/srv/irclogs.linaro.org/2011/11/23/#linaro-android.txt

bhojpfefferz, around ?00:50
asac_bhoj: good morning :)00:51
asac_what time is it at your place? 10 to 8am ?00:51
bhojasac_, morning :) . 621 am :)00:51
asac_ouch00:52
asac_thats tough00:52
asac_so status is we might have a panda build that works, but zach is having dinner :)00:52
asac_and lava didnt digest and plars was not there to resubmit with fixed download paths :)00:52
asac_https://android-build.linaro.org/builds/~pfefferz/ics-staging-panda/00:53
bhojasac_, okay  cool :)00:53
asac_that uses default google toolchain. seems 4.4.0 wasnt the default anymore00:53
asac_bhoj: so no guarantee that it boots ... if you want to have fun try it out. i dont have a panda here unfortunately00:53
bhojoh great ... Ill try it ...00:54
asac_bhoj: saw you said you figured that gralloc was missing?00:54
asac_for snowball00:54
asac_?00:54
bhojthere is a new component hwcomposer ...which doesn build by default .That was missing.Building a new one with it00:55
asac_cool00:55
asac_bhoj: but guess you didnt push anything yet? e.g. should we setup a build :)?00:55
bhojasac_, haven't pushed the patches.Once I have working  snowball locally Ill then push the patches.should be anytime soon :)00:56
asac_cool00:56
plarsasac_: hi01:18
plarsasac_: sure, one moment01:18
plarsasac_, pfefferz, bhoj: http://validation.linaro.org/lava-server/scheduler/job/419201:21
plarshmm, strange01:25
plarssomething messed up the console01:25
plarsasac_, pfefferz, bhoj: http://validation.linaro.org/lava-server/scheduler/job/419301:27
plarsah, looks like fsck ran and outputted some characters that made the console go crazy01:28
plarshmm01:28
plarswell, 4193 ought to be alright at least, just ignore the prior one01:29
asac_plars: look reasonable still, right?01:36
pfefferzhere01:36
pfefferztesting https://android-build.linaro.org/builds/~pfefferz/ics-staging-panda/#build=901:36
asac_good01:37
asac_pfefferz: lava is also running it :)01:37
asac_not sure what happens there though01:37
asac_I really like this stuff. Having this as a great full desktop solution with ambitions to expand into tablet and handset markets in the not so long future is reason enough to start embracing this technology.01:37
asac_But what folks often miss is that you get a sophisticated accelerated window manager middleware stack here that has potential to empower a magnitude of innovations.01:37
asac_oops01:37
asac_http://validation.linaro.org/lava-server/scheduler/job/419301:37
asac_thats the job01:37
pfefferzhackbox is out of space01:37
pfefferzbtw01:37
plarsasac_: hmm... it booted01:38
plarsasac_: but the prompt is funny01:38
plarssec01:38
asac_pfefferz: maybe time to reboot and wipe :)01:38
asac_bette rnot01:38
asac_with mpoirier keeping his stuff there :(01:38
bhojpfefferz, cleaning up my stuff01:39
pfefferzme too01:39
asac_plars: guess that prompt comes unexpected for the dispatcher commands?01:40
asac_i doubt that we have 0xbench or anything in that image anyway01:40
asac_so if it really booted ok thats probably good01:40
plarsasac_: well, we change the prompt at deployment time to something we expect01:40
plarsasac_: let me cancel it and check something01:40
plarsin init.rc01:40
asac_we dont get a complete logcat though, right? e.g. because serial doesnt see its a prompt01:40
asac_plars: why cancel :)01:41
asac_seems it died anyway01:41
asac_Freezing user space processes ...01:41
asac_wonder if that is bad01:41
plarsasac_: so that I can debug, it's not going to go any further, it will just timeout and assume we didn't boot01:42
bhojasac_, that is PM stuff ...looks fine01:43
asac_plars: kk01:44
asac_wodner if we have any 0xbench or something in it01:44
asac_seems its in repo at least01:44
asac_err wrong build01:45
pfefferzFUCK YEAH!01:45
pfefferzhttps://android-build.linaro.org/builds/~pfefferz/ics-staging-panda/#build=9 comes up01:45
pfefferzwhith inverted colors01:45
asac_cool01:45
pfefferzand is sllllllloooooowwwww01:45
asac_:)01:45
asac_maybe rumbling in the background?01:45
pfefferzand I get6 a billion W/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_WAKE failed (Bad file number)01:45
pfefferzW/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_WAKE failed (Bad file number)01:45
pfefferzW/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_WAKE failed (Bad file number)01:45
pfefferzW/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_WAKE failed (Bad file number)01:45
pfefferzand there a console...oh this is slow01:46
pfefferzbut it works!01:46
pfefferzbhoj, whats the hack for 640x48001:46
pfefferzwe're going to need it01:46
asac_good start01:46
pfefferzugg....and we need to turn off this log spam....W/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_WAKE failed (Bad file number)01:47
