linux-stable.git/kernel/bpf, branch linux-5.8.y

bpf: Limit caller's stack depth 256 for subprogs with tailcalls

2020-10-29T09:08:24+00:00

[ Upstream commit 7f6e4312e15a5c370e84eaa685879b6bdcc717e4 ]

Protect against potential stack overflow that might happen when bpf2bpf
calls get combined with tailcalls. Limit the caller's stack depth for
such case down to 256 so that the worst case scenario would result in 8k
stack size (32 which is tailcall limit * 256 = 8k).

Suggested-by: Alexei Starovoitov 
Signed-off-by: Maciej Fijalkowski 
Signed-off-by: Alexei Starovoitov 
Signed-off-by: Sasha Levin

bpf: Use raw_spin_trylock() for pcpu_freelist_push/pop in NMI

2020-10-29T09:08:21+00:00

[ Upstream commit 39d8f0d1026a990604770a658708f5845f7dbec0 ]

Recent improvements in LOCKDEP highlighted a potential A-A deadlock with
pcpu_freelist in NMI:

./tools/testing/selftests/bpf/test_progs -t stacktrace_build_id_nmi

[   18.984807] ================================
[   18.984807] WARNING: inconsistent lock state
[   18.984808] 5.9.0-rc6-01771-g1466de1330e1 #2967 Not tainted
[   18.984809] --------------------------------
[   18.984809] inconsistent {INITIAL USE} -> {IN-NMI} usage.
[   18.984810] test_progs/1990 [HC2[2]:SC0[0]:HE0:SE1] takes:
[   18.984810] ffffe8ffffc219c0 (&head->lock){....}-{2:2}, at: __pcpu_freelist_pop+0xe3/0x180
[   18.984813] {INITIAL USE} state was registered at:
[   18.984814]   lock_acquire+0x175/0x7c0
[   18.984814]   _raw_spin_lock+0x2c/0x40
[   18.984815]   __pcpu_freelist_pop+0xe3/0x180
[   18.984815]   pcpu_freelist_pop+0x31/0x40
[   18.984816]   htab_map_alloc+0xbbf/0xf40
[   18.984816]   __do_sys_bpf+0x5aa/0x3ed0
[   18.984817]   do_syscall_64+0x2d/0x40
[   18.984818]   entry_SYSCALL_64_after_hwframe+0x44/0xa9
[   18.984818] irq event stamp: 12
[...]
[   18.984822] other info that might help us debug this:
[   18.984823]  Possible unsafe locking scenario:
[   18.984823]
[   18.984824]        CPU0
[   18.984824]        ----
[   18.984824]   lock(&head->lock);
[   18.984826]   
[   18.984826]     lock(&head->lock);
[   18.984827]
[   18.984828]  *** DEADLOCK ***
[   18.984828]
[   18.984829] 2 locks held by test_progs/1990:
[...]
[   18.984838]  
[   18.984838]  dump_stack+0x9a/0xd0
[   18.984839]  lock_acquire+0x5c9/0x7c0
[   18.984839]  ? lock_release+0x6f0/0x6f0
[   18.984840]  ? __pcpu_freelist_pop+0xe3/0x180
[   18.984840]  _raw_spin_lock+0x2c/0x40
[   18.984841]  ? __pcpu_freelist_pop+0xe3/0x180
[   18.984841]  __pcpu_freelist_pop+0xe3/0x180
[   18.984842]  pcpu_freelist_pop+0x17/0x40
[   18.984842]  ? lock_release+0x6f0/0x6f0
[   18.984843]  __bpf_get_stackid+0x534/0xaf0
[   18.984843]  bpf_prog_1fd9e30e1438d3c5_oncpu+0x73/0x350
[   18.984844]  bpf_overflow_handler+0x12f/0x3f0

This is because pcpu_freelist_head.lock is accessed in both NMI and
non-NMI context. Fix this issue by using raw_spin_trylock() in NMI.

Since NMI interrupts non-NMI context, when NMI context tries to lock the
raw_spinlock, non-NMI context of the same CPU may already have locked a
lock and is blocked from unlocking the lock. For a system with N CPUs,
there could be N NMIs at the same time, and they may block N non-NMI
raw_spinlocks. This is tricky for pcpu_freelist_push(), where unlike
_pop(), failing _push() means leaking memory. This issue is more likely to
trigger in non-SMP system.

