/srv/irclogs.linaro.org/2018/01/08/#linaro-android.txt

liuyqping bero bero|208:37
minseongping bero bero|208:37
bero|2liuyq: minseong: hi, sorry, had to catch a dog outside... back now08:40
bero|2are you still there?08:40
minseongbero|2: I am here~08:41
bero|2great, thanks08:42
liuyqhi, bero08:42
liuyq and bero|2 :)08:42
bero|2hi liuyq08:43
bero|2Let's get this thing started...08:43
bero|2#startmeeting08:43
linarobotMeeting started Mon Jan  8 08:43:17 2018 UTC.  The chair is bero|2. Information about MeetBot at http://wiki.debian.org/MeetBot.08:43
linarobotUseful Commands: #action #agreed #help #info #idea #link #topic.08:43
bero|2No big news today, but let's talk about Connect a bit after the status updates...08:43
bero|2minseong: do you want to go first with updates?08:44
minseong=== Plans for This Week ===08:44
minseongContinue the maintenance of the two CIs and LAVA boot-up test08:44
minseong1. Follow up errors while building AOSP images (occured at 5 Jan).08:44
minseongError building images08:44
minseongart/tools/ahat/src/main/com/android/ahat/HeapTable.java:69: error: unmappable character (0xCE) for encoding US-ASCII08:44
minseong      subcols.add(new Column("??", Column.Align.RIGHT, snapshot.isDiffed()));08:44
minseong.....08:44
minseong--> Not reproducible in my local env. This looks like the docker's Java environment issue regrading the default Java file encoding not being UTF-8. Will report this issue to the CI personell to fix this issue in the docker.08:44
minseong2. Follow up infrastructure errors on "Building AOSP master using the upstream clang"08:44
minseongOn going: infrastructure errors while uploading built images (from 27 Dec)08:44
minseonghttps://projects.linaro.org/browse/CTT-80408:44
minseong3. Follow up image flashing errors on "LAVA boot-up test"08:44
minseongOn going: Caused by firmware update at 28 Dec: currently being handled by IT service.08:44
bero|2yes, agree with 1. -- chances are it's a matter of setting LANG, LC_ALL and friends, I think if they're set to C, the default file encoding is ASCII08:46
minseongbero|2: yes~ precisely thanks08:47
bero|2Is 2. still happening? Comments on the report seem to think it should be fixed08:47
minseongYes still uploading error until 4th Jan.08:47
bero|2ok... Guess it's not our bug08:48
bero|2do you have anything else?08:48
minseongbero|2: that's all for now. Thanks08:48
bero|2thanks08:49
bero|2liuyq: do you want to go next?08:49
liuyqbero|2, here is my plan for this week:08:49
liuyqPlan for next week:08:49
liuyq1. Check the result on lava for builds with android tag 8.1.0_r7 tag08:49
liuyq2. Continue on the hikey UI crash problem investigation08:49
liuyq    - try to cherry pick other art changes to see if that help08:49
liuyq    - try to check changes related icu/Cleaner/NativeAllocationRegistry/ReferenceQueue in libcore/frameworks project08:49
liuyq3. LCR report management system08:49
liuyq    - update to support new cts result output08:49
liuyqlast week tried with many changes on art/runtime/gc  cherry picked from master version, but the problem still there08:50
bero|2I think it probably isn't icu's fault because then we'd see the problem on other hardware... But let's make sure08:51
liuyqtried with changes cherry pick  on libnativehelper, luni/src/main/native/libcore_icu_XXX from master too, but not much help.08:51
bero|2I put a current build on my Nexus 6P in the weekend, not seeing the problem there08:51
bero|2but of course  can't rule out anything...08:53
liuyqbero|2, I still not got how java.math.NativeBN.BN_copy and ucnv_close are called there, not sure if you have any idea about that08:53
liuyqbero|2, btw, the problem could be reproduced with monkey command like this: adb shell monkey -s 3 --pct-syskeys 0 -p com.android.settings 100008:54
liuyqbero|2, you could try it with your Nexus 6P as well, but guess the problem won't be reproduced08:55
bero|2yes, will try that on the 6P, thanks for the hint. So far I was just using the 6P as a normal user08:55
bero|2Not sure where NativeBN and uconv_close would be called08:56
liuyqbero|2, ok, thanks, that's all from my side I think08:57