pfefferzW/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_SLEEP failed (Bad file number)01:47
pfefferzW/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_WAKE failed (Bad file number)01:47
pfefferzW/SurfaceFlinger( 1210): ANDROID_WAIT_FOR_FB_SLEEP failed (Bad file number)01:47
plarsI'm not sure why this didn't work... we have to do a lot of uglyness to modify the uInitrd, extract the cpio archive, modify init.rc, put it all back together01:47
asac_plars: ?01:47
plarsor else things won't get mounted in the proper way, prompts won't get set right, etc01:47
plarsin this case, it all worked though01:47
asac_plars: 0xbench is not even on this build01:47
plarsinit.rc in the uInitrd even has PS1 set properly01:47
asac_plars: cant we do it better? are we using adb?01:47
asac_maybe we can just send through adb and repeat until it works?01:48
plarsso I'm not sure where it picked up this root@android:/ # prompt01:48
asac_hmm. guess we get adb before the prompt01:48
plarsasac_: this is pre-adb01:48
asac_what was the previous prompt?01:48
plarsasac_: we can't use adb until we get a shell prompt, bring up the network, get the ip, etc01:48
asac_sure01:48
plarsasac_: previously not set01:48
plarsI was checking the init.rc to see if maybe it was set somewhere else there01:49
asac_is it hard to recognize multiple prompts (general question)01:49
asac_plars: ?01:49
plarsasac_: well, we have to know what we're looking for01:50
plarsasac_: easiest way to do that is to set it to what we expect01:50
plarsasac_: it has worked with all the other android images so far, not sure what's different here01:50
asac_sure. but you can either reprogram/patch images to do what you expect or have the code so that it just has a list of options01:50
bhojpfefferz, bootargs need to be changed in boot.scr01:50
plarsasac_: that's fragile though as we proved with the linaro images - everytime something changed like hostname, default user, anything, we get a change in prompt01:51
pfefferzcool, I can do that01:51
plarsasac_: which is why we set it now01:51
pfefferzwould you send the boot args you used?01:51
asac_plars: hmm. ok. have to think a bit about this01:51
asac_plars: i see your point though01:51
bhojpfefferz, can you change the kernel to 3.1 I think EARLYSUSPEND and FB_EARLYSUSPEND are enabled in it .01:51
bhojtehn ANDROID_WAIT_FOR_FB_WAKE this should not happen01:52
asac_yeah ... i was suggesting to go for 3.1 as well01:52
bhojbut let me first confirm its enabled01:52
pfefferzoklay we're on https://android-build.linaro.org/builds/~pfefferz/ics-staging-panda/#build=901:52
pfefferzI mean, <project path="kernel" name="people/andygreen/repo-branch-archive" revision="tilt-linaro-android-3.0" remote="linaro-other"/>01:53
plarsasac_: aha, I think I found it01:53
plarsasac_: this image seems to include a /etc/mkshrc that is getting executed after the init.rc in the uInitrd01:53
bhojyes that needs to move to tilt-android-3.101:53
pfefferzroll to tilt-android-3.101:54
plarswhich contains...01:54
plarsmkshrc:PS1='$(precmd)$USER@$HOSTNAME:${PWD:-?} '"$PS1 "01:54
pfefferzbhoj, k, making the chang, would you approve?01:54
bhojpfefferz,okay ....  I have upstream panda manifest upgraded ..so triggered that build https://android-build.linaro.org/builds/~vishalbhoj/ics-panda/01:54
asac_plars: interesting01:55
asac_plars: so we can just put a file in /somewhere/roothome/.mkshrc setting it to our value?01:55
asac_plars: maybe thats the best long term solution if there is a roothome01:55
asac_plars: is there a $HOME set or something?01:56
asac_on android?01:56
pfefferzbhoj,  would you look at: http://review.android.git.linaro.org/#change,97001:58
bhojpfefferz, FB_EARLYSUSPEND is enabled in 3.1 http://git.linaro.org/gitweb?p=landing-teams/working/ti/kernel.git;a=blob;f=arch/arm/configs/android_omap4_defconfig;h=21a77ab44f9d1125d100d699dc6b795b702ed4a1;hb=HEAD01:58
pfefferzahh01:59
pfefferzon their tip tip01:59
pfefferzwhat's the commit called?01:59
bhojI don't know I think it came along the 3.1 tip I guess . But there is one more problem ...the boot.scr changes for resolution are not being picked properly by kernel ...02:00
asac_http://git.linaro.org/gitweb?p=landing-teams/working/ti/kernel.git;a=blob;f=arch/arm/configs/android_omap4_defconfig;h=24fa0723ecf5c98c59bbb03c50b77d0caf957237;hb=ce1b928e651923e4aa1caf3e644b94b54902f3e702:00
asac_seems its in last release02:00
pfefferzhmmm02:00
bhojnever got it working fine with tracking-panda kernel ...but worked fine with mainline kernel02:00
asac_android-release-2011-11-302:01
pfefferzwe could switch to john's mainline02:01
pfefferzclone it and update what we need02:01
