07:34:03 <xruxa> #startmeeting
07:34:03 <linarobot> Meeting started Thu Oct 15 07:34:03 2015 UTC.  The chair is xruxa. Information about MeetBot at http://wiki.debian.org/MeetBot.
07:34:03 <linarobot> Useful Commands: #action #agreed #help #info #idea #link #topic.
07:34:09 <xruxa> #chair bero
07:34:09 <linarobot> Current chairs: bero xruxa
07:34:10 <bero> Qian: no problem
07:34:17 <xruxa> morning everyone
07:34:43 <Qian> hello everyone
07:34:49 <bero> hi everyone
07:34:56 <xavierhsu> Hi Bero
07:35:07 <yspan> hi
07:35:11 <xruxa> bero, why don't you go first, you have lots on the doc :)
07:35:58 <bero> ok...
07:36:17 <bero> What did you do?
07:36:18 <bero> Updated binutils in our gcc 5.1 toolchain builds for compatibility with M
07:36:18 <bero> Built M with gcc 5.1, ran tests (it boots and works perfectly)
07:36:18 <bero> Updated gcc 5.1 based toolchains in prebuilts/ to gcc-linaro 5.1-2015.08
07:36:18 <bero> Started bisecting toolchains to find out what causes 5.1 to work and 5.2 to break
07:36:18 <bero> Added 7 optimization opportunity cards to JIRA in case new assignees start coming in
07:36:18 <bero> Submitted gcc 5.x related patches for M upstream
07:36:19 <bero> Improved libc++ configuration for recent versions of gcc
07:36:19 <bero> What are you going to do?
07:36:20 <bero> Continue bisecting toolchains
07:36:20 <bero> Start working on LTO with gcc 5.1, can update to 5.2 later
07:36:21 <bero> Check if M can be built with clang
07:36:21 <bero> What problems do you have?
07:36:22 <bero> The “Earth” app is buggy… Can’t configure a day to have more than 24 hours! Doesn’t seem to be Open Source and they don’t seem to take bug reports though.
07:37:30 <xruxa> bero, thanks for the new optimizations items - team feel free to have a look and volunteer.
07:38:45 <xruxa> yspan,would you go next?
07:38:51 <bero> I'm "volunteering" everyone for https://projects.linaro.org/browse/LMG-915
07:38:53 <yspan> sure
07:39:17 <bero> I'd like everyone to keep that one in mind and come back with at least 1 idea next week
07:39:19 <yspan> i discussed with xavierhsu how to segment encoding part of libjpeg
07:39:28 <xruxa> +1 on that bero - perhaps we should have a go after the stats round? Perhaps people have ideas already
07:39:55 <yspan> I am going to continue to separate encoding part of skia
07:40:03 <Qian> i have one question about libjepg-turbo
07:40:18 <bero> xruxa: yes, let's have a first run today and another one next week
07:40:27 <xruxa> bero, ok
07:40:30 <bero> Qian:  go ahead
07:40:47 <Qian> the improvement of performance is NOT remarkable through my test case on MEIZU MX5 and PRO 5
07:41:06 <Qian> I don't know why
07:41:30 <bero> That's odd, we've seen pretty good improvements on Nexus devices and Juno. What did you do to test the performance?
07:41:42 <yspan> Qian, what benchmarks did U use?
07:41:45 <Qian> i can send you the test case
07:42:06 <Qian> i just test cjpeg and djepg
07:43:10 <bero> were MX5 and PRO5 using regular libjpeg before? If their stacks were based on someone's BSP instead of AOSP, there's some chance they were using libjpeg-turbo or something similar already
07:44:08 <Qian> the libjpeg is the same as AOSP
07:44:29 <xruxa> Qian, email your finding to bero and yspan please.
07:44:39 <xruxa> lets finish the status round
07:44:43 <xruxa> yspan, anything else?
07:44:45 <Qian> OK
07:45:17 <xruxa> #action Qian to email to bero and yspan about his findings about libjpeg-turbo on Meizu platforms
07:45:31 <xruxa> hearing none, lets move to xavierhsu
07:45:38 <yspan> no, that's all
07:45:42 <xavierhsu> What did you do? #Separating libjpeg into encoder and decoder.
07:46:08 <xavierhsu> What are you going to do?Investigating and modifying the code of Android will call libjpeg.so.
07:46:41 <Qian> i want to transplant libjepg-turbo  on MX%
07:46:44 <xavierhsu> What problems do you have? No, because I compile now.
07:46:48 <xruxa> xavierhsu, how is that separating going? You know how to do it?
