bug-grub
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Issues with GRUB-gfxboot


From: chaac
Subject: Re: Issues with GRUB-gfxboot
Date: Mon, 22 Mar 2004 00:02:29 +0200
User-agent: Mozilla Thunderbird 0.5 (Windows/20040312)

J. Silverman wrote:

Hi Vesa,

Thank you for offering to take a look at the source code! The link to the code is:

http://neverborn.org/files/grub-gfxboot/

The person who made the patch created a source and a binary directory. In the .deb files under the binary dir, there are source files which I use to create the "message" which is used to create the menu. You might want to look into that as well as the source code under the source directory.

Thanks again,
JS

Just looked at it. In actual patched grub there is no problem. But custom gfxboot utils that make actual image to be show on screen or what ever also it does. Actually this image seems to contain assembler made source code and grub stub just calls to code in image that is commanded to be loaded.

As this file is in assembler and whole drawing to screen is spread all over the file I don't have _that_ much of time to fix it's problems. Maybe you could look for RedHat's version as that seems to be more compatible. But probably doesn't have that fancy MOD player......

But anyway. Some notes about for person fixing it:

- get & read http://www.vesa.org/vbe3.pdf :)

- VBE2/3 support can make your life easier when using linear frame buffer. Only really old cards supports only < VBE2. Of course you could also support banked modes.

- make virtual screen buffer for storing on screen gfx.

- make some functionality to draw that buffer to actual screen, to make
it more easier to support multiple cards with different settings.

- if that is not option, make sure when calculating position in screen to use formula (y * bytes in scan line) + x and _NOT_ currently used (y * width) + x. And then set appropriate window segment.

- DO NOT use set logical scanline length (some cards BIOS has bugs with it), instead use GET logical scanline length if needed and use that information when drawing to screen.

Thanks,
Vesa Jääskeläinen





reply via email to

[Prev in Thread] Current Thread [Next in Thread]