/srv/irclogs.linaro.org/2012/01/05/#linaro-android.txt

vk41hi and good morning all03:27
tgall_fooah-ha!03:58
tgall_fooskia_bench on ICS 4.0.3 on panda is segv03:58
tgall_fooI went back to a plain libjpeg and it still fails so I know it's not the libjpeg-turbo code03:58
tgall_foobug #912035  linaro-android now copied04:02
ubot5Launchpad bug 912035 in libjpeg-turbo "android : skia-bench segvs " [High,Confirmed] https://launchpad.net/bugs/91203504:02
botao_sunping pfefferz05:01
botao_sunpfefferz: If you're around, take a look at the email.05:05
Dr_Whobotao_sun: I think he's on vacation yet05:33
pundiramitbhoj, did you push android-x86 wifi patch?05:38
nicksydneyHi...09:15
pundiramithi nicksydney09:16
nicksydneyI managed to successfully built the android toolchain  following the wiki https://wiki.linaro.org/Platform/Android/Toolchain but when i use the toolchain to build Gingerbread AOSP it throws error but when i revert back using the prebuilt GCC toolchain from AOSP it worked fine..does this mean that the latest toolchains are not backward compatible with old android AOSP code ?09:16
pundiramitAOSP or Linaro modified AOSP?09:20
nicksydneypundiramit: AOSP09:49
pundiramitnicksydney, it won compile then09:49
nicksydneydo this means that the AOSP and "Linaro AOSP" have got different code base ..different in a sense that it is modified to make it work/compiled using the Linaro toolchain ?09:50
pundiramitlinaro android toolchain is meant to be used for linaro modified AOSP.. upstreaming linaro toolchain modifications to AOSP is in TODO list09:50
pundiramitsame framework codebase but modified to support gcc-4.609:51
pundiramitfeature/framework wise it is almost same09:52
pundiramitnicksydney, yes you got the point.. modified to make it work/compiled using the Linaro toolchain09:52
nicksydneyso say I have the AOSP code (Gingerbread) and I want to compare it with Linaro modified AOSP code (Gingerbread) too...where can i check out the code from ?09:54
pundiramitnicksydney, which GB version? 2.3.4/2.3.7 ?09:57
nicksydneyin AOSP the tag is called 'gingerbread' so I assume that would be the latest one 2.3.7 ?09:59
pundiramitnicksydney, ok10:28
pundiramitcheckout linaro-android-2.3.7 code "repo init -u git://android.git.linaro.org/platform/manifest.git -b linaro_android_2.3.7"10:29
pundiramitrepo sync10:29
nicksydneyok thanks10:57
=== ramana_away is now known as ramana
doanacnicksydney: you may want to subscribe to this bug: https://bugs.launchpad.net/linaro-android/+bug/87897314:52
ubot5Ubuntu bug 878973 in Linaro Android "Builds can not be compiled with the original AOSP GCC 4.4 toolchain" [High,In progress]14:52
nicksydneydoanac: thanks15:01
mpoirierbero|2: good day15:29
bero|2hi mpoirier, *15:29
mpoirierbero|2: I'm sending you a pastebin... very short.  See if you can make something out of it.15:30
mpoirierhttp://paste.ubuntu.com/793844/15:30
mpoirierbuild on ubuntu 10.04, 64 bit.15:30
mpoirier10.04 is very old...15:31
mpoirierbero|2: I will request that they send me all they steps they're taking for the compilation.15:32
bero|2mpoirier: I've seen that before...15:33
bero|2The short answer is "Ubuntu sucks"15:33
mpoirierbero|2: fabulous.15:33
mpoirierI knew it.15:34
mpoiriershould they upgrade to 11.10 ?15:34
bero|2The longer answer is "They broke the host compiler by making it define _FORTIFY_SOURCE by default in a way that conflicts with defining it manually"15:34
bero|2Quick fix: Use gcc-4.5 as host compiler15:34
bero|211.10 probably won't help, I think they never fixed that bug15:34
mpoirierso how does it work on hackbox ?15:35
