Skip to contentSkip to conversation

possible deadlock in proc_pid_personality

PATCH Applied syzbot 3 participants 6 messages Mar 6, 2020 – Apr 15, 2020

Subsystems (2)
lore
Share & tools
lore

syzbot

author syzbot
#0 lore raw

Hello,

syzbot found the following crash on:

HEAD commit:    63623fd4 Merge tag 'for-linus' of git://git.kernel.org/pub..
git tree:       upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=17345181e00000
kernel config:  https://syzkaller.appspot.com/x/.config?x=5d2e033af114153f
dashboard link: https://syzkaller.appspot.com/bug?extid=d9ae59d4662c941e39c6
compiler:       clang version 10.0.0 (https://github.com/llvm/llvm-project/ c2443155a0fb245c8f17f2c1c72b6ea391e86e81)
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=1374670de00000

IMPORTANT: if you fix the bug, please add the following tag to the commit:

Reported-by [email protected]

======================================================
WARNING: possible circular locking dependency detected
5.6.0-rc3-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor.0/9659 is trying to acquire lock:
ffff88807bd9dc90 (&sig->cred_guard_mutex){+.+.}, at: lock_trace fs/proc/base.c:408 [inline] ffff88807bd9dc90 (&sig->cred_guard_mutex){+.+.}, at: proc_pid_personality+0x4f/0x110 fs/proc/base.c:3052

but task is already holding lock:
ffff88808b48d640 (&p->lock){+.+.}, at: seq_read+0x6b/0xdb0 fs/seq_file.c:161

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #3 (&p->lock){+.+.}:

       lock_acquire+0x154/0x250 kernel/locking/lockdep.c:4484
       __mutex_lock_common+0x16e/0x2f30 kernel/locking/mutex.c:956
       __mutex_lock kernel/locking/mutex.c:1103 [inline]
       mutex_lock_nested+0x1b/0x30 kernel/locking/mutex.c:1118
       seq_read+0x6b/0xdb0 fs/seq_file.c:161
       do_loop_readv_writev fs/read_write.c:714 [inline]
       do_iter_read+0x4a2/0x5b0 fs/read_write.c:935
       vfs_readv+0xc2/0x120 fs/read_write.c:1053
       kernel_readv fs/splice.c:365 [inline]
       default_file_splice_read+0x579/0xa40 fs/splice.c:422
       do_splice_to fs/splice.c:892 [inline]
       splice_direct_to_actor+0x3c9/0xb90 fs/splice.c:971
       do_splice_direct+0x200/0x330 fs/splice.c:1080
       do_sendfile+0x7e4/0xfd0 fs/read_write.c:1520
       __do_sys_sendfile64 fs/read_write.c:1581 [inline]
       __se_sys_sendfile64 fs/read_write.c:1567 [inline]
       __x64_sys_sendfile64+0x176/0x1b0 fs/read_write.c:1567
       do_syscall_64+0xf7/0x1c0 arch/x86/entry/common.c:294
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #2 (sb_writers#3){.+.+}:

       lock_acquire+0x154/0x250 kernel/locking/lockdep.c:4484
       percpu_down_read include/linux/percpu-rwsem.h:40 [inline]
       __sb_start_write+0x189/0x420 fs/super.c:1674
       sb_start_write include/linux/fs.h:1649 [inline]
       mnt_want_write+0x4a/0xa0 fs/namespace.c:354
       ovl_want_write+0x77/0x80 fs/overlayfs/util.c:21
       ovl_setattr+0x55/0x790 fs/overlayfs/inode.c:27
       notify_change+0xb6e/0xfb0 fs/attr.c:336
       do_truncate+0x194/0x230 fs/open.c:64
       handle_truncate fs/namei.c:3083 [inline]
       do_last fs/namei.c:3496 [inline]
       path_openat+0x271d/0x4380 fs/namei.c:3607
       do_filp_open+0x192/0x3d0 fs/namei.c:3637
       do_sys_openat2+0x42b/0x6f0 fs/open.c:1149
       do_sys_open fs/open.c:1165 [inline]
       ksys_open include/linux/syscalls.h:1386 [inline]
       __do_sys_creat fs/open.c:1233 [inline]
       __se_sys_creat fs/open.c:1231 [inline]
       __x64_sys_creat+0xd5/0x100 fs/open.c:1231
       do_syscall_64+0xf7/0x1c0 arch/x86/entry/common.c:294
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #1 (&ovl_i_mutex_key[depth]){+.+.}:

