Date: Fri, 15 Jul 2005 00:52:14 +0100 (BST) From: Robert Watson <rwatson@FreeBSD.org> To: Jeff Roberson <jroberson@chesapeake.net> Cc: cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/vm uma_core.c uma_int.h Message-ID: <20050715004327.B46671@fledge.watson.org> In-Reply-To: <20050714164138.X574@10.0.0.1> References: <200507141617.j6EGHL2I088805@repoman.freebsd.org> <20050714164138.X574@10.0.0.1>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 14 Jul 2005, Jeff Roberson wrote: > Are you interested in getting those cache improvements into the tree? > The alloc/free to the same bucket change? I am quite interested in them, but have not yet been able to measure a statistically significant improvement with them for the couple of micro-benchmarks I had easily on-hand. There was some measurable change in variance though. I hope to have a chance to look at them more closely in the next week or two. With Joseph Koshy's hwpmc support now working with gprof, we should be able to look a lot more closely at how the caches are interacting with various bits of code. Robert N M Watson > > On Thu, 14 Jul 2005, Robert Watson wrote: > >> rwatson 2005-07-14 16:17:21 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/vm uma_core.c uma_int.h >> Log: >> In addition to tracking allocs in the zone, also track frees. Add >> a zone free counter, as well as a cache free counter. >> >> MFC after: 1 week >> >> Revision Changes Path >> 1.121 +5 -0 src/sys/vm/uma_core.c >> 1.32 +2 -0 src/sys/vm/uma_int.h >> >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050715004327.B46671>