[igt-dev] [PATCH i-g-t 6/6] tests/i915/gem_exec_reloc: Don't attempt active relocations with the command parser

Daniel Vetter daniel at ffwll.ch
Tue Jul 13 09:24:40 UTC 2021


On Mon, Jul 12, 2021 at 6:02 PM Jason Ekstrand <jason at jlekstrand.net> wrote:
> On Mon, Jul 12, 2021 at 10:00 AM Daniel Vetter <daniel at ffwll.ch> wrote:
> > On Sat, Jul 10, 2021 at 10:52:04PM -0500, Jason Ekstrand wrote:
> > > The command parser stalls waiting for the batch to be idle.  If it's
> > > blocked by a spinner, execbuf will timeout.
> > >
> > > Signed-off-by: Jason Ekstrand <jason at jlekstrand.net>
> >
> > I think we should just ditch these tests as invalid. Can you pick the
> > right set of reverts from my series and just push them?
> >
> > https://patchwork.freedesktop.org/series/91160/
>
> Yeah, we can do that too.  Kind-of invalidates the need for the first
> 4.  I went this way because I was trying to be minimalist and wasn't
> sure what happened with your reloc series.

Just me being burried happened, so I'm happy if you pick whatever works.

And I thought your first 4 patches where about other test issues, not
about synchronous relocs specifically? Like the igt spinner stuff. I'm
not planning to touch anything in that area.
-Daniel

> --Jason
>
> > Cheers, Daniel
> >
> > > ---
> > >  tests/i915/gem_exec_reloc.c | 15 +++++++++++++++
> > >  1 file changed, 15 insertions(+)
> > >
> > > diff --git a/tests/i915/gem_exec_reloc.c b/tests/i915/gem_exec_reloc.c
> > > index d54473341..0c2f389ea 100644
> > > --- a/tests/i915/gem_exec_reloc.c
> > > +++ b/tests/i915/gem_exec_reloc.c
> > > @@ -415,6 +415,11 @@ static void many_active(int i915, const intel_ctx_t *ctx, unsigned engine)
> > >       const uint64_t max = 2048;
> > >       unsigned long count = 256;
> > >
> > > +     /* Relocating in an active batch buffer doesn't work with the
> > > +      * command parser
> > > +      */
> > > +     igt_require(!gem_engine_has_cmdparser(i915, &ctx->cfg, engine));
> > > +
> > >       igt_until_timeout(2) {
> > >               uint64_t required, total;
> > >
> > > @@ -488,6 +493,11 @@ static void wide_active(int i915, const intel_ctx_t *ctx, unsigned engine)
> > >       const uint64_t max = gem_aperture_size(i915) / 4096 / 2;
> > >       unsigned long count = 256;
> > >
> > > +     /* Relocating in an active batch buffer doesn't work with the
> > > +      * command parser
> > > +      */
> > > +     igt_require(!gem_engine_has_cmdparser(i915, &ctx->cfg, engine));
> > > +
> > >       igt_until_timeout(2) {
> > >               uint64_t required, total;
> > >
> > > @@ -517,6 +527,11 @@ static void active_spin(int fd, const intel_ctx_t *ctx, unsigned engine)
> > >       struct drm_i915_gem_execbuffer2 execbuf;
> > >       igt_spin_t *spin;
> > >
> > > +     /* Relocating in an active batch buffer doesn't work with the
> > > +      * command parser
> > > +      */
> > > +     igt_require(!gem_engine_has_cmdparser(fd, &ctx->cfg, engine));
> > > +
> > >       spin = igt_spin_new(fd,
> > >                           .ctx = ctx,
> > >                           .engine = engine,
> > > --
> > > 2.31.1
> > >
> > > _______________________________________________
> > > igt-dev mailing list
> > > igt-dev at lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/igt-dev
> >
> > --
> > Daniel Vetter
> > Software Engineer, Intel Corporation
> > http://blog.ffwll.ch



-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch


More information about the igt-dev mailing list