Messages in this thread | | | Date | Fri, 21 Mar 2008 10:27:34 +0100 | From | Stefan Richter <> | Subject | Re: use of preempt_count instead of in_atomic() at leds-gpio.c |
| |
I wrote: > Andrew Morton wrote: >> ./drivers/ieee1394/ieee1394_transactions.c >> >> Possibly buggy: deadlockable > > That's in hpsb_get_tlabel(), an exported symbol of the ieee1394 core. > > The in_atomic() there didn't cause problems yet and is unlikely to do so > in the future, because there are no plans for substantial changes to the > whole drivers/ieee1394/ anymore (because of drivers/firewire/). > > Nevertheless I shall look into replacing the in_atomic() by in_softirq() > or something like that.
Or extend the API to have separate calls for callers which can sleep and callers which can't. But that may be thwarted by deep call chains.
> Touching this legacy code is dangerous though. -- Stefan Richter -=====-==--- --== =-=-= http://arcgraph.de/sr/
| |