[OE-core] [PATCH 0/8] Various QA fixes

Richard Purdie richard.purdie at linuxfoundation.org
Tue Jul 5 10:25:35 UTC 2011


On Mon, 2011-07-04 at 17:46 +0100, Richard Purdie wrote:
> The following changes since commit 2b3bf5350861f62435e2fdf1c56c8a02f4b1b4ac fix
> a number of QA warnings/errors. There are a couple of RFC style commits in this mix:
> 
> A key change is that functionality is added to insane.bbclass to allow 
> skipping of individual QA tests by name. It needs two existing users (elfutils and u-boot)
> to specify which QA tests they want to skip (I'm working on a patch). Any external 
> layer using that variable would need to update and I can't decide if that is a 
> drawback or a feature.
> 
> Also a gettext change I'm testing to see which approach performs best
> is included. A final decision on that will depend on the performance
> test results (thanks go to paul for hightlighting the impact of that
> git-native depenedency).
> 
> Also included is a gcc libiberty fix (yocto #1199).
> 
> They are available in the git repository at:
>   git://git.openembedded.org/openembedded-core-contrib rpurdie/master
>   http://cgit.openembedded.org/cgit.cgi/openembedded-core-contrib/log/?h=rpurdie/master
> 
> Richard  Purdie (5):
>   gcc: Fix removal of libiberty.a
>   gettext: Disable both git and cvs for autopoint's archive format.
>   gcc: Remove unneeded module .la file and .so link
>   insane.bbclass: Allow INSANE_SKIP to work on a per test basis
>   lttng-viewer: Fixup various QA warnings and a false positive
> 
> Richard Purdie (3):
>   oprofile: Fix QA warnings
>   libgsmd: Fix QA warnings
>   gcc-package-cross: Switch to using pattern matching to detect when to
>     stash libgcc into the sysroot

I've merged these to master. Whilst not a 10% build time improvement on
my test machine, the gettext change did show a clear 2% gain.

I also added the INSANE_SKIP tweaks for db, qemu, u-boot and elfutils so
the majority of QA warnings should be squashed apart from some issues
with ldflags which I'll discuss in another email...

Cheers,

Richard





More information about the Openembedded-core mailing list