       lock_acquire+0x154/0x250 kernel/locking/lockdep.c:4484
       down_write+0x57/0x140 kernel/locking/rwsem.c:1534
       inode_lock include/linux/fs.h:791 [inline]
       process_measurement+0x2d7/0x18d0 security/integrity/ima/ima_main.c:230
       ima_file_check+0x9b/0xe0 security/integrity/ima/ima_main.c:442
       do_last fs/namei.c:3494 [inline]
       path_openat+0x1def/0x4380 fs/namei.c:3607
       do_filp_open+0x192/0x3d0 fs/namei.c:3637
       do_open_execat+0xff/0x620 fs/exec.c:860
       __do_execve_file+0x758/0x1ca0 fs/exec.c:1765
       do_execveat_common fs/exec.c:1871 [inline]
       do_execve fs/exec.c:1888 [inline]
       __do_sys_execve fs/exec.c:1964 [inline]
       __se_sys_execve fs/exec.c:1959 [inline]
       __x64_sys_execve+0x94/0xb0 fs/exec.c:1959
       do_syscall_64+0xf7/0x1c0 arch/x86/entry/common.c:294
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&sig->cred_guard_mutex){+.+.}:

       check_prev_add kernel/locking/lockdep.c:2475 [inline]
       check_prevs_add kernel/locking/lockdep.c:2580 [inline]
       validate_chain+0x1507/0x7be0 kernel/locking/lockdep.c:2970
       __lock_acquire+0xc5a/0x1bc0 kernel/locking/lockdep.c:3954
       lock_acquire+0x154/0x250 kernel/locking/lockdep.c:4484
       __mutex_lock_common+0x16e/0x2f30 kernel/locking/mutex.c:956
       __mutex_lock kernel/locking/mutex.c:1103 [inline]
       mutex_lock_killable_nested+0x1b/0x30 kernel/locking/mutex.c:1133
       lock_trace fs/proc/base.c:408 [inline]
       proc_pid_personality+0x4f/0x110 fs/proc/base.c:3052
       proc_single_show+0xe7/0x180 fs/proc/base.c:758
       seq_read+0x4d8/0xdb0 fs/seq_file.c:229
       do_loop_readv_writev fs/read_write.c:714 [inline]
       do_iter_read+0x4a2/0x5b0 fs/read_write.c:935
       vfs_readv+0xc2/0x120 fs/read_write.c:1053
       kernel_readv fs/splice.c:365 [inline]
       default_file_splice_read+0x579/0xa40 fs/splice.c:422
       do_splice_to fs/splice.c:892 [inline]
       splice_direct_to_actor+0x3c9/0xb90 fs/splice.c:971
       do_splice_direct+0x200/0x330 fs/splice.c:1080
       do_sendfile+0x7e4/0xfd0 fs/read_write.c:1520
       __do_sys_sendfile64 fs/read_write.c:1581 [inline]
       __se_sys_sendfile64 fs/read_write.c:1567 [inline]
       __x64_sys_sendfile64+0x176/0x1b0 fs/read_write.c:1567
       do_syscall_64+0xf7/0x1c0 arch/x86/entry/common.c:294
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

other info that might help us debug this:

Chain exists of:
  &sig->cred_guard_mutex --> sb_writers#3 --> &p->lock

 Possible unsafe locking scenario:
       CPU0                    CPU1
       ----                    ----
  lock(&p->lock);
                               lock(sb_writers#3);
                               lock(&p->lock);

lock(&sig->cred_guard_mutex);

*** DEADLOCK ***

2 locks held by syz-executor.0/9659:
#0: ffff88807b422428 (sb_writers#12){.+.+}, at: file_start_write include/linux/fs.h:2903 [inline] #0: ffff88807b422428 (sb_writers#12){.+.+}, at: do_sendfile+0x7c2/0xfd0 fs/read_write.c:1519 #1: ffff88808b48d640 (&p->lock){+.+.}, at: seq_read+0x6b/0xdb0 fs/seq_file.c:161

