Red Hat USERS:
On Sun, 14 Jun 1998, Lawrence Houston wrote:
> This week I upgraded a RH 4.2 Installtion with Red Hat's recommended
> XFree86 3.3.2-1 for Intel and now XDM is failing to start with Systems's
> "onboard" S3 Trio64V2/DX. The Trio64V2/DX had been working with S3 Server
> included in 3.3.1 (I understood the V2 variety of the Trio64 had been
> unsupported with the XFree86 3.2 included in RH 4.2 the Distribution)!
> Symptoms: VTs 1-7 (including the one for the X-11 Server) are filled with
> virtical lines of junk and the XDM Log reports:
>
> xdm error (pid 280): Hung in XOpenDisplay(:0), aborting
> xdm error (pid 280): server open failed for :0, giving up
> xdm error (pid 275): Display :0 cannot be opened
>
> Found XFree86 3.3.1-1.2 on an FTP Site in Japan (all the Mirrors I tried
> and my own copies had overwritten with 3.3.2-1) and the Trio64V2/DX
> appears to "happy" starting XDM at boot time (AGAIN)?
>
> Might there a "bug" in XFree86 3.3.2's X3 support for the Trio64V2/DX
> Chipset, or is there something "new" in the mix since this version was
> built under 2.0.34 and is now being wrapped with "Xwrapper" instead of
> "xserver-wrapper". I am NOT comfortable moving away from 3.3.2 since it
> addressed some Security Issues, but 3.3.1 is the presently the ONLY way
> this baby will "fly"!
Following up to my own post:
My original observations were NOT entirely accurate, downgrading XFree86
had NOT cleared the problem with XDM failing to start at boot time.
Whether using XFree86 3.3.1 or 3.3.2 the S3 Server would get "stuck"
during the initialization process, pulling just under 100% of the CPU's
Time (a "slower" Linux still being available for remote sessions)! The
XDM Error Messages would NOT always be recorded, although CPU getting
"stuck" at 100% happened in all cases.
The computer in question is a 200MMX HP Vectra which dual boots between
WIN95 OSR2 and RH 4.2. Although this "problem" appeared within days of
the XFree86 3.3.2 Security Upgrade being applied, presently it looks like
a warm boot from WIN95 to RH 4.2 might have left the Trio64V2/DX in a
state XFree86's S3 Server could NOT initialize! Unfortunately these HP
Vectras have NO Reset Button, hence the ONLY way to get a non-warm reboot
is to cycle the power. After which RH 4.2 with the XFree86 3.3.1 and
3.3.2 S3 Servers started correctly (and run without hogging the CPU).
Might the WIN95 OSR2 Driver (or possibly Starnet's V4.0 X-WIN32 Server)
have left the Trio64V2/DX Chipset in a state which the XFree86's S3 Server
could NOT handle and looped endlessly attempting to do so? I've had
similar problems with my SB16 Sound Card, on occasion failing to be
detected when Linux is warm booted from a Windows 3.X/95 Session (cold
boot or reset are often requried)!
Lawrence Houston - ([EMAIL PROTECTED])
--
PLEASE read the Red Hat FAQ, Tips, Errata and the MAILING LIST ARCHIVES!
http://www.redhat.com/RedHat-FAQ /RedHat-Errata /RedHat-Tips /mailing-lists
To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject.
