Re: test patch for computed gotos

[Richard Henderson <rth at redhat dot com>]

On Fri, Feb 14, 2003 at 12:22:34AM -0500, Brad Lucier wrote:
> http://www.math.purdue.edu/~lucier/scheme.i.gz

The problem here is in the cross-jumping pass.  Previously, we'd
not do much here, because we don't crossjump across computed goto.
With the patch in question, we've hordes of incoming direct jumps
to the computed goto block.

With -fno-crossjumping, the time to compile drops from 84 seconds
to 18 seconds.

I'll investigate further.  Probably we want to drop the cutoff
from 100 edges to like 10 edges, and make it a --param.


r~