]> git.ipfire.org Git - thirdparty/openembedded/openembedded-core-contrib.git/commit
multilib_global: Handle PREFERRED_RPROVIDER
authorRichard Purdie <richard.purdie@linuxfoundation.org>
Sun, 10 Dec 2017 17:31:48 +0000 (17:31 +0000)
committerRichard Purdie <richard.purdie@linuxfoundation.org>
Sun, 10 Dec 2017 22:45:21 +0000 (22:45 +0000)
commit4137923598e2c1f9e4e54a3bf0f98b823fd54ff1
tree6a0469d535ebe593cab77467b86811dfc00156f6
parent2a7a02c513a55284ef5af0580f5393225d60ba85
multilib_global: Handle PREFERRED_RPROVIDER

Running:

$ oe-selftest -r sstatetests.SStateTests.test_sstate_sametune_samesigs

after commit cdcebd81c872cb7386c658998e27cf24e1d0447c results in:

NOTE: Resolving any missing task queue dependencies
NOTE: Multiple providers are available for runtime lib32-initd-functions (lib32-initscripts, lib32-lsbinitscripts)
Consider defining a PREFERRED_RPROVIDER entry to match lib32-initd-functions

and will occasionally pick a different value on the second stamps run
causing a test failure. Update the multilib code to handle
PREFERRED_RPROVIDER too.

There is a bigger worry here which is why the builds aren't deterministic. This is
caused by a bug in bitbake's providers.py and a separate fix will be sent for that
which would cause this test to always pass or always fail.

(From OE-Core rev: ced4ac760926ce43a937dad2be3b873b1beec6aa)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meta/classes/multilib_global.bbclass