From: Oleg Nesterov Date: Tue, 3 Oct 2017 23:15:48 +0000 (-0700) Subject: exec: binfmt_misc: shift filp_close(interp_file) from kill_node() to bm_evict_inode() X-Git-Url: https://git.stricted.de/?a=commitdiff_plain;h=83f918274e4b;p=GitHub%2FLineageOS%2Fandroid_kernel_motorola_exynos9610.git exec: binfmt_misc: shift filp_close(interp_file) from kill_node() to bm_evict_inode() To ensure that load_misc_binary() can't use the partially destroyed Node, see also the next patch. The current logic looks wrong in any case, once we close interp_file it doesn't make any sense to delay kfree(inode->i_private), this Node is no longer valid. Even if the MISC_FMT_OPEN_FILE/interp_file checks were not racy (they are), load_misc_binary() should not try to reopen ->interpreter if MISC_FMT_OPEN_FILE is set but ->interp_file is NULL. And I can't understand why do we use filp_close(), not fput(). Link: http://lkml.kernel.org/r/20170922143644.GA17216@redhat.com Signed-off-by: Oleg Nesterov Acked-by: Kees Cook Cc: Al Viro Cc: Ben Woodard Cc: James Bottomley Cc: Jim Foraker Cc: Cc: Travis Gummels Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 6451e7520e05..203598ccb40a 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -594,8 +594,13 @@ static struct inode *bm_get_inode(struct super_block *sb, int mode) static void bm_evict_inode(struct inode *inode) { + Node *e = inode->i_private; + + if ((e->flags & MISC_FMT_OPEN_FILE) && e->interp_file) + filp_close(e->interp_file, NULL); + clear_inode(inode); - kfree(inode->i_private); + kfree(e); } static void kill_node(Node *e) @@ -606,11 +611,6 @@ static void kill_node(Node *e) list_del_init(&e->list); write_unlock(&entries_lock); - if ((e->flags & MISC_FMT_OPEN_FILE) && e->interp_file) { - filp_close(e->interp_file, NULL); - e->interp_file = NULL; - } - dentry = e->dentry; drop_nlink(d_inode(dentry)); d_drop(dentry);