Oleg Nesterov <oleg@xxxxxxxxxx> writes: > On 03/25, Eric W. Biederman wrote: >> >> So looking quickly the flip side of the coin is gdb (and other >> debuggers) needs a way to know these threads are special, so it can know >> not to attach. > > may be, > >> I suspect getting -EPERM (or possibly a different error code) when >> attempting attach is the right was to know that a thread is not >> available to be debugged. > > may be. > > But I don't think we can blame gdb. The kernel changed the rules, and this > broke gdb. IOW, I don't agree this is gdb bug. My point would be it is not strictly a regression either. It is gdb not handling new functionality. If we can be backwards compatible and make ptrace_attach work that is preferable. If we can't saying the handful of ptrace using applications need an upgrade to support processes that use io_uring may be acceptable. I don't see any easy to implement path that is guaranteed to work. Eric