asac_why wouldnt the 3.1 not work?02:01
asac_FB_EARLYSUSPEND -> we dont want that?02:02
bhojpfefferz, that is what is there in default.xml ...  https://android-build.linaro.org/builds/~vishalbhoj/ics-panda/02:02
pfefferzbhoj, okay, we'll test that one02:02
pfefferzrather us tip anyway02:02
pfefferzrather use tip anyway02:02
pfefferzat this point02:02
bhojpfefferz,  http://android.git.linaro.org/gitweb?p=kernel/linaro-android.git;a=shortlog;h=refs/heads/linaro-android-3.1-agreen-rebase02:03
pfefferzright02:03
pfefferzbhoj, is that what default.xml is pointing to?02:03
bhojyes.02:03
bhojhttp://review.android.git.linaro.org/#change,95102:04
pfefferzcool...well lets see if http://review.android.git.linaro.org/#change,970,patchset=1 works as well02:04
pfefferzand we'll do a tracking02:04
pfefferzthen TI will be done02:04
pfefferzand we can do snowball02:04
asac_yay02:04
asac_maybe time to move the official staging-X to ICS?02:05
asac_instead of dodging in the private namespace?02:05
asac_:)02:05
pfefferzasac_, I'd like to make some official builds called staging-panda-ics02:05
pfefferztracking-panda-ics02:05
pfefferzetc..02:05
asac_why not just move over?02:05
pfefferzokay with you?02:05
pfefferzbecause we need to do a little 2.3.7 clean up02:05
bhojpfefferz, snowball I am compiling it ...will send the patches once it is working ...snowball uses device/linaro/snowball build setup where as pandas are using standard device/ti/panda build setup02:05
pfefferzthat we're going to use in ICS02:05
pfefferzbhoj, right02:05
pfefferzbhoj, heard Origen is also okay02:06
plarsok, let's see if this works02:06
asac_pfefferz: the branch will still exist ... i would rather move the 2.3.7 to some other place and show that ics is now primary focus of developmenet by giving them the proper name :)02:06
plarshttp://validation.linaro.org/lava-server/scheduler/job/419502:06
asac_but your call ... in the end i dont care ... as long as we move over eventually02:06
pfefferzasac_, okay02:06
asac_and dont loose the history02:06
plarshave to go afk a few02:06
pfefferzasac_, you got it02:06
asac_plars: thanks02:06
bhojpfefferz, I had booted up once at 1080P with panda fs .... but Amit is handling the manifest and buildfile upsgrade02:06
bhojit had come up fine02:06
pfefferzcool02:07
asac_wow02:07
pfefferzis Amit around?02:07
asac_!seen amitpundir02:07
ubot5I have no seen command02:07
pfefferzshoot we may have a full deck of ICS by tomorrow morning02:07
asac_:)02:07
asac_damn02:07
bhojasac_, its just 7:20 am in India :) .there is some time for him to come up :) .He is a family man :002:07
pfefferz:)02:08
asac_i didnt complain02:08
pfefferzI like that...engineer boot up time02:08
pfefferzfrom when you wake up, till you're on the computer...02:08
pfefferz:)02:08
asac_yeah maybe upgrading to SSD can speed things up there02:08
asac_thats for me roughly 2 seconds02:08
asac_well 502:08
pfefferzasac_, 10 with a smoke02:09
pfefferz;)02:09
asac_heh02:09
asac_well i usually check ... then go off breakfast and coffee and other things02:09
pfefferzbhoj, would you ack http://review.android.git.linaro.org/#change,970,patchset=102:09
pfefferzand I'll get that cooking02:09
asac_others read news .. i read IRC :)02:09
bhojpfefferz, okay ill merge it02:09
pfefferzcool02:09
asac_and mail :(02:09
pfefferzI'll get tracking going as well02:09
pfefferzasac_, hah!02:09
* asac_ out02:11
pfefferzbhoj, do you have a build jstultz_vm could load up on his panda with his kernel02:12
pfefferzjstultz_vm, not sure...looks like EARLYSUSPEND and FB_EARLYSUSPEND02:13
bhojpfefferz, sadly not currently ...but replacing the kernel from your build will be sufficient02:13
pfefferzbhoj, okay...lets see if your build works end to end02:13
jstultz_vmpfefferz:  i enabled those on all my defconfigs for 11.11...02:13
=== liuyq-afk is now known as liuyq
pfefferzjstultz_vm, yeah, thanks for that!02:13
bhojpfefferz, I had built it yesterday. and verified it ..but I shutdown the AMI so no images from that build.02:14
plarsok, back02:27
plarswell, we got further, it just failed to run any of the tests?02:29
plarspfefferz: is 0xbench/monkey included in these?02:29
plarslooks like it went all the way through otherwise02:29
pfefferzplars, that is a good question02:29
plarsnetworking worked, adb attached, etc02:29
plarspfefferz: hopefully that's all it is, in which case this worked as good as we could hope for02:29
pfefferznope02:29
plarspfefferz: nope as in, they are not installed?02:30
pfefferzthey're not install02:30
pfefferzed02:30
plarscool02:30
plars\o/02:31
plarswe had a successful boot test of it in lava then :)02:31
pfefferzsweet02:31
pfefferzMiller Time02:31
* plars thinks it's insanely awesome how fast you guys had ics up and running for linaro :)02:32
