... literally. The X display immediately corrupts itself on the Disk Druid
setup screen. Chunks of the display are repeated randomly across the top
half of the screen. The installer still functions, but it is near
impossible to make out anything remotely resembling the installer unless
you continually click the Show Help/Hide Help button, which redraws the
window enough so that you can see it. Also, there is black space along the
edges of the window where there shouldn't be.
Machine is a IBM Intellistation M Pro model 6889-78U, 128MB, 4GB SCSI
drive, aic7xxx SCSI adapter (the on-board 2940UW, I forget the
designation). The X server in use is SVGA, utilizing the s3_savage driver.
Also, on the first virtual console, I have this error repeated 3 times:
Gtk-CRITICAL **: file gtkaccelgroup.c: line 188 (gtk_accel_group_attach):
assertion 'g_slist_find (accel_group->attach_objects, object) == NULL' failed
I don't know if that is related or not. The machine is available for
examination at your convenience.
This might be related ....
I am having problems with an S3 Savage4 too. The machine is an IBM 300GL
workstation. Installed in text mode over ftp so I didnt have the installatin
problem. The card and monitor got detected accurately. Selected 1024x768
resolution with 24 bit color. X starts up with a gdm screen correctly. But
display does seemingly "blow up into chunks" when I log in (right after the
"gnome hint" appears) tried reducing the resolution and color depth. Didnt help.
The only way I found to make the system usable was turn of X accelaration with
Option no_accel in XF86config. But of course the performance is then very very
bad. I had no similar problems with Mandrake 7.1 that was running on the system
I also got the video corruption after booting up and logging in. I used:
Under the S3 Savage4 (generic) Graphics device section in XF86Config, and that
worked, similar to keeda's fix (does it matter, no_accel or noaccel?).
Performance, of course, is not that great now.
Ruminating, I'd wager this is related to running XFree 3.3.6 against the 4.0
No this will be a server problem. You can however test if its a problem with the
3.x or 4.x server by
running Xconfigurator and choosing the server you wish (see man Xconfigurator).
It may be there is a bug in the 3.x or 4.x server we ship.
AFAIK there is no support for Savage4 in 4.x yet. Xconfigurator seems to be
doing the right thing. The 3.3.6 server is being used. I have never had problems
with the 3.3.6 server and this card earlier.
Yep, the server seems to be the problem ....
Copied the /usr/X11R6/bin/XF86_SVGA binary over from another machine (debian
actually, I know I'm not supposed to do that :-) ) And all problems seem to have
Also, bug 17991 seems to be a duplicate of this
Thanks for the info. That implies we picked up a patch along the way that broke
this card (and probably fixed something else in the process) or that we have a
build error. Glad to know the binary copy solved it for you.
I just want to add that I have the same problem... for the stats.... and
I am also seeing this problem on machines with cards detected as Savage 3D.
Among the set of machines in use, some will not show any graphics mode
video. others show the chunky behavior described. the cards with 1MB vram
give chunks. the cards with 8MB vram crash and burn completely (Xconfigurator
fails to start the X server). Different variations on the card type were
with no success.
Please refer to the following link for a workaround for these cards:
Fix for Savage should be ready soon if it checks out in our labs.
*** This bug has been marked as a duplicate of 17991 ***
I have the same problem, but I have not been able to solve it using any of the
two methods specified in the gotchas-7-4.html#ss4.7
SuperProbe tells me I'm running the following card:
S3 ViRGE/GX (PCI Probed)
Ramdac: Generic 8-bit pseudo color DAC (with 6-bit wide lookup tables (or in 6-
I'm a complete "newbie" in this Linuxworld, so I could indeed be doing
Will it help upgrading from 7.0 -> 7.1???