vgacon disable again

Hi,

The much easier fix is this one ;)

cheers,
Gerd

Gerd Hoffmann <[email protected]>

---
 drivers/video/console/vgacon.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Index: paravirt-2.6.20-hg749/drivers/video/console/vgacon.c
===================================================================
@@ -372,7 +372,8 @@ static const char *vgacon_startup(void)
 	}
 
 	/* VGA16 modes are not handled by VGACON */
-	if ((ORIG_VIDEO_MODE == 0x0D) ||	/* 320x200/4 */
+	if ((ORIG_VIDEO_MODE == 0x00) ||	/* not set */
+	    (ORIG_VIDEO_MODE == 0x0D) ||	/* 320x200/4 */
 	    (ORIG_VIDEO_MODE == 0x0E) ||	/* 640x200/4 */
 	    (ORIG_VIDEO_MODE == 0x10) ||	/* 640x350/4 */
 	    (ORIG_VIDEO_MODE == 0x12) ||	/* 640x480/4 */

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.osdl.org/mailman/listinfo/virtualization

Oh, and it obsoletes vgacon-fix-zero-size.patch btw ...

cheers,
Gerd

Gerd Hoffmann <[email protected]>

Well, I think the zero-size allocation oops is a bug anyway. But is it really necessary to disable VGA beyond that?

J

Yes, it is. Otherwise it oopses as soon as something is printed on /dev/tty<any>. Quite early at boot with "console=tty1", otherwise as soon as /sbin/getty starts on /dev/tty1. Oh, and the reason is it thinks it has an EGA card with 0 lines and 0 columns because we never ran the real-mode boot code and thus SCREEN_INFO isn't initialized. The 0x0 screen probably leads to the zero-sized buffer you are catching in that patch.

cheers,

Gerd

Gerd Hoffmann <[email protected]>

OK, I'll add your other patch to the series.

J