Comment 19 for bug 1258837

Just based on the last test that I did, it would seem that the drivers in use in kernel 3.5.0-31 are different to the generic mouse ones in kernel 3.5.0-30 (looking at the output from xinput) - rather than that a bug was introduced into the same set of drivers.

If the generic mouse drivers work and the Alps-specific drivers cause problems - I'm not sure how it would help to bisect the commits; you would either have to revert to the generic drivers as a solution or try to debug the Alps driver without having a previous working version of those drivers as a starting point for investigation.