On Mon, 2014-09-29 at 20:24 +0100, Jiong Wang wrote:
> On 29/09/14 19:32, Richard Henderson wrote:
> > On 09/29/2014 11:12 AM, Jiong Wang wrote:
> >> +inline rtx single_set_no_clobber_use (const rtx_insn *insn)
> >> +{
> >> + if (!INSN_P (insn))
> >> + return NULL_RTX;
> >> +
> >> + if (GET_CODE (PATTERN (insn)) == SET)
> >> + return PATTERN (insn);
> >> +
> >> + /* Defer to the more expensive case, and return NULL_RTX if there is
> >> + USE or CLOBBER. */
> >> + return single_set_2 (insn, PATTERN (insn), true);
> >> }
>
> Richard,
>
> thanks for review.
>
> > What more expensive case?
>
> single_set_no_clobber_use is just a clone of single_set, I copied the
> comments with
> only minor modifications.
I introduced that comment to single_set, in r215089, when making
single_set into an inline function (so that it could check that it
received an rtx_insn *, rather than an rtx).
> I think the "more expensive case" here means the case where there are
> PARALLEL that
> we need to check the inner rtx.
My comment may have been misleading, sorry. IIRC, what I was thinking
that the old implementation had split single_set into a macro and a
function. This was by Honza (CCed), 14 years ago to the day back in
r36664 (on 2000-09-29):
https://gcc.gnu.org/ml/gcc-patches/2000-09/msg00893.html
/* Single set is implemented as macro for performance reasons. */
#define single_set(I) (INSN_P (I) \
? (GET_CODE (PATTERN (I)) == SET \
? PATTERN (I) : single_set_1 (I)) \
: NULL_RTX)
I think by "the more expensive case" I meant having to make a function
call to handle the less-common cases (which indeed covers the PARALLEL
case), rather than having logic inline; preserving that inlined vs
not-inlined split was one of my aims for r215089.
Perhaps it should be rewritten to "Defer to a function call to handle
the less common cases", or somesuch?
[...snip rest of post...]
Dave