stack backtrace:
CPU: 0 PID: 9659 Comm: syz-executor.0 Not tainted 5.6.0-rc3-syzkaller #0
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+0x1fb/0x318 lib/dump_stack.c:118
 print_circular_bug+0xc3f/0xe70 kernel/locking/lockdep.c:1684
 check_noncircular+0x206/0x3a0 kernel/locking/lockdep.c:1808
 check_prev_add kernel/locking/lockdep.c:2475 [inline]
 check_prevs_add kernel/locking/lockdep.c:2580 [inline]
 validate_chain+0x1507/0x7be0 kernel/locking/lockdep.c:2970
 __lock_acquire+0xc5a/0x1bc0 kernel/locking/lockdep.c:3954
 lock_acquire+0x154/0x250 kernel/locking/lockdep.c:4484
 __mutex_lock_common+0x16e/0x2f30 kernel/locking/mutex.c:956
 __mutex_lock kernel/locking/mutex.c:1103 [inline]
 mutex_lock_killable_nested+0x1b/0x30 kernel/locking/mutex.c:1133
 lock_trace fs/proc/base.c:408 [inline]
 proc_pid_personality+0x4f/0x110 fs/proc/base.c:3052
 proc_single_show+0xe7/0x180 fs/proc/base.c:758
 seq_read+0x4d8/0xdb0 fs/seq_file.c:229
 do_loop_readv_writev fs/read_write.c:714 [inline]
 do_iter_read+0x4a2/0x5b0 fs/read_write.c:935
 vfs_readv+0xc2/0x120 fs/read_write.c:1053
 kernel_readv fs/splice.c:365 [inline]
 default_file_splice_read+0x579/0xa40 fs/splice.c:422
 do_splice_to fs/splice.c:892 [inline]
 splice_direct_to_actor+0x3c9/0xb90 fs/splice.c:971
 do_splice_direct+0x200/0x330 fs/splice.c:1080
 do_sendfile+0x7e4/0xfd0 fs/read_write.c:1520
 __do_sys_sendfile64 fs/read_write.c:1581 [inline]
 __se_sys_sendfile64 fs/read_write.c:1567 [inline]
 __x64_sys_sendfile64+0x176/0x1b0 fs/read_write.c:1567
 do_syscall_64+0xf7/0x1c0 arch/x86/entry/common.c:294
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x45c479
Code: ad b6 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 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f853eeaac78 EFLAGS: 00000246 ORIG_RAX: 0000000000000028
RAX: ffffffffffffffda RBX: 00007f853eeab6d4 RCX: 000000000045c479
RDX: 0000000000000000 RSI: 0000000000000004 RDI: 0000000000000003
RBP: 000000000076bfc0 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000283 R11: 0000000000000246 R12: 00000000ffffffff
R13: 00000000000008d1 R14: 00000000004cb364 R15: 000000000076bfcc

---
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 [email protected].

syzbot will keep track of this bug report. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot. syzbot can test patches for this bug, for details see:
https://goo.gl/tpsmEJ#testing-patches

author syzbot
#1 lore raw
syzbot suspects this bug was fixed by commit:

syzbot suspects this bug was fixed by commit:

commit 2db9dbf71bf98d02a0bf33e798e5bfd2a9944696

Author: Bernd Edlinger <[email protected]>
Date:   Fri Mar 20 20:27:24 2020 +0000

proc: Use new infrastructure to fix deadlocks in execve

bisection log:  https://syzkaller.appspot.com/x/bisect.txt?x=136aea00100000
start commit:   63623fd4 Merge tag 'for-linus' of git://git.kernel.org/pub..
git tree:       upstream
kernel config:  https://syzkaller.appspot.com/x/.config?x=5d2e033af114153f
dashboard link: https://syzkaller.appspot.com/bug?extid=d9ae59d4662c941e39c6
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=1374670de00000

If the result looks correct, please mark the bug fixed by replying with:

#syz fix: proc: Use new infrastructure to fix deadlocks in execve

For information about bisection process see: https://goo.gl/tpsmEJ#bisection

#2 lore raw

PATCH 1/1 proc: Handle umounts cleanly

Quoted 29 lines · syzbot
KASAN: use-after-free Read in dput (2)

proc_fill_super: allocate dentry failed
==================================================================

BUG: KASAN: use-after-free in fast_dput fs/dcache.c:727 [inline]
BUG: KASAN: use-after-free in dput+0x53e/0xdf0 fs/dcache.c:846
Read of size 4 at addr ffff88808a618cf0 by task syz-executor.0/8426

