Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
XF86_SVGA -quiet fails to find PCI base address, XF86_SVGA works
4 views
Skip to first unread message
Wolfgang Molzer
Nov 9, 1998, 3:00:00 AM11/9/98
to
Hi,
I use a Scenic Mobile 700, i.e. SIEMENS NIXDORF Laptop with S.u.S.E Linux 5.3.
HW is a intel Pentium 166MHz MMX, 32MB, 2.1 GB HD, 20x CD-ROM, Synaptics touchpad ...
The graphics HW is Chips & Technologies 65550 2MB; 800x600, 11.3' TFT.
S.u.S.E. 5.3 includes XFree86 3.3.2.3.
I have succeeded in installing XF86_SVGA (package xsvga, series x) using SaX.
Some notes about my configuration can be found in the appendix.
Here i report some strange behavior of the XF86_SVGA x server.
Using
startx
to start X seems working, see attached .X.err renamed to .X.noerr.
As i understand the sh program startx, it is in fact
X 0:
(or even more expanded
XF86_SVGA 0:
) via
xinit ~/.xinitrc
(the 'actual i.e. expanded command line' in the startx sh-program) that starts the X server.
Using
startx -- -quiet
i.e. in fact
XF86_SVGA 0: -quiet
the X-server does not find the memory where to display, at least this is how i interprete the error message in .X.err.
I found this when tracking down the problem, why xdm continiously tries and fails to start the X server and leaves me with nothing else working but Ctrl-Alt-Del to reboot.
After having read the approriate man-pages, scanned through the files, i found the following line in <XRoot>/lib/X11/xdm/Xservers -- being responsible for the xdm version of starting the X server
:0 local /usr/X11R6/bin/X :0 vt07 -quiet
Testing the various combinations for the arguments to the X server with startx as above, i found:
-quiet is the source for the problem.
Has anybody else seen this behavior?
Can it be reproduced on other machines, e.g. using c&t65550 etc.?
A bug in XF86_SVGA?
Leaving .../Xservers as it is but adding
# xdm doesn't like to run without it
MemBase 0xFD000000
in the Device-Section of the file /etc/XF86Config also leads to a working xdm.
I have take this address value right form ~/.X.noerr.
Appendix:
To get an acceptable X configuration I had to decelerate the 'default' dotclock setting in the 800x600 modeline.
('Default' means, i basically did what is recommended at http://www.suse.de/Support/sdb/maddin_notebook.html)
All 800x600 TFT-displays in the database of SaX, i.e. S.u.S.E X-configuration program, i found, used something around 60MHz.
This seems to be too fast for the TFT built in my laptop (or is there some other reason for the following problems).
The gray X start screen is flickering, a typical windows screen after startup (with any window manager i tried) also is very unquiet, i.e. some pixels change color/brightness, fonts near edges become invisible.
At 48MHz the situation looks much better, in fact pretty good.
As i don't know much about the display (still haven't got an answer from technical support at SIEMENS), i use VESA standard -- as found in the XFree86 Documentation -- with 40MHz which looks very much the same as with 48MHz.
Wolfgang Molzer
I use a Scenic Mobile 700, i.e. SIEMENS NIXDORF Laptop with S.u.S.E Linux 5.3.
HW is a intel Pentium 166MHz MMX, 32MB, 2.1 GB HD, 20x CD-ROM, Synaptics touchpad ...
The graphics HW is Chips & Technologies 65550 2MB; 800x600, 11.3' TFT.
S.u.S.E. 5.3 includes XFree86 3.3.2.3.
I have succeeded in installing XF86_SVGA (package xsvga, series x) using SaX.
Some notes about my configuration can be found in the appendix.
Here i report some strange behavior of the XF86_SVGA x server.
Using
startx
to start X seems working, see attached .X.err renamed to .X.noerr.
As i understand the sh program startx, it is in fact
X 0:
(or even more expanded
XF86_SVGA 0:
) via
xinit ~/.xinitrc
(the 'actual i.e. expanded command line' in the startx sh-program) that starts the X server.
Using
startx -- -quiet
i.e. in fact
XF86_SVGA 0: -quiet
the X-server does not find the memory where to display, at least this is how i interprete the error message in .X.err.
I found this when tracking down the problem, why xdm continiously tries and fails to start the X server and leaves me with nothing else working but Ctrl-Alt-Del to reboot.
After having read the approriate man-pages, scanned through the files, i found the following line in <XRoot>/lib/X11/xdm/Xservers -- being responsible for the xdm version of starting the X server
:0 local /usr/X11R6/bin/X :0 vt07 -quiet
Testing the various combinations for the arguments to the X server with startx as above, i found:
-quiet is the source for the problem.
Has anybody else seen this behavior?
Can it be reproduced on other machines, e.g. using c&t65550 etc.?
A bug in XF86_SVGA?
Leaving .../Xservers as it is but adding
# xdm doesn't like to run without it
MemBase 0xFD000000
in the Device-Section of the file /etc/XF86Config also leads to a working xdm.
I have take this address value right form ~/.X.noerr.
Appendix:
To get an acceptable X configuration I had to decelerate the 'default' dotclock setting in the 800x600 modeline.
('Default' means, i basically did what is recommended at http://www.suse.de/Support/sdb/maddin_notebook.html)
All 800x600 TFT-displays in the database of SaX, i.e. S.u.S.E X-configuration program, i found, used something around 60MHz.
This seems to be too fast for the TFT built in my laptop (or is there some other reason for the following problems).
The gray X start screen is flickering, a typical windows screen after startup (with any window manager i tried) also is very unquiet, i.e. some pixels change color/brightness, fonts near edges become invisible.
At 48MHz the situation looks much better, in fact pretty good.
As i don't know much about the display (still haven't got an answer from technical support at SIEMENS), i use VESA standard -- as found in the XFree86 Documentation -- with 40MHz which looks very much the same as with 48MHz.
Wolfgang Molzer
--
================================================================ Wolfgang Molzer Semiconductor Technology CAD engineer Mail: | Phone: +49 89 636 42232 SIEMENS AG | Fax : +49 89 636 45411 HL SIM | email: wolfgan...@siemens-scg.com D-81730 Munich | Germany | ================================================================
0 new messages