> Btw, is it allowed to use utrace_control(UTRACE_DETACH) from ->report_any() ? > If yes, then we need a bit more fixes...
It's kind of a stupid thing to do, but yes, it should be allowed. I think it's fine either if it always returns -EINPROGRESS or if it returns 0 for the target==current case. It doesn't have to be handled optimally, but it should have about the same semantics as returning UTRACE_DETACH from the callback. (The differences being that later utrace_* calls inside that callback could return -ESRCH.) Thanks, Roland
