This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: INTEGRATE_THRESHOLD ever changed?
- To: Stan Shebs <shebs at apple dot com>
- Subject: Re: INTEGRATE_THRESHOLD ever changed?
- From: Richard Henderson <rth at cygnus dot com>
- Date: Tue, 14 Mar 2000 17:02:01 -0800
- Cc: gcc at gcc dot gnu dot org
- References: <38CED544.8DB519AE@apple.com>
On Tue, Mar 14, 2000 at 04:11:49PM -0800, Stan Shebs wrote:
> The reason I ask is that Apple's code has a version that calculates
> the actual number of call-used registers at the point where the
> compiler is thinking about whether to inline ...
That information is not availible during inlining.
One would be able to do some educated guesswork based on number of
registers live across function calls -- which is not really the same
thing, though related. But to compute this value in less than 200
lines requires infrastructure only recently written.
I'm guessing the patch isn't this big, and thus is largely bogus.
r~