bero|2We need to mention this + the workaround on the wiki somewhere15:35
bero|2I think the hackbox is old enough to have gcc 4.5 as the main host compiler15:35
bero|2They only broke 4.615:35
bero|2Let me check15:35
mpoirierpls.15:35
mpoirierbero|2: I run 11.10 and it is 4.6.1...15:36
bero|2The hackbox is even older. Here's why it works:15:37
mpoirierbero|2: hackbox is 4.4.515:37
bero|2ubuntu@ip-10-204-73-140:~$ gcc -v 2>&1 |tail -n115:37
bero|2gcc version 4.4.5 (Ubuntu/Linaro 4.4.4-14ubuntu5)15:37
bero|2They only started breaking it with 4.6 builds15:37
bero|2So anything older is ok15:37
bero|2So is compiling 4.6 from source15:37
bero|2It's just ubuntu packaging that causes that breakage15:38
mpoirierbero|2: so the work around is to install gcc > 4.6 ?15:38
bero|2No, it is to install either gcc < 4.6 from their packages, or to install any gcc from source15:38
bero|2The key thing is "don't use any Ubuntu packaged gcc >= 4.6"15:38
mpoirierya, brain bug here...15:39
bero|2mpoirier: run "apt-get dist-upgrade" on your brain, that should fix it ;)15:39
mpoirierbero|2: cool - I'll communicate the new to them.15:39
mpoirierbero|2: my brain is unfixable...15:39
bero|2The other workaround is to remove any manual definition of _FORTIFY_SOURCE from the source tree15:40
mpoirierbero|2: beurk....15:40
bero|2But that will cause builds on non-Ubuntu compilers to not have fortify features built in, so I'm somewhat reluctant to do that15:40
bero|2I never like punishing others for Ubuntu's bugs15:40
mpoirierbero|2: thanks for your time.15:41
botao_sunpfefferz: Hi Zach, I'm going to sleep and I have sent 3 emails to you. One of them is important because it will determine the end date of our event and I haven't bought the ticket yet.16:04
pfefferzbotao_sun, I'll take a look16:05
pfefferzbotao_sun, stay through ELC16:06
pfefferzbotao_sun, sound good?16:06
botao_sunpfefferz: OK, so the end date will be the 16 Feb.16:06
botao_sunpfefferz: It's OK for me.16:06
botao_sunpfefferz: Then I will buy the ticket tomorrow.16:06
botao_sunpfefferz: And book the hotel room. Thank you.16:06
pfefferzbotao_sun, it goes through Feb 17th16:07
botao_sunpfefferz: Hold on16:07
bhojpfefferz, same here . I have not booked the tickets yet . wanted to confirm  the list of events to attend . ELC also right ?16:07
pfefferzbhoj, yes16:07
botao_sunpfefferz: From our website, it will be from 14 to 16. Does it extend to 17?16:08
botao_sunpfefferz: http://connect.linaro.org/events/event/embedded-linux-conference/16:08
pfefferzhttps://events.linuxfoundation.org/events/embedded-linux-conference16:09
botao_sunpfefferz: OK, got it!16:10
botao_sunpfefferz: Another 2 emails are related to the task in this month and our event, not hurry to reply.16:11
botao_sunpfefferz: Thank you!16:11
=== AndChat- is now known as bhoj-2g
pfefferzbero|2, hey hey16:31
bero|2hi pfefferz16:31
pfefferzhey bero|2, hows it going?16:31
bero|2pfefferz: Quite well, and for you?16:31
pfefferzwell16:32
pfefferzbero|2, did the 4.0.3 switch over go okay?16:32
bero|2yes, no problems, all done16:32
pfefferzbero|2, has someone tested all the baselines yet?16:34
bero|2Yes, various people have tested them16:36
pfefferzbero|2, okay cool. Did people record which build they tested and what they tested in the target spreadsheets?16:37
bero|2I don't think so16:38
bero|2I think it's just been informal testing16:38
doanachey guys - we may want to update the documentation for android-build on landing-panda builds.16:38
doanacthe install-binaries.sh script grabs 4.0.1 code not 4.0.3.16:38
pfefferzbero|2, okay I'll ask Abhishek to run through the tips16:39
