The Following User Says Thank You to AapoRantalainen For This Useful Post: | ||
![]() |
2013-08-15
, 17:12
|
Posts: 3,074 |
Thanked: 12,964 times |
Joined on Mar 2010
@ Sofia,Bulgaria
|
#32
|
I have tried to put this fixed version to the extras-devel, but it silently rejects it. Can you fetch some logs for me?
I should use bigger version number for fixed version (and not same than non-fixed version).
The Following User Says Thank You to freemangordon For This Useful Post: | ||
![]() |
2013-08-16
, 07:28
|
Posts: 839 |
Thanked: 3,386 times |
Joined on Mar 2009
|
#33
|
profile-opt: @echo "Building with support for profile generation:" $(MAKE) clean $(MAKE) build_all_generate_profile @echo "Running benchmark to generate profile data:" $(MAKE) profile-removal $(MAKE) run_profile_task @echo "Rebuilding with profile guided optimizations:" $(MAKE) clean $(MAKE) build_all_use_profile
The Following User Says Thank You to AapoRantalainen For This Useful Post: | ||
![]() |
2013-08-16
, 08:56
|
Posts: 3,074 |
Thanked: 12,964 times |
Joined on Mar 2010
@ Sofia,Bulgaria
|
#34
|
Now I can see builder logs (you did something?)
armel build ok: 04-Aug-2013 21:17
i386 failed: 15-Aug-2013 18:07
(Reading from logs) During tests on arm qemu seg faults and then skips tests and success with packages.
On i386 it runs tests and fails with "Name or service not known" during test_docxmlrpc (SimpleXMLRPCServer.py+SocketServer.py) which looks something that builder might block for reason.
And actually it is not tests, but run_profile_task on
And debian/rules is explicitly asking profiled build (so it can easily disabled).Code:profile-opt: @echo "Building with support for profile generation:" $(MAKE) clean $(MAKE) build_all_generate_profile @echo "Running benchmark to generate profile data:" $(MAKE) profile-removal $(MAKE) run_profile_task @echo "Rebuilding with profile guided optimizations:" $(MAKE) clean $(MAKE) build_all_use_profile
The Following 2 Users Say Thank You to freemangordon For This Useful Post: | ||
![]() |
2013-08-16
, 10:31
|
Posts: 3,074 |
Thanked: 12,964 times |
Joined on Mar 2010
@ Sofia,Bulgaria
|
#35
|
Exception in thread Thread-268: Traceback (most recent call last): File "/home/builder1/maemo-fremantle-i386-extras-devel/work/python2.7-2.7.3~rc2-2.1/Lib/threading.py", line 551, in __bootstrap_inner self.run() File "/home/builder1/maemo-fremantle-i386-extras-devel/work/python2.7-2.7.3~rc2-2.1/Lib/logging/config.py", line 883, in run ready=self.ready) File "/home/builder1/maemo-fremantle-i386-extras-devel/work/python2.7-2.7.3~rc2-2.1/Lib/logging/config.py", line 851, in __init__ ThreadingTCPServer.__init__(self, (host, port), handler) File "/home/builder1/maemo-fremantle-i386-extras-devel/work/python2.7-2.7.3~rc2-2.1/Lib/SocketServer.py", line 408, in __init__ self.server_bind() File "/home/builder1/maemo-fremantle-i386-extras-devel/work/python2.7-2.7.3~rc2-2.1/Lib/SocketServer.py", line 419, in server_bind self.socket.bind(self.server_address) File "/home/builder1/maemo-fremantle-i386-extras-devel/work/python2.7-2.7.3~rc2-2.1/Lib/socket.py", line 224, in meth return getattr(self._sock,name)(*args) gaierror: [Errno -2] Name or service not known make[2]: *** [run_profile_task] Terminated
![]() |
2013-08-16
, 11:05
|
Posts: 839 |
Thanked: 3,386 times |
Joined on Mar 2009
|
#36
|
@aapo - Definitely it is not autobuilder to blame, for some reason a python process got stuck.
test_exceptions qemu: uncaught target signal 11 (Segmentation fault) - core dumped make[2]: [run_profile_task] Error 126 (ignored)
The Following User Says Thank You to AapoRantalainen For This Useful Post: | ||
![]() |
2013-08-16
, 11:14
|
Posts: 3,074 |
Thanked: 12,964 times |
Joined on Mar 2010
@ Sofia,Bulgaria
|
#37
|
Built success. Waiting package coming to the repository for testing.
How about this, happens on builder for ARM, but not my local SB.
Code:test_exceptions qemu: uncaught target signal 11 (Segmentation fault) - core dumped make[2]: [run_profile_task] Error 126 (ignored)
I have tried to do that, but this time that profiled build was slipped out of my radar.
ii scratchbox-core:i386 1.0.18 Scratchbox base system ii scratchbox-devkit-apt-https:i386 1.0.11 APT HTTPS devkit for Scratchbox ii scratchbox-devkit-debian-squeeze:i386 1.0.4 Debian Squeeze devkit for Scratchbox ii scratchbox-devkit-doctools:i386 1.0.14 Doctools for Scratchbox ii scratchbox-devkit-git:i386 1.0.2 Git for Scratchbox ii scratchbox-devkit-perl:i386 1.0.5 Perl modules for Scratchbox ii scratchbox-devkit-qemu:i386 0.12.50-0sb3 Qemu scratchbox devkit ii scratchbox-libs:i386 1.0.18 Scratchbox libraries ii scratchbox-toolchain-cs2005q3.2-glibc2.5-arm:i386 1.0.7.2 cs2005q3.2-glibc2.5-arm compiler for Scratchbox ii scratchbox-toolchain-cs2005q3.2-glibc2.5-i386:i386 1.0.7 cs2005q3.2-glibc2.5-i386 compiler for Scratchbox ii scratchbox-toolchain-cs2007q3-glibc2.5-arm7:i386 1.0.16-2 cs2007q3-glibc2.5-arm7 compiler for Scratchbox ii scratchbox-toolchain-cs2007q3-glibc2.5-i486:i386 1.0.16-1 cs2007q3-glibc2.5-i486 compiler for Scratchbox ii scratchbox-toolchain-host-gcc:i386 1.0.18 Scratchbox host-gcc toolchain
The Following User Says Thank You to freemangordon For This Useful Post: | ||
![]() |
2013-08-16
, 19:06
|
Posts: 839 |
Thanked: 3,386 times |
Joined on Mar 2009
|
#38
|
The Following User Says Thank You to AapoRantalainen For This Useful Post: | ||
![]() |
2013-08-18
, 19:45
|
Posts: 3,074 |
Thanked: 12,964 times |
Joined on Mar 2010
@ Sofia,Bulgaria
|
#39
|
Builder got -maemo5 built (arm+i386) , but there were also stucked -maemo6, so it didn't push it to the repository.
I bumped version number and reuploaded, but seems i386 version is again exhausted, armel deps are already in results: https://garage.maemo.org/builder/fre...c2-2.1-maemo6/
The Following User Says Thank You to freemangordon For This Useful Post: | ||
![]() |
2013-08-19
, 06:26
|
Posts: 839 |
Thanked: 3,386 times |
Joined on Mar 2009
|
#40
|
I guess next time you should disable regression tests in the build scripts
There seems to be another problem with python2.7, similar to the KP problem - newest version does not appear in the package interface as the latest version. I'll look at that problem as soon as I have some spare time.
I should use bigger version number for fixed version (and not same than non-fixed version).