Bug 8155 - defaultpkgname completion_function failed: "usr/bin/gdk-pixbuf-query-loaders: No such file or directory"
Summary: defaultpkgname completion_function failed: "usr/bin/gdk-pixbuf-query-loaders:...
Status: VERIFIED INVALID
Alias: None
Product: OE-Core
Classification: Build System, Metadata & Runtime
Component: core (show other bugs)
Version: unspecified
Hardware: x86 Multiple
: Medium+ normal
Target Milestone: 1.9 M3
Assignee: Jackie Huang
QA Contact: Alexandru Georgescu
URL:
Whiteboard:
Depends on:
Blocks:
 
Reported: 2015-08-13 09:00 UTC by Jackie Huang
Modified: 2015-08-28 13:01 UTC (History)
6 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: ---
Verified:
Documentation change: Don't know


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Jackie Huang 2015-08-13 09:00:52 UTC
It's rare case that I don't know it can be reproduced yet, it happened 3 times these two days in our handreds of builds:

Seems like a race condition issue.

Here is the error logs:
=======================
[2015-08-12 11:50:40.990666860-04:00] /home/wrlbuild/builds/builds-2015-08-12-114224-g7q0/qemux86-64_world_oe_bd/bitbake_build/tmp/work/core2-64-wrs-linux/defaultpkgname/1.0-r0/temp/completion_function/run.completion_function.17504: line 95: /home/wrlbuild/builds/builds-2015-08-12-114224-g7q0/qemux86-64_world_oe_bd/bitbake_build/tmp/sysroots/x86_64-linux/usr/bin/gdk-pixbuf-query-loaders: No such file or directory
[2015-08-12 11:50:40.991696597-04:00] WARNING: /home/wrlbuild/builds/builds-2015-08-12-114224-g7q0/qemux86-64_world_oe_bd/bitbake_build/tmp/work/core2-64-wrs-linux/defaultpkgname/1.0-r0/temp/completion_function/run.completion_function.17504:1 exit 1 from
[2015-08-12 11:50:40.993286731-04:00]   exit 1
[2015-08-12 11:50:41.054167800-04:00] ERROR: Execution of event handler 'base_eventhandler' failed
[2015-08-12 11:50:41.055101622-04:00] Traceback (most recent call last):
[2015-08-12 11:50:41.055882324-04:00]   File "base_eventhandler(e)", line 48, in base_eventhandler(e=<bb.runqueue.sceneQueueComplete object at 0x7ba9fd0>)
[2015-08-12 11:50:41.056865819-04:00]   File "/home/wrlbuild/builds/builds-2015-08-12-114224-g7q0/qemux86-64_world_oe_bd/bitbake/lib/bb/build.py", line 229, in exec_func(func='completion_function', d=<bb.data_smart.DataSmart object at 0x3853990>, dirs=None):
--
[2015-08-12 11:50:41.061752175-04:00]              finally:
[2015-08-12 11:50:41.062608671-04:00] FuncFailed: Function failed: completion_function
[2015-08-12 11:50:41.063398486-04:00]
[2015-08-12 11:50:41.064230917-04:00] ERROR: An uncaught exception occured in runqueue, please see the failure below:
[2015-08-12 11:50:41.226016617-04:00] ERROR: Running idle function
[2015-08-12 11:50:41.227409868-04:00] Traceback (most recent call last):
[2015-08-12 11:50:41.228246423-04:00]   File "/home/wrlbuild/builds/builds-2015-08-12-114224-g7q0/qemux86-64_world_oe_bd/bitbake/lib/bb/event.py", line 108, in fire_class_handlers(event=<bb.runqueue.sceneQueueComplete object at 0x7ba9fd0>, d=<bb.data_smart.DataSmart object at 0x3853990>):
[2015-08-12 11:50:41.228886714-04:00]                          continue
Comment 1 Jackie Huang 2015-08-13 09:03:15 UTC
It's likeyly that this is intrudeced by:

commit b411085fefc8582d24c172db5f6610705eb44c2f
Author: Richard Purdie <richard.purdie@linuxfoundation.org>
Date:   Tue Jul 28 17:51:42 2015 +0100

    pixbufcache: Use sceneQueueComplete event to simplify usage
    
    Ensuring the native pixbuf cache is correct after new loaders have
    been installed is tricky. This needs to be done without races
    and work regardless of whether the build is from sstate or freshly
    built for one or more modules.
    
    This adds a hook into base.bbclass which is then triggered by the
    code from pixbufcache. This patch is an improved version which
    means base.bbclass has no pixbuf knowledge and the mechanism can
    be reused in other cases.
    
    Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Comment 2 Randy MacLeod 2015-08-13 20:45:12 UTC
This happened 4 times out of ~1000 builds on Aug 12 between 6 am and 7 pm ET after we update oe-core at 2015-08-10 11:00 PM. The builders can take as much as 12 hours to update. All the build failures occurred in our docker-base builders rather than on bare hardware so that could be a factor.

These failures stopped today so I suspect that it was a commit/network glitch or an sstate problem. Jackie will investigate.
Comment 3 Jackie Huang 2015-08-19 01:52:25 UTC
I can't reproduce this and it and there haven't been any more failures in our daily builds these days, so it might be a network glitch or somthing instead of a real bug.
Comment 4 Alexandru Georgescu 2015-08-19 12:00:13 UTC
verified as per above comments. Please reopen if it reproduces.