doanacI lost a couple of hours making sense of that :)16:39
pfefferzdoanac, would you send a note to bhoj about it16:39
pfefferz?16:39
doanacmaybe we should version the script ie - install-binaries-4.0.1.sh and install-binaries-4.0.3.sh16:39
doanacpfefferz: sure. thanks16:39
pfefferzdoanac, yeah, the version would be a good idea16:39
doanacpfefferz: good news is I have 4.0.3 running with and prebuilt 4.4 compiler again16:40
pfefferzdoanac, hey cool16:40
pfefferzdoanac, are you going to push the changes to Gerrit?16:41
doanacpfefferz: will be doing it shortly. I was trying to make sure my changes weren't what was causing gfx to fail :)16:41
pfefferzdoanac, okay, bero|2 would you take a look at doanac's changes?16:42
bero|2pfefferz: sure16:42
bero|2doanac: Please make sure you put me in the reviewer list so I get notified when you submitted them16:43
doanacbero|2: will do.16:44
doanacbero|2: BTW - that's a cool thing you did in the makefiles to see if the compiler supports certain flags. I wish I'd known about that trick about 4 years ago! :)16:45
bero|2doanac: It didn't exist 4 years ago - I added the macros we're using for that just a couple of months ago16:45
doanacbero|2: no i mean just the general idea. In my last job I did some ugly hacks to try and provide something similar16:46
pfefferztixy, I've scheduled board specific sessions for all the boards at connect except Versatile Express, like https://blueprints.launchpad.net/linaro-android/+spec/linaro-platforms-q112-android-ics-imx5316:51
* tixy looking16:51
tixypfefferz: Might be good to have one for vexpress as well?16:53
pfefferztixy, aye, just wanted to get your input16:55
pfefferztixy, I'll file it16:56
tixypfefferz: We'll be getting UEFI working for android over this month and into next, and there is another board which we will be working on, so it's probably more like an ARM LT Android session rather than just Vexpress16:57
pfefferzright16:58
pfefferztixy, can I assign the session to you?16:58
tixypfefferz: yes16:59
pfefferztixy, done https://blueprints.launchpad.net/linaro-android/+spec/linaro-platforms-q112-android-ics-versatile17:00
tixypfefferz: thanks17:01
tixypfefferz: can I subscribe you?17:01
pfefferztixy, yeah17:03
pundiramitpfefferz, hello there17:17
pundiramitwelcome back :)17:17
pfefferzhey pundiramit !17:21
pfefferz:)17:21
pfefferzthanks17:21
tgall_foopfefferz, if you haven't seen : https://bugs.launchpad.net/libjpeg-turbo/+bug/912035,  you probably want to pick up this fix :  https://github.com/CyanogenMod/android_external_skia/commit/f25df1f7228ef8151ec6633c649e5081d4c28f6717:31
ubot5Ubuntu bug 912035 in Linaro Android "android : skia-bench segvs " [Undecided,New]17:31
pfefferztgall_foo, okay, since we track, <project path="external/jpeg" name="people/tomgall/libjpeg-turbo/libjpeg-turbo" revision="android" remote="linaro-other"/>17:34
pfefferzwe should be cool right?17:34
tgall_foono, the fix is in skia-bench itself17:34
tgall_foocool part is I will have libjpeg v libjpeg-turbo skia_bench numbers on panda here in a few17:35
pfefferztgall_foo, ahhh17:38
pfefferztgall_foo, we'll have to create a branch then17:38
pfefferzthanks for the heads up17:38
doanactgall_foo: regarding that commit. I guess I only checked it into 4.0.1 and did it after 4.0.3 was created17:42
doanachttp://review.android.git.linaro.org/#change,130817:42
doanacI'm doing some 4.0.3 changes right now, so I'll do a submission for it.17:42
tgall_foodoanac, awesome thanks17:43
doanactgall_foo: http://review.android.git.linaro.org/135017:50
