Return to BSD News archive
Xref: sserve comp.os.386bsd.development:1767 comp.unix.bsd:13365 Newsgroups: comp.os.386bsd.development,comp.unix.bsd Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!decwrl!decwrl!olivea!spool.mu.edu!howland.reston.ans.net!xlink.net!subnet.sub.net!flatlin!bad From: bad@flatlin.ka.sub.org (Christoph Badura) Subject: Re: Late comments on the DMA problem. Date: Sat, 5 Feb 1994 19:41:29 GMT Message-ID: <CKro16.JBM@flatlin.ka.sub.org> References: <1994Jan24.103912.8591@cc.usu.edu> <940131281@rwsys.wimsey.bc.ca> <jmonroyCKLDAu.CK@netcom.com> Organization: Guru Systems/Funware Department Lines: 27 In <jmonroyCKLDAu.CK@netcom.com> jmonroy@netcom.com (Jesus Monroy Jr) writes: >Randy Wright (rw@rwsys.wimsey.bc.ca) wrote: >: ivie@cc.usu.edu () writes: >: > In article <jmonroyCK2t6n.1pM@netcom.com>, jmonroy@netcom.com (Jesus Monroy Jr) writes: >: > > ivie@cc.usu.edu wrote: >: > > : Since you can't get an interrupt to tell you that the DMA controller has >: > > : finished doing a page, you can't reload the DMA addresses when the end of >: > > : a page is reached. >: > > please define page. your message is ambiguous. >: > Page = virtual memory page. I'm most familiar with the VAX, where a page >: > is 512 bytes; I don't know how big it is on these newfangled 386 thingies. >: 386 virtual memory paging deals in 4-kb pages. > Can you please define this 4-kb page a bit more with a qualifier? The definition of page is completely irrelevant since you can't get an interrupt from the DMA controller when it finishes a page of any size. The point is that the DMA controller knows zilch about pages, so you can't get him to do something when he's finished with a page he doesn't know about. Anyone how thinks the DMA controller would know anything about the MMU's pages is simply not competent enough to discuss DMA issues. -- Christoph Badura bad@flatlin.ka.sub.org +49 721 606137 Es genuegt nicht, keine Gedanken zu haben; man muss auch unfaehig sein, sie auszudruecken. - Karl Kraus