pfefferz:)02:33
pfefferzyeah...it has been slightly crazy02:33
pfefferzits really most bhoj's awesome work02:33
pfefferzwe're all just support him02:33
plarsok, back to context switching... the stack just got deeper02:33
plarswell, good work bhoj!02:33
plarsand bhoj-2 also02:34
plarsmaybe that's how he does it, he clones himself :)02:34
bhoj-2plars this is on my phone02:34
bhoj-2Right now I came out02:35
pfefferzbhoj, yeah, I see it in my reference tree: http://git.linaro.org/gitweb?p=people/andygreen/repo-branch-archive.git;a=blob;f=arch/arm/configs/android_omap4_defconfig;h=21a77ab44f9d1125d100d699dc6b795b702ed4a1;hb=tilt-android-3.102:38
pfefferzbhoj-2, lets make snowball the first offical build02:39
pfefferzbhoj-2, then we'll post to all the stuff02:39
bhoj-2pfefferz I am outside .I have kept for build .I am confident tht it will boot.once that is done.ill push the patches and snowball can roll :)02:41
pfefferzbhoj-2, okay02:41
pfefferzbhoj-2, I'll be up for a bit02:42
pfefferzbhoj-2, are you able to give a +1 to a Gerrit review?02:42
pfefferzhttp://review.android.git.linaro.org/#change,97102:43
bhoj-2pfefferz,for snowball I had to modify build/core for snowball uboot build.it shouldn't affect other builds02:43
bhoj-2I can02:43
pfefferzbhoj-2, okay02:43
pfefferzbhoj-2, np02:43
pfefferzI copied staging-panda.xml to tracking-panda.xml and updated the kernel to point to tracking02:44
pfefferzbhoj-2, thanks02:47
pfefferzkicking build off now02:47
bhoj-2np :)02:47
pfefferzbhoj-2, seeded builds change the whole build paradigm02:51
bhoj-2True02:52
pfefferzpersonal tracking panda launched02:53
pfefferz:)03:11
bhoj:)03:11
pfefferzsomething just finished in jenkins03:11
pfefferzjust daily staging-panda03:12
bhojpfefferz, can you try these http://ec2-107-22-90-152.compute-1.amazonaws.com/ICS/snowball/03:18
bhojfor snowball ?03:18
pfefferzbhoj, yup03:18
bhojthanks03:18
pfefferzbhoj, no, thank you03:19
bhoj:)03:19
pfefferzi see the android03:28
pfefferz:)03:28
bhojany crashes ?03:28
pfefferznot yet03:28
pfefferzI see some of: [   95.032501] request_suspend_state: wakeup (0->0) at 85492521351 (2000-01-01 00:01:43.796945119 UTC)03:28
pfefferz[   95.045806] init: untracked pid 1592 exited03:29
pfefferz[   95.059722] init: untracked pid 1593 exited03:29
bhojcool03:29
bhojthats sad :( ...03:29
bhojcan you send me the logs03:29
pfefferzyeh...03:29
pfefferzlooks like rolling reboot again03:29
pfefferzI'll send you the logs03:29
pfefferzlibskia blowing up03:30
bhojthanks :( ...Ill take abhishek's help to test the  images which I build next ....wondering why libskia is blowing ... that was the case even yesterday03:30
pfefferzI'll do one more boot03:30
bhojokay .03:32
pfefferzsame thing03:34
bhojpfefferz, I think there is some more work to be done with Boardconfig file configuration ...let me see what device/ti/panda/Boardconfig.mk has enabled ... thanks :)03:34
bhojpfefferz, can you review http://review.android.git.linaro.org/#change,972 http://review.android.git.linaro.org/#change,97303:52
pfefferzbhoj, yup03:54
pfefferz972 Merged03:54
bhojthanks.03:55
pfefferzbhoj, what's 973 doing?03:55
pfefferzoh got it from text03:56
pfefferz973 Merged03:56
bhojpfefferz, okay .verified it on panda build . doesn't harm other builds ...03:56
pfefferzcool03:56
bhojrunning one more snowball build with few more bits from panda BoardConfig.mk . will ask abhishek to try that ..03:57
pfefferzcool03:58
pfefferzhey pabhishek03:58
pfefferzbhoj, can you kick off official builds?03:59
bhojpfefferz, okay ...should it be ics-staging-* ?03:59
pfefferzjust switch the snowball build over04:00
pfefferzjust put out an official snowball build04:00
pfefferzwe'll wait for the panda builds04:00
pabhishekpfefferz, Hi04:00
bhojyou mean ics-staging-snowball ?04:01
pfefferzthen send a note on G+ igloo, linaro-dev when snowball comes out of the offcial build04:01
pfefferzplease send a note to Patrik Klinger as well04:01
pfefferzpabhishek, hey!04:01
pabhishekpfefferz, just booted up ics panda images (from your mail)04:01
pfefferzthen we'll switch the panda's over04:01
bhojthere is one more last bit... device/linaro/snowball changes ... which is not yet tested ..04:01
pfefferzpabhishek, :)04:01
bhojpfefferz,  okay ..got it ..04:02
pfefferzbhoj, cool04:02
pfefferzthen I'm going to start working on Patrik Klinger to get us a PoC for graphics04:03
bhojcool04:03
pfefferzand nipuna for SGX04:03
pfefferzetc...04:03
bhojokay04:03
pfefferzokay, bhoj and pabhishek you guys good?04:06
pfefferzokay, bhoj and pundiramit  you guys good?04:07
pundiramithi pfefferz04:08
pabhishekpfefferz, yes04:08
bhojgood04:08
pfefferzpabhishek, would you go through all the bugs and make sure things look good04:08
