Re: Bug report - svgalib 1.9.11 - r128 driver

Search this archive.

From: Matan Ziv-Av (matan@svgalib.org)
Date: Fri 28 Sep 2001 - 10:03:16 IST


On Mon, 24 Sep 2001, Andrea Mazzoleni wrote:

> svgalib 1.9.11 - file r128.c
> 
> The registers "amcgpio_en_reg" "amcgpio_mask" are used not initialized.
> They are saved and restored correctly, but not initialized before the mode
> set. They are used as they are after the allocation :

Thanks for this and the other bug reports:

The r128 bug that you finds seems to be the cause for the long standing
"sync lost" problem. The driver works fine after adding the lines 
 
    save->amcgpio_en_reg   = INREG(R128_AMCGPIO_EN_REG);
    save->amcgpio_mask     = INREG(R128_AMCGPIO_MASK);
 
In the function R128InitCommonRegisters(). Michal, can you check if it
also works for you?


The g400 bug is actually a problem, in that I don't want to depend on
the bios. There is a way to probe the clock without depending on the
bios, but it might be dangerous. Meanwhile I changed the default to
27MHz, since the G200 doc I have only mentions this possible reference
frequency. 


The trident bug - I added the ids to the range, but the driver is still
checked only on one 9685 card, so I don't know if it works on
newer/older cards.


The vgaregs bug - It does not affecr standard vga modes, since they set
all registers from a saved array, and it should not affect super vga
modes, since the driver should disable overscan (if possible), or set it
to black using this register or extended registers, so drivers can not
depend on the value of this register.


-- 
Matan Ziv-Av.                         matan@svgalib.org





------------------------------------------------------------------
Unsubscribe:  To:   listbot@svgalib.org
              Body: unsubscribe linux-svgalib


This archive was generated by hypermail 2.1.4 : Wed 21 Jan 2004 - 22:10:24 IST