Fix this issue with an extra list, pcpu_freelist.extralist. The extralist
is primarily used to take _push() when raw_spin_trylock() failed on all
the per CPU lists. It should be empty most of the time. The following
table summarizes the behavior of pcpu_freelist in NMI and non-NMI:

non-NMI pop(): 	use _lock(); check per CPU lists first;
                if all per CPU lists are empty, check extralist;
                if extralist is empty, return NULL.

non-NMI push(): use _lock(); only push to per CPU lists.

NMI pop():    use _trylock(); check per CPU lists first;
              if all per CPU lists are locked or empty, check extralist;
              if extralist is locked or empty, return NULL.

NMI push():   use _trylock(); check per CPU lists first;
              if all per CPU lists are locked; try push to extralist;
              if extralist is also locked, keep trying on per CPU lists.

Reported-by: Alexei Starovoitov 
Signed-off-by: Song Liu 
Signed-off-by: Daniel Borkmann 
Acked-by: Martin KaFai Lau 
Link: https://lore.kernel.org/bpf/20201005165838.3735218-1-songliubraving@fb.com
Signed-off-by: Sasha Levin

bpf: disallow attaching modify_return tracing functions to other BPF programs

2020-10-29T09:07:46+00:00

[ Upstream commit 1af9270e908cd50a4f5d815c9b6f794c7d57ed07 ]

From the checks and commit messages for modify_return, it seems it was
never the intention that it should be possible to attach a tracing program
with expected_attach_type == BPF_MODIFY_RETURN to another BPF program.
However, check_attach_modify_return() will only look at the function name,
so if the target function starts with "security_", the attach will be
allowed even for bpf2bpf attachment.

Fix this oversight by also blocking the modification if a target program is
supplied.

Fixes: 18644cec714a ("bpf: Fix use-after-free in fmod_ret check")
Fixes: 6ba43b761c41 ("bpf: Attachment verification for BPF_MODIFY_RETURN")
Acked-by: Andrii Nakryiko 
Signed-off-by: Toke Høiland-Jørgensen 
Signed-off-by: Alexei Starovoitov 
Signed-off-by: Sasha Levin

bpf: Fix sysfs export of empty BTF section

2020-10-14T09:55:47+00:00

commit e23bb04b0c938588eae41b7f4712b722290ed2b8 upstream.

If BTF data is missing or removed from the ELF section it is still exported
via sysfs as a zero-length file:

  root@OpenWrt:/# ls -l /sys/kernel/btf/vmlinux
  -r--r--r--    1 root    root    0 Jul 18 02:59 /sys/kernel/btf/vmlinux

Moreover, reads from this file succeed and leak kernel data:

  root@OpenWrt:/# hexdump -C /sys/kernel/btf/vmlinux|head -10
  000000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
  *
  000cc0 00 00 00 00 00 00 00 00 00 00 00 00 80 83 b0 80 |................|
  000cd0 00 10 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
  000ce0 00 00 00 00 00 00 00 00 00 00 00 00 57 ac 6e 9d |............W.n.|
  000cf0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
  *
  002650 00 00 00 00 00 00 00 10 00 00 00 01 00 00 00 01 |................|
  002660 80 82 9a c4 80 85 97 80 81 a9 51 68 00 00 00 02 |..........Qh....|
  002670 80 25 44 dc 80 85 97 80 81 a9 50 24 81 ab c4 60 |.%D.......P$...`|

This situation was first observed with kernel 5.4.x, cross-compiled for a
MIPS target system. Fix by adding a sanity-check for export of zero-length
data sections.

Fixes: 341dfcf8d78e ("btf: expose BTF info through sysfs")
Signed-off-by: Tony Ambardar 
Signed-off-by: Daniel Borkmann 
Acked-by: John Fastabend 
Acked-by: Andrii Nakryiko 
Link: https://lore.kernel.org/bpf/b38db205a66238f70823039a8c531535864eaac5.1600417359.git.Tony.Ambardar@gmail.com
Signed-off-by: Greg Kroah-Hartman

bpf: Fix scalar32_min_max_or bounds tracking

2020-10-14T09:55:45+00:00

commit 5b9fbeb75b6a98955f628e205ac26689bcb1383e upstream.

