Cedric Le Goater [clg at fr.ibm.com] wrote: | comments below, | | > autofs_hash_nuke(sbi); | > for ( n = 0 ; n < AUTOFS_MAX_SYMLINKS ; n++ ) { | > if ( test_bit(n, sbi->symlink_bitmap) ) | | too many white spaces. may be you could fix those at the same time ? There are too many of those errors in the autofs/autofs4 files. I am moving all those changes to a separate patch so the main patch stays small. Let me know if thats a bad idea. | > + if ( parse_options(data, &pipefd, &root_inode->i_uid, | > + &root_inode->i_gid, &sbi->oz_pgrp, &minproto, | > + &maxproto) ) { | | I would keep the old prototype of parse_options | | > printk("autofs: called with bogus options\n"); | > goto fail_dput; | > } | | and do the find_get_pid() here. it looks simpler for me. Agree. | > + DPRINTK(("autofs_lookup: pid = %u, pgrp = %u, catatonic = %d, " | > + "oz_mode = %d\n", pid_nr(task_pid(current)), | > + process_group(current), sbi->catatonic, | > + oz_mode)); | | | you can keep current->pid I think | I am not sure. If a process unshares its pid-ns and then used to automount in the child ns, won't we print the wrong pid ?