On Tue, Jan 27, 2004 at 10:00:15PM -0500, Kevin E Martin wrote:
>On Tue, Jan 27, 2004 at 01:07:00PM -0500, David Dawes wrote:
>> Does anyone have a reason to not apply the patch attached to
>> <http://bugs.xfree86.org/show_bug.cgi?id=935>?
>
>I don't think this patch is correct. What it does is enable the CRTC
>while writing the new mode -- generally that is bad. I would suggest
>either (1) removing the "R128_CRTC_VSYNC_DIS | R128_CRTC_HSYNC_DIS" from
>the R128RestoreCrtcRegisters code or (2) add those bits to the
>R128Unblank code. #1 will change the behavior from putting the monitor
>in suspend mode during the CRTC changes to just blanking the screen,
>while #2 will automatically enable the H&V sync signals when the screen
>is unblanked after a mode switch. Both should be safe.
Thanks Kevin, I'll commit option 2:
Index: r128_driver.c
===================================================================
RCS file: /home/x-cvs/xc/programs/Xserver/hw/xfree86/drivers/ati/r128_driver.c,v
retrieving revision 1.87
diff -u -r1.87 r128_driver.c
--- r128_driver.c 10 Nov 2003 18:41:21 -0000 1.87
+++ r128_driver.c 29 Jan 2004 02:41:02 -0000
@@ -500,7 +500,9 @@
if(info->isDFP)
OUTREGP(R128_FP_GEN_CNTL, 0, ~R128_FP_BLANK_DIS);
else
- OUTREGP(R128_CRTC_EXT_CNTL, 0, ~R128_CRTC_DISPLAY_DIS);
+ OUTREGP(R128_CRTC_EXT_CNTL, 0, ~(R128_CRTC_DISPLAY_DIS |
+ R128_CRTC_VSYNC_DIS |
+ R128_CRTC_HSYNC_DIS);
}
/* Compute log base 2 of val. */
David
--
David Dawes
developer/release engineer The XFree86 Project
www.XFree86.org/~dawes
_______________________________________________
Devel mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/devel
