On Thu, 16 Apr 2015 11:14:11 +0300 Konstantin Khlebnikov <koct9i@xxxxxxxxx> wrote: > Keeping pointer to priviledged task is a bad idea. > There is no easy way to drop it when task exits and this doesn't work > for threads. > I think it's better to keep pointer to priveledged struct file and > drop it in method > f_op->release() when task closes fd or exits. Server task could obtain second > non-priveledged fd and struct file for that inode via > open(/proc/../fd/), dup3(), > openat() or something else and send it to read-only users. Thank you, I was hoping someone would suggest a different authentication method, I will look into this idea. What is the thread concern? I have not run in to any problems yet while testing, but have been more focused on getting my user space memfd transport daemon up and running before I put it through the torture test. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>