[PATCH i-g-t 2/5] tests/fbdev: Add testplan documentation

[Kamil Konieczny]

Hi Bhanuprakash,
On 2024-05-29 at 17:04:07 +0530, Modem, Bhanuprakash wrote:
> Hi Kamil,
> 
> On 28-05-2024 09:09 pm, Kamil Konieczny wrote:
> > Hi Bhanuprakash,
> > On 2024-05-28 at 11:12:20 +0530, Bhanuprakash Modem wrote:
> > > Add testplan documentation to fbdev tests.
> > > 
> > > Cc: Kamil Konieczny <kamil.konieczny at linux.intel.com>
> > > Signed-off-by: Bhanuprakash Modem <bhanuprakash.modem at intel.com>
> > > ---
> > >   tests/fbdev.c | 39 +++++++++++++++++++++++++++++++++++++++
> > >   1 file changed, 39 insertions(+)
> > > 
> > > diff --git a/tests/fbdev.c b/tests/fbdev.c
> > > index 794882a6a..f329e6ac6 100644
> > > --- a/tests/fbdev.c
> > > +++ b/tests/fbdev.c
> > > @@ -21,6 +21,15 @@
> > >    * IN THE SOFTWARE.
> > >    */
> > > +/**
> > > + * TEST: fbdev
> > > + * Category: Display
> > > + * Description: Sanity test for framebuffer device.
> > > + * Driver requirement: i915, xe
> > 
> > I am not sure we need it? I mean it can run on all gpus,
> > not only Intel?
> 
> "Driver requirement" is Intel thing to differentiate between i915 & xe
> specific tests (As few kms tests contains both i915 & xe specific subtests
> in same test file).
> 

Ok.

> > 
> > > + * Functionality: fbdev
> > > + * Mega feature: General Display Features
> > > + * Test category: functionality test
> > 
> > Please sort these fields after other tests, see recent Kasia
> > patchseries or look into other kms tests.
> 
> These fields are already sorted alphabetically. We are using the same style
> for entire KMS tests.
> 

Ok.

> > 
> > Could you also fix this test? After merging locally your
> > patchset I got errors from 'meson test -C build':
> 
> This issue is not seen in my local/CI:
> https://gitlab.freedesktop.org/gfx-ci/igt-ci-tags/-/jobs/59210298
> 
> Ok:                 423
> Expected Fail:      4
> Fail:               0
> Unexpected Pass:    0
> Skipped:            0
> Timeout:            0
> 

Hmm, something strange happens on my system, now I see timouts.

> > 
> > It turns out test uses in igt_main:
> > 
> > igt_subtest_group {
> >      mode_tests(fd);
> > }
> > 
> > and then defines subtests in functions. That is causing test failing.
> 
> I guess, that kind of implementation is legal in IGT.
> 
> - Bhanu
> 

Yes, sorry for wrong alarm.

Patch itself looks ok so:
Reviewed-by: Kamil Konieczny <kamil.konieczny at linux.intel.com>

Regards,
Kamil

> > 
> > Regards,
> > Kamil
> > 
> > 
> > > + */
> > >   #include "config.h"
> > >   #include "igt.h"
> > > @@ -39,6 +48,36 @@
> > >   #include "igt.h"
> > > +/**
> > > + * SUBTEST: eof
> > > + * Description: Check framebuffer access near EOF
> > > + *
> > > + * SUBTEST: info
> > > + * Description: Check if screeninfo is valid
> > > + *
> > > + * SUBTEST: nullptr
> > > + * Description: Check framebuffer access with NULL
> > > + *
> > > + * SUBTEST: pan
> > > + * Description: Check panning / page flipping
> > > + *
> > > + * SUBTEST: read
> > > + * Description: Check read operations on framebuffer memory
> > > + *
> > > + * SUBTEST: write
> > > + * Description: Check write operations on framebuffer memory
> > > + */
> > > +
> > > +/**
> > > + * SUBTEST: unaligned-%s
> > > + * Description: Check %arg[1] operations on unaligned locations in framebuffer memory
> > > + *
> > > + * arg[1]:
> > > + *
> > > + * @read:	read
> > > + * @write:	write
> > > + */
> > > +
> > >   #define PANSTEP(panstep_) \
> > >   	((panstep_) ? (panstep_) : 1)
> > > -- 
> > > 2.43.2
> > >