pfefferzfor the release04:08
pfefferzbhoj, cool04:09
pfefferzpabhishek, also, please go through the spread sheets and fill in anything that got missed, links etc...04:09
pfefferzokay...off for sleep04:09
bhojpfefferz, night ..04:09
pfefferzsee you guys for the sync up04:09
pfefferz:)04:09
pfefferzgreat work!04:09
bhojpabhishek, can you please try these images http://ec2-107-22-90-152.compute-1.amazonaws.com/ICS/snowball/04:19
pundiramitbhoj are you using Linaro's toolchain for ICS build or stock toolchain?04:20
bhojpundiramit, I am using the default prebuilt toolchain to build all the image ...the plan is for tomorrow we will get them build on 4.4.x toolchain and then upgrade to 4.6 toolchain04:22
pundiramitwhat do you mean by 4.4.x? Linaro-4.4.x or stock-4.4.x?04:24
bhojstock 4.4.x present in prebuilt/linux_x86/toolchain04:24
pundiramitthat is what get used by default right? 4.4.3?04:26
pabhishekbhoj, downloading...04:28
bhojpabhishek, thanks .04:30
pabhishekbhoj, ok04:31
panda|x201bero, ping05:55
panda|x201pfefferz, ping05:56
ldesrochesHi, does the Origen evaluation build (or another one) supports the mali 3D acceleration?08:04
pundiramitldesroches Mali integration is in process08:34
bhojpabhishek, patrikryd cyang  can someone try snowball ics build  https://android-build.linaro.org/builds/~vishalbhoj/ics-staging-snowball/08:42
pabhishekbhoj, i'll chk - downloading...08:44
bhojpabhishek, okay coo;08:45
patrikrydbhoj, I'll have a look at it08:46
bhojpatrikryd, thanks08:46
ldesrochespundiramit, thanks for the answer, so I will have to work on it08:50
davepigottHi all, I fixed a problem on the beagle (https://bugs.launchpad.net/lava-lab/+bug/893368) that didn't support ext4. I then ran a previously failing android test through it and it was fine with ext4, and android booted, but we never hit a command prompt, even after waiting 15 minutes. Any idea why that might be? Test log at: http://validation.linaro.org/lava-server/scheduler/job/418308:58
ubot5Ubuntu bug 893368 in LAVA Validation Lab "beaglexm boards need an upgraded kernel with ext4 support" [High,In progress]08:58
springzdavepigott, no. what's the next line after usb-hub in normal boot up case09:06
davepigottspringz: Not sure. I haven't dealt much with Android before. But it was hung there for the whole of the timeout, which is why it failed.09:07
pabhishekbhoj,  boots up- gets ip - not able to browse & resume not working properly, mouse speed is reasonable,09:15
pabhishekreboot works,09:16
pabhishekbhoj,  soft reboot09:16
bhojpabhishek, cool .for the record.. Do a hard boot ..just for our records if we have known issues on it.and  Thanks for all the helping for debugging the image :) .09:17
bhojpatrikryd, pfefferz I am switching staging-snowball to 4.0.109:18
bhojpatrikryd, pfefferz I am switching official staging-snowball to 4.0.109:21
pabhishekbhoj, hard boot works, mouse right click not enabled for back keyevent , thoguh keyevent are recognized va serial console09:23
bhojpabhishek, thanks . right click is not mapped at the moment.you can use escape on keyboard or the back button displayed on the left bottom .09:24
pabhishekbhoj,  sd card partition not mounted09:28
bhojpabhishek, okay :) ...09:28
patrikrydbhoj, I have not managed to boot it yet on my V2 snowball. What snowball version have you used? It that some u-boot magic needed?09:39
patrikryds/It/Is09:40
bhojpatrikryd, I don't have a snowball to test it :) .I have built from whatever uboot master from igloocommunity points to ...09:40
bhojpatrikryd, does 2.3.7 boot on that board ?09:41
bhojif yes then this one should also boot since I use the same kernel and uboot from 2.3.7 manifest in 4.0.1 branch09:41
patrikrydbhoj, It might be my monitor. It has caused problems before.09:41
pabhishekpabhishek,  I used V509:41
pabhishekpatrikryd, , snowball V509:42
patrikrydpabhishek, ok09:42
patrikryd:)09:42
pabhishek:)09:42
patrikrydbhoj, I have the same problem when booting 2.3.7 on my hw. I have also tried your build on a V12 and a V5 at friends desk, but we did not manage to get it running there either.10:16
patrikrydprobly wrong u-boot config in his case. He work in the ST-Ericsson snowball project.10:16
myrjolahi, I've tried to create a bootable sd card for a Snowball board from the ICS tarballs at http://people.canonical.com/~mpoirier/android10:29
myrjolathe short version is that the result freezes at the point where kernel and initrd have been loaded into memory and kernel should start10:30
myrjolai.e. like this: http://pastebin.com/3HqJiTT010:31
myrjolaI'm happy to provide more details is someone's interested to give advice about this10:32
patrikrydmyrjola,  mpoirier had the following config for u-boot. http://pastebin.com/CU0kHswz10:33
patrikrydmyrjola, Look exactly the same. :(10:34
* patrikryd lunch10:34
bhojmyrjola, Are you setting the bootargs ? what is the uInitrd.new ?10:36
bhojmyrjola, which version of snowball did you try it on ?10:36
myrjolabhoj: the boot tarball contains uInitrd.new instead of uInitrd10:37
myrjolaso I replaced that in the bootargs10:37
myrjolaand there was no uImage originally, instead it had just a kernel that hadn't gone through mkimage10:38
myrjolabasically I wonder whether I'm using a wrong version of the linaro-android-media-create, or if the boot tarball in mpoirier's URL is somehow not the right one, although of course the problem could be something completely different than those guesses...10:43
* fgiff prepares an sd card with this image10:45
bhojmyrjola, you are trying the wrong images .Can you try this   https://android-build.linaro.org/builds/~vishalbhoj/ics-staging-snowball/  ?10:49
bhojfgiff, I hope you are using   https://android-build.linaro.org/builds/~vishalbhoj/ics-staging-snowball/10:50
fgiffbhoj: Does it work now? Zach reported a problem with skia10:51
bhojfgiff, It works now . pabhishek has verified it10:51
myrjolabhoj: ah, ok, that explains it then10:53
myrjolabhoj: I'll give the ones from your URL a try, thanks!10:53
* fgiff wants to see what LAVA does to ICS10:54
fgiffbhoj: Works for me too on Snowball V211:19
bhojfgiff, V2 ? cool ... patrikryd  was having some problem on V2 . Any idea on how text on the webpage  https://android-build.linaro.org/builds/~linaro-android/staging-snowball/  can be updated ?11:21
fgiffbhoj: I've never changed the text on android-build. bero knows how to do this I think.11:23
bhojfgiff, okay thanks :)11:23
beroYou have to go to the jenkins page for the build (locate it in https://android-build.linaro.org/jenkins)11:43
beroFrom there you can edit the text IF you're the actual owner (<--- you can't modify the test for ~linaro-android, even if you can build there)11:43
bhojbero,  okay thanks11:44
fgiffpatrikryd: Vishal's ICS build works for me on V2 u-boot env is here https://pastebin.linaro.org/328/11:53
patrikrydfgiff, My problem is probably caused by my monitor (or lack of support in kernel) because the 2.3.7 build doesn't work either.11:54
fgiffpatrikryd: OK. I need some sysfs magic to get my DVI monitor to work properly.11:56
patrikrydfgiff, ?11:56
patrikrydfgiff, what do you do?11:56
fgiffpatrikryd: echo 2 > /sys/devices/av8100_hdmi.0/hdmisdtvswitch11:56
fgiffpatrikryd: Assuming you have a working console and the framebuffer has been created.11:57
patrikrydfgiff, thanks!!! :)12:04
patrikrydfgiff, John here in Lund just mentioned the we could add git://igloocommunity.org/git/android/platform/vendor/st-ericsson/multimedia/linux/HDMIservice.git to the snowball build handle it12:07
bhojpatrikryd, were you able to boot it ?12:08
patrikrydNow I got 2.3.7 up, will try the 4.0.1 image now.12:08
bhojokay12:08
botao_sunpatrikryd: Hi, There should be a space between "TARGET_USE_GATOR" and ":="12:11
patrikrydbotao_sun, did I just solve that... ?12:13
* patrikryd checking12:13
patrikrydbotao_sun, forgot to do git add. :(12:14
patrikrydsorry12:14
patrikrydpushing a new patch12:14
patrikrydbhoj, ICS booted on snowball12:15
bhojpatrikryd, cool . are you able to browse with ethernet ? on panda browsing on ICS works but abhishek reported that he is not able to browse on ICS on snowball12:16
myrjolain my case I'm having different kind of problems with a card prepared from the stuff in the new URL (initialization of sd card fails with error -110, or alternatively I get a kernel oops right in the beginning of the boot)12:17
myrjolaI've tried with two memory cards12:17
botao_sunpatrikryd: It's OK. I'm care about it because I remember that Vishal and I tried to fix a bug and the root cause we finally found is related to the "space" in the code...12:18
bhojmyrjola, I am not sure ...but mpoirier said that there were some issues with sdcards....12:18
myrjolabhoj: the cards I've used are both Kingston cards, perhaps I should try a third one from a different manufacturer just in case12:19
patrikrydbotao_sun, That ok. I prefer that people review my code.12:19
botao_sunpatrikryd: :D12:20
botao_sunpatrikryd: Merged. Are you going to do this again on the 2.3.5 branch?12:22
patrikrydbotao_sun, I think we can ignore the 2.3.5 branch now. We have 2.3.7 and soon everyone will just look at 4.0.1 :)12:23
botao_sunpatrikryd: Understood.12:38
fgiffpatrikryd: Yes, we should definitely look at pulling in the hdmiservice.12:39
myrjolabhoj: I could get ICS to boot with a third card12:46
bhojmyrjola, cool :)12:46
beropfalcon: Around?13:17
pfalconbero: Hi!13:17
