linaro-android August 2012

linaro-android@lists.linaro.org

40 participants
50 discussions

Re: [BUILD] ANDROID ICS 4.0.4 build problem

by Zygmunt Krynicki

W dniu 08.08.2012 17:33, Parin Choganwala pisze: > Ok thank you very much for your help! > > i did report a bug on launchpad and trying your scripts. syncing the > repo right now. > > Actually i wanted to make few configuration changes in linux kernel > itself. Is there any way to run menuconfig during the build process? > How can i build kernel and android source separately? I think you can change the kernel config easily but I have not attempted this with android yet. I'm sure others here can help you. Best regards ZK -- Zygmunt Krynicki Linaro Validation Team s/Validation/Android/

12 years, 4 months

[BUILD] ANDROID ICS 4.0.4 build problem

by Parin Choganwala

Hi, I am trying to compile Android from source for Pandabaord from https://android-build.linaro.org/builds/~linaro-android/panda-ics-gcc47-til… when executing make command I get the following error and build stops. It looks like some simple copy issue. How to fix it? ============================================ PLATFORM_VERSION_CODENAME=REL PLATFORM_VERSION=4.0.4 TARGET_PRODUCT=full TARGET_BUILD_VARIANT=eng TARGET_BUILD_TYPE=release TARGET_BUILD_APPS= TARGET_ARCH=arm TARGET_ARCH_VARIANT=armv7-a HOST_ARCH=x86 HOST_OS=linux HOST_BUILD_TYPE=release BUILD_ID=IMM76L ============================================ mkdir -p out/target/product/generic/boot echo "Target boot fs tarball: out/target/product/generic/boot.tar.bz2" Target boot fs tarball: out/target/product/generic/boot.tar.bz2 mkdir -p out/target/product/generic/boot cp -f out/target/product/generic/boot/. cp: missing destination file operand after `out/target/product/generic/boot/.' Try `cp --help' for more information. make: *** [out/target/product/generic/boot.tar.bz2] Error 1 Please help! Thank you Parin

12 years, 4 months

Linaro 2012/8/8 Android Platform Team Agenda Posted

by Zach Pfeffer

The agenda has been posted to: https://wiki.linaro.org/Platform/Android/Meetings/2012-08-08 Please take a look and add to it if you need to. See everyone tomorrow in #linaro-meeting on irc.freenode.net at 13:00 UTC. -- Zach Pfeffer Android Platform Team Lead, Linaro Platform Teams Linaro.org | Open source software for ARM SoCs Follow Linaro: http://www.facebook.com/pages/Linaro http://twitter.com/#!/linaroorg - http://www.linaro.org/linaro-blog

12 years, 4 months

Analysis of JB tests and comparison with ICS

by Zygmunt Krynicki

Hey folks. I've ran a few builds to compare the tests that are present in JB builds as compared to its older ICS brother. Here are the results (I use 'maguro' to mean the JB build for nexus below) Tests present in both 'eng' and 'tests' variants of maguro: https://github.com/zyga/android-test-tools/blob/master/common-test-execs-ma… Additional executables present in 'tests' variant of maguro: https://github.com/zyga/android-test-tools/blob/master/extra-execs-maguro.t… Additional APKs present in 'tests' variant of maguro: https://github.com/zyga/android-test-tools/blob/master/extra-apks-maguro.txt Now having that I compared maguro to ICS panda builds with a new script i wrote (everything is in my git repo: https://github.com/zyga/android-test-tools the script is called compare-ics-and-jb.py) Delta between additional apks added by the 'tests' variant of JB: https://github.com/zyga/android-test-tools/blob/master/diff-ics-jb-apks.txt Delta between additional executables added by the 'tests' variant of JB (this delta is quite big): https://github.com/zyga/android-test-tools/blob/master/diff-ics-jb-execs.txt Delta between executables present in both 'eng' and 'tests' variant of JB: https://github.com/zyga/android-test-tools/blob/master/diff-ics-jb-common-t… So there you have it. BIG REQUEST: I would appreciate if anyone with access to JB and nexus could run the new test executables for me and send me the logs back. Thanks -- Zygmunt Krynicki Linaro Validation Team s/Validation/Android/

