[Intel-gfx] [RFC] Convert INTEL_INFO(...)->gen to INTEL_GEN(...)

[Dave Gordon]

On 08/04/16 09:24, Jani Nikula wrote:
> On Thu, 07 Apr 2016, Dave Gordon <david.s.gordon at intel.com> wrote:
>> Since Jani has given us this macro, I thought I'd make use of it by
>> converting all existing instances of this construct with a really
>> simple little Coccinelle script:
>>
>> @intel_gen@
>>      expression E;
>> @@
>>      <...
>> -   INTEL_INFO(E)->gen
>> +   INTEL_GEN(E)
>>      ...>
>
> I intentionally did not do this because I think it causes more trouble
> than it's worth. Basically this will conflict with roughly all patches
> currently in flight. Yes, I admit I did tell people to go wild, but
> please let's do this piecemeal.
>
> BR,
> Jani.

The point of including the actual Cocci code in the commit message is 
that then anyone who has a set of not-yet-submitted changes can apply 
the Cocci script to their own codebase before attempting to rebase onto 
the new version of nightly, thus eliminating from the conflicts all 
those which are simply due to the patch applied to upstream.

For this reason, I'd like to recommend that anyone doing this sort of 
bulk transformation with Cocci or awk or just sed should /always/ 
include the transformation script.

As for doing it piecemeal, what are the chances that there will ever be 
a time when there are no changes in flight in, say, intel_display.c?

$ git churn drivers/gpu/drm/i915/*.c | tail -6

918	drivers/gpu/drm/i915/intel_ringbuffer.c
1027	drivers/gpu/drm/i915/intel_dp.c
1108	drivers/gpu/drm/i915/i915_irq.c
1448	drivers/gpu/drm/i915/intel_pm.c
1583	drivers/gpu/drm/i915/i915_gem.c
3290	drivers/gpu/drm/i915/intel_display.c

Perhaps the maintainers could target one file at a time, any time it 
looks like we're having a quiet period?

.Dave.