On Thu, Apr 07, 2016 at 03:05:38PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin <[email protected]> > > Knowledge of which register per platform belonds in which > forcewake domain was embedded in the MMIO accessors themselves. > > Extract it into standalone macros so they can be used from > new code in the following patches. > > This causes GCC to compile some of the MMIO accessors slightly > differently and grows the code a tiny amount. But none of the > growth is on the fast-path so it does not matter hugely. > > Affected sizes before: > > 00000000000026f0 00000000000001a5 t gen6_read16 > 0000000000002390 00000000000001a5 t gen6_read32 > 00000000000028a0 00000000000001a5 t gen6_read64 > > 00000000000061d0 000000000000019e t gen8_write16 > 0000000000006510 000000000000019d t gen8_write32 > 0000000000006370 000000000000019d t gen8_write64 > 00000000000021f0 000000000000019d t gen8_write8 > > Affected sizes after: > > 0000000000002840 00000000000001aa t gen6_read16 > 00000000000024e0 00000000000001a9 t gen6_read32 > 00000000000029f0 00000000000001a9 t gen6_read64 > > 0000000000004f20 00000000000001b5 t gen8_write16 > 0000000000004ba0 00000000000001b4 t gen8_write32 > 00000000000050e0 00000000000001b4 t gen8_write64 > 0000000000004d60 00000000000001b4 t gen8_write8 > > Other MMIO accessors are not affected in size. > > Signed-off-by: Tvrtko Ursulin <[email protected]> > Acked-by: Chris Wilson <[email protected]>
Went through the macros side-by-side and confirmed it was supposed to only be code motion. Silly compiler. Reviewed-by: Chris Wilson <[email protected]> -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
