Date: Mon, 5 Nov 2012 04:22:33 -0800 (PST) From: "Rodney W. Grimes" <freebsd@pdx.rh.CN85.ChatUSA.com> To: Eitan Adler <lists@eitanadler.com> Cc: Juli Mallett <juli@clockworksquid.com>, "freebsd-mips@FreeBSD.org" <freebsd-mips@freebsd.org> Subject: Re: CACHE_LINE_SIZE macro. Message-ID: <201211051222.qA5CMYNF095111@pdx.rh.CN85.ChatUSA.com> In-Reply-To: <CAF6rxgn-bNJOuvdiRj_UUGQUTRaeOt54OdzHOioNz5f566hoig@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[ Charset UTF-8 unsupported, converting... ] > On 4 November 2012 13:28, Rodney W. Grimes > <freebsd@pdx.rh.cn85.chatusa.com> wrote: > > > > What ABI is exposing anything about cache parameters or may be come dependent > > on such information? > > There has been some discussion recently about padding lock mutexs to > the cache line size in order to avoid false sharing of CPUs. Some have > claimed to see significant performance increases as a result. I actually just went and read the diffs in attilio r242402 commit and it appears to me that he is actually REMOVING the CACHE_LINE_SIZE abuse and using a mtx_padalign struct that must already exist cause his commit didnt create it. Independent of weither using cache line size alignment is evil, at least this commit reduces the abuse of the constant. -- Rod Grimes freebsd@freebsd.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201211051222.qA5CMYNF095111>