> On Wed, Jun 25, 2008 at 18:59, Axel Dörfler <axeld@xxxxxxxxxxxxxxxx> > wrote: > > Exactly -- someone has to brand that semaphore in the same way that > > someone has to choose the thread semaphore; there is just no > > advantage. > > Uh, I guess I didn't make myself clear... > > On my first message I wrote: "We could use this knowledge to boost > processes that we *know* to be interactive, because they blocked on a > locking primitive that was *declared on a code path that has to do > with interactivity*. [...] Not to mention that there's zero code > overhead except for an extra *private* call that creates a branded > semaphore" (emphasis added) > > Meaning: I'm talking about branding semaphores in the app_server, > input_server and the like, not application code. Then what about a semaphore flag on *_sem_etc() to activate priotity inheritance? Not all of them will want to do that anyway. François.