On Fri, 3 Jun 2005, chas williams - CONTRACTOR wrote:
In message <[EMAIL PROTECTED]>,Jason McCormick w
rites:
So far so good. I assume this patch is the last bit of the cache
performance improvement? .83 is significantly faster (in unscientific
observations) than <= .82.
my patch doesnt have anything to do with performance. i think someone
goofed a bit when implementimg splitdcache. hopefully the author of
the split dcache stuff will double check my patch.
That'd be me, and I'm still looking.
What were phases 0 and 1 should be phases 4 and 5, and when you have split
dcache disabled you should be doing 2 phases with different numbers but
that still work as before. So switching those should not do what you want.
But I'm trying to see what might be happening.
_______________________________________________
OpenAFS-devel mailing list
[email protected]
https://lists.openafs.org/mailman/listinfo/openafs-devel
