== Progress ==
* upload the schedule for ll updates in 2014.06, did the updates to llct and ll trees (CARD-788, 50%)
* bug 1282076 - tried to narrow down the issue, but failed to reproduce it even using the older, "known to fail" Image.
Maybe the kernel update on the host machine is the cause (bug 1282076, 10%)
* time off (40%)
== Plans ==
* time off 60%
* usual updates to ll / llct (CARD-788)
* try the recently posted patch set for the "Add ILP32 support for ARM64 to Linaro engineering builds" card (CARD-1161)
* update the wiki doc for DEVPLAT-204
== Issues ==
* None
== Progress ==
* Finally got QT4 built and uploaded (DEVPLAT-217: 20%)
* Gettext bug filed, maintainer chivied. Now uploaded (DEVPLAT-217: 10%)
* 'Nothing to build on buildds'. Worked on better ways to find out what blockers are (DEVPLAT-217: 15%)
* R built, dependencies building. repo passed 6000 packages. (5%)
* Buildd jammed up. PDU not working. Got that sorted. (DEVPLAT-221: 10%)
* Feedback on Chen's arm64 fixes until good. Built and uploaded ~20 packages (DEVPLAT-221: 15%)
* Build, Tag 0.1 release, upload cross-binutils package to debian (DEVPLAT-220: 15%)
* Got access to new arm64 platform for testing. (DEVPLAT-219: 5%)
* More broken timesheeting faff (2%)
* Provided some internal crossbuilding advice (5%)
* Chivied people on bootstrap sprint org (3%)
== Plans ==
* Build poppler. (DEVPLAT-221)
* Build webkit (DEVPLAT-217)
* Build kdelibs (DEVPLAT-217)
== Issues ==
* Too many packages are buildable but not rebuildable.
* Having to build NMUs twice in slightly different formats for
debian-ports unreleased and debian proper bug/upload is a pain.
Wookey
--
Principal hats: Linaro, Emdebian, Wookware, Balloonboard, ARM
http://wookware.org/
== Progress ==
* Added libatomic to Android toolchain (DEVPLAT-101(?), 30%)
* Revamped Android toolchain build to use Bionic from master branch (needed
for v8 support), adapt to changes there (DEVPLAT-101(?), 30%)
* Cleaned up Android toolchain build scripts while at it (DEVPLAT-101, 20%)
* Hunted down fixes for 2 gcc 4.9 bugs that are showstoppers for Android
(DEVPLAT-101(?), 10%)
* Helped with "Android Toolchain: Member Expectations" document
(DEVPLAT-101(?), 10%)
== Plans ==
* Toolchain binary release (DEVPLAT-101)
* Android toolchain release (DEVPLAT-101(?))
* Update NDK build scripts to match Android toolchain improvements
(DEVPLAT-101(?))
* Investigate any compiler bugs that (might) show up with the new Android
toolchain
* Finish porting AArch64 Bionic optimizations to ARMv7 (DEVPLAT-104)
== Issues ==
* Not sure the scope of DEVPLAT-101 is wide enough to include all work done
last week and upcoming work (esp. NDK), may want to create a few related
cards.
Patch to fix the below two issues -
1. Fix the part where parent waits to fetch child status. In case
SIGALRM is received, fetch new status after killing child so that
subsequent check for termination status passes.
2. In case of normal termination, just check if program exited
normally (checking exit status is not necessary)
diff --git a/idlestat.c b/idlestat.c
index 787d7b0..179e7a7 100644
--- a/idlestat.c
+++ b/idlestat.c
@@ -1196,12 +1196,12 @@ static int execute(int argc, char *argv[],
char *const envp[],
alarm(options->duration);
again:
if (waitpid(pid, &status, 0) < 0) {
- if (errno != EINTR || !sigalrm)
- goto again;
- kill(pid, SIGTERM);
+ if (errno == EINTR && sigalrm)
+ kill(pid, SIGTERM);
+ goto again;
}
- if (WIFEXITED(status) && !WEXITSTATUS(status)) {
+ if (WIFEXITED(status)) {
/*
* Cancel the timer in case the program
* finished before the timeout
--
Thanks,
-Meraj