CPU: 0 PID: 8426 Comm: syz-executor.0 Not tainted 5.6.0-next-20200412-syzkaller #0
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+0x188/0x20d lib/dump_stack.c:118
 print_address_description.constprop.0.cold+0xd3/0x315 mm/kasan/report.c:382
 __kasan_report.cold+0x35/0x4d mm/kasan/report.c:511
 kasan_report+0x33/0x50 mm/kasan/common.c:625
 fast_dput fs/dcache.c:727 [inline]
 dput+0x53e/0xdf0 fs/dcache.c:846
 proc_kill_sb+0x73/0xf0 fs/proc/root.c:195
 deactivate_locked_super+0x8c/0xf0 fs/super.c:335
 vfs_get_super+0x258/0x2d0 fs/super.c:1212
 vfs_get_tree+0x89/0x2f0 fs/super.c:1547
 do_new_mount fs/namespace.c:2813 [inline]
 do_mount+0x1306/0x1b30 fs/namespace.c:3138
 __do_sys_mount fs/namespace.c:3347 [inline]
 __se_sys_mount fs/namespace.c:3324 [inline]
 __x64_sys_mount+0x18f/0x230 fs/namespace.c:3324
 do_syscall_64+0xf6/0x7d0 arch/x86/entry/common.c:295
 entry_SYSCALL_64_after_hwframe+0x49/0xb3
Quoted 8 lines
RIP: 0033:0x45c889
Code: ad b6 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 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffc1930ec48 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 0000000001324914 RCX: 000000000045c889
RDX: 0000000020000140 RSI: 0000000020000040 RDI: 0000000000000000
RBP: 000000000076bf00 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000003
R13: 0000000000000749 R14: 00000000004ca15a R15: 0000000000000013

Looking at the code now that it the internal mount of proc is no
longer used it is possible to unmount proc. If proc is unmounted
the fields of the pid namespace that were used for filesystem specific state are not reinitialized.

Which means that proc_self and proc_thread_self can be pointers to already freed dentries.

The reported user after free appears to be from mounting and unmounting proc followed by mounting proc again and using error injection to cause the new root dentry allocation to fail. This in turn results in proc_kill_sb running with proc_self and proc_thread_self still retaining their values from the previous mount of proc. Then calling dput on either proc_self of proc_thread_self will result in double put. Which KASAN sees as a use after free.

Solve this by always reinitializing the filesystem state stored in the struct pid_namespace, when proc is unmounted.

Reported-by [email protected]
Fixes 69879c01a0c3 ("proc: Remove the now unnecessary internal mount of proc")
---
 fs/proc/root.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/proc/root.c +7 -0
index 2633f10446c3..fbf084a7d14c 100644
@@ -196,6 +196,13 @@ static void proc_kill_sb(struct super_block *sb)
 	if (ns->proc_thread_self)
 		dput(ns->proc_thread_self);
 	kill_anon_super(sb);
+
+	/* Make the pid namespace safe for a new mount of proc */
+	ns->proc_self = NULL;
+	ns->proc_thread_self = NULL;
+	ns->pid_gid = GLOBAL_ROOT_GID;
+	ns->hide_pid = 0;
+
 	put_pid_ns(ns);
 }
 
Signature · 3 lines
2.20.1

This should fix the reported syzbot failure.
Acked-by on v1 1/1
#3 lore raw

Re: [PATCH] proc: Handle umounts cleanly

Quoted 55 lines · Eric W. Biederman
Quoted 29 lines · syzbot
KASAN: use-after-free Read in dput (2)

proc_fill_super: allocate dentry failed
==================================================================

BUG: KASAN: use-after-free in fast_dput fs/dcache.c:727 [inline]
BUG: KASAN: use-after-free in dput+0x53e/0xdf0 fs/dcache.c:846
Read of size 4 at addr ffff88808a618cf0 by task syz-executor.0/8426