doanacpfefferz: ^^^ can you merge that change?17:50
pfefferzdoanac, let me take a look?17:56
pfefferzdoanac, it got pushed to the wrong branch17:57
pfefferzdoanac, let me take a look at the current manifest17:57
doanacpfefferz: I was using landing-panda.xml. I think you are right about the manifest being off18:00
doanachttp://android.git.linaro.org/gitweb?p=platform/manifest.git;a=blob;f=landing-panda.xml;h=05c5f49d46544ff04ed39840da6ee8eaf4c2f379;hb=linaro_android_4.0.318:00
pfefferzyeah, we're still using the AOSP default18:01
pfefferz<project path="external/skia" name="platform/external/skia"/>18:01
pfefferzso we'll need to create a linaro_android_4.0.3 branch18:02
pfefferzand put the change on it18:02
doanacpfefferz: given the whole manifest is specifying the wrong branch, can we just rename the current branch?18:04
doanacn/m18:04
pfefferzlet me take a look at things18:04
doanacthis is why you should never go on vacation! :)18:05
pfefferzhehe18:05
pfefferz:)18:05
pfefferzhmmm18:06
pfefferzjust this huh, #elif defined(SK_BUILD_FOR_UNIX) || defined(ANDROID)18:06
doanacpfefferz: yeah. the commit message describes why18:07
Dr_Whoyup simple as that pfefferz18:10
pfefferzhmmm18:10
pfefferzso this came in because of libjpeg-turbo's upgrade right?18:11
Dr_Whoso at some point is the panda image going to have the icons and such back?   the 4.0.3 code I have just displays "android" and that's it18:11
pfefferzyeah, everything will be working again18:12
pfefferzjust need to sort it all out18:12
Dr_Whowe want it because I'm running skis-bench both with and without ltj18:12
doanacpfefferz: it has nothing to do with libjpeg turbo actually18:12
doanacskia_bench just flat out didn't work18:12
Dr_Whoexactly18:12
pfefferzokay18:13
pfefferzhave we sent the change to Google?18:13
doanacpfefferz: I don't think so. I'm not really sure how that process has been working for this team.18:13
doanactgall_foo: you may need the right binary bits for your 4.0.3 panda. take a look at: http://people.linaro.org/~vishalbhoj/18:14
doanacvishal uploaded a new script that does the right install18:14
Dr_Whook I'll have a look …  for what I'm doing right now I'm ok18:15
pfefferzDr_Who, take a look at the instructions listed on https://android-build.linaro.org/builds/~linaro-android/landing-panda/#build=1118:21
mpoirierpundiramit: ping18:26
pundiramithi mpoirier18:27
mpoirierpundiramit: congrats on the wifi18:27
pundiramitsorry I missed you yesterday18:27
mpoirierpundiramit: so did I - my apologies for that.18:27
pundiramitmpoirier, thanks.. but still lot to do18:27
mpoirierindeed - but you may be chasing  a ghost.18:28
pundiramitmpoirier, we need to improve on chip powerup failure18:28
pundiramitfailure rate18:28
mpoirierindeed.18:28
mpoirierlet me think a little.18:28
pundiramitthanks18:28
mpoirierpundiramit: let me fix a pastebin...18:29
pundiramitmpoirier, it gets 3 types of failure.. one is fatal upon which we need to reboot the board, to use wifi other 2 failures can be ignored for Wifi18:30
pundiramitmpoirier, i will shoot you one mail on this and what I have done so far18:30
mpoirierpundiramit: http://paste.ubuntu.com/794033/18:30
mpoirieryou're ahead of me now.18:31
mpoirieron line 27-29 is what i think is the power failure.18:31
pundiramitmpoirier, both of these failure can be ignored for wifi18:31
mpoirierok18:31
pundiramitwait18:31
mpoirierforget line 26, I'm chasing after something else.18:32
mpoirierpundiramit: still with me ?18:35
pundiramityes18:36
pundiramitcreating a pastebin of errors18:36
pundiramitgive me 5 more minutes18:36
mpoirierack18:36
