Date: Sat, 8 Mar 2003 13:51:16 -0800 From: David Schultz <das@FreeBSD.org> To: "Marc G. Fournier" <scrappy@hub.org> Cc: Wes Peters <wes@softweyr.com>, Vallo Kallaste <kalts@estpak.ee>, freebsd-stable@FreeBSD.org Subject: Re: "leak" in softupdates? Message-ID: <20030308215116.GA70447@HAL9000.homeunix.com> In-Reply-To: <20030307203014.D66674@hub.org> References: <20030305204526.T38115@hub.org> <20030307101718.GA1908@kevad.internal> <20030307081643.B15693@hub.org> <200303070648.26984.wes@softweyr.com> <20030307152045.P18433@hub.org> <20030307214306.GB63881@HAL9000.homeunix.com> <20030307203014.D66674@hub.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Thus spake Marc G. Fournier <scrappy@hub.org>: > On Fri, 7 Mar 2003, David Schultz wrote: > > > Thus spake Marc G. Fournier <scrappy@hub.org>: > > > > Being worked on. Not so hard to do, much harder to do right. Guess who > > > > took the easy sleazy path? ;^) The other good news is that the intel > > > > network cards, both 10/100 (fxp) and 10/100/1000 (em) support 64-bit > > > > addressing, even in 32-bit PCI slots, so you'll have at least ONE enet > > > > interface that'll work reasonably fast. > > > > > > Yes, I don't recall who it was that explained it to me (Terry, maybe?), > > > but I understand the problem with going above 4gig under ia32, and was > > > personally just sitting back and waiting for Intel to go full steam ahead > > > on the ia64 stuff ... but they just sacked it :( Man, did that ever throw > > > a shiver up my back ... > > > > It's amazing how many times bank switching has been reinvented, eh? > > Just curious here, but with the speed of CPUs nowadays, how much impact > would bank switching have on performance? I don't know exactly what impact PAE would have on performance, but an 8 GB PAE machine can't be worse than a 4 GB machine that's swapping like mad. I do know that I/O performance will suffer if your I/O devices can't address more than 4 GB of RAM, since that means much of your data will have to be copied to bounce buffers below 4 GB. Does anyone have Actual Data on how well it works? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030308215116.GA70447>