Return to BSD News archive
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!bunyip.cc.uq.oz.au!munnari.oz.au!spool.mu.edu!howland.reston.ans.net!Germany.EU.net!news.dfn.de!RRZ.Uni-Koeln.DE!usenet From: se@parc03.MI.Uni-Koeln.DE (Stefan Esser) Newsgroups: comp.os.386bsd.questions Subject: Re: Nagging FreeBSD 2.0 questions Date: 17 Feb 1995 23:25:21 GMT Organization: Institute for Mathematics, University of Cologne, Germany Lines: 25 Distribution: world Message-ID: <3i3b91$2is@news.rrz.uni-koeln.de> References: <1995Feb16.025123.8370@wdl.loral.com> <1995Feb16.225325.41987@cc.usu.edu> NNTP-Posting-Host: parc03.mi.uni-koeln.de In article <1995Feb16.225325.41987@cc.usu.edu>, kurto@cc.usu.edu (Kurt Olsen) writes: |> I had this same problem with a DX2/50 (according to Cyrix all their DX2s |> leave the factor with a large green heat sink and DX2 printed on it.) |> If you are willing to rip your chip out there should be a lot number |> on the bottom (mine was FLxxxxx.) |> |> However 2.0R identified my chip as a DLC as well. I sent a note to the |> hackers list, and Bruce Evans (or Rod Grimes, can't remember which) replied |> that the DLC has some problems with cache and DMA. To fix the cache/dma |> problem FreeBSD disables the cache on the I/O channel. |> |> I think they misunderstood my problem, which was the chip identification. |> Anyway I hacked out the code to identify the DLC with a 'jmp 2f' just before |> the section that checks for a DLC. This had the net effect of increasing |> my system performance by about 250%. You don't have any bus master controllers, I suppose ? Regards, -- Stefan Esser Internet: <se@ZPR.Uni-Koeln.DE> Zentrum fuer Paralleles Rechnen Tel: +49 221 4706019 Universitaet zu Koeln FAX: +49 221 4705160 Weyertal 80 50931 Koeln