c/7344: performance regression on huge case statements

[Daniel Berlin]

On Thursday, October 10, 2002, at 10:04  PM, Nathanael Nerode wrote:

> http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit- 
> trail&database=gcc&pr=7344
>
> I'm tired of investigating this, but it seems likely that the problem
> was introduced with the introduction of et-forest.c, since that's  
> where the loop is.  This was introduced by Pavel Nejedly and committed  
> to mainline by Jan Hubicka (along with most of the surrounding code).
>
> (So that's why I'm ccing you; it looks like you caused it, so maybe  
> you  can figure out how to fix it. :-/)
>
As I mentioned, this is likely a known problem, with a known fix.

Jan, if you don't have plans to make it constant time soon, i suggest  
we cache the info using the patch that is on the tree-ssa branch.

> --Nathanael
>
>