| From: Srikanth Konjarla <srikanth.konjarla@xxxxxxxxx> | I have similar kind of freeze issues. However, i could ssh into the box and | run gdm-restart to bring back the sanity to the frozen laptop. So, this leads | me to think that i have issues with nvidia proprietary driver (beryl is | running too). I have a laptop (Tecra M5) with Core 2 Duo, 3G RAM with nvidia | Quadro NVS 110M video Card. | | D. Hugh Redelmeier wrote: | > Your symptoms are non-specific: lots of diseases have the same | > symptom. So you need more information. Srikanth Konjarla: Your problem must be very different from mine since the ssh server is still working. This is good news for you: your kernel is not hung, but some process or processes are in a bad state. Once you ssh in, you have a lot of tools available to you for figuring out what is wrong. ps(1) can tell you lots of things about running processes. top(1) can tell you what is taking the most CPU (sometimes a hang is in the form of a process caught in a cpu-bound loop). If you figure out what process is hanging, you can even use gdb on it. As an extreme example of the usefulness of ssh: last month I had a problem with ndiswrapper. In the end, I let the author of ndiswrapper ssh into my machine and he fixed the code! He then checked the fix into the official source so future versions should have that bug fixed for everyone. A true kernel hang leaves you with a much poorer choice of tools. That is the case that my previous message focused upon. (As you know, official Fedora channels are not too supportive of proprietary drives. For understandable reasons. There seems to be an nVidia driver person who pops up on this list every once in a while.)
