ANDROID: Incremental fs: Fix dentry get/put imbalance on vfs_mkdir() failure

Syz{bot,kaller} reports[0]:

  BUG: Dentry ffff888119d8a000{i=0,n=.index}  still in use (1) [unmount of ramfs ramfs]
  ------------[ cut here ]------------
  WARNING: CPU: 0 PID: 367 at fs/dcache.c:1616 umount_check+0x18d/0x1d0 fs/dcache.c:1607
  Modules linked in:
  CPU: 0 PID: 367 Comm: syz-executor388 Not tainted 5.10.75-syzkaller-01082-g234d53d2bb60 #0
  Hardware name: Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
  RIP: 0010:umount_check+0x18d/0x1d0 fs/dcache.c:1607
  Code: 8b 0b 49 81 c6 f8 03 00 00 48 c7 c7 00 40 2e 85 4c 89 e6 48 8b 55 d0 4c 89 e1 45 89 f8 31 c0 41 56 e8 ae d9 9e ff 48 83 c4 08 <0f> 0b e9 f1 fe ff ff 89 d9 80 e1 07 80 c1 03 38 c1 0f 8c c9 fe ff
  RSP: 0018:ffffc9000096f770 EFLAGS: 00010292
  RAX: 0000000000000055 RBX: ffffffff866af200 RCX: 1ad6b89836e5b500
  RDX: 0000000000000000 RSI: 0000000000000002 RDI: 0000000000000000
  RBP: ffffc9000096f7a0 R08: ffffffff81545368 R09: 0000000000000003
  R10: fffff5200012de41 R11: 0000000000000004 R12: ffff888119d8a000
  R13: dffffc0000000000 R14: ffff88811d7373f8 R15: 0000000000000001
  FS:  0000000000000000(0000) GS:ffff8881f7000000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f01b7bddb68 CR3: 000000010c4f0000 CR4: 00000000003506b0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Call Trace:
   d_walk+0x309/0x540 fs/dcache.c:1326
   do_one_tree fs/dcache.c:1623 [inline]
   shrink_dcache_for_umount+0x8e/0x1b0 fs/dcache.c:1639
   generic_shutdown_super+0x66/0x2c0 fs/super.c:447
   kill_anon_super fs/super.c:1108 [inline]
   kill_litter_super+0x75/0xa0 fs/super.c:1117
   ramfs_kill_sb+0x44/0x50 fs/ramfs/inode.c:270
   deactivate_locked_super+0xb0/0x100 fs/super.c:335
   deactivate_super+0xa5/0xd0 fs/super.c:366
   cleanup_mnt+0x45f/0x510 fs/namespace.c:1118
   __cleanup_mnt+0x19/0x20 fs/namespace.c:1125
   task_work_run+0x147/0x1b0 kernel/task_work.c:154
   exit_task_work include/linux/task_work.h:30 [inline]
   do_exit+0x70e/0x23a0 kernel/exit.c:813
   do_group_exit+0x16a/0x2d0 kernel/exit.c:910
   get_signal+0x133e/0x1f80 kernel/signal.c:2790
   arch_do_signal+0x8d/0x620 arch/x86/kernel/signal.c:805
   exit_to_user_mode_loop kernel/entry/common.c:161 [inline]
   exit_to_user_mode_prepare+0xaa/0xe0 kernel/entry/common.c:191
   syscall_exit_to_user_mode+0x24/0x40 kernel/entry/common.c:266
   do_syscall_64+0x3d/0x70 arch/x86/entry/common.c:56
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x7f01b7b884f9
  Code: Unable to access opcode bytes at RIP 0x7f01b7b884cf.
  RSP: 002b:00007f01b7b19308 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca
  RAX: fffffffffffffe00 RBX: 00007f01b7c103f8 RCX: 00007f

Which was due to a missing dput() before returning from a vfs_mkdir() failure.

Bug: 203827798
Link: [0] https://syzkaller.appspot.com/bug?extid=81b5ca9b2848f4dad8fa
Reported-by: syzbot+81b5ca9b2848f4dad8fa@syzkaller.appspotmail.com
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Change-Id: Iaef9aa0aecc964645aaca5fe8d79388ae28527bd
diff --git a/fs/incfs/vfs.c b/fs/incfs/vfs.c
index 0650ee9..ea7866f 100644
--- a/fs/incfs/vfs.c
+++ b/fs/incfs/vfs.c
@@ -458,8 +458,10 @@
 	err = vfs_mkdir(backing_inode, index_dentry, 0777);
 	inode_unlock(backing_inode);
 
-	if (err)
+	if (err) {
+		dput(index_dentry);
 		return ERR_PTR(err);
+	}
 
 	if (!d_really_is_positive(index_dentry) ||
 		unlikely(d_unhashed(index_dentry))) {