On 1018643803 seconds since the Beginning of the UNIX epoch
Graham Allan wrote:
>

>I've never used an Alphastation 600 specifically, but some alphas have
>a "tga_sync_green" SRM console variable you can set.

Ah, yes, I remember that from my AS 200/4 233 and multias.  I do
not think that those settings control the workings of the TGA2---since
the SRM doesn't initialise the card.  But, I was thinking that it
would be an interesting idea for the tga(4) driver to check for
the presence of that SRM var (only on alpha, probably) and set the
sync-on-green bit in the RAMDAC appropriately.

Right now the tga(4) driver does not force sync-on-green on the
TGA2 and in fact cannot change the sync-on-green setting, so I
think that this probably isn't the problem (I just scanned the src
for that one).  But, it does sound similar to the symptoms described.

Although, perhaps the card is defaulting differently than mine and
the sync-on-green is on by default..  Perhaps the enclosed patch
applied to src/sys/dev/ic/bt485.c will help?  (You'll need to
rebuild the kernel.)

 == Roland Dowdeswell                      http://www.Imrryr.ORG/~elric/  ==
 == The Unofficial NetBSD Web Pages        http://www.Imrryr.ORG/NetBSD/  ==
 == The NetBSD Project                            http://www.NetBSD.ORG/  ==
Index: bt485.c
===================================================================
RCS file: /cvsroot/syssrc/sys/dev/ic/bt485.c,v
retrieving revision 1.8
diff -u -u -r1.8 bt485.c
--- bt485.c	2001/12/12 07:47:46	1.8
+++ bt485.c	2002/04/12 21:04:19
@@ -210,6 +210,10 @@
 	 * resolution.
 	 */
 	regval |= 0x02;
+	/*
+	 * Force the RAMDAC out of sync on green mode.
+	 */
+	regval &= ~0x08;
 	data->ramdac_wr(data->cookie, BT485_REG_COMMAND_0, regval);
 
 	/* Set the RAMDAC to 8BPP (no interestion options). */