12 years, 4 months

Re: Using git for storing configs of all -tests builds

by Zygmunt Krynicki

W dniu 07.08.2012 10:30, Zygmunt Krynicki pisze: CCing the mailing list. > Hello > > In accordance with blueprint: eng-and-test-builds [1] > > To implement the final work item of my blueprint I'd have to manually > create (and later keep in sync) of the following configurations. > > * > https://android-build.linaro.org/builds/~linaro-android/vexpress-ics-gcc47-… > > * > https://android-build.linaro.org/builds/~linaro-android/vexpress-rtsm-ics-g… > > * > https://android-build.linaro.org/builds/~linaro-android/origen-ics-gcc47-sa… > > * > https://android-build.linaro.org/builds/~linaro-android/snowball-ics-gcc46-… > > * > https://android-build.linaro.org/builds/~linaro-android/snowball-ics-gcc47-… > > * > https://android-build.linaro.org/builds/~linaro-android/panda-ics-gcc47-til… > > * > https://android-build.linaro.org/builds/~linaro-android/panda-ics-gcc44-kwg… > > * > https://android-build.linaro.org/builds/~linaro-android/panda-ics-gcc47-oma… > > > * > https://android-build.linaro.org/builds/~linaro-android/vexpress-jb-gcc47-a… > > * > https://android-build.linaro.org/builds/~linaro-android/origen-jb-gcc47-sam… > > * > https://android-build.linaro.org/builds/~linaro-android/snowball-jb-gcc47-i… > > * > https://android-build.linaro.org/builds/~linaro-android/panda-jb-gcc47-tilt… > > * > https://android-build.linaro.org/builds/~linaro-android/panda-jb-gcc47-tilt… > > * > https://android-build.linaro.org/builds/~linaro-android/galaxynexus-jb-gcc4… > > > This screams as insane in my head so I went looking and recalled that > Infrastructure has implemented feature where we can keep the configs in > git. That would dramatically simplify everything. > > So, with that in context I'd like to propose three options and I'd like > to ask you to pick the one you'd like to see implemented now: > > 1) Ignore git configs, duplicate everything manually and later check if > it changes once a week to update -tests builds. > > 2) Create -tests builds manually but pull the configs from git. We still > need to keep them in sync manually (with the base configs) but at least > updates are easy to manage (one commit instead of a messy edit-and-save > on dozen pages) > > 3) Move all existing configs to git, have a flag day, move all builds to > git. Derive -tests builds from -eng builds. This is by far the most > 'lazy' solution (as in I'm too lazy to do all the manual work all the > time in the future) but has the most risk. > > pfefferz, asac: please decide and let me know > > pfalcon: for something like -tests builds it would be perfect if we had > a include feature, where one config can include some other (presumably > 'base') config and just override something. It's not a strict > requirement for implementing -tests builds sensibly but it would be an > advantage, is that supported by any chance? (are those shell files I > could just source with something like '. base.conf') > > Best regards > ZK > > [1] > https://blueprints.launchpad.net/linaro-android/+spec/eng-and-test-builds > -- Zygmunt Krynicki Linaro Validation Team s/Validation/Android/

12 years, 4 months

what's the story of the AOSP tests in our builds

by Alexander Sack

Hi, AFAIK, we reverted our experiments to use =eng instead of =test ... I understand this was done to prevent ICS/JB from slowing down due to hundreds of apps getting registered and displayed in the launcher... Now, what's the plan on how to get the AOSP tests produced in a way that we can run/use them? -- Alexander Sack Technical Director, Linaro Platform Teams http://www.linaro.org | Open source software for ARM SoCs http://twitter.com/#!/linaroorg - http://www.linaro.org/linaro-blog

