*BSD News Article 29152


Return to BSD News archive

Xref: sserve comp.os.os2.misc:90162 comp.windows.x.i386unix:8353 comp.os.linux.misc:12398 comp.os.linux.help:27924 comp.os.386bsd.misc:2223 comp.os.386bsd.questions:9625 comp.sys.next.software:10495 comp.sys.next:17044
Newsgroups: comp.os.os2.misc,comp.windows.x.i386unix,comp.os.linux.misc,comp.os.linux.help,comp.os.386bsd.misc,comp.os.386bsd.questions,comp.sys.next.software,comp.sys.next
Path: sserve!newshost.anu.edu.au!munnari.oz.au!foxhound.dsto.gov.au!fang.dsto.gov.au!yoyo.aarnet.edu.au!news.adelaide.edu.au!news.cs.su.oz.au!harbinger.cc.monash.edu.au!yeshua.marcam.com!news.kei.com!hookup!ames!pacbell.com!amdahl!juts.ccc.amdahl.com!orca!haw30
From: haw30@eng.amdahl.com (Henry A Worth)
Subject: Re: PCI bus cards (graphics and SCSI) which work?
Message-ID: <1994Apr4.173523.11421@ccc.amdahl.com>
Sender: netnews@ccc.amdahl.com (UTS Tech Support)
Organization: Amdahl
References: <2mslqq$1vl8@rs560.cl.msu.edu> <michaelv.764537928@ponderous.cc.iastate.edu> <DAVEH.94Mar28165649@canopus.commodore.com>
Date: Mon, 4 Apr 1994 17:35:23 GMT
Lines: 45

In article <DAVEH.94Mar28165649@canopus.commodore.com>, daveh@canopus.commodore.com (Dave Haynie) writes:
|> 
|> In article <1994Mar25.175046.28527@ccc.amdahl.com> haw30@eng.amdahl.com (Henry A Worth) writes:
|>    In article <michaelv.764537928@ponderous.cc.iastate.edu>, michaelv@iastate.edu (Michael L. VanLoon) writes:
|> 
|>    |> Do NOT buy ANY Diamond products!  They are not and will never be
|>    |> supported by XFree86 (although a few adventerous souls have gotten
|>    |> them to limp along at times).  Diamond will not allow their drivers to
|>    |> be public and doesn't care at all that this may cost them a sale
|>    |> because you run a free "unix".  They are quite indifferent to the
|>    |> pleas of free "unix" users.
|> 
|>    A small point, none of the chip/board vendors provide driver code,
|>    the issue is whether they will provide hardware documentation, without
|>    non-disclosure agreements, that will allow a 3rd party, like the
|>    XFree86[TM] project, write X drivers that can be released to the
|>    public in source form. A few vendors go the extra mile and provide
|>    some tech support, loaner boards, or bits of example code, but such
|>    cooperation is not a pre-req to XFree86 support.      
|> 
|> Some companies will provide register-level information on their chips
|> without extra cost or NDAs.  If you can get this, it matters very
|> little what the card's manufacturer will or won't supply.  We're not
|> talking about a custom design here -- all of the graphic cards based
|> on a handful of controller chips.  At most, the card vendor gets to
|> pick a LUT/DAC, clock synthesizer, and DRAM configuration.
|> 
 
As the level of on-chip integration gets higher, the role of the 
board designer is reduced. But, there are still some design
decisions that are left up to the board designers, i.e. how the
extended control registers in the newer RAMDAC's and external
progammable clock generators are mapped into the I/O address space, 
and by what convoluted sequences they are accessed. 

I'm currently having to work thru a board vendor to get such 
information for the server I'm working on. The accelerator
chip only provides one extra RAMDAC address bit. That one bit,
combined with the standard VGA LUT addresses, is enough to access 
only half the control registers on the RAMDAC used, leaving it up
to the board designer to handle the mapping of the RAMDAC's registers.
 
-- 
Henry Worth - haw30@eng.amdahl.com
No, I don't speak for Amdahl... I'm not even sure I speak for myself.