2010/12/29 Philippe Houdoin <philippe.houdoin@xxxxxxxxx>: >> Precedence wasn't the issue here, though. >> The semantics of the post-increment operator, which always returns the old >> value, was. The pre-increment operator would have worked. > I'm *so* sorry about this one. > Let's hope it doesn't explain any ACPI issues we had since long... > > There is the same broken code with queue->head in dpc_thread(), BTW. It was fixed in the commit after. No need to be sorry, I'd probably do a worse job if I wrote the DPC-module. /Fredrik Holmqvist, TQH