
| ogra_ | 0: SDP4430 | 14:27 |
| ogra_ | ogra@panda:/$ | 14:27 |
| ogra_ | diwic, that seems to do something | 14:27 |
| ogra_ | (no desktop installed on that image though) | 14:27 |
| diwic | alejandro-gm, hi there, I'm not even sure what patches to test | 14:28 |
| ogra_ | alejandro-gm, hey, good to see you, we are fighting with the patches atm, diwic is one of our audio gusy | 14:28 |
| ogra_ | *guys | 14:28 |
| * hrw rebuilds alsa-utils | 14:28 | |
| diwic | ogra_, can you file a bug against alsa-utils for the dpkg problem | 14:29 |
| alejandro-gm | what patches? I'm not sure what do you mean? | 14:29 |
| hrw | rootfs on my panda is 5.3GB now | 14:29 |
| ogra_ | diwic, will do | 14:29 |
| hrw | I have ugly postinst change which want to test first | 14:29 |
| diwic | alejandro-gm, we have three sets of patches to choose from, two different tarballs and one posted to the ML | 14:29 |
| ogra_ | alejandro-gm, final freeze of ubuntu is on monday, initially we were told we would only get configs and two patches for panda to make sound work, now we are struggling with the amount of patches you provide and try to shrink down the patchset to at least get approval for a minimal set to get soujnd working | 14:30 |
| ogra_ | else it wont make the release at all | 14:30 |
| alejandro-gm | diwic: ho, ok, well they are all basically the same, but the ones I send are rebased for the 0.9.22 PA version | 14:31 |
| ogra_ | alejandro-gm, have these patches been tested against non omap hardware at all ?` | 14:31 |
| alejandro-gm | and the ones on the PA mailing list are on a more updated PA version | 14:31 |
| alejandro-gm | ogra_ : as far as I know, yes, Margarita do the implementation on x86 | 14:32 |
| ogra_ | alejandro-gm, at this late time of release we need to be 100% sure it doesnt break the x86 world | 14:32 |
| diwic | alejandro-gm, e g, in patch 5/6, the function jack_insert_new_hook_callback differs: in the tarball, it's a stub, in the PA mailinglist, it seems to contain test code. Neither looks like good enough to release... | 14:32 |
| alejandro-gm | diwic: mmmm, let me look at it, it may be referring to 'new' PA features (added after 0.9.22) | 14:34 |
| diwic | alejandro-gm, for this particular thing also see https://tango.0pointer.de/pipermail/pulseaudio-discuss/2011-April/009758.html | 14:35 |
| ogra_ | diwic, bug 753544 for you | 14:36 |
| * ogra_ slaps the bot | 14:37 | |
| hrw | ogra_: http://paste.ubuntu.com/590766/ is fix | 14:37 |
| ogra_ | https://bugs.launchpad.net/ubuntu/+source/alsa-utils/+bug/753544 | 14:37 |
| ogra_ | hrw, add it to the bug ;) | 14:37 |
| hrw | will do | 14:37 |
| ogra_ | thx | 14:37 |
| hrw | added patch, tagged patch, confirmed | 14:38 |
| hrw | ogra_: final freeze == no new uploads for all repos? | 14:40 |
| ogra_ | hrw, only with exceptions and your bug needs to be very very serious to get one | 14:40 |
| hrw | and no NEW packages? | 14:41 |
| ogra_ | not sure about universe but usually not | 14:41 |
| hrw | I have two packages waiting for me to get upload rights - so probably one_eric will see them | 14:42 |
| ogra_ | well, its actually not final freeze but beta2 freeze which turns into final freeze three days later | 14:43 |
| ogra_ | so there is a small window where you can upload stuff that doesnt affect the images | 14:43 |
| alejandro-gm | diwic: It looks to me that the patches on the mailist and the ones on the tarball differ only on some alignment (new lines and spaces), can you point me to the code you identify to be missing please? | 14:53 |
