Return to BSD News archive
Xref: sserve comp.os.386bsd.development:1705 comp.os.linux.development:4832 comp.sys.ibm.pc.hardware.chips:1829 comp.sys.ibm.pc.hardware.storage:2025 comp.periphs:4928 comp.realtime:4440 comp.arch.storage:2302
Newsgroups: comp.os.386bsd.development,comp.os.linux.development,comp.sys.ibm.pc.hardware.chips,comp.sys.ibm.pc.hardware.storage,comp.periphs,comp.realtime,comp.arch.storage
Path: sserve!newshost.anu.edu.au!munnari.oz.au!bunyip.cc.uq.oz.au!harbinger.cc.monash.edu.au!msuinfo!agate!howland.reston.ans.net!cs.utexas.edu!uunet!olivea!decwrl!decwrl!netcomsv!netcom.com!jmonroy
From: jmonroy@netcom.com (Jesus Monroy Jr)
Subject: Late comments on the DMA problem.
Message-ID: <jmonroyCJy610.HB7@netcom.com>
Followup-To: comp.os.386bsd.development,comp.unix.bsd
Keywords: DMA problem
Organization: NETCOM On-line Communication Services (408 241-9760 guest)
Date: Thu, 20 Jan 1994 21:22:11 GMT
Lines: 40
Late comments on the DMA problem
--------------------------------
Please be aware that the programming of the DMA and the
system buffering techinques are not under my direct control.
These comments were sent to me very recently on the DMA
problem. Your comments, as always, are quite welcome.
==============================================================================
>> B) It is possible that they way you are setting up DMA may be a problem.
>>
I have not considered this.
However, I did look into the problem at some extent early on.
I do not remember my results.
>> For example, it is not possible to implment "scatter/gather" with the FDC by
>> ^^^
|||
I think he meant to say that "it is possible", because I know
it is. Evidence is the the FDC (flexible drive controller)
has a command for read-a-track.
>> telling the FDC to transfer a large block of data, but telling the DMA
>> controller
>> to transfer say one file buffer, and then reprogram the DMA for a 2nd file
>> buffer. The interrupt service lattencies and service times far exceed the
>> FDC's limited buffering capability. This can be done for DMA-SCSI,
>> because SCSI devices have very large buffer capacities (except for some
>> older SCSI tape drives many years ago) that can "coast through" these
>> delays.
>>
--
Jesus Monroy Jr jmonroy@netcom.com
Zebra Research
/386BSD/device-drivers /fd /qic /clock /documentation
___________________________________________________________________________