Comment 13 for bug 1041594

Interesting. So that means the bug was introduced in one of the 3.5 stable updates. That commit was probably reverted in 3.6-rc1.

We might be better off performing a regular kernel bisect on v3.5 to find the offending commit. We can then see how it was fixed upstream, and request the same fix in linux-stable upstream.

Can you test the following two kernels and post which is the first to have the bug:

v3.5.1: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.5.1-quantal/
v3.5.2: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.5.2-quantal/

Thanks again!