This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Inlining and estimate_num_insns
- From: Paul Schlie <schlie at comcast dot net>
- To: Richard Guenther <rguenth at tat dot physik dot uni-tuebingen dot de>
- Cc: Steven Bosscher <stevenb at suse dot de>,<gcc at gcc dot gnu dot org>,Jan Hubicka <hubicka at ucw dot cz>,<mark at codesourcery dot com>
- Date: Sun, 27 Feb 2005 11:24:05 -0500
- Subject: Re: Inlining and estimate_num_insns
> Richard Guenther writes:
> ... I also think our in-lining limits are way too high,
> at least unnecessary high with the patch applied...
- If function-call overhead could be more accurately estimated, possibly
by enabling the inclusion of set/move instructions in rtl cost estimates to
more accurately account for a function's variable and/or constant parameter
call overhead based on the target's machine description, then the in-lining
limit may be more productively expressed as a ratio ranging from 0 to 2-3x
code-growth limit based on the estimated in-lined-code/call-overhead based
on optimization goals; thereby enabling in-lining limits to be expressed
similarly to loop-unrolling limit's, as it's this ratio which would seem to
most likely estimates it's benefit?