beropfalcon: Any idea why the builds 2 and 3 in https://android-build.linaro.org/builds/~berolinux/ics-imx53/ are failing? Looks like they're just marked FAILURE instantly13:18
pfalconbero: that's ec2 slave issue. I killed it, new one should be spawned soon13:21
beropfalcon: thanks13:21
liuyqhimamura, is there anyone who know whether there are source to force the monitor to display full screen in our linaro-android?13:28
mansson_Patrikryd: On Gator 5.8 "Verify that it actually works on Pandaboard (assign to ARM LT): TODO" Does this mean that you are BLOCKED waiting for the results?13:55
patrikrydNo I got a confirmation that it worked on RC113:56
patrikrydmansson_,   I received it today. I should update the BP13:56
mansson_patrikryd: You are reading my mind :-)13:56
asachello14:49
asacanybody here has a snowball v3?14:49
asacmansson_: patrikryd: fgiff: pabhishek: ?14:49
fgiffasac: No, just v214:49
mansson_We're all in meeting. later?14:50
davepigottasac: We have two in the val lab14:50
pabhishekasac, i have V514:50
asacthanks14:52
patrikrydasac, I have a v215:07
bhoj2pfefferz, https://android-build.linaro.org/builds/~linaro-android/staging-snowball/ is updated to 4.0.1  ..the text needs to be updated .15:10
pundiramithi pfefferz15:30
pundiramithi mansson_ patrikryd I dont have permissions to do a new build at https://android-build.linaro.org/15:35
pundiramitwhom to ping for help?15:35
mansson_Not sure... hang on15:37
pundiramitok15:40
patrikrydmaybe danilo15:40
patrikrydwho is probably called danilos15:41
mansson_Yeah. Or pfefferz himself. Try danilo first.15:41
pundiramitthanks :)15:42
pundiramithi danilos, I dont have permissions to do a new build at https://android-build.linaro.org/ please help.15:44
danilospfalcon, hi, can you perhaps help pundiramit? I am busy atm15:45
danilospundiramit, if pfalcon can't help you right now, I'd be happy to do that in an hour or so, so please re-ping me15:47
pundiramitsure thanks danilos15:47
pfefferzpundiramit, I can help you15:50
pfefferzpundiramit, are you going to switch the origen build over15:51
pundiramityes15:51
pundiramitpfefferz, yes15:51
pfefferz:)15:51
pfefferzpundiramit, you should be good to go15:52
pfefferz:)15:52
pfefferzthanks for switch the build over15:52
pundiramitlets give it a try15:52
pfefferzfabo, do you need anything else for the release15:52
pfefferz:)15:52
pundiramiti need to try ICS staging-origen on lcd15:52
pfefferzpundiramit, yeah, lets switch it over!15:53
pundiramitit works OK on HDMI15:53
pfefferzthats cool15:53
fabopfefferz: no, I don't need anything special from Android Team. It's ok afaik.15:55
pundiramitpfefferz, i think it is gonna take a while before I can actually trigger any build. still no build permissions for me.15:56
bhoj-2gpfefferz staging-snowball page on android-build needs an update abt 4.0.1 then we can announce it15:57
pundiramitbhoj-2g, can you start an staging-origen build for me?15:58
bhoj-2gpundiramit okay...15:58
pundiramitbhoj-2g thanks16:00
pfefferzpundiramit, it should work for you, try closing and reopening your browser16:01
pfefferzbhoj-2g, cool16:01
bhoj-2gpundiramit let me knw if it doesn work for u..ill then trigger a origen build16:03
pfefferzis anyone blocked an anything from me16:05
bhoj-2gpfefferz,nothing from my side :)16:08
pfefferzcool16:09
patrikrydbhoj-2g, I never managed to browse with Snowball ICS :(16:12
pfefferzpatrikryd, would you take some data and file a bug16:12
pfefferzpatrikryd, please assign it to yourself16:12
pfefferz:)16:12
bhoj-2gpatrikryd ,its working for me on panda though. pabhishek also mentioned the same thig16:14
mansson_Announcing a "Community page" on the wiki. It's not much and it shouldn't really be Android-specific but it has to start somewhere. Use it as a reference and by all means improve it. Here: https://wiki.linaro.org/Platform/Android/WorkWithCommunitie16:47
mansson_s16:47
pfefferzthanks mansson_17:25
pfefferzmansson_, let akgraner know about your community page17:27
pfefferzpundiramit were you able to kick off the build?17:28
pfefferzman the snowball really needs to change its mini-a connector18:37
pfefferzdoes anyone have a snowball20:15
beroI could go outside and make one.... Oh, wait, that's not the sort of snowball you mean ;-)20:15
pfefferzbero, :)20:32
beropfefferz: Some progress: https://android-build.linaro.org/builds/~berolinux/ics-imx53/#build=9 <--- boots, logcat shows all the DexOpt bits etc., but no output on VGA after booting the kernel, probably the graphics driver is missing20:33
pfefferzbero, hey cool!20:39
beroI see the android text on iMX53 ICS20:53
bero:)20:53
beroUI is up! :)20:55
beroLooks like ICS is totally useless without an egl.cfg20:55
* bero adds egl.cfg, but maybe we should just fix the defaults, that should make it easier for the other boards20:56
