Re: Scanning devices from inside the kernel
Well after a string of kernel panics, I spotted the following comment above the declaration of IORegistryIterator::getNextObject(): "The entry returned is retained while the iterator is pointing at it (its the current entry), or recursing into it. The caller should not release it." So be sure to remove the call to service->release() if you try out this code! I am surprised to see that in the Darwin source there are examples of an object returned by an iterator being accessed after the iterator has been released. If the iterator only retains the object whilst it is pointing at it, surely the caller would need to retain the object before releasing the iterator? See IOPMPagingPlexus::setAggressiveness for an example. Sam On Friday, May 30, 2003, at 15:00 Australia/Melbourne, darwin-kernel-request@lists.apple.com wrote: OSDictionary* services = IOService::serviceMatching(kIOMediaClass); if (!services) return; OSIterator* iterator = IOService::getMatchingServices(services); if (!iterator) return; while (true) { IOService* service = OSDynamicCast(IOService, iterator->getNextObject()); if (!service) break; printf("name: %s\n", service->getName()); // service->release(); NO! - see IORegistryIterator::getNextObject() } iterator->release(); services->release(); _______________________________________________ darwin-kernel mailing list | darwin-kernel@lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.
participants (1)
-
Sam Vaughan