This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Sigh. Inlining heuristics.
- To: dewar at gnat dot com
- Subject: Re: Sigh. Inlining heuristics.
- From: "Janis Johnson" <janis at us dot ibm dot com>
- Date: Wed, 11 Jul 2001 09:13:03 -0700
- Cc: aoliva at redhat dot com, torvalds at transmeta dot com, dan at cgsoftware dot com, dewar at gnat dot com, gcc at gcc dot gnu dot org, mark at codesourcery dot com
> The way gcc currently generates code, you'll often get _worse_ code from
> inlining, because it often adds register pressure when gcc decides to
keep
> stuff live for long stretches. Very noticeable on x86 with the few
> registers - you can get into stack slot reload hell, causing much worse
> code generation than if you didn't inline.
Does, or could, inlining use different parameters for different platforms?
Inlining on platforms with a lot of registers shouldn't be constrained by
rules designed for those with few registers.
Can inlining use profile information to determine whether to inline some
functions that would otherwise not be inlined? This was brought up at the
GCC IA-64 Summit.
Janis