CPU: 0 PID: 8426 Comm: syz-executor.0 Not tainted 5.6.0-next-20200412-syzkaller #0
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+0x188/0x20d lib/dump_stack.c:118
 print_address_description.constprop.0.cold+0xd3/0x315 mm/kasan/report.c:382
 __kasan_report.cold+0x35/0x4d mm/kasan/report.c:511
 kasan_report+0x33/0x50 mm/kasan/common.c:625
 fast_dput fs/dcache.c:727 [inline]
 dput+0x53e/0xdf0 fs/dcache.c:846
 proc_kill_sb+0x73/0xf0 fs/proc/root.c:195
 deactivate_locked_super+0x8c/0xf0 fs/super.c:335
 vfs_get_super+0x258/0x2d0 fs/super.c:1212
 vfs_get_tree+0x89/0x2f0 fs/super.c:1547
 do_new_mount fs/namespace.c:2813 [inline]
 do_mount+0x1306/0x1b30 fs/namespace.c:3138
 __do_sys_mount fs/namespace.c:3347 [inline]
 __se_sys_mount fs/namespace.c:3324 [inline]
 __x64_sys_mount+0x18f/0x230 fs/namespace.c:3324
 do_syscall_64+0xf6/0x7d0 arch/x86/entry/common.c:295
 entry_SYSCALL_64_after_hwframe+0x49/0xb3
Quoted 8 lines
RIP: 0033:0x45c889
Code: ad b6 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 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffc1930ec48 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 0000000001324914 RCX: 000000000045c889
RDX: 0000000020000140 RSI: 0000000020000040 RDI: 0000000000000000
RBP: 000000000076bf00 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000003
R13: 0000000000000749 R14: 00000000004ca15a R15: 0000000000000013

Looking at the code now that it the internal mount of proc is no
longer used it is possible to unmount proc. If proc is unmounted
the fields of the pid namespace that were used for filesystem specific state are not reinitialized.

Which means that proc_self and proc_thread_self can be pointers to already freed dentries.

The reported user after free appears to be from mounting and unmounting proc followed by mounting proc again and using error injection to cause the new root dentry allocation to fail. This in turn results in proc_kill_sb running with proc_self and proc_thread_self still retaining their values from the previous mount of proc. Then calling dput on either proc_self of proc_thread_self will result in double put. Which KASAN sees as a use after free.

Quoted 6 lines

Solve this by always reinitializing the filesystem state stored in the struct pid_namespace, when proc is unmounted.

Reported-by [email protected]
Fixes 69879c01a0c3 ("proc: Remove the now unnecessary internal mount of proc")

Was looking at that earlier right before eod briefly here as well.

Thanks!
Christian

#4 lore raw

Re: [PATCH] proc: Handle umounts cleanly

Quoted 59 lines · Christian Brauner
Quoted 58 lines · Eric W. Biederman
Quoted 29 lines · syzbot
KASAN: use-after-free Read in dput (2)

proc_fill_super: allocate dentry failed
==================================================================

BUG: KASAN: use-after-free in fast_dput fs/dcache.c:727 [inline]
BUG: KASAN: use-after-free in dput+0x53e/0xdf0 fs/dcache.c:846
Read of size 4 at addr ffff88808a618cf0 by task syz-executor.0/8426

CPU: 0 PID: 8426 Comm: syz-executor.0 Not tainted 5.6.0-next-20200412-syzkaller #0
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+0x188/0x20d lib/dump_stack.c:118
 print_address_description.constprop.0.cold+0xd3/0x315 mm/kasan/report.c:382
 __kasan_report.cold+0x35/0x4d mm/kasan/report.c:511
 kasan_report+0x33/0x50 mm/kasan/common.c:625
 fast_dput fs/dcache.c:727 [inline]
 dput+0x53e/0xdf0 fs/dcache.c:846
 proc_kill_sb+0x73/0xf0 fs/proc/root.c:195
 deactivate_locked_super+0x8c/0xf0 fs/super.c:335
 vfs_get_super+0x258/0x2d0 fs/super.c:1212
 vfs_get_tree+0x89/0x2f0 fs/super.c:1547
 do_new_mount fs/namespace.c:2813 [inline]
 do_mount+0x1306/0x1b30 fs/namespace.c:3138
 __do_sys_mount fs/namespace.c:3347 [inline]
 __se_sys_mount fs/namespace.c:3324 [inline]
 __x64_sys_mount+0x18f/0x230 fs/namespace.c:3324
 do_syscall_64+0xf6/0x7d0 arch/x86/entry/common.c:295
 entry_SYSCALL_64_after_hwframe+0x49/0xb3
Quoted 8 lines
RIP: 0033:0x45c889
Code: ad b6 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 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffc1930ec48 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 0000000001324914 RCX: 000000000045c889
RDX: 0000000020000140 RSI: 0000000020000040 RDI: 0000000000000000
RBP: 000000000076bf00 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000003
R13: 0000000000000749 R14: 00000000004ca15a R15: 0000000000000013

Looking at the code now that it the internal mount of proc is no
longer used it is possible to unmount proc. If proc is unmounted
the fields of the pid namespace that were used for filesystem specific state are not reinitialized.

