To add to my previous report: of the workrounds described above, the 11n_disable suggestion does absolutely nothing for me. However, pcie_aspm=off as a boot option does greatly reduce the extent of the problem. (Sadly it doesn't eliminate it completely -- I just had a repeat of the problem apparently triggered by moving from mains to battery power. But that's the first time in a week of use since I made the change to the command line.)
I'd suggest therefore that the recent change to the kernel in stable (which I believe to be the only plausible explanation for my experience) is one that affects the PCIe power saving controlled by this command line option. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org