| diwic | alejandro-gm, http://afuera.cortijodelrio.net/~ddiaz/paucm/UCM-patches-commit-message.tar.bz2 - v2/0005-alsa-create-ucm-profiles.patch, row 736, function jack_insert_new_hook_callback is a stub | 14:58 |
| diwic | alejandro-gm, in total, that function is 8 lines and only outputs a debug message | 14:58 |
| diwic | alejandro-gm, compare that to https://tango.0pointer.de/pipermail/pulseaudio-discuss/2011-April/009758.html where the same function is 25-30 lines | 15:02 |
| alejandro-gm | diwic: regarding the PA discussion, what I can tell is 1) need to look at the 'pa_hook_cb_t' definition I couldn't find it at fist glance but it may be overloaded, otherwise I really dobut it would compile in any environment. 2) about the match card-event, every card has its own set of events and in PA each sink/source is attached to one card so I doubt there could be a conflict on that, but I'll ask Margarita, she has a setup with more than one | 15:02 |
| alejandro-gm | card. 3) Headset is the device the UC is attached to the '.0' is the identifier, you may have more than one, but here the are using it only a default in case there is no device specified | 15:02 |
| === prpplague^2 is now known as prpplague | ||
| === zyga is now known as zyga-food | ||
| diwic | alejandro-gm, for 2) the hook is fired with an additional parameter containing the card to attach to but that is ignored due to 1). | 15:07 |
| diwic | alejandro-gm, in 3) how do you specify a device | 15:08 |
| alejandro-gm | diwic: this jack_insert_new_hook_callback function confuse me, let me see what goes wrong, or why it is this way on the patch of the tarball | 15:13 |
| alejandro-gm | diwic: you specify the device of the UC on the configuration files, there is a file per verb=UC | 15:14 |
| hrw | ndec: tiwlan-wl1271 will also be ready after 11.04 release? | 15:30 |
| hrw | ok | 15:30 |
| alejandro-gm | diwic, I check with maggie_ and the reason that function is different is because she add new code after I send you the patches, it was the initial code for the jack detection, she is just about to answer to the list explaining the points we are concerned about | 15:31 |
| alejandro-gm | ogra_, also, for being able to test the UCM even at Alsa level (alsaucm) you need the configuration files I send: 'PA_UCM_cofigFiles.tar.gz' (just remember is it really a .bz2 file, I need to change the extention) | 15:35 |
| ndec | hrw: there is no tiwlan-wl1271 anymore. the .38 kernel which is already in natty has built in support for wl127x | 15:41 |
| hrw | so only fw is lacking? | 15:44 |
| hrw | [508763.406768] wl1271_sdio: probe of mmc1:0001:2 failed with error -22 | 15:44 |
| sebjan | hrw: the Natty firmware package contains the good one for wl1271. Not sure about this error... | 15:49 |
| alejandro-gm | diwic, ogra_, What happens is that the patches we send to you contains UCM + 1st part of jack-detection. Now this are 2 different features, and jack-detection interacts with UCM that was this was necessary and you see some 'dummy' functions like 'jack_insert_new_hook_callback', but UCM do not depends on the jack-detection feature, so even though maggie_ will add more code, it will be related to the 2nd part of the jack detection | 15:50 |
| hrw | ok | 15:51 |
| ppisati | GrueMaster: installed natty-beta1 on panda, added pvr-omap4, but i still cannot reproduce your video issue | 15:51 |
| hrw | rebooting | 15:51 |
| alejandro-gm | diwic, ogra_, but the UCM implementation is complete | 15:51 |
| ppisati | GrueMaster: do you have a panda A1 board? | 15:51 |
| hrw | [ 4111.341583] wl1271: ERROR sdio read failed (-110) | 15:55 |
