On Wed, Apr 24, 2019 at 12:02:29PM +0000, Vladimir Sementsov-Ogievskiy wrote: > 24.04.2019 13:59, Thomas Huth wrote: > > On 24/04/2019 12.48, Daniel P. Berrangé wrote: > >> On Wed, Apr 24, 2019 at 12:37:46PM +0200, Thomas Huth wrote: > >>> People often forget to run the iotests before submitting patches or > >>> pull requests - this is likely due to the fact that we do not run the > >>> tests during our mandatory "make check" tests yet. Now that we've got > >>> a new "ci" group of iotests that should be fine to run in every environ- > >>> ment, it should be OK to enable the iotests during "make check" again. > >>> Thus we now run the "ci" tests by default from the qemu-iotests-quick.sh > >>> script, and only use the former "quick" group (that contains some tests > >>> that are failing in some environments) when the user decided to run > >>> "make check SPEED=thorough" or something similar. > >>> > >>> Signed-off-by: Thomas Huth <th...@redhat.com> > >>> --- > >>> tests/Makefile.include | 2 +- > >>> tests/qemu-iotests-quick.sh | 17 ++++++++++++++++- > >>> 2 files changed, 17 insertions(+), 2 deletions(-) > >> > >>> diff --git a/tests/qemu-iotests-quick.sh b/tests/qemu-iotests-quick.sh > >>> index 0e554bb972..416b3fc48b 100755 > >>> --- a/tests/qemu-iotests-quick.sh > >>> +++ b/tests/qemu-iotests-quick.sh > >> > >> s/-quick// perhaps, since it no longer only runs the "quick" tests. > >> > >>> @@ -1,8 +1,23 @@ > >>> #!/bin/sh > >>> > >>> +# Honor the SPEED environment variable, just like we do it for the > >>> qtests. > >>> +# The default is to run all tests that still work fine in a CI > >>> environments, > >>> +# but if the user set SPEED=slow or SPEED=thorough, we also run all other > >>> +# tests that are still marked as "quick" > >>> +if [ "$SPEED" = "slow" -o "$SPEED" = "thorough" ]; then > >>> + group=quick > >> > >> When user asks for "slow" tests we run the "quick" tests, which > >> presumably implies... > >> > >>> +else > >>> + group=ci > >> > >> ..."ci" is quicker than the "quick" tests ! > >> > >> Confused ? We certainly try to ensure that you will be :-) > > > > ... and there is also tests/check-block.sh (which is not used by "make > > check-block", by the way, but still talks about "make check-block"), > > which is even slower than the "ci" and "quick" group... > > > > I guess that could be cleaned up, too (why do we need two scripts?) ... > > just not sure how the final result should look like and if it should be > > part of this patch series or rather something separate. Max, Kevin, what > > do you think? > > > > Maybe we should merge the two scripts, and depending on the SPEED > > variable, we run: > > > > quick => "ci" group > > thorough => "quick" group > > slow => old behaviour of the check-block.sh script > > May be, instead of SPEED, introduce GROUP variable? To make it plain
"$SPEED" is something used/defined by the gtester harness, so other bits run by "make check" already honour that, hence it is desirable fro the iotests todo the same. Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
