Re: New X problems...
Yes, I have no NFS, I have no NFS today... <sung to "yes we have no
bannanas"> ;-)
Ah, the joys of living under an IT department. I hear tales from my son
all the time. He works in development and has to defeat a lot of IT stuff
that keeps finding its way onto his machine every time he docks it at
work. (Better him than me ;-)
My xterm and bash windows come up just fine (don't crash like yours), but
they don't have any content. No prompt, no cursor, nothing but black.
With mozilla I could log into http://www.linuxbase.org/spec/ just fine,
but I couldn't download anything. The dialog box for choosing an alternate
file name and saving it, comes up just fine, but when you click on OK, it
reports "Connection refused ..."
After bumping into never before seen problems with an installation attempt
(I have other issues I'd like to resolve as well as this one) I did the
more obvious thing, and reverted to my previous 2.2.17 kernel.
Everything works peachy keen fine!
David, from your comments below, you seem to be using kpkg, while I use
make. I got my 2.2.19 source from ftp.kernel.org, and, after a bit of
unusual bother untarring it, did 'make mrproper ; make menuconfig'. At
this point I had to go get ncurses4-dev ... somehow it failed to upgrade
last time...
Anyway, I usually work through every option in the config menu. I like the
menuconfig option because it is easy to go back and forth between various
areas of configuration. I make as much as I can with modules (keeping only
enough to boot with as "built in"), and these days almost always have to
go back at least once to trim some fat out so I can get a zImage. (this
appears to be just a perverse bit of stubbornness on my part ;-)
So it looks like I need to go back and look at some of that "fat" I
trimmed and see what happens with a less anorexic kernel. I was just
hoping that someone would see the symptoms, and say "Oh, those things
depend upon bla-bla-bla so that's where the problem is." But I guess I'll
just have to put in the time...
David, see additional comments below:
On Thu, 26 Apr 2001, David A. Greene wrote:
> Dale Scheetz wrote:
>
> > Now, when I bring up and xterm or a bash window, I get no cursor and no
> > keystrokes appear in the window. I'm also having very flaky problems with
>
> I have had problems with this as well. I tracked it down to
> NFS/amd not working properly. This has been a problem on my
> laptop for about a year now but NFS is not so critical that
> I've had time to go fix the problem. It seems odd, though,
> that maps provided by our IT department don't work on my
> woody laptop, but work on other machines on the network (running
> potato). I'm guessing this is a configuration problem on my end,
> though. In any event, disabling amd solved it for me.
I don't use NFS myself, but from what others have said it is not the most
robust system because of various issues. YMMV
>
> Hmm...After re-reading your message, I think I misunderstood you.
> My problem is that the terminal session hangs (nothing is displayed
> at all, and no, there are no references to NFS mounts in my
> startup files). It sounds like you're having strictly input
> problems.
Well, we may actually have the same problem for different reasons. It
sounds like your symptoms match mine, although I'm not sure I'd use the
word "hangs", as I can close the window. There is just nothing in the
window, and not action at the keyboard changes that...
Switching kernels fixes it for me, so this isn't an X config issue.
>
> > mozilla not being able to get into simple web pages and download material.
> > (I'm trying to get the current LSB spec)
>
> Now _this_ I just recently (i.e. today) started seeing more
> often. My laptop (Dell Latitude CPx/J) has always had problems
> with the Vortex card after a BIOS resume. For one thing, the
> routes take forever to come back (i.e. 'route' hangs for a
> minute or more before being able to resolve to DNS). Today,
> however, I saw something new. I dist-upgraded testing and
> rebooted with a freshly-compiled 2.2.17 (no configuration changes,
> just a kpkg --revision change so it doesn't conflict with
> the standard packages). After a BIOS suspend/resume cycle,
> the network was extremely flaky. As usual, route took forever.
> Once I was able to ping a host, I left it running and used mozilla
> to access the network through both the WWW and IMAP clients.
> In both cases, as soon as mozilla took any action, the network
> was lost and the pings stopped going through.
>
It sounds like you have a combination of complex local net that you must
"get out of", and possibly some flakey hardware. You also may not have
power management properly configured in your kernel.
> Strangely enough, I just now resumed the machine again and everything
> seems fine. In general, the Vortex/Cardbus/SOMETHING has always
> been very flaky for me, but today was the first time I could
> _consistently_ bring the network down. Until I suspended/resumed
> again, of course. :(
>
Sounds more and more like either some collision interaction with your
local net, or flakey hardware.
> > At this point I can't tell whether this has something to do with my new
> > kernel, or is caused by something else. I'm pretty sure that once I got
> > things working the last time, I used an xterm with no problems, so my
> > suspicions lie heavily with the new kernel.
>
> It's suspicious that I had a change in behavior after a kernel
> recompile as well, but since I didn't change any configurations,
> I'm assuming it lies with something in testing. I suspect it
> might be with the new (for me, at least) pcmcia-cs.
>
Well, pcmcia-cs has always installed ok on my compaq laptop, so if you
used the Debian provided modules, that probably isn't it. (remember, I'm a
sample of one ;-)
I remember back in the dark ages ;-) when one release of a kernel would
fail on feature (for me it was an IDE controler), the next release it
would work (but something else was busted), and the following release the
controler issue was fragged again. This went back and forth for several
releases because it was a poorly understood race condition. This was
finally fixed and that particular controler code has been stable ever
since.
Sorry for the ramble, but I think at least some of your problem may be
related to an actual kernel bug, but you need to fiddle a bit more.
Luck,
Dwarf
--
_-_-_-_-_- Author of "Dwarf's Guide to Debian GNU/Linux" _-_-_-_-_-_-
_- _-
_- aka Dale Scheetz Phone: 1 (850) 656-9769 _-
_- Flexible Software 11000 McCrackin Road _-
_- e-mail: dwarf@polaris.net Tallahassee, FL 32308 _-
_- _-
_-_-_-_-_- Released under the GNU Free Documentation License _-_-_-_-
available at: http://www.polaris.net/~dwarf/
Reply to: