Return to BSD News archive
Xref: sserve comp.os.386bsd.development:275 comp.os.386bsd.questions:989 comp.os.linux:30931 comp.windows.x.i386unix:949 comp.unix.pc-clone.32bit:2022 Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!osuunx.ucc.okstate.edu!moe.ksu.ksu.edu!zaphod.mps.ohio-state.edu!sdd.hp.com!network.ucsd.edu!yossarian!bill From: bill@yossarian.ucsd.edu (Bill Reynolds) Newsgroups: comp.os.386bsd.development,comp.os.386bsd.questions,comp.os.linux,comp.windows.x.i386unix,comp.unix.pc-clone.32bit Subject: Cirrus XFree86 Driver: Alpha testers wanted Date: 24 Mar 93 00:52:59 Organization: Institute for Nonlinear Science, UCSD Lines: 68 Distribution: world Message-ID: <BILL.93Mar24005259@yossarian.ucsd.edu> NNTP-Posting-Host: louise.ucsd.edu I've got a color driver up and running for the Cirrus 5426 chip. So far, it has only been tested on my board: a Genoa 8500. If you're interested in being an alpha tester, send me mail at the address below (not the one in the From line). Note that you need to be able to build the XFree server on your system. Here is the README: $Header: /usr/local/src/Xaccel/cirrus/RCS/README,v 1.3 1993/03/24 07:03:36 bill Exp $ This is the alpha version of the Cirrus Chipset driver for XFree86. The code is best described in terms of what it's not rather than what it is. WHAT IT DOESN'T DO -It doesn't use the Hardware BitBLT -It doesn't use the Extended Write Modes for Text -It doesn't use the hardware cursor -It doesn't do monochrome (use the generic server) -It doesn't do interlace (Probably not to bad to add, how badly is this needed?) -It doesn't support any chip other than the 5426 (probably easy to fix, although getting less easy as we add goodies) WHAT IT DOES DO -It supports the following twelve dot clocks Clocks 25.2 28.3 41.2 36.1 31.5 40.0 44.9 49.87 65.0 72.2 75.0 80.0 (in that order) Currently it's just code for a dumb framebuffer. It's not fast, but I find it tolerable on my machine (which is probably bad news for someone else who finds it intolerable on theirs :-). Most of the fishy code is now gone. Not to say that it won't reappear as we add more features. BUGS I see the following problems: -Sparklies. Could be a hardware problem on my bus, could be a bogus register someplace. I'm open to suggestions. -Diagonal Lines. When a client (like xv) really beats on the colormap, diagonal lines appear on the display, going from the lower left to the upper right. This is undoubtedly some kind of software problem. As to what that problem is, I'm clueless. -The amount of video memory is hardwired in. Cirrus claims that you can't determine the amount of memory without a bios call. I, feeling that this is bullshit, am waiting for a call from the Cirrus bios guy. -The clock probe routine finds different clocks than what the Cirrus databook says should be there. Still seems to work. Not clear if it's the clock probe hack or the chip doesn't run at the speed it thinks it runs at. FIXED BUGS The problems with moving virtual screen sizes around seems to be fixed. Send Bug Reports, Flames and Suggestions to me: ______________________________________________________________________________ Bill Reynolds | Hey Rocky! Watch me pull a rabbit out of my hat! bill@goshawk.lanl.gov | -Bullwinkle -- ______________________________________________________________________________ Bill Reynolds | Hey Rocky! Watch me pull a rabbit out of my hat! bill@goshawk.lanl.gov | -Bullwinkle