On Sat, Dec 21, 2013 at 12:42, Mark Kettenis wrote: > As Ted pointed out, the size of the struct (or union) becomes part of > the ABI. So we should probably reserve a bit more space than we > actually need for the implementation. That way, we'll leave some room > for experimentation. Wasting a bit more memory shouldn't be a big > issue as typical code that uses semaphores won't be constructing > thousands of these. And rounding up to the nearest power of two > should be pretty much free.
Something I've also belatedly realized is we are in effect already exposing internals via sem_open. Two programs with two different majors of libpthread.so can open the same named semaphore, but it may be different internally. We should add a version number inside the semaphore to prevent that from happening.
