Somebody in the thread at some point said: > I figure that, for the next kernel update, I'll manually remove the > kernel with the problem, and do the update. The known good kernel should > be the second kernel, and therefore retained. I can then test the > updated kernel. > > Does that sound right? Sounds good, and you have the opportunity to check what yum plans before okaying it as well. > Does it make sense to keep testing new kernels, even if there's been no > specific response to this issue? Yes, definitely. The upstream kernel folks are constantly fixing and meddling with stuff that could cause this. Even if the changelog doesn't mention e100, for example maybe this problem is actually provoked by something in the detail of Interrupt routing or some other subsystem. So I would keep trying new kernels. -Andy