Simon reported an issue with the current scalar32_min_max_or() implementation.
That is, compared to the other 32 bit subreg tracking functions, the code in
scalar32_min_max_or() stands out that it's using the 64 bit registers instead
of 32 bit ones. This leads to bounds tracking issues, for example:

  [...]
  8: R0=map_value(id=0,off=0,ks=4,vs=48,imm=0) R10=fp0 fp-8=mmmmmmmm
  8: (79) r1 = *(u64 *)(r0 +0)
   R0=map_value(id=0,off=0,ks=4,vs=48,imm=0) R10=fp0 fp-8=mmmmmmmm
  9: R0=map_value(id=0,off=0,ks=4,vs=48,imm=0) R1_w=inv(id=0) R10=fp0 fp-8=mmmmmmmm
  9: (b7) r0 = 1
  10: R0_w=inv1 R1_w=inv(id=0) R10=fp0 fp-8=mmmmmmmm
  10: (18) r2 = 0x600000002
  12: R0_w=inv1 R1_w=inv(id=0) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  12: (ad) if r1 < r2 goto pc+1
   R0_w=inv1 R1_w=inv(id=0,umin_value=25769803778) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  13: R0_w=inv1 R1_w=inv(id=0,umin_value=25769803778) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  13: (95) exit
  14: R0_w=inv1 R1_w=inv(id=0,umax_value=25769803777,var_off=(0x0; 0x7ffffffff)) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  14: (25) if r1 > 0x0 goto pc+1
   R0_w=inv1 R1_w=inv(id=0,umax_value=0,var_off=(0x0; 0x7fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  15: R0_w=inv1 R1_w=inv(id=0,umax_value=0,var_off=(0x0; 0x7fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  15: (95) exit
  16: R0_w=inv1 R1_w=inv(id=0,umin_value=1,umax_value=25769803777,var_off=(0x0; 0x77fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  16: (47) r1 |= 0
  17: R0_w=inv1 R1_w=inv(id=0,umin_value=1,umax_value=32212254719,var_off=(0x1; 0x700000000),s32_max_value=1,u32_max_value=1) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  [...]

The bound tests on the map value force the upper unsigned bound to be 25769803777
in 64 bit (0b11000000000000000000000000000000001) and then lower one to be 1. By
using OR they are truncated and thus result in the range [1,1] for the 32 bit reg
tracker. This is incorrect given the only thing we know is that the value must be
positive and thus 2147483647 (0b1111111111111111111111111111111) at max for the
subregs. Fix it by using the {u,s}32_{min,max}_value vars instead. This also makes
sense, for example, for the case where we update dst_reg->s32_{min,max}_value in
the else branch we need to use the newly computed dst_reg->u32_{min,max}_value as
we know that these are positive. Previously, in the else branch the 64 bit values
of umin_value=1 and umax_value=32212254719 were used and latter got truncated to
be 1 as upper bound there. After the fix the subreg range is now correct:

  [...]
  8: R0=map_value(id=0,off=0,ks=4,vs=48,imm=0) R10=fp0 fp-8=mmmmmmmm
  8: (79) r1 = *(u64 *)(r0 +0)
   R0=map_value(id=0,off=0,ks=4,vs=48,imm=0) R10=fp0 fp-8=mmmmmmmm
  9: R0=map_value(id=0,off=0,ks=4,vs=48,imm=0) R1_w=inv(id=0) R10=fp0 fp-8=mmmmmmmm
  9: (b7) r0 = 1
  10: R0_w=inv1 R1_w=inv(id=0) R10=fp0 fp-8=mmmmmmmm
  10: (18) r2 = 0x600000002
  12: R0_w=inv1 R1_w=inv(id=0) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  12: (ad) if r1 < r2 goto pc+1
   R0_w=inv1 R1_w=inv(id=0,umin_value=25769803778) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  13: R0_w=inv1 R1_w=inv(id=0,umin_value=25769803778) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  13: (95) exit
  14: R0_w=inv1 R1_w=inv(id=0,umax_value=25769803777,var_off=(0x0; 0x7ffffffff)) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  14: (25) if r1 > 0x0 goto pc+1
   R0_w=inv1 R1_w=inv(id=0,umax_value=0,var_off=(0x0; 0x7fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  15: R0_w=inv1 R1_w=inv(id=0,umax_value=0,var_off=(0x0; 0x7fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  15: (95) exit
  16: R0_w=inv1 R1_w=inv(id=0,umin_value=1,umax_value=25769803777,var_off=(0x0; 0x77fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  16: (47) r1 |= 0
  17: R0_w=inv1 R1_w=inv(id=0,umin_value=1,umax_value=32212254719,var_off=(0x0; 0x77fffffff),u32_max_value=2147483647) R2_w=inv25769803778 R10=fp0 fp-8=mmmmmmmm
  [...]

