home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!uwm.edu!rutgers!njitgw.njit.edu!hertz.njit.edu!dic5340
- From: dic5340@hertz.njit.edu (David Charlap)
- Newsgroups: comp.os.os2.misc
- Subject: ATI VGA Wonder (Old model, revision 5 chipset) and TRI32New drivers
- Message-ID: <1993Jan10.230015.27050@njitgw.njit.edu>
- Date: 10 Jan 93 23:00:15 GMT
- Sender: news@njit.edu
- Organization: New Jersey Institute of Technology, Newark, N.J.
- Lines: 39
- Nntp-Posting-Host: hertz.njit.edu
-
-
- I downloaded the TRI32NEW.ZIP package and tried to install the
- 800x600x256 driver on my 512K VGA-Wonder board. As everyone else
- reported, the top 88 lines were duplicated on the bottom of the
- screen, and vice versa. AFter spending much time with a friend going
- over the SVGADATA.PMI file, we found a bug in it: CRT register 9 was a
- 0, causing the Line Compare Register to contain 511 - causing a split
- screen at row 512, exactly what happened. According to my specs, bit
- 6 of the Maximum Scan Line register (R9) should be set to a 1, to
- yield a Line Compare value of 1023 - below the bottom of the screen.
-
- Unfortunately, changing the value of this in the SVGADATA.PMI file did
- nothing. As a matter of fact, nothing happened after chaning ANY of
- the normal IBM regusters (sequencer, CRT, GDC, and attributes). The
- only registers that had any effect were the ATI-specific ones (at
- 1ce/1cf on my system), and they only made things worse.
-
- Since Joel Armengaud's ATI256C2 driver uses the SVGADATA.PMI file
- without this bug, I assumed that the BVHSVGA.DLL file is ignoring the
- IBM-standard register listings in the SVGADATA.PMI file, and is only
- reading the manufacturer-specific portions - leaving the device driver
- to itself when setting the registers.
-
- Anyway, does this seem to make sense? Is there anything I can do
- about it, short of getting source code to the driver? Interestingly
- enough, other chipsets (like TSENG) correctly set the Line Compare
- register to 1023, so the Trident drivers don't have this problems on
- those systems.
-
- I tried searching through the SVGA600.DLL file for a parameter table
- to patch, but this is nearly impossible. I ran EXEHDR on the .DLL,
- but IBM hasn't documented the output (for 32-bit DLL's) completely
- enough, so I can't locate any of the "prime-candidate" modules in the
- .DLL file to look at them.
- --
- |) David Charlap | .signature confiscated by FBI due to
- /|_ dic5340@hertz.njit.edu | an ongoing investigation into the
- ((|,) | source of these .signature virusses
- ~|~
-