| hrw | [ 4111.346893] wl1271: ERROR chip id doesn't match after firmware boot | 15:55 |
| hrw | ea1 board | 15:55 |
| GrueMaster | ppisati: I'll get back to you after the meeting. | 15:59 |
| hrw | after reboot it works | 15:59 |
| hrw | BT is seen, but down. "hciconfig hci0 up' fails | 16:02 |
| GrueMaster | WIP. Link is https://wiki.canonical.com/UbuntuPlatform/ARM/NattyReleaseNotes | 16:04 |
| sebjan | hrw: for BT, you have to start 'uim' (http://gitorious.org/uim) during startup (quite fast after BT driver init). As a hack we did it in /etc/rc.local. We miss a clean udev rule or something for doing that. | 16:05 |
| hrw | sebjan: what does uim has to bt? | 16:07 |
| hrw | Description: Simple and flexible input method collection and library | 16:07 |
| === robbiew1 is now known as robbiew | ||
| diwic | alejandro-gm, are you saying the jack detection feature does not work, and if so, why should we merge it? Would it instead be possible to rip out the jack detection entirely and merge UCM only | 16:13 |
| alejandro-gm | diwic, yes, the jack detection is not working it is only the 'empty shell' | 16:16 |
| sebjan | hrw: here are some references to uim: http://www.omappedia.org/wiki/Porting_Guide_BT_FM_GPS | 16:16 |
| alejandro-gm | diwic, I'm not sure if we can just rip it out I guess there is a reason why maggie_ add the 'prototype' of the funtions in the same patch, but let's ask her | 16:18 |
| alejandro-gm | maggie_, any comments? | 16:18 |
| diwic | alejandro-gm, because that would also rip out the invasive changes to module-udev-detect e g | 16:19 |
| diwic | alejandro-gm, so if it does not add extra value to the end user we should not merge it | 16:20 |
| ogra_ | well, release team wouldnt approve nonfunctional additions anyway i think | 16:20 |
| hrw | sebjan: thx | 16:22 |
| sebjan | np | 16:23 |
| maggie_ | diwic: I will send some updates today, jack detection will be working, e.g PA will be able to detect jack insertion and removal, and UCM will be used to set the Headset as device | 16:24 |
| maggie_ | diwic: will you be able to take those changes? or do you prefere to remove jack detection code? | 16:25 |
| diwic | maggie_, hi there! | 16:26 |
| maggie_ | hi :) | 16:26 |
| diwic | maggie_, well, first I was trying to figure out which patches to review, but you missed that conversation. | 16:27 |
| maggie_ | diwic: yes I missed that part | 16:28 |
| diwic | maggie_, how possible would it be to add a module parameter to module-udev-detect that turns on the jack detection stuff? | 16:28 |
| diwic | maggie_, I'm thinking that way we could have it turned off by default and only turn it on for omap4, thereby minimizing regression risk | 16:29 |
| maggie_ | diwic: I see, I would need to look at that, there are some flags that I can use | 16:29 |
| diwic | maggie_, as for the other, I was asked to review today, the jack detection stuff I looked at earlier was more out of interest | 16:30 |
| diwic | maggie_, so I haven't looked that much at the UCM part of it yet | 16:30 |
| maggie_ | diwic: ok, yes your comments were mainly related to jack detection | 16:32 |
| diwic | maggie_, but basically, these changes are on the large side for being so late in the development cycle, so if we take them, we want to take every precaution we can that it doesn't blow up anything else. | 16:32 |
| diwic | maggie_, do you have any advice on how we can minimize risks for other platforms? | 16:32 |
| maggie_ | diwic: sorry bit confuse, risk related to jack detection or UCM integration? | 16:33 |
