[igt-dev] [PATCH i-g-t 2/2] tests/intel-ci/xe-fast-feedback: add create-invalid-mbz

Kamil Konieczny kamil.konieczny at linux.intel.com
Thu Nov 30 17:07:28 UTC 2023


Hi Matthew,
On 2023-11-30 at 15:00:32 +0000, Matthew Auld wrote:
> On 30/11/2023 13:55, Kamil Konieczny wrote:
> > Hi Matthew,
> > On 2023-11-30 at 11:17:35 +0000, Matthew Auld wrote:
> > > Ensure we have some coverage in BAT for gem_create MBZ fields.
> > > 
> > > Signed-off-by: Matthew Auld <matthew.auld at intel.com>
> > > Cc: Pallavi Mishra <pallavi.mishra at intel.com>
> > > Cc: Francois Dugast <francois.dugast at intel.com>
> > > ---
> > >   tests/intel-ci/xe-fast-feedback.testlist | 1 +
> > >   1 file changed, 1 insertion(+)
> > > 
> > > diff --git a/tests/intel-ci/xe-fast-feedback.testlist b/tests/intel-ci/xe-fast-feedback.testlist
> > > index 5c53e249d..a83c17de7 100644
> > > --- a/tests/intel-ci/xe-fast-feedback.testlist
> > > +++ b/tests/intel-ci/xe-fast-feedback.testlist
> > > @@ -2,6 +2,7 @@
> > >   igt at xe_module_load@load
> > >   igt at xe_compute@compute-square
> > > +igt at xe_create@create-invalid-mbz
> > 
> > Please move it before @create-massive-size,
> > it should be sorted alphabetically.
> 
> Ok, I didn't realise that was a requirement. I can move it.
> 
> Note that there are a bunch of tests in this list not following any kind of
> alphabetical sorting, but is more ordered/grouped/structured by
> functionality, which I think makes more sense IMO. Like here
> create-invalid-mbz is just a really simple sanity check for gem_create so
> running it first made sense to me. A better example is maybe xe_vm at shared-*,
> since for that it makes sense to run the pte subtest first, with every next
> test being the next level up in the paging structures:
> 
> igt at xe_vm@shared-pte-page
> igt at xe_vm@shared-pde-page
> igt at xe_vm@shared-pde2-page
> igt at xe_vm@shared-pde3-page
> 
> So here it rather ordered by functionally. Anyway, I think if we intend to
> sort this alphabetically we should put a comment at the top of the file or
> something, and then also send a patch to sort everything?
> 

imho when we put something out-of-order we should write comment, like:

# A comment why this should run before other tests

or

# A comment why this should run after other tests

Then with such comment you can have non-alphabetical order.

Regards,
Kamil

> > 
> > Regards,
> > Kamil
> > 
> > >   igt at xe_create@create-execqueues-noleak
> > >   igt at xe_create@create-execqueues-leak
> > >   igt at xe_create@create-massive-size
> > > -- 
> > > 2.43.0
> > > 


More information about the igt-dev mailing list