Exactly, well, not exactly, but I'd give it a >95% correlation that X on fgconsole when rmmod e1000 is typed will cause a panic shortly thereafter, noting that the rmmod itself does complete successfully and exits, it's a short time afterwards (long enough for the suspend scripts to chvt typically) that the panic occurs. The chvt does not have to happen to cause a panic, but it can be acheived. My first workaround was in fact just to change vts earlier in the suspend process. Someone seemed to indicate they at least once managed to suspend even without a workaround, so it might not be 100%.
I have not seen the panic at all myself after rmmod e1000 if the current VT is text. See my third comment near the beginning. -- IBM T41,42 Regression: Kernel Panic for Suspend/Hibernate https://launchpad.net/bugs/62196 -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
