On Wed, 18 Feb 2009, Kjetil Barvik wrote: > > To fix this, and use an O(N) algorithm instead, where N is the number > of cache_entries, we delete/remove all entries in one loop through all > entries. Ack. We've had things like this before. I'm somewhat surprised that it is noticeable even in the "o->update" case (since I would have expected the cost of actually doing an unlink() to swamp the memmove overhead), but the patch looks obviously correct. Looking at the numbers, I guess it's not _really_ noticeable - it's 1.8% of git user-space time, which in turn is about 75% of total time, so it's not a huge improvement, but I'll certainly ack it anyway as being the right thing to do. Every little bit helps. Linus -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
