memcg: css_alloc should return an ERR_PTR value on error

mem_cgroup_css_alloc() was returning NULL on failure while cgroup core
expected it to return an ERR_PTR value leading to the following NULL
deref after a css allocation failure.  Fix it by return
ERR_PTR(-ENOMEM) instead.  I'll also update cgroup core so that it
can handle NULL returns.

  mkdir: page allocation failure: order:6, mode:0x240c0c0(GFP_KERNEL|__GFP_COMP|__GFP_ZERO)
  CPU: 0 PID: 8738 Comm: mkdir Not tainted 4.7.0-rc3+ torvalds#123
  ...
  Call Trace:
   [<ffffffff81321937>] dump_stack+0x68/0xa1
   [<ffffffff811522f6>] warn_alloc_failed+0xd6/0x130
   [<ffffffff81152816>] __alloc_pages_nodemask+0x4c6/0xf20
   [<ffffffff8119df86>] alloc_pages_current+0x66/0xe0
   [<ffffffff81153564>] alloc_kmem_pages+0x14/0x80
   [<ffffffff811705ca>] kmalloc_order_trace+0x2a/0x1a0
   [<ffffffff811a7a61>] __kmalloc+0x291/0x310
   [<ffffffff811718dc>] memcg_update_all_caches+0x6c/0x130
   [<ffffffff818d0290>] mem_cgroup_css_alloc+0x590/0x610
   [<ffffffff810f4c7b>] cgroup_apply_control_enable+0x18b/0x370
   [<ffffffff810f8afe>] cgroup_mkdir+0x1de/0x2e0
   [<ffffffff8123cf35>] kernfs_iop_mkdir+0x55/0x80
   [<ffffffff811c6599>] vfs_mkdir+0xb9/0x150
   [<ffffffff811cc666>] SyS_mkdir+0x66/0xd0
   [<ffffffff81002df3>] do_syscall_64+0x53/0x120
   [<ffffffff818d719a>] entry_SYSCALL64_slow_path+0x25/0x25
  ...
  BUG: unable to handle kernel NULL pointer dereference at 00000000000000d0
  IP: [<ffffffff810f2ca7>] init_and_link_css+0x37/0x220
  PGD 34b1e067 PUD 3a109067 PMD 0
  Oops: 0002 [#1] SMP
  Modules linked in:
  CPU: 0 PID: 8738 Comm: mkdir Not tainted 4.7.0-rc3+ torvalds#123
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.9.2-20160422_131301-anatol 04/01/2014
  task: ffff88007cbc5200 ti: ffff8800666d4000 task.ti: ffff8800666d4000
  RIP: 0010:[<ffffffff810f2ca7>]  [<ffffffff810f2ca7>] init_and_link_css+0x37/0x220
  RSP: 0018:ffff8800666d7d90  EFLAGS: 00010246
  RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
  RDX: ffffffff810f2499 RSI: 0000000000000000 RDI: 0000000000000008
  RBP: ffff8800666d7db8 R08: 0000000000000003 R09: 0000000000000000
  R10: 0000000000000001 R11: 0000000000000000 R12: ffff88005a5fb400
  R13: ffffffff81f0f8a0 R14: ffff88005a5fb400 R15: 0000000000000010
  FS:  00007fc944689700(0000) GS:ffff88007fc00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f3aed0d2b80 CR3: 000000003a1e8000 CR4: 00000000000006f0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Stack:
   ffffffff81f0f8a0 ffffffff81f0f8a0 ffff88005a5fb400 0000000000000000
   ffff88005a5fb400 ffff8800666d7e18 ffffffff810f4c9c ffff88005a5fb400
   ffffffff82d23470 ffff88007cbc57f0 ffff88007cbc5200 ffff88007d013000
  Call Trace:
   [<ffffffff810f4c9c>] cgroup_apply_control_enable+0x1ac/0x370
   [<ffffffff810f8afe>] cgroup_mkdir+0x1de/0x2e0
   [<ffffffff8123cf35>] kernfs_iop_mkdir+0x55/0x80
   [<ffffffff811c6599>] vfs_mkdir+0xb9/0x150
   [<ffffffff811cc666>] SyS_mkdir+0x66/0xd0
   [<ffffffff81002df3>] do_syscall_64+0x53/0x120
   [<ffffffff818d719a>] entry_SYSCALL64_slow_path+0x25/0x25
  Code: 89 f5 48 89 fb 49 89 d4 48 83 ec 08 8b 05 72 3b d8 00 85 c0 0f 85 60 01 00 00 4c 89 e7 e8 72 f7 ff ff 48 8d 7b 08 48 89 d9 31 c0 <48> c7 83 d0 00 00 00 00 00 00 00 48 83 e7 f8 48 29 f9 81 c1 d8
  RIP  [<ffffffff810f2ca7>] init_and_link_css+0x37/0x220
   RSP <ffff8800666d7d90>
  CR2: 00000000000000d0
  ---[ end trace a2d8836ae1e852d1 ]---

Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Tejun Heo <[email protected]>
Reported-by: Johannes Weiner <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>

mm/memcontrol.c

@@ -4203,7 +4203,7 @@ mem_cgroup_css_alloc(struct cgroup_subsys_state *parent_css)
 	return &memcg->css;
 fail:
 	mem_cgroup_free(memcg);
-	return NULL;
+	return ERR_PTR(-ENOMEM);
 }
 
 static int