Hi everyone,<br>
<br>
I'm running Fedora 8, kernel 2.6.24.3-50.fc8. I was experiencing<br>
frequent kernel hangs, so to get to the bottom of it I setup kexec/<br>
kdump with this kernel. I did it according to the documentation in<br>
<kernel_src>/Documentation/<div id=":ai" class="ii gt">kdump/kdump.txt. I'm running Xorg with the<br>
closed-source binary nvidia driver. I did the following:<br>
<br>
1) Ensured that my running kernel supported kexec (as given in the<br>
documentation above), obtained the vmlinux for it from the fedora<br>
debug yum repos.<br>
2) Built a new crash-dump kernel (linux-2.6.23.17-kdump).<br>
3) Appended the following lines to my /etc/rc.sysinit:<br>
<br>
921 echo "1" > /proc/sys/kernel/sysrq<br>
922 savetextmode<br>
923 # Tell kexec which kernel and ramdisk to use<br>
924 kexec -p /usr/src/kernels/linux-2.6.23.17/vmlinux --initrd=/boot/<br>
initrd-2.6.23.17-kdump.img --args-linux --append="root=LABEL=/ 1<br>
irqpoll maxcpus=1"<br>
925<br>
926 DATE=`date +%Y-%m-%d-%T`<br>
927 NAME=`uname -r`<br>
928 if [ "$NAME" == "2.6.23.17-kdump" ]; then<br>
929 echo -e "Creating crash dump..\n"<br>
930 mkdir -p /var/crash/127.0.0.1-$DATE<br>
931 cp /proc/vmcore /var/crash/127.0.0.1-$DATE/vmcore-incomplete<br>
932 if [ $? == 0 ]<br>
933 then<br>
934 mv /var/crash/127.0.0.1-$DATE/vmcore-incomplete /var/crash/<br>
127.0.0.1-$DATE/vmcore<br>
935 echo -e "Done, rebooting..\n"<br>
936 reboot -f<br>
937 fi<br>
938 fi<br>
<br>
The above is supposed to dump the vmcore to my /var/crash/ directory<br>
as soon as the kdump kernel boots after a hang. The motivation for<br>
doing this was that I get kernel hangs while running X, and pressing<br>
<Alt>+<Sysrq> causes the screen to get garbled when the kdump kernel<br>
boots, so I don't have a chance of doing this manually.<br>
<br>
I've tested the above while in runlevel 3 while I'm inside a VT,<br>
everything works as expected (I see the kdump kernel boot up and copy<br>
the kdump), and I can see the vmcore end up inside /var/crash/. But<br>
when I try the same while X is running, the screen gets garbled while<br>
the kdump kernel boots, but nothing happens! The kernel does not<br>
reboot, and neither does it copy the vmcore to /var/crash. It is<br>
responsive though, I can toggle the CapsLock key, but the screen is<br>
blank.<br>
<br>
Is it something in the nvidia driver that's causing this to not work?<br>
Any clues on what's going wrong, or how I can debug this further?<br>
<br>(Please CC me in the reply, I'm not on this list)<br><br>
Thanks!<br>
-Mayank</div><br>