12 years, 4 months

memcpy() failure: bionic optimizations

by Annamalai Lakshmanan

Hi Bero, It is seen that latest bionic patches are causing Origen MM to crash, I just verified by removing the stream of bionic patches for optimization. There was a specific failure from memcpy(), this was due to 32 bit un-aligned address (Source Pointer) passed as an argument , this addess is obtained from mmap (device space) It looks like the latest optmizations are not taking this scenario into consideration and thus causing issues for specific case as mentioned above. I have not tried disabling NEON flag though. With Regards, Annamalai

12 years, 4 months

Can't reach source depot

by Chu, Max

Hi, I'm trying to build from linaro source but can't reach your server. D:\wa\java>ping android.git.linaro.org Pinging android.git.linaro.org [91.189.95.33] with 32 bytes of data: Request timed out. Request timed out. Request timed out. Request timed out. Ping statistics for 91.189.95.33: Packets: Sent = 4, Received = 0, Lost = 4 (100% loss), Please advise. Thanks, /max chu

12 years, 4 months

Snowball: in rare cases, wpa_supplicant is started before the wlan driver could create the interface wlan0

by BENIGOT Yann

Hi, While working on a chrooted Android (the host distribution being a Debian one) stored in a tmpfs (in RAM), I noticed that, sometimes, enabling WLAN fails (the button goes straight from "On" to "Off"). On the console, the module was loaded ("CW1200 WSM init done"), but logcat showed that wpa_supplicant could not find the interface: E/wpa_supplicant( 8015): Could not read interface wlan0 flags: No such device E/wpa_supplicant( 8015): WEXT: Could not set interface 'wlan0' UP E/wpa_supplicant( 8015): wlan0: Failed to initialize driver interface D/wpa_supplicant( 8015): Failed to add interface wlan0 E/WifiStateMachine( 2292): Failed to start supplicant! When I put udev (in the host system) in charge of firmware loading, this bug always happens ; when I disable udev and let uevent (in the Android chroot) do the job, it only happens one time out of fifteen I would say. The problem seems to be that libhardware_legacy does not wait for the interface to be created to state that the driver has been loaded, leading the Wifi state machine in Android to load the supplicant before the interface was created. The patch at the end of this mail (concerning libhardware_legacy) enables libhardware_legacy to wait for the interface to be created in /sys/class/net, thus solving the problem. However, since this problem only occurs in rare cases when not using Linaro Android as intended (from emmc or sd card), I do not know if this is worth having a look at. Regards, Yann Benigot PATCH: diff --git a/wifi/wifi.c b/wifi/wifi.c index 301d91b..c2661f2 100644 --- a/wifi/wifi.c +++ b/wifi/wifi.c @@ -200,6 +200,8 @@ int wifi_load_driver() #ifdef WIFI_DRIVER_MODULE_PATH char driver_status[PROPERTY_VALUE_MAX]; int count = 100; /* wait at most 20 seconds for completion */ + DIR *ifaceDir; + char ifaceDirPath[16+PROPERTY_VALUE_MAX]; if (is_wifi_driver_loaded()) { return 0; @@ -218,10 +220,21 @@ int wifi_load_driver() property_set("ctl.start", FIRMWARE_LOADER); } sched_yield(); + property_get("wifi.interface", iface, WIFI_TEST_INTERFACE); + sprintf(ifaceDirPath, "/sys/class/net/%s", iface); while (count-- > 0) { if (property_get(DRIVER_PROP_NAME, driver_status, NULL)) { if (strcmp(driver_status, "ok") == 0) - return 0; + { + ifaceDir = opendir(ifaceDirPath); + if(ifaceDir != NULL) { + closedir(ifaceDir); + return 0; + } + else { + LOGI("%s not found, waiting for wifi driver to complete initialization. Will try again %d times.", ifaceDirPath, count-1); + } + } else if (strcmp(DRIVER_PROP_NAME, "failed") == 0) { wifi_unload_driver(); return -1;

12 years, 4 months

Linaro android build for Beagleboard

by yan zhang

Hi~: I'm a fan of linaro,and I'd like to build the linaro android for beagleboard.But I can't find any successful cases based on the lasted release. So,Can you show me any ideas? Thanks~

12 years, 4 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

linaro-android August 2012