test patch for computed gotos

[Richard Henderson <rth at redhat dot com>]

Would you benchmark this patch vs your scheme->c compiler?

It's a collection of Joseph's bb-reorder rewrite plus tweaks, plus stuff.
It implements the idea I mentioned the other day about doing computed goto
with a common computed goto block in order to reduce edge count, but with
bb-reorder duplicating the block with the indirect jump at the last moment.

You should be able to run this with -O2, including gcse.  I tested vs
conform.i, and get compile times of 18.53 seconds before the patch (with
gcse auto-disabled), and 12.02 seconds after the patch (with gcse *not*
auto-disabled).

I saw what appeared to be better code generation with -fno-crossjumping.
Hard to tell if that's real or imaginary though.


r~

Attachment: d-comp-goto.gz
Description: GNU Zip compressed data