Bug 9319

Summary: bitbake hangs on early ^C
Product: [Build System, Metadata & Runtime] BitBake Reporter: Gary Thomas <gary>
Component: bitbakeAssignee: Markus Lehtonen <markus.lehtonen>
Status: VERIFIED FIXED QA Contact: Alexandru Roman <alexandru.costinx.roman>
Severity: normal    
Priority: Medium+ CC: alexandru.costinx.roman, otavio, poky.bs.watcher, poky.watcher
Version: unspecified   
Target Milestone: 2.1 M4   
Hardware: All   
OS: Multiple   
Whiteboard:
OS type for building Yocto: --- Type of Regression: ---
Verified: Documentation change: No (bug/feature does not impact docs)

Description Gary Thomas 2016-03-22 05:09:38 UTC
If you try to abort bitbake (with ^C) very early on, it will hang and external forces are required to kill it.

>> $ bitbake some-image
>> NOTE: Started PRServer with DBfile:
>> /local/p7619_2016-02-23/cache/prserv.sqlite3, IP: 127.0.0.1, PORT: 43616,
>> PID: 14386 Parsing recipes:  11% |###########
>>
>>           | ETA:  00:00:21
>>
>> Keyboard Interrupt, closing down...
>>
>> WARNING: Timeout while attempting to communicate with bitbake server
>> WARNING: Timeout while attempting to communicate with bitbake server
>> WARNING: Timeout while attempting to communicate with bitbake server
>> WARNING: Timeout while attempting to communicate with bitbake server
>> ERROR: Gave up; Too many tries: timeout while attempting to communicate with
>> bitbake server
>>
>> Tested using poky:8debfea81e69d038bd2d56314b272cb74f5582ed

This seems to only happen when there is no [initial] cache and may be very ephemeral as I was unable to make it happen on demand.
Comment 1 Markus Lehtonen 2016-03-30 12:56:39 UTC
This was a hard one. Especially because it was hard for me to replicate it. Anyway, I think I fixed it:
http://patchwork.openembedded.org/patch/119029/
Comment 3 Alexandru Roman 2016-04-18 10:20:48 UTC
Verified on master.