bero|2thanks08:57
bero|2LiXingHe: are you around? Do you have any news?08:57
bero|2ok, let's change topics for now -- Connect08:59
bero|2The deadline for talk and demo submissions is coming up...08:59
bero|2Does anyone have ideas for talk topics and/or demos?09:00
bero|2I've been thinking we may want to do another lightning talk primarily so we sync up the various LMG subteams - we should try harder to stay aware of what the ART team is doing (and vice versa)09:04
liuyqmight be better if we could talk about how to quickly setup builds with latest new android tag, and with the problems fixed, from the UI hikey crash problem and the hw vsync problem we had before, it's really not a easy thing to do..09:06
bero|2true09:07
bero|2I hope some interesting stuff will come up at Bootcamp, maybe we can talk about and/or something they're introducing about P there09:09
LiXingHebero|2: no update.09:10
bero|2stuff like Codec 2.0 might finally come around09:10
bero|2LiXingHe: ok, thanks -- let me know if you need anything from our side09:10
liuyqbero|2, when is the Bootcamp?09:10
minseongbero|2: true09:11
LiXingHebero|2 thanks bero。09:11
bero|2liuyq: February 5 to 909:11
liuyqbero|2, oh, it's one month earlier than our connect, should be some interesting things we could talk at connect09:13
bero|2assuming we get anything there that we can talk about without NDA09:14
bero|2I'm also hoping we can demo the AOSP TV stuff, but that's more on the LHG side of things... But we can turn that into a joint demo if we get to a point where e.g. AOSP TV can be enabled in any of our builds (working on that)09:16
minseongbero|2: sounds good.09:17
minseongbero|2: this NDA stuff might well be there for all of the sutff presented there.09:17
bero|2minseong: yes, Bootcamp is usually NDA unless we can find independent sources on any information we want to share and/or get the Android guys to explicitly allow it09:20
minseongbero|2: that's true ~ ^^09:21
bero|2unfortunately Bootcamp us quite a bit after the deadline for Connect talk submissions, so getting a new session based on some material we find out about there on the agenda may be a bit hard09:21
bero|2but probably if it's a good topic we can do it.09:22
bero|2Does anyone have more suggestions?09:22
minseongbero|2: Sorry not being extremely helpful here.09:24
bero|2No problem, it's currently a little hard to find new things to talk about and demo for everyone...09:25
bero|2Let's keep thinking about it and bring it up again in the next meeting09:25
bero|2will also sync up with Tom09:25
bero|2That's all I had for today... Have a great week everyone!09:26
bero|2#endmeeting09:26
griffinppraneeth: hey18:17
praneethgriffinp,  hello18:31
griffinppraneeth: are you praneeth from ti?18:31
praneethyes18:32
griffinppraneeth: ok great :) I was told earlier that am57x had been updated to use v4.9 kernel18:33
praneethgriffinp,  are you peter from LHG ?18:34
praneethdepends on what feature you are looking for.18:34
griffinppraneeth: yes :)18:34
praneethhave issues with graphics .18:34
griffinppraneeth: I did yes..but that seems to be a build issue..a clean re-build and the powervr issue has gone18:35
praneethlinux or android ?18:36
griffinppraneeth: I noticed there is a http://git.ti.com/android/kernel/commits/ti-android-linux-4.9.y, but it is missing some files like android_am57x.cfg which existed in v4.4 tree18:36
griffinppraneeth: android18:36
griffinppraneeth: just wondering what current status of v4.9 with android is for am57x18:37
praneethas it stands now ... you will need the following patches in kernel to make it atleast boot on android o with commandline18:37
praneethhttp://review.omapzoom.org/#/q/status:open+project:kernel/omap+branch:p-ti-android-linux-4.9.y+topic:4.9_dec_update18:37
praneethwith oreo18:37
praneethand 2017.01 uboot .18:38
griffinppraneeth: I also now have an account on http://review.omapzoom.org as well18:38
praneethcool18:38
praneethgraphics is stil la issue though ..18:38
griffinppraneeth: what are the graphics issue?18:39

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