KASAN: null-ptr-deref Read in css_task_iter_advance

KASAN: null-ptr-deref Read in css_task_iter_advance

[syzbot]

Hello,

syzbot found the following crash on:

HEAD commit:    3f310e51 Add linux-next specific files for 20190607
git tree:       linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=170acfa6a00000
kernel config:  https://syzkaller.appspot.com/x/.config?x=5d176e1849bbc45
dashboard link: https://syzkaller.appspot.com/bug?extid=d4bba5ccd4f9a2a68681
compiler:       gcc (GCC) 9.0.0 20181231 (experimental)

Unfortunately, I don't have any reproducer for this crash yet.

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+d4bba5ccd4f9a2a68681@syzkaller.appspotmail.com

==================================================================
BUG: KASAN: null-ptr-deref in atomic_read  
include/asm-generic/atomic-instrumented.h:26 [inline]
BUG: KASAN: null-ptr-deref in css_task_iter_advance+0x240/0x540  
kernel/cgroup/cgroup.c:4503
Read of size 4 at addr 0000000000000004 by task syz-executor.2/26575

CPU: 1 PID: 26575 Comm: syz-executor.2 Not tainted 5.2.0-rc3-next-20190607  
#11
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS  
Google 01/01/2011
Call Trace:
  __dump_stack lib/dump_stack.c:77 [inline]
  dump_stack+0x172/0x1f0 lib/dump_stack.c:113
  __kasan_report.cold+0x5/0x36 mm/kasan/report.c:486
  kasan_report+0x12/0x20 mm/kasan/common.c:614
  check_memory_region_inline mm/kasan/generic.c:185 [inline]
  check_memory_region+0x123/0x190 mm/kasan/generic.c:191
  kasan_check_read+0x11/0x20 mm/kasan/common.c:94
  atomic_read include/asm-generic/atomic-instrumented.h:26 [inline]
  css_task_iter_advance+0x240/0x540 kernel/cgroup/cgroup.c:4503
  css_task_iter_start+0x18b/0x230 kernel/cgroup/cgroup.c:4543
  __cgroup_procs_start.isra.0+0x32f/0x400 kernel/cgroup/cgroup.c:4638
  cgroup_procs_start kernel/cgroup/cgroup.c:4660 [inline]
  cgroup_procs_start+0x1e7/0x260 kernel/cgroup/cgroup.c:4647
  cgroup_seqfile_start+0xa4/0xd0 kernel/cgroup/cgroup.c:3752
  kernfs_seq_start+0xdc/0x190 fs/kernfs/file.c:118
  seq_read+0x2a7/0x1110 fs/seq_file.c:224
  kernfs_fop_read+0xed/0x560 fs/kernfs/file.c:252
  do_loop_readv_writev fs/read_write.c:714 [inline]
  do_loop_readv_writev fs/read_write.c:701 [inline]
  do_iter_read+0x4a4/0x660 fs/read_write.c:935
  vfs_readv+0xf0/0x160 fs/read_write.c:997
  do_preadv+0x1c4/0x280 fs/read_write.c:1089
  __do_sys_preadv fs/read_write.c:1139 [inline]
  __se_sys_preadv fs/read_write.c:1134 [inline]
  __x64_sys_preadv+0x9a/0xf0 fs/read_write.c:1134
  do_syscall_64+0xfd/0x680 arch/x86/entry/common.c:301
  entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x459279
Code: fd b7 fb ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 48 89 f8 48 89 f7  
48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff  
ff 0f 83 cb b7 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f4ee9fa8c78 EFLAGS: 00000246 ORIG_RAX: 0000000000000127
RAX: ffffffffffffffda RBX: 0000000000000004 RCX: 0000000000459279
RDX: 0000000000000001 RSI: 0000000020000180 RDI: 0000000000000005
RBP: 000000000075bf20 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00007f4ee9fa96d4
R13: 00000000004c6376 R14: 00000000004dae78 R15: 00000000ffffffff
==================================================================


---
This bug is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this bug report. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

Re: KASAN: null-ptr-deref Read in css_task_iter_advance

[Tejun Heo]

Hello, Hillf.

On Tue, Jun 11, 2019 at 12:59:23AM +0800, Hillf Danton wrote:
> >syzbot will keep track of this bug report. See:
> >https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
> >
> Ignore my noise if you have no interest seeing the syzbot report.

They're awesome.

> The following tiny diff, made in the hope that it may help you perhaps
> reproduce the crash, waits css lock holders and then pick task off
> the list of dying tasks.

I already posted a fix.  Well, at least I think it's the fix.

  https://www.spinics.net/lists/cgroups/msg22468.html

Thanks.

-- 
tejun