Re: [patches] Re: more robust purge_dead_edges

[Richard Henderson <rth at redhat dot com>]

On Tue, Jul 24, 2001 at 11:59:26PM +0200, Jan Hubicka wrote:
> GCSE never converts conditional jumps to other condtiional jumps,
> so the function returns early and executes only on the uncondjump
> path.

Hum.  Well, I still think it's a bad idea.  Leave it out for now.

At some later non-emergency-bug-fixing point we decide to verify
that we havn't dropped any prediction data, then we can add a global
variable that is true after prediction data is added.  Then 
purge_dead_edges can not try to update what shouldn't be there.


r~