Alright - there's a known regression affecting only pae-kernels, where 2.6.32-31.61 is good and 2.6.32-32.62 is bad (see bug #805209). You said that "any kernel later" than 2.6.32-30.59 is causing this problem - was 2.6.32-31.61 part of your tests?
Alright - there's a known regression affecting only pae-kernels, where 2.6.32-31.61 is good and 2.6.32-32.62 is bad (see bug #805209).
You said that "any kernel later" than 2.6.32-30.59 is causing this problem - was 2.6.32-31.61 part of your tests?