| hrw | sebjan: looks like this TI uim tool is used to set BT ADDR etc but it also looks that it expects old modules | 16:33 |
| hrw | ndec: ? | 16:34 |
| diwic | maggie_, well, both, if we decide to take them. | 16:35 |
| sebjan | hrw: are you testing with the kernel coming with current daily images? | 16:36 |
| maggie_ | diwic: for UCM integration e.g to generate the profiles using the UCM data instead of the current guessing mode there is no risk | 16:36 |
| ndec | hrw: yes | 16:37 |
| maggie_ | diwic: if your platform doesn't have the UCM configuration file PA will use the current method, e.g UCM will no be used | 16:37 |
| maggie_ | diwic: for jack detection we are basically listening dev/input/event in all cases but I can look into add some code in order to skip loading module-alsa-jack in case the UCM is not present | 16:39 |
| diwic | maggie_, I think I'll have to look at the code being touched by the UCM integration a little bit more closely and see what parts of common code that is changed. I will do this later this evening as I have to go and make dinner now. | 16:41 |
| hrw | sebjan: 2.6.38-1207 | 16:42 |
| hrw | ndec: so is there any docs for getting BT working with 2.6.38? | 16:42 |
| maggie_ | diwic: ok, good. please send me your comments/questions | 16:42 |
| ndec | hrw: no. not yet | 16:43 |
| sebjan | hrw: the current Natty tree misses a patch series for supporting BT | 16:43 |
| hrw | thx | 16:43 |
| micahg | hi, I upgraded my arm netbook last night, tried unity and it kicked me back to classic desktop instead of unity 2D, is this known? | 17:06 |
| ogra_ | yes, thats not implemented yet | 17:07 |
| ogra_ | select unity-2d from gdm for now | 17:07 |
| ogra_ | its impressing to hear the fallback worked for you :) | 17:07 |
| rsalveti | GrueMaster: new debs for edid testing posted at bug 753071 | 17:08 |
| ogra_ | (our arm images all default to unity-2d) | 17:08 |
| NCommander | rsalveti: (GrueMaster wanted me to let you knew that he saw you post before his system went down) | 17:09 |
| rsalveti | NCommander: ok, thanks :-) | 17:10 |
| ndec | rsalveti: the debs for edid, that's for OMAP3 or 4? | 17:17 |
| === diwic is now known as diwic_afk | ||
| rsalveti | ndec: that's for omap 3, for omap 4 I sent the patches to sebjan | 17:18 |
| ndec | rsalveti: ok. just curious what did you do? | 17:18 |
| ndec | rsalveti: for some reasons i thought we had this in OMAP4 already, no? | 17:18 |
| ogra_ | ndec, not for dvi | 17:19 |
| rsalveti | ndec: http://kernel.ubuntu.com/git?p=rsalveti/ubuntu-natty.git;a=shortlog;h=refs/heads/ti-omap4-drm | 17:19 |
| rsalveti | the edid parsing we had already, but implemented by the hdmi driver | 17:19 |
| rsalveti | the drm driver now is the responsible for the edid parsing and setting | 17:19 |
| ndec | rsalveti: oh... you've taken rob's DRM changes? | 17:19 |
| rsalveti | ndec: yes, and improved it to have it properly working also for dvi | 17:20 |
| ndec | rsalveti: that's cool... didn't know. | 17:20 |
| ndec | rsalveti: are you going to make the DDK changes too for XRandR? | 17:20 |
| rsalveti | ndec: that's the idea | 17:20 |
| rsalveti | that's why I was first working to get the driver in | 17:21 |
| ndec | rsalveti: did you see this: http://adventuresinsilicon.blogspot.com/2011/04/ubuntu-1104-to-be-released-in-days.html? | 17:22 |
| rsalveti | ndec: yeah :-) | 17:22 |
| ndec | rsalveti: i was wondering why the auto detect monitor was mentioned ;-) | 17:23 |
| ndec | i guess I know now... | 17:23 |
| GrueMaster | rsalveti: there are two kernels in your directory. | 17:23 |