Fixes: 3f50f132d840 ("bpf: Verifier, do explicit ALU32 bounds tracking")
Reported-by: Simon Scannell 
Signed-off-by: Daniel Borkmann 
Reviewed-by: John Fastabend 
Acked-by: Alexei Starovoitov 
Signed-off-by: Greg Kroah-Hartman

bpf: Fix a rcu warning for bpffs map pretty-print

2020-10-01T15:36:29+00:00

[ Upstream commit ce880cb825fcc22d4e39046a6c3a3a7f6603883d ]

Running selftest
  ./btf_btf -p
the kernel had the following warning:
  [   51.528185] WARNING: CPU: 3 PID: 1756 at kernel/bpf/hashtab.c:717 htab_map_get_next_key+0x2eb/0x300
  [   51.529217] Modules linked in:
  [   51.529583] CPU: 3 PID: 1756 Comm: test_btf Not tainted 5.9.0-rc1+ #878
  [   51.530346] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.9.3-1.el7.centos 04/01/2014
  [   51.531410] RIP: 0010:htab_map_get_next_key+0x2eb/0x300
  ...
  [   51.542826] Call Trace:
  [   51.543119]  map_seq_next+0x53/0x80
  [   51.543528]  seq_read+0x263/0x400
  [   51.543932]  vfs_read+0xad/0x1c0
  [   51.544311]  ksys_read+0x5f/0xe0
  [   51.544689]  do_syscall_64+0x33/0x40
  [   51.545116]  entry_SYSCALL_64_after_hwframe+0x44/0xa9

The related source code in kernel/bpf/hashtab.c:
  709 static int htab_map_get_next_key(struct bpf_map *map, void *key, void *next_key)
  710 {
  711         struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
  712         struct hlist_nulls_head *head;
  713         struct htab_elem *l, *next_l;
  714         u32 hash, key_size;
  715         int i = 0;
  716
  717         WARN_ON_ONCE(!rcu_read_lock_held());

In kernel/bpf/inode.c, bpffs map pretty print calls map->ops->map_get_next_key()
without holding a rcu_read_lock(), hence causing the above warning.
To fix the issue, just surrounding map->ops->map_get_next_key() with rcu read lock.

Fixes: a26ca7c982cb ("bpf: btf: Add pretty print support to the basic arraymap")
Reported-by: Alexei Starovoitov 
Signed-off-by: Yonghong Song 
Signed-off-by: Alexei Starovoitov 
Acked-by: Andrii Nakryiko 
Cc: Martin KaFai Lau 
Link: https://lore.kernel.org/bpf/20200916004401.146277-1-yhs@fb.com
Signed-off-by: Sasha Levin

bpf: Fix a buffer out-of-bound access when filling raw_tp link_info

2020-09-09T17:14:16+00:00

[ Upstream commit b474959d5afda6e341a02c85f9595d85d39189ae ]

Commit f2e10bff16a0 ("bpf: Add support for BPF_OBJ_GET_INFO_BY_FD for bpf_link")
added link query for raw_tp. One of fields in link_info is to
fill a user buffer with tp_name. The Scurrent checking only
declares "ulen && !ubuf" as invalid. So "!ulen && ubuf" will be
valid. Later on, we do "copy_to_user(ubuf, tp_name, ulen - 1)" which
may overwrite user memory incorrectly.

This patch fixed the problem by disallowing "!ulen && ubuf" case as well.

Fixes: f2e10bff16a0 ("bpf: Add support for BPF_OBJ_GET_INFO_BY_FD for bpf_link")
Signed-off-by: Yonghong Song 
Signed-off-by: Alexei Starovoitov 
Acked-by: Andrii Nakryiko 
Link: https://lore.kernel.org/bpf/20200821191054.714731-1-yhs@fb.com
Signed-off-by: Sasha Levin

bpf: Avoid visit same object multiple times

2020-09-03T09:29:37+00:00

[ Upstream commit e60572b8d4c39572be6857d1ec91fdf979f8775f ]

Currently when traversing all tasks, the next tid
is always increased by one. This may result in
visiting the same task multiple times in a
pid namespace.

This patch fixed the issue by seting the next
tid as pid_nr_ns(pid, ns) + 1, similar to
funciton next_tgid().

