[OE-core] Broken "bitbake -S printdiff" in current OE

[Ulf Magnusson]

On Tue, Aug 23, 2016 at 8:18 AM, Mike Looijmans <mike.looijmans at topic.nl>
wrote:

> Was attempting to figure out why something wasn't being grabbed from the
> sstate-cache, but running "bitbake -S printdiff qt4-embedded" only yields
> tons
> of lines like this:
>
> WARNING: /home/mike/.../bitbake/lib/bb/runqueue.py:1347: ResourceWarning:
> unclosed file <_io.BufferedReader
> name='/home/mike/.../build/tmp-glibc/stamps/cortexa9hf-
> neon-oe-linux-gnueabi/tiff/4.0.6-r0.do_populate_sysroot.sigdata.
> 868460a5d532ef809db9bfa7ce2a89e4'>
>   out2 = bb.siggen.compare_sigfiles(hashfiles[hash1], hashfiles[hash2],
> recursecb)
>
> Left it running for about fifteen minutes, but nothing useful came out.
>

Not a fix for the original issue, but for what it's worth, I just added
more comprehensive documentation on how to view task signatures here:

http://www.yoctoproject.org/docs/2.2/ref-manual/ref-manual.html#usingpoky-viewing-task-variable-dependencies

Running bitbake-diffsigs instead of 'bitbake -S diffsigs' would be an
alternative.

I also had trouble getting 'bitbake -S diffsigs' and 'bitbake -S none'
to work as advertised. They write a file locked-sigs.inc too, which
seems to be undocumented. I relegated the bitbake -S documentation to
the end of the new section and kept it as it was before.

Cheers,
Ulf
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openembedded.org/pipermail/openembedded-core/attachments/20160823/6ebc6a28/attachment-0002.html>