| rsalveti | GrueMaster: yup, get the latest | 17:23 |
| rsalveti | GrueMaster: 2.6.38-8.41 | 17:23 |
| GrueMaster | ok | 17:24 |
| ogra_ | GrueMaster, cant have enough kernels ;) | 17:24 |
| GrueMaster | ogra_: netbook omap image is mia. WTF? | 17:45 |
| ogra_ | omap didnt build | 17:45 |
| ogra_ | acorn is up since this morning only | 17:45 |
| GrueMaster | I didn't get an email. | 17:45 |
| GrueMaster | Ah. | 17:46 |
| ogra_ | i got one at 3:55 | 17:46 |
| ogra_ | 8or 1:55 UTC) | 17:46 |
| GrueMaster | I have a headless image. Just no netbook. | 17:47 |
| ogra_ | weird, i have mail for both | 17:47 |
| ogra_ | Tue, 5 Apr 2011 23:45:07 +0000 (UTC) for headless | 17:48 |
| ogra_ | Wed, 6 Apr 2011 01:55:07 +0000 (UTC) for netbook | 17:48 |
| GrueMaster | Last email I have is for 6 Apr 2011 01:55:07 +0000 (UTC) for the netbook image. | 17:49 |
| GrueMaster | I have a headless image for today. | 17:49 |
| ogra_ | yeah, headless gets built before 0:00 UTC | 17:49 |
| * ogra_ checks | 17:49 | |
| GrueMaster | rsalveti: Does the edid data get exported to userspace in the new kernel? THought I'd ask while I download. | 17:52 |
| ogra_ | it was build, weird | 17:53 |
| GrueMaster | ogra_: Which? | 17:53 |
| jhobbs | /wg 2 | 17:55 |
| ogra_ | No filesystem for armel+omap! | 17:55 |
| ogra_ | + exit 1 | 17:55 |
| ogra_ | make: *** [bin-preinstalled_images] Error 1 | 17:55 |
| ogra_ | hmm | 17:55 |
| ogra_ | hmm, seems it wasnt copied or something | 17:58 |
| rsalveti | GrueMaster: probably like before, I'm checking | 17:59 |
| ogra_ | GrueMaster, i have fired off a new manual build, should be ready in 90min, if not, shout | 18:01 |
| * ogra_ goes for dinner | 18:02 | |
| micahg | ogra_: re fallback> yeah, I was shocked, it dropped me right into classic | 18:02 |
| micahg | well, took about a minute, but no further input required | 18:02 |
| rsalveti | GrueMaster: $ sudo modprobe eeprom | 18:07 |
| rsalveti | GrueMaster: $ decode-edid 3 | 18:07 |
| GrueMaster | ok | 18:07 |
| GrueMaster | Will try it out once I get to that point. Currently bringing the image to current. | 18:07 |
| rsalveti | GrueMaster: decode-edid comes from the i2c-tools package | 18:07 |
| rsalveti | sure | 18:08 |
| * GrueMaster has had the hiccups for 3 days. Very frustrating when added to leg being wrapped like a mummy. | 18:09 | |
| * rsalveti lunch | 18:14 | |
| GrueMaster | sigh. Jasper is broken...again. | 18:18 |
| ogra_ | GrueMaster, whats wrong ? | 18:23 |
| GrueMaster | It isn't rebooting after resizing, and other things. I am looking into it now. | 18:24 |
| GrueMaster | The bad edid data from my switch makes it more complicated. | 18:24 |
| ogra_ | not rebooting usually means the script exits before | 18:25 |
| GrueMaster | On reboot, it starts over, so I think you are right. | 18:27 |
| GrueMaster | Not easy to watch, though. | 18:28 |
| ogra_ | ARGH ! | 18:28 |
| * ogra_ sees the typo | 18:28 | |
| ogra_ | missing closing single quote, grmpf | 18:28 |
| ogra_ | thats what you get if you work with vim without syntax on grmbl | 18:29 |
| GrueMaster | I am very limited in mobility atm, and respectfully request that basic code be checked before it is inserted into the images. | 18:30 |
| GrueMaster | I can't afford to waste what limited time I can sit in my office looking at boot issues that shouldn't exist. I need to be able to do image testing. | 18:31 |
| ogra_ | fixed and uploaded | 18:35 |
| ogra_ | sorry, i tested that but with a different branch | 18:36 |
| GrueMaster | Well, I have ~20 minutes left before I need to return to the couch. I'll finish testing the kernel for omap, then I'm back to what I can test remotely. | 18:37 |