Which means that proc_self and proc_thread_self can be pointers to already freed dentries.

The reported user after free appears to be from mounting and unmounting proc followed by mounting proc again and using error injection to cause the new root dentry allocation to fail. This in turn results in proc_kill_sb running with proc_self and proc_thread_self still retaining their values from the previous mount of proc. Then calling dput on either proc_self of proc_thread_self will result in double put. Which KASAN sees as a use after free.

Solve this by always reinitializing the filesystem state stored in the struct pid_namespace, when proc is unmounted.

Quoted 6 lines
Quoted 3 lines
Reported-by [email protected]
Fixes 69879c01a0c3 ("proc: Remove the now unnecessary internal mount of proc")

Was looking at that earlier right before eod briefly here as well.

The syzbot report or did you see the failure another way?

Eric

#5 lore raw

Re: [PATCH] proc: Handle umounts cleanly

Quoted 61 lines · Eric W. Biederman
Quoted 59 lines · Christian Brauner
Quoted 58 lines · Eric W. Biederman
Quoted 29 lines · syzbot
KASAN: use-after-free Read in dput (2)

proc_fill_super: allocate dentry failed
==================================================================

BUG: KASAN: use-after-free in fast_dput fs/dcache.c:727 [inline]
BUG: KASAN: use-after-free in dput+0x53e/0xdf0 fs/dcache.c:846
Read of size 4 at addr ffff88808a618cf0 by task syz-executor.0/8426

CPU: 0 PID: 8426 Comm: syz-executor.0 Not tainted 5.6.0-next-20200412-syzkaller #0
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+0x188/0x20d lib/dump_stack.c:118
 print_address_description.constprop.0.cold+0xd3/0x315 mm/kasan/report.c:382
 __kasan_report.cold+0x35/0x4d mm/kasan/report.c:511
 kasan_report+0x33/0x50 mm/kasan/common.c:625
 fast_dput fs/dcache.c:727 [inline]
 dput+0x53e/0xdf0 fs/dcache.c:846
 proc_kill_sb+0x73/0xf0 fs/proc/root.c:195
 deactivate_locked_super+0x8c/0xf0 fs/super.c:335
 vfs_get_super+0x258/0x2d0 fs/super.c:1212
 vfs_get_tree+0x89/0x2f0 fs/super.c:1547
 do_new_mount fs/namespace.c:2813 [inline]
 do_mount+0x1306/0x1b30 fs/namespace.c:3138
 __do_sys_mount fs/namespace.c:3347 [inline]
 __se_sys_mount fs/namespace.c:3324 [inline]
 __x64_sys_mount+0x18f/0x230 fs/namespace.c:3324
 do_syscall_64+0xf6/0x7d0 arch/x86/entry/common.c:295
 entry_SYSCALL_64_after_hwframe+0x49/0xb3
Quoted 8 lines
RIP: 0033:0x45c889
Code: ad b6 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 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffc1930ec48 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 0000000001324914 RCX: 000000000045c889
RDX: 0000000020000140 RSI: 0000000020000040 RDI: 0000000000000000
RBP: 000000000076bf00 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000003
R13: 0000000000000749 R14: 00000000004ca15a R15: 0000000000000013

Looking at the code now that it the internal mount of proc is no
longer used it is possible to unmount proc. If proc is unmounted
the fields of the pid namespace that were used for filesystem specific state are not reinitialized.

Which means that proc_self and proc_thread_self can be pointers to already freed dentries.

The reported user after free appears to be from mounting and unmounting proc followed by mounting proc again and using error injection to cause the new root dentry allocation to fail. This in turn results in proc_kill_sb running with proc_self and proc_thread_self still retaining their values from the previous mount of proc. Then calling dput on either proc_self of proc_thread_self will result in double put. Which KASAN sees as a use after free.

Solve this by always reinitializing the filesystem state stored in the struct pid_namespace, when proc is unmounted.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
Reported-by [email protected]
Fixes 69879c01a0c3 ("proc: Remove the now unnecessary internal mount of proc")

Was looking at that earlier right before eod briefly here as well.

The syzbot report or did you see the failure another way?

Yep, the syzbot report. I haven't seen other issues so far.

Christian