> OK, so admittedly I still don't understand what this patch is doing Me neither, but my guess is that upstream don't understand it too, since lock_flag is unused variable (greped in whole src tree):
./main/streams/plain_wrapper.c: int lock_flag; /* stores the lock state */ ./main/streams/plain_wrapper.c: self->lock_flag = LOCK_UN; ./main/streams/plain_wrapper.c: self->lock_flag = LOCK_UN; ./main/streams/plain_wrapper.c: self->lock_flag = LOCK_UN; ./main/streams/plain_wrapper.c: self->lock_flag = LOCK_UN; ./main/streams/plain_wrapper.c: data->lock_flag = value; So unless there is some hidden C magick, they just set it and never use its value. So I think the underlying logic for this patch was: "Ah, we don't need to set it to LOCK_UN here, since we are closing anyway." If it bothers you, somebody could write to Tony or Ilya and ask them (I'm not doing that). -- FFE for php 5.3.2 https://bugs.launchpad.net/bugs/539663 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
