Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!nic.hookup.net!swrinde!cs.utexas.edu!howland.reston.ans.net!agate!dog.ee.lbl.gov!hellgate.utah.edu!cc.usu.edu!ivie From: ivie@cc.usu.edu Newsgroups: comp.os.386bsd.questions Subject: Re: micro-code vs DMA (was Re: CDROM support) Message-ID: <1994Jan18.115506.8075@cc.usu.edu> Date: 18 Jan 94 11:55:06 MDT References: <2g5qaq$ng0@u.cc.utah.edu> <DykkFc3w165w@oasys.pc.my><BLYMN.94Jan9192240@siren.awadi.com.au><2gs11p$58g@delphinium.cig.mot.com> <BLYMN.94Jan17230950@siren.awadi.com.au> Organization: Utah State University Lines: 24 In article <BLYMN.94Jan17230950@siren.awadi.com.au>, blymn@awadi.com.au (Brett Lymn (master of the siren)) writes: > 680x0 as well), for a start you lose big time if the target buffer > does not exist in the TLB since you will suffer the overhead of a > cache miss for each byte in the buffer. Also if you ignore the REP > You should only pay for a cache miss for one byte on each cache line. And you should only pay for a TLB miss for one byte on each page. I don't know how the x86 crowd works the REP prefix, but the Z80 does block moves by not advancing the PC if it's not done yet; in other words, a Z80 will fetch and decode the instruction for each byte in the move. > Dan> The BIOS for the original IBM AT uses NO DMA to interface to the > Dan> drive controller. the 286 was faster thyan DMA for DOS (Unix is a > Dan> different story). > > A lot of programmers have a mistrust of DMA, must be because it's a In the case of the IBM AT, it's because the DMA engine runs at a whopping 3 MHz. Block moves really are faster than DMA on an IBM AT. -- ----------------+------------------------------------------------------ Roger Ivie | Don't think of it as a 'new' computer, think of it as ivie@cc.usu.edu | 'obsolete-ready'