| Message ID | 20181115120126.9545-3-jani.nikula@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [1/5] drm/i915: extract fixed point math to i915_fixed.h | expand |
Quoting Jani Nikula (2018-11-15 14:01:24) > No need to use a compound statement enclosed in parenthesis where a C99 > compound literal will do. No functional changes. > > Signed-off-by: Jani Nikula <jani.nikula@intel.com> Out of curiosity, did this have an effect on asm generation? Presumably not. Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Regards, Joonas
Quoting Jani Nikula (2018-11-15 12:01:24) > No need to use a compound statement enclosed in parenthesis where a C99 > compound literal will do. No functional changes. > > Signed-off-by: Jani Nikula <jani.nikula@intel.com> > --- > drivers/gpu/drm/i915/i915_fixed.h | 6 +----- > 1 file changed, 1 insertion(+), 5 deletions(-) > > diff --git a/drivers/gpu/drm/i915/i915_fixed.h b/drivers/gpu/drm/i915/i915_fixed.h > index 08316e50167a..927c59395569 100644 > --- a/drivers/gpu/drm/i915/i915_fixed.h > +++ b/drivers/gpu/drm/i915/i915_fixed.h > @@ -7,11 +7,7 @@ typedef struct { > u32 val; > } uint_fixed_16_16_t; > > -#define FP_16_16_MAX ({ \ > - uint_fixed_16_16_t fp; \ > - fp.val = UINT_MAX; \ > - fp; \ > -}) > +#define FP_16_16_MAX ((uint_fixed_16_16_t){ .val = UINT_MAX }) Following the standard set by pgprot_t #define u16_16(x) ((u16_16_t){ .val = (x) }) #define U16_16_MAX u16_16(U32_MAX) -Chris
On Fri, 16 Nov 2018, Joonas Lahtinen <joonas.lahtinen@linux.intel.com> wrote: > Quoting Jani Nikula (2018-11-15 14:01:24) >> No need to use a compound statement enclosed in parenthesis where a C99 >> compound literal will do. No functional changes. >> >> Signed-off-by: Jani Nikula <jani.nikula@intel.com> > > Out of curiosity, did this have an effect on asm generation? > Presumably not. For some reason it caused tons of label etc. changes so it was hard to check, but I didn't spot any functional asm changes. BR, Jani. > > Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> > > Regards, Joonas
On Fri, 16 Nov 2018, Chris Wilson <chris@chris-wilson.co.uk> wrote: > Quoting Jani Nikula (2018-11-15 12:01:24) >> No need to use a compound statement enclosed in parenthesis where a C99 >> compound literal will do. No functional changes. >> >> Signed-off-by: Jani Nikula <jani.nikula@intel.com> >> --- >> drivers/gpu/drm/i915/i915_fixed.h | 6 +----- >> 1 file changed, 1 insertion(+), 5 deletions(-) >> >> diff --git a/drivers/gpu/drm/i915/i915_fixed.h b/drivers/gpu/drm/i915/i915_fixed.h >> index 08316e50167a..927c59395569 100644 >> --- a/drivers/gpu/drm/i915/i915_fixed.h >> +++ b/drivers/gpu/drm/i915/i915_fixed.h >> @@ -7,11 +7,7 @@ typedef struct { >> u32 val; >> } uint_fixed_16_16_t; >> >> -#define FP_16_16_MAX ({ \ >> - uint_fixed_16_16_t fp; \ >> - fp.val = UINT_MAX; \ >> - fp; \ >> -}) >> +#define FP_16_16_MAX ((uint_fixed_16_16_t){ .val = UINT_MAX }) > > Following the standard set by pgprot_t > > #define u16_16(x) ((u16_16_t){ .val = (x) }) > #define U16_16_MAX u16_16(U32_MAX) I left that for follow-up, and pushed v2 of the series. I think the uint_fixed_16_16_t type name is a bit unwieldy, should we rename that while at it... BR, Jani.
diff --git a/drivers/gpu/drm/i915/i915_fixed.h b/drivers/gpu/drm/i915/i915_fixed.h index 08316e50167a..927c59395569 100644 --- a/drivers/gpu/drm/i915/i915_fixed.h +++ b/drivers/gpu/drm/i915/i915_fixed.h @@ -7,11 +7,7 @@ typedef struct { u32 val; } uint_fixed_16_16_t; -#define FP_16_16_MAX ({ \ - uint_fixed_16_16_t fp; \ - fp.val = UINT_MAX; \ - fp; \ -}) +#define FP_16_16_MAX ((uint_fixed_16_16_t){ .val = UINT_MAX }) static inline bool is_fixed16_zero(uint_fixed_16_16_t val) {
No need to use a compound statement enclosed in parenthesis where a C99 compound literal will do. No functional changes. Signed-off-by: Jani Nikula <jani.nikula@intel.com> --- drivers/gpu/drm/i915/i915_fixed.h | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-)