remove reverse_die_lists

Mark Mitchell mark@codesourcery.com
Sun Apr 23 03:36:00 GMT 2006


David Edelsohn wrote:
>>>>>> Richard Guenther writes:
> 
>> Are these patches somehow fix bugs or regressions?
> 
> 	GCC has goals for compile time performance, not just performance
> of generated code, specific transformations, and new features.  Overall
> compile time regressions are regressions and any patches to improve that
> area are appreciated.

Certainly, as RM, I am happy to have improvements in compile-time
performance made at this point; I certainly consider our gradual
degredation in compile-time performance over time to be a bug.  On the
other hand, if Geoff's patches causes problems, we'll back them out
rather aggressively, since, as you say, they do not affect the core
correctness of the compiler.

-- 
Mark Mitchell
CodeSourcery
mark@codesourcery.com
(650) 331-3385 x713



More information about the Gcc-patches mailing list