Re: ✗ GitLab.Pipeline: warning for intel: igt_draw and intel_bufops improvements

["Ville Syrjälä" <ville.syrjala@linux.intel.com>]

On Fri, Oct 11, 2024 at 08:37:27AM +0200, Zbigniew Kempczyński wrote:
> On Thu, Oct 10, 2024 at 10:41:13PM +0300, Ville Syrjälä wrote:
> > On Mon, Oct 07, 2024 at 03:39:24PM -0000, Patchwork wrote:
> > > == Series Details ==
> > > 
> > > Series: intel: igt_draw and intel_bufops improvements
> > > URL   : https://patchwork.freedesktop.org/series/139556/
> > > State : warning
> > > 
> > > == Summary ==
> > > 
> > > Pipeline status: FAILED.
> > > 
> > > see https://gitlab.freedesktop.org/gfx-ci/igt-ci-tags/-/pipelines/1285113 for the overview.
> > > 
> > > build:tests-debian-meson has failed (https://gitlab.freedesktop.org/gfx-ci/igt-ci-tags/-/jobs/64744864):
> > >   [1900/1907] Generating xe_tests.rst with a custom command.
> > >   [1901/1907] Generating kms_tests.rst with a custom command.
> > >   [1902/1907] Compiling C object 'runner/527aa9f@@runner_test@exe/runner_tests.c.o'.
> > >   ninja: build stopped: subcommand failed.
> > >   ninja: Entering directory `build'
> > >   [1/848] Generating version.h with a custom command.
> > >   [2/8] Linking target runner/runner_test.
> > >   [3/8] Linking target assembler/intel-gen4asm.
> > >   [4/8] Generating intel-ci-tests with a custom command.
> > >   FAILED: docs/testplan/intel-ci-tests 
> > >   /builds/gfx-ci/igt-ci-tags/scripts/igt_doc.py --config /builds/gfx-ci/igt-ci-tags/tests/intel/i915_test_config.json /builds/gfx-ci/igt-ci-tags/tests/intel/kms_test_config.json /builds/gfx-ci/igt-ci-tags/tests/intel/xe_test_config.json --intelci-testlist docs/testplan/intel-ci-tests
> > >   /builds/gfx-ci/igt-ci-tags/tests/intel/gem_draw.c:10: Error: unrecognized line. Need to add field at /builds/gfx-ci/igt-ci-tags/tests/intel/i915_test_config.json?
> > 
> > Do I need to add something? Who the heck knows?
> > Can the person who added all this crap provide
> > the patch to fix it?
> > 
> > We've reached a point where making any changes to igt means
> > spending 10% of time on actual work, and the remaining 90%
> > is spent dealing with the idiotic "documentation" stuff.
> 
> There're people who are using this so my personal opinion about
> this doesn't matter.

Who are those people? They need to step up and fix this mess.

At the *very least* the inane arg[1] stuff needs to be nuked
from orbit and instead the test names must be made to support
real wildcards of some sort (ie. without having to painstakingly
list each and every value for every damn wildcard).

Hmm. Seems if I stick a '*' into a test name then the
documentation build scripts just seem to ignore it or
something. I'll probably just start using that even if it
doesn't do what some people want. They can come in and
fix their crazy scripts themselves.

-- 
Ville Syrjälä
Intel