nanogui: Testing the 4 planes Vga Driver - Bare hardware


Previous by date: 6 Oct 1999 01:05:27 -0000 Re: A small request, Robert J.Lawrence
Next by date: 6 Oct 1999 01:05:27 -0000 Re: Testing the 4 planes Vga Driver - Bare hardware, Greg Haerr
Previous in thread: 6 Oct 1999 01:05:27 -0000 Re: Testing the 4 planes Vga Driver - Bare hardware, Greg Haerr
Next in thread: 6 Oct 1999 01:05:27 -0000 Re: Testing the 4 planes Vga Driver - Bare hardware, Greg Haerr

Subject: RE: Testing the 4 planes Vga Driver - Bare hardware
From: Rosimildo daSilva ####@####.####
Date: 6 Oct 1999 01:05:27 -0000
Message-Id: <199910060058.RAA21350@www2.xoommail.com>

 > : The code runs in protected mode and it does not have BIOS access.
 > 
 > 	The protected mode is no problem, since the vgaplan4 driver
 > runs under Linux ELKS.  But the scr_bios wrapper requires that the
 > bios be called to init and deinit the graphics card.  I never wrote
 > the routines to init the vga card, since they vary so much from
 > card to card.  I have some routines that you could fiddle with
 > that attempt to program the vga hardware directly for initialization, if you want.
 > They are in David Bell's original mini-x code, in a file called ega.s, or something
 > like it.  See ftp://microwindows.censoft.com/pub/mini-x  get the file mini-x-new.tar.gz.
 > 


I linked vgainit.c, and it seems to have the routines there.

My initial problem was that I thought outp() and outb() had the
same signature. :-).



 > 	It would be fairly easy to write a #ifdef TEST main() routine
 > for the video driver.  Perhaps you could do it for the working linux screen
 > driver, and then integrate Bell's ega startup code.  If you send me the changes,
 > I'll add them to the release, and test your test code with other drivers.
 > 


I'll try as soon as I get something testable.


 > 	Like I mentioned, the scr_bios wrapper around vgaplan.c needs to be
 > slightly rewritten.  If you set #define HWINIT 1, then you can just write two
 > replacement bios routiones, ega_hwinit(), and ega_hwterm().  Other than that,
 > the driver is used under the ELKS build currently.
 > 


Ok, I guess I found the problem with the initialization.
I had defined outp() and outb() with the same signature.


Now, I need to understand how should I setup
rom_char_addr. My target does not have BIOS.
I have tried this:

     extern FONT font_rom8x16;  /* font generated with convrom.c */
     FONT *font_ptr = &font_rom8x16;
     rom_char_addr = (FARADDR)( font_ptr->bits );

When I execute the code, the result does not look correct.

Do I need to use the convrom.exe utility ?

Any hints ?

Rosimildo.


 


______________________________________________________
Get your free web-based email at http://www.xoom.com
Birthday? Anniversary? Send FREE animated greeting
cards for any occasion at http://greetings.xoom.com



Previous by date: 6 Oct 1999 01:05:27 -0000 Re: A small request, Robert J.Lawrence
Next by date: 6 Oct 1999 01:05:27 -0000 Re: Testing the 4 planes Vga Driver - Bare hardware, Greg Haerr
Previous in thread: 6 Oct 1999 01:05:27 -0000 Re: Testing the 4 planes Vga Driver - Bare hardware, Greg Haerr
Next in thread: 6 Oct 1999 01:05:27 -0000 Re: Testing the 4 planes Vga Driver - Bare hardware, Greg Haerr


Powered by ezmlm-browse 0.20.