| ogra_ | sure | 18:37 |
| === diwic_afk is now known as diwic | ||
| diwic | maggie_, still around? | 18:42 |
| diwic | ...seems not | 18:43 |
| alejandro-gm | diwic, I have her contact on gtalk, do you want me to ping her? | 18:53 |
| diwic | alejandro-gm, hm, I'm currently writing up my review and have a few suggestions but I could just as well write that in the review | 18:55 |
| diwic | maggie_, hi there | 18:56 |
| maggie_ | diwic: hi, I'm back | 18:57 |
| diwic | maggie_, working hard with the PA stuff - how's it coming along? | 18:58 |
| maggie_ | I'm checking the card_name part | 18:58 |
| diwic | maggie_, you were talking about adding more jack detection code today, are you still thinking of that or should we skip that part for the time being? | 18:59 |
| diwic | maggie_, you were talking about adding more jack detection code today, are you still thinking of that or should we skip that part for the time being? | 19:01 |
| maggie__ | diwic: sorry I'm having connection issues | 19:01 |
| maggie__ | diwic: yes, I'll send some code today, it is early here | 19:01 |
| diwic | maggie__, ok. Are you located in west US? | 19:03 |
| maggie__ | diwic: if you can wait, based on the progress I made today you can decide to skip or take the jack detection code | 19:03 |
| maggie__ | yeap | 19:03 |
| maggie__ | dwic: same as Dallas time | 19:03 |
| diwic | maggie__, I'll write my review based on current code as I probably will be asleep by then (I'm in Sweden). | 19:03 |
| maggie__ | diwic: (I'm in mexico) | 19:04 |
| diwic | right | 19:04 |
| maggie__ | diwic: that is ok, the UCM part has been tested more, you can review that part | 19:05 |
| GrueMaster | rsalveti: Tested your kernel and updated bug 753071 with the results. Looks good. Ship it. | 19:14 |
| diwic | maggie__, ogra_, here' | 19:14 |
| GrueMaster | Hmmm. bugbot is mia. | 19:15 |
| diwic | here's the review: bug 746023 | 19:15 |
| diwic | https://bugs.launchpad.net/ubuntu/+source/alsa-lib/+bug/746023/comments/7 | 19:15 |
| maggie__ | diwic: thanks, reading your review | 19:18 |
| diwic | maggie__, do ask if you have any question regading the review | 19:20 |
| diwic | regarding | 19:20 |
| maggie__ | diwic: I could move some ucm functionallity (like parsing the verbs) to a new .c file, but we still have the main code in module-alsa-card | 19:25 |
| diwic | maggie__, hm, ok, I guess the more you could move the better | 19:32 |
| maggie__ | diwic: ok, I'll check | 19:33 |
| diwic | maggie__, also since we're more likely to take the UCM code than the jack detection code make sure it's easy to see which patch belongs to which functionality (no patch should include a little of each) | 19:34 |
| maggie__ | diwic: ok, then I will basically split the create_ucm_profiles patche into ucm code and jack detection code | 19:37 |
| diwic | yes | 19:37 |
| maggie__ | diwic: I will send the changes to alejandro | 19:38 |
| rsalveti | GrueMaster: cool, thanks | 19:44 |
| === zyga-food is now known as zyga | ||
| diwic | maggie__, do you need me for anything? Otherwise I'll log off for today. | 19:51 |
| diwic | maggie__, I hope you felt the review made sense; otherwise please tell me | 19:51 |
| maggie_ | diwic: its ok | 19:52 |
| diwic | maggie__, I hope you felt the review made sense; otherwise please tell me | 19:52 |
| maggie_ | diwic: thanks I'll let you know | 19:54 |
| Martyn1 | Hey all | 21:14 |
| === prpplague^2 is now known as prpplague | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!