pfefferzpundiramit, your ABS/ELC session looks great18:39
pfefferzpundiramit, feel free to submit it18:39
pundiramitpfefferz, thanks18:39
pundiramitmpoirier, list of driver failures http://paste.ubuntu.com/794041/ which are very common and random18:46
mpoirierpundiramit: let me see.18:46
pundiramitmpoirier, we seriously need to deal them to get stable release out18:47
pundiramitmpoirier, i will start looking into it from tomorrow. user-space is almost ready.18:47
mpoirierpundiramit: 0 is common and different from ubuntu.18:48
mpoirierubuntu doesn't have line 33 and 35.18:48
pundiramit0: is not a failure18:48
pundiramitok18:48
mpoirierat least i've never seen it on that side.18:48
pundiramitok18:48
mpoirierit may not be a failure but I've noticed it.18:49
mpoirier1 and 2 I've never seen.18:49
mpoirier6 I've seen, and a lot18:49
pundiramit1 is the fatal error18:50
pundiramitno way out18:50
pundiramithave to reboot18:50
mpoirierok. ok.18:50
mpoirierlet me think.18:50
pundiramit2 and 3 can be ignored for wifi18:50
pundiramitI also get last error often while inserting cw1200_wlan module18:51
pundiramitworkaround is to insert the module again after some time18:51
mpoirierlast is because sys_clk_out has not been initialized by BT.18:52
mpoirierdo something for me.  grep SKE .config18:52
pundiramitok18:53
pundiramitmpoirier, CONFIG_KEYBOARD_NOMADIK_SKE=y18:54
mpoiriercrap.18:54
pundiramitmpoirier, I think you once told me to get rid of that iirc18:54
mpoirierI did.18:54
mpoirieractually,18:54
mpoirierI don't understand how it can work at all with it on.18:55
pundiramiti have not updated my tree for a while18:55
pundiramiti'm on 4.0.118:55
pundiramitwill move to 4.0.3 tomorrow18:55
mpoirierit has nothing to do with 4.0.1.18:55
mpoirierit's a kernel thing.18:55
pundiramityes but you said you fixed it18:56
pundiramiti thought you fixed it in 4.0.3 branch18:56
mpoirierlet me verify.18:56
mpoirierpundiramit: before you get the guns out on this one,18:59
mpoirier1. disable CONFIG_KEYBOARD_NOMADIK_SKE and test again.18:59
pundiramitsure18:59
mpoirier2. clone stable-android-ux500-3.1-mali-v9 from git://igloocommunity.org/git/kernel/igloo-kernel.git19:00
mpoirierrecompile the kernel using u8500_android_defconfig.19:00
mpoirierand drop the uImage under /boot/19:00
pundiramitso you want me to move to 3.1 from now onwards19:00
mpoirierit's brand new, did it 2 days ago.19:00
pundiramitmpoirier, cool19:01
mpoirierit's the same in functionaility.19:01
pundiramitmpoirier, will move to it19:01
mpoirier*but* I've never seen the errors you showed me.19:01
pundiramitdo I still have to do "1." on 3.0-stable branch?19:02
mpoiriersorry I don't understand.19:02
mpoirierthe "1.".19:02
pundiramit disable CONFIG_KEYBOARD_NOMADIK_SKE and test again.19:03
mpoirieroh,19:03
mpoirierno, it's off by default.19:03
pundiramitok19:04
mpoirieryou see, on 3.1 CONFIG_KEYBOARD_NOMADIK_SKE *has* to be off for BT to work.19:04
pundiramityes I will make that sure19:04
mpoirierand on that 3.1, I've never seen error 1 and 2.19:04
mpoiriererror 3 is persistant.19:04
mpoiriersorry, let me take that back.19:05
pundiramitthats good to know19:05
mpoirier0 and 3 are the normal.19:05
pundiramitas long as I'm not seeing error 1 I'm good19:05
mpoirierI've never seen it and i've been booting BT about 50 times in the last 2 days.19:06
pundiramitawesome19:06
mpoirierthere is a feature interaction issue with CONFIG_KEYBOARD_NOMADIK_SKE19:06
pundiramitok19:06
mpoirierpin 161 is sinking everything, hence disabling it.19:06
mpoirierbut it is still a huge problem.19:07
mpoiriergimme 30 seconds...19:07
