[Intel-gfx] [PATCH] drm/i915: fix an error code in intel_modeset_all_tiles()

[Jani Nikula]

On Tue, 07 Jan 2020, Chris Wilson <chris at chris-wilson.co.uk> wrote:
> Quoting Souza, Jose (2020-01-07 14:01:40)
>> On Tue, 2020-01-07 at 16:03 +0300, Dan Carpenter wrote:
>> > There is a cut and paste bug so we return the wrong error code.
>> > 
>> > Fixes: a603f5bd1691 ("drm/i915/dp: Make sure all tiled connectors get
>> > added to the state with full modeset")
>> 
>> Reviewed-by: José Roberto de Souza <jose.souza at intel.com>
>> 
>> > Signed-off-by: Dan Carpenter <dan.carpenter at oracle.com>
>
> Pushed. Thanks for the patch,

Hrmh, we are still supposed to wait for CI results for every patch, no
exceptions.

If it's not "every patch", it's going to be a subjective assessment, and
I really don't like where that would lead. Who's going to say what's
going to be just fine. Even if it's "obviously correct".

As it is, this patch won't get the CI results at all now because BAT
fails with "patch already applied". Tracing any issues back to this
patch would mean bisecting the results from drm-tip runs.

---

If the problem is that CI is too heavy, slow and itself error prone for
small patches, then that's what we need to address instead of just
bypassing CI.

I also don't like the prospect of spending time on hacking dim to
double-check the patches passed CI before pushing.


BR,
Jani.


-- 
Jani Nikula, Intel Open Source Graphics Center