Re: [Monotone-devel] mtn status speed

[Jack Lloyd]

On Wed, Dec 10, 2008 at 10:29:57AM +0100, Markus Wanner wrote:

> Yeah, this is known and has been discussed before. Unfortunately I don't
> remember the reason for calling stat() twice. Caching of the stat()
> result has been considered, but argued against, because stat() should be
> fast and simple enough.

Yes the first stat will (or at least should) bring the inode info into
the OS cache, and since the second immediately follows it should
simply be the context switch for the syscall (which admittedly has a
nonzero cost though syscall entry is quite well optimized these days)
plus the cache read.