mpoirierI want to make sure i give you rock solid information.19:07
pundiramitok19:10
mpoirierpundiramit: 100% sure now.19:11
pundiramitok19:11
mpoirieronly #3, nothing else.19:11
pundiramitthats good to know19:11
pundiramitI will jump to 4.0.3 and 3.1 kernel tomorrow19:11
mpoirierpundiramit: I'm on 4.0.1 too.19:11
mpoirierupdating AOSP probably won't do much for you.19:12
mpoirierbut your choice really.19:12
pundiramitbut we need to move to 4.0.3 for this release19:12
mpoirierhumm....19:12
mpoirierI agree but it's already done.19:12
mpoiriermake wifi work on 4.0.1 first.19:13
mpoirierthat way you don't spend time with the upgrade.19:13
mpoirierbut again, up to you.19:13
mpoirierI'm under the opinion that if it work in 4.0.1, it will work in 4.0.3.19:13
mpoirierchip power failures are caused by the kernel, not AOSP.19:14
pundiramityes it should work on 4.0.3 without much concerns19:14
mpoirierexactly.19:14
mpoirierand if it comes up to be in the 11th hours, we can all pitch in and move to 4.0.3.19:14
mpoirierthat is not a problem.19:14
mpoirieras long as we get it working in 4.0.1.19:14
pundiramiti wasted my day today on an Android WiFi feature (or bug?) which doesn't let me connect to APs which are not connected to Internet19:15
pundiramitso I will try on 3.1 tomorrow19:15
mpoirierya.19:15
pundiramitand will keep you posted19:16
mpoirierand don't worry about connection drops.19:16
mpoirieryou'll need an antenna.19:16
mpoirierit was the same on my side when I first started working on it.19:16
pundiramitmpoirier, thats good to know19:16
mpoirierif you want me to test connectivity push the images somewhere and I'll test them.19:17
pundiramitwhat about wifi connection settings window/activity? i dont see it either on snowball19:17
mpoirierno clue what that is about - sorry.19:17
mpoirierask master Vishal.19:17
pundiramitmpoirier, everything is a mess right now.. I will send you the images soon19:17
mpoirierof course it's a mess... it's enablement.19:18
pundiramitVishal told me he can see the settings on clicking clock19:18
pundiramitI dont see it there19:18
pundiramitit tries to open a window but fails19:18
mpoirierdo you get gles errors in the logs ?19:18
pundiramitmpoirier, who is the graphics poc for ST? yes I have the logs19:19
mpoiriergood question - we don't have one.19:19
pundiramitmpoirier, should I mail it to you19:19
pundiramitand you fwd it to concern people/department19:19
mpoirierok, shoot it my way.19:19
pundiramitcool will cc patrik too19:20
mpoirierI won't be able to do anything about it until you check stuff in.19:20
mpoirierthat will be a bug.19:20
mpoirierif the feature work and we get wifi than it a simple bug.19:20
mpoirierstabilize things a little and check in.19:20
pundiramityes thats the plan19:21
mpoirierwe'll file bugs.  much easier to get help that way.19:21
pundiramitok19:21
mpoirierotherwise STE friends can't help us.19:21
mpoirierbtw, I'm still waiting for those famous SHAs.19:22
mpoirierbut they shouldn't be required if we use 3.119:22
mpoirierpundiramit: anything else we should talk about ?19:23
pundiramitmpoirier, ohh didn't I tell you that all those patches are merged in 3.0 stable.. my bad19:23
pundiramitmpoirier, no we are good19:23
mpoiriergood and good.19:24
mpoirierI'm going to step out for 30 minutes.19:24
pundiramitsure.. i will move to bed too19:24
pundiramitsee you tomorrow19:24
mpoirieryes, you should.19:24
mpoiriersee you tomorrow and thanks for your time.19:24
pundiramitthank you too :)19:24
* mpoirier will be back in 30 minutes.19:25
tgall_fooso what profiling tool should one use on panda android ?19:47