07:47:24 <Qian> i didn't seperate the encoder and decoder
07:47:51 <Qian> just test the performance of cjepg and djpeg
07:48:03 <bero> Qian: we know, this is about something else (our oogoing effort to reduce memory usage)
07:48:16 <xavierhsu> xuxa, according to yspan tell me, I try to separate libjpeg.
07:48:54 <xavierhsu> I seperate libjpeg into libjpeg.so and libjpeg-enc.so.
07:49:05 <Qian> bero: yes, but my company wants to transplant them
07:49:50 <Qian> if i can get the perfermance data quickly, then i can spend more on reducing the memory work
07:49:56 <bero> Qian: yes, we'll talk about that when the status round is done
07:50:21 <xruxa> xavierhsu, thanks. LEt us know tomorrow how far you got with that code separation.
07:50:53 <xruxa> luther, want to go next?
07:51:20 <xavierhsu> xruxa, ok. I will.
07:51:51 <luther> ok
07:51:58 * luther What did you do?
07:51:58 <luther> 1. Google Doc can be used now
07:51:58 * luther What are you going to do?
07:51:58 <luther> 1. Setup eclipse to peruse java code
07:51:58 * luther What problems do you have?
07:51:59 <luther> 1. None till now
07:53:22 <xruxa> luther, ok.
07:53:27 <xruxa> sunao, you are next please.
07:53:35 <sunao> What did you do?
07:53:37 <sunao> 1. take a look at libogg and libraries depend on it, such as libopus, libvorbis, libvorbisidec(tremolo)...
07:53:38 <sunao> what are you going to do?
07:53:40 <sunao> 1、continue look at the libogg/libvorbis/libvorbisidec.
07:53:41 <sunao> what problems do you have?
07:53:43 <sunao> I still can't use google related website in my office, I will update the status home.
07:54:22 <xruxa> sunao, oh, you moved on from giflib to other libs. Lets update the JIRA tasks accordingly. Good progress. Thanks.
07:54:50 <xruxa> and Qian - anything on top of what you wrote already? bero - would you have some task to assign to Qian ?
07:55:35 <bero> yes -- any of the new ones. Got to see what he's most interested in
07:56:06 <xruxa> bero, do you agree that we are done with giflib or still some tasks left there?
07:56:13 <Qian> What did you do?
07:56:13 <Qian> Collect the performance data on MEIZU MX5 and PRO 5
07:56:13 <Qian> What are you going to do?
07:56:15 <Qian> I will try with different pictures
07:56:17 <Qian> What problems do you have?
07:56:19 <Qian> The improvement on MEIZU MX5 and PRO 5 is not remarkable, I need some help
07:56:21 <Qian> why this happens?
07:56:36 <sunao> xruxa, do we have this meeting tomorrow? the agenda said we don't have meeting on Monday and Friday
07:57:12 <xruxa> sunao, yes, Tom should send invitation for the meeting early morning Chinese time.
07:57:23 <xruxa> if not, I will be here instead at the usual time
07:57:39 <xruxa> #action tgall_foo to send Monday and Friday meeting invitations
07:57:57 <xruxa> #info new item in sprint LMG-912 for sunao
07:58:03 <sunao> OK
07:59:39 <xruxa> sunao, thanks. zhizhoutian_ - do you have update for us?
07:59:46 <zhizhoutian_> What did you do?
07:59:46 <zhizhoutian_> Fix juno aosp build
07:59:46 <zhizhoutian_> What are you going to do?
07:59:46 <zhizhoutian_> Fix juno aosp build
07:59:46 <zhizhoutian_> What problems do you have?
07:59:46 <zhizhoutian_> Investigate how to reduce android build warning-error level, because i have got a lot of warning error.
08:00:19 <zhizhoutian_> xruxa: i am investigating this thing.
08:00:51 <xruxa> zhizhoutian_, still on the Juno build invetigations? Ok, hope you crack that solved soon :)
08:00:55 <zhizhoutian_> xruxa: there are some warning errors in juno builds. some upstream build issue has been fixed.
08:01:30 <zhizhoutian_> xruxa: ok;)
08:01:41 <xruxa> #action Qian to check new tasks in JIRA (https://projects.linaro.org/secure/RapidBoard.jspa?rapidView=38&view=planning.nodetail) and see which one he wants to work on (from backlog)
08:01:47 <xruxa> zhizhoutian_, thanks.
08:01:50 <xruxa> Did I forget someone?
08:02:55 <Qian> OK, i will
08:04:32 <bero> I think we got everyone
08:05:04 <xruxa> alright, lets close the status part of the meeting and any discussion that is on your mind can happily continue
08:05:07 <xruxa> #endmeeting