pfefferzbero, that's not a bad idea21:10
pfefferz:)21:10
beroOddly enough, the code already there to do a default in frameworks/base/opengl/libs/EGL/Loader.cpp Loader::Loader() looks right, that's exactly what I'd have put there21:14
beroMaybe it actually requires the presence of 0 1 something (I put 0 1 default in the test egl.cfg)21:15
pfefferzlooks like our local compile needs work21:43
pfefferzmake -C /u-boot ARCH=arm CROSS_COMPILE= O=21:43
pfefferzDirectory: out/target/product/snowball/data21:43
pfefferzmake: *** /u-boot: No such file or directory.  Stop.21:43
pfefferzmake: *** [build-uboot] Error 221:43
pfefferzmake: *** Waiting for unfinished jobs....21:43
beropfefferz: That looks like you didn't give it a TARGET_TOOLS_PREFIX22:05
* bero compiles locally all the time, works fine as long as I specify TARGET_TOOLS_PREFIX22:05
pfefferzh,mmm22:10
pfefferzdidn't like that: make TARGET_PRODUCT=snowball TARGET_TOOLS_PREFIX=./prebuilt/linux-x86/toolchain/arm-linux-androideabi-4.4.x/bin/arm-linux-androideabi- -j822:10
beroThat's odd... I don't see how that could cause that u-boot command line to be run22:11
pfefferzoh ick!22:11
pfefferzhow does that even work22:12
beroActually something EXTREMELY weird must be going on22:12
berobecause uboot doesn't do ARCH=arm22:12
pfefferzyeah...22:12
beroARCH=arm is a kernel thing22:12
beroSo it looks like somehow your kernel.mk is being run inside the u-boot tree, which doesn't make sense at all22:12
pfefferzI just sync'd snowball down22:12
* bero wonders if there's anything wrong w/ snowball and its custom uboot version...22:13
beroBut it does use the same uboot.mk we're using everywhere else, so I doubt it...22:13
pfefferzright22:13
pfefferzlet me try something22:13
beroBut that could explain why something adds ARCH=arm22:13
beroMaybe our uboot.mk calls u-boot's Makefile, and that in turn acts up on unexpected input22:13
mpoirierpfefferz: good day22:13
bero(but our u-boot's Makefile is ok)22:14
mpoirierpfefferz: have time for a little mumble ?22:14
pfefferzbero, I wonder if you source build/envsetup if it breaks things22:14
pfefferzmpoirier, sur, of course22:14
pfefferzactually22:14
beroI typically don't source build/envsetup... I just do make TARGET_PRODUCT=whatever TARGET_TOOLS_PREFIX=whatever systemtarball userdatatarball boottarball22:15
pfefferzmpoirier, got 10 min to make a movie and post it?22:15
mpoirierya sure.22:15
mpoiriermovie of ?22:15
beroBtw, what's up w/ the toolchain sync meeting that's supposed to be going on right now? Canceled?22:15
pfefferzbero, ahh22:16
pfefferzbero, missed that22:16
pfefferzbero, that bi-weekly meeting always trips me up22:16
beropfefferz: Looks like you're not the only one ;)22:16
pfefferz:)22:16
pfefferzbero, so if you just type make without systemtarball userdatatarball boottarball22:17
pfefferzyou get the error I'm seeing22:17
pfefferzhttps://android-build.linaro.org/builds/~linaro-android/staging-snowball/#build=8422:19
beromake without the *tarball bits doesn't work anyway (at least for me) because it tries to run all sorts of ABI checks that break22:29
berohttps://android-build.linaro.org/builds/~berolinux/ics-imx53/#build=10 <--- works22:35
beroiMX53 looks like it's in 256 color mode or something22:39
berobut other than that seems to work22:39
mpoirierpfefferz: ping22:49
mpoirierpfefferz: the build doesn't work and I think I know of at least one reason.22:49
mpoirierthe kernel video driver is not set to 640x480.22:50
mpoirierI'll fix that, push and kick another build.22:50
* bero starts an origen ICS build just for fun23:07
mpoirierpfefferz: problem is more severe...  surface flinger is not starting at all.23:14
mpoirierpfefferz: will investigate.23:14
mpoirierbero: ping23:23
pfefferzmpoirier, hmmm worked for me23:27
mpoirierpfefferz: what ?23:27
mpoirierhow could it...23:27
pfefferzthis build: https://android-build.linaro.org/builds/~linaro-android/staging-snowball/#build=8423:27
pfefferzworked for me23:27
mpoirierexactly, that's what I used.23:27
pfefferzit took a long time to boot23:27
pfefferzbut eventually it did work23:27
pfefferzmaybe try a reboot23:27
mpoirierhow long are we talking about here ?23:28
pfefferz10 min23:28
mpoirierreally, 10 minutes ?23:28
mpoirierok.23:28
pfefferzfabo, is also see a slow boot23:28
mpoirierlet me try again.23:28
mpoirierthat would imply the resolution doesn't have to be 640x480...23:29
mpoirierok, I'll try again.23:29
pfefferzmpoirier, no I saw it at 1280x72023:40
mpoirierpfefferz: you're right23:40
mpoirierit just came up23:40
mpoirierafter... 7 minutes.23:40
mpoirierthat is horrible.23:40
mpoirierok, time to shoot the horror movie23:40

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!