tgall_fooanyone?  anyone ?20:37
doanactgall_foo: I think bero|2 might be best to ask. I haven't done any profiling on android before.20:56
tgall_foook thanks doanac20:56
doanacpfefferz: android-build looks to be out of disk space: https://android-build.linaro.org/jenkins/job/doanac_panda-with-google4.4/3/console21:45
doanacis there someone I should notify to get it fixed?21:45
pfefferzcool21:47
pfefferztalk to danilos and pfalcon21:47
pfefferzand asac21:47
doanack. thanks.21:47
pfefferzbero|2, do you think you'll be able to enable strict aliasing and -O3 in ICS builds by 12.01?22:23
Dr_Whopfefferz: FYI I just pushed my ICS updates into git for libjpeg-turbo22:38
pfefferzDr_Who, cool22:39
Dr_Whopfefferz: I'm going to see what I can do to get the skia_bench numbers up22:39
Dr_Whothat's my goal for the rest of 12.0122:39
pfefferzcool22:39
pfefferzdo you have a BP?22:39
Dr_WhoI will … this is a snap decision I made about 30 minutes ago22:40
Dr_Whoif skia-bench isn't better than stock I don't think those in the plea would be impressed22:40
Dr_Whos/plea/plex22:40
Dr_Whoso time to do something about that22:40
pfefferzright22:41
pfefferzwould you assign me to the BP?22:41
Dr_Whoyes I will22:41
pfefferzthanks22:43
pfefferzDr_Who, you'd talked to doanac about his benchmark right: https://wiki.linaro.org/Platform/Android/AndroidToolchainBenchmarking22:44
Dr_Whono I hadn't actually22:44
pfefferztake a look at his work22:46
pfefferzdoanac runs the benchmarking stuff22:46
pfefferzand has recorded our benchmarks over the last 7 cycles22:47
Dr_Whocool22:47
pfefferzdoanac, I put together https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-benchmark-12.01-toolchain22:49
pfefferzdoanac, would you mind reprising your awesome benchmarking role22:49
pfefferz?22:49
doanacpfefferz: just saw them. i'll not only do them, but they'll be done with red-hot LAVA22:50
pfefferzoh awesome blossom22:53
pfefferz:)22:53
pfefferzfeel free to change the WI's I spec'd22:54
pfefferzdoanac, are you working on basically: https://blueprints.launchpad.net/linaro-android/+spec/lavatize-toolchain-benchmarking23:07
doanacpfefferz: yep23:08
doanacfeel free to assign that to me as well23:09
pfefferzcool, would you rework the WIs?23:09
doanacpfefferz: yeah. they need some cleaning up23:10
doanacpfefferz: that blueprint is really superseded by https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-bmark-process23:26
doanacyou mind if I mark it superseded and then update the other blueprint accordingly?23:27
doanacpfefferz: also can you assign these two bp's to me as well:23:35
doanachttps://blueprints.launchpad.net/linaro-android/+spec/update-monthly-toolchain-benchmark-in-lava23:35
doanachttps://blueprints.launchpad.net/linaro-android/+spec/toolchain-benchmark-report23:35
pfefferzdoanac, I don't mind23:43
pfefferzhttps://blueprints.launchpad.net/linaro-android/+spec/update-monthly-toolchain-benchmark-in-lava DONE23:44
pfefferzhttps://blueprints.launchpad.net/linaro-android/+spec/toolchain-benchmark-report DONE23:45
pfefferzthanks!23:45
doanacthank you.23:45
doanacpfefferz: I'm going to move this bp: https://blueprints.launchpad.net/linaro-android/+spec/toolchain-benchmark-report23:46
doanacout to 12.0223:46
doanacI'll have reports generated this month, but they won't live on validation.l.o as indicated by that bp23:47
doanacthe code I'm writing is in python though, so hopefully we'll be able to get it running on the validation server in the future without too much sweat.23:47

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