On 12/26/2012 09:27, Théophile Ranquet wrote:
>I am attaching the testcase, that runs in 11 sec with my patch, and in >maybe 50X that without.I suppose you meant "about fifty seconds", and not "50 times that", right? This is what I observed before and after applying your changes:
Sorry for my late response. This is what happens when issues are tracked via the mailing list.
No, I meant exactly "50 times faster", this is what I observed. So is this bottleneck fix checked in now? Yuri
