Kernel upgrade followed by Palm sync freezing X?
Could that be more convoluted?
In order to fix my already-posted-about problem with ripping CD's, I
finally upgraded from kernel-image-2.6.7-1-k7 to
kernel-image-2.6.8-1-k7. It worked -- paranoia can read CD's again.
Unfortunately, the first time I ran jpilot, X froze solidly. No
mouse movement, no reaction to any keyboard input including
ctrl-backspace, ctrl-alt-Fn, and ctrl-alt-del, no nothing. I had to
use the reset button.
Same thing the SECOND time I ran jpilot.
As long as I didn't actually try to sync, no problem. That is, the
program can sit open as long as it doesn't try to communicate with
the Palm device.
So I tried kpilot. Same thing. It took longer to freeze -- over a
minute vs. seconds. (The sync was long because it was the first time
kpilot had run, so it had to do a full backup.) By this time I had
installed telnetd so I could connect via my laptop and just kill X
rather than reboot. Except I couldn't. Even kill -9 wouldn't stop
either X or xdm.
Kpilot froze during an AvantGo sync. If I use malsync (command-line
AvantGo sync) in a VT, no freeze.
It's clearly a bug, but I'm torn between reporting it for XFree86 and
xPilot. No application should be able to freeze X so badly it can't
be killed. (At all, really, but surely not so badly it can't be
killed.)
I tend to think it isn't a USB bug (I use USB sync) because my USB
mouse and ifp continue to work fine with the new kernel, but I
welcome other opinions.
In fact, I welcome any advice on how to further debug this before
bothering the maintainers with possibly-misaimed bug reports.
BTW, why did I use telnetd? Because I can't seem to make a ssh
connection to my tower from the laptop. I just get refusals from my
laptop ssh client, while the "VERBOSE" logging on the tower shows
nothing whatever after the "connect from" message. PAM problem? Who
knows?
I thank you for reading this far.
--
Carl Fink carl@fink.to
Jabootu's Minister of Proofreading
http://www.jabootu.com
Reply to: