This is the mail archive of the
mailing list for the GCC project.
Re: Inlining heuristics
- From: Michael Matz <matz at suse dot de>
- To: Brad Lucier <lucier at math dot purdue dot edu>
- Cc: gcc-patches at gcc dot gnu dot org, <feeley at iro dot umontreal dot ca>
- Date: Tue, 13 May 2003 22:10:45 +0200 (CEST)
- Subject: Re: Inlining heuristics
On Tue, 13 May 2003, Brad Lucier wrote:
> The heuristic is applied at the expression level. The default
> inlining-limit is 300%, i.e., inlining is allowed to triple the size of
> the function.
I feel that this is indeed also a fairly useful heuristic. I.e. making
the allowable size of inlined functions depend on the initial size of
the calling function. It would better reflect the relative saving of
inlining compared to the overall overhead of the caller.