Signed-off-by: Yonghong Song 
Signed-off-by: Alexei Starovoitov 
Cc: Rik van Riel 
Link: https://lore.kernel.org/bpf/20200818222310.2181500-1-yhs@fb.com
Signed-off-by: Sasha Levin

bpf: Fix a rcu_sched stall issue with bpf task/task_file iterator

2020-09-03T09:29:37+00:00

[ Upstream commit e679654a704e5bd676ea6446fa7b764cbabf168a ]

In our production system, we observed rcu stalls when
'bpftool prog` is running.
  rcu: INFO: rcu_sched self-detected stall on CPU
  rcu: \x097-....: (20999 ticks this GP) idle=302/1/0x4000000000000000 softirq=1508852/1508852 fqs=4913
  \x09(t=21031 jiffies g=2534773 q=179750)
  NMI backtrace for cpu 7
  CPU: 7 PID: 184195 Comm: bpftool Kdump: loaded Tainted: G        W         5.8.0-00004-g68bfc7f8c1b4 #6
  Hardware name: Quanta Twin Lakes MP/Twin Lakes Passive MP, BIOS F09_3A17 05/03/2019
  Call Trace:
  
  dump_stack+0x57/0x70
  nmi_cpu_backtrace.cold+0x14/0x53
  ? lapic_can_unplug_cpu.cold+0x39/0x39
  nmi_trigger_cpumask_backtrace+0xb7/0xc7
  rcu_dump_cpu_stacks+0xa2/0xd0
  rcu_sched_clock_irq.cold+0x1ff/0x3d9
  ? tick_nohz_handler+0x100/0x100
  update_process_times+0x5b/0x90
  tick_sched_timer+0x5e/0xf0
  __hrtimer_run_queues+0x12a/0x2a0
  hrtimer_interrupt+0x10e/0x280
  __sysvec_apic_timer_interrupt+0x51/0xe0
  asm_call_on_stack+0xf/0x20
  
  sysvec_apic_timer_interrupt+0x6f/0x80
  asm_sysvec_apic_timer_interrupt+0x12/0x20
  RIP: 0010:task_file_seq_get_next+0x71/0x220
  Code: 00 00 8b 53 1c 49 8b 7d 00 89 d6 48 8b 47 20 44 8b 18 41 39 d3 76 75 48 8b 4f 20 8b 01 39 d0 76 61 41 89 d1 49 39 c1 48 19 c0 <48> 8b 49 08 21 d0 48 8d 04 c1 4c 8b 08 4d 85 c9 74 46 49 8b 41 38
  RSP: 0018:ffffc90006223e10 EFLAGS: 00000297
  RAX: ffffffffffffffff RBX: ffff888f0d172388 RCX: ffff888c8c07c1c0
  RDX: 00000000000f017b RSI: 00000000000f017b RDI: ffff888c254702c0
  RBP: ffffc90006223e68 R08: ffff888be2a1c140 R09: 00000000000f017b
  R10: 0000000000000002 R11: 0000000000100000 R12: ffff888f23c24118
  R13: ffffc90006223e60 R14: ffffffff828509a0 R15: 00000000ffffffff
  task_file_seq_next+0x52/0xa0
  bpf_seq_read+0xb9/0x320
  vfs_read+0x9d/0x180
  ksys_read+0x5f/0xe0
  do_syscall_64+0x38/0x60
  entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x7f8815f4f76e
  Code: c0 e9 f6 fe ff ff 55 48 8d 3d 76 70 0a 00 48 89 e5 e8 36 06 02 00 66 0f 1f 44 00 00 64 8b 04 25 18 00 00 00 85 c0 75 14 0f 05 <48> 3d 00 f0 ff ff 77 52 c3 66 0f 1f 84 00 00 00 00 00 55 48 89 e5
  RSP: 002b:00007fff8f9df578 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
  RAX: ffffffffffffffda RBX: 000000000170b9c0 RCX: 00007f8815f4f76e
  RDX: 0000000000001000 RSI: 00007fff8f9df5b0 RDI: 0000000000000007
  RBP: 00007fff8f9e05f0 R08: 0000000000000049 R09: 0000000000000010
  R10: 00007f881601fa40 R11: 0000000000000246 R12: 00007fff8f9e05a8
  R13: 00007fff8f9e05a8 R14: 0000000001917f90 R15: 000000000000e22e

Note that `bpftool prog` actually calls a task_file bpf iterator
program to establish an association between prog/map/link/btf anon
files and processes.

In the case where the above rcu stall occured, we had a process
having 1587 tasks and each task having roughly 81305 files.
This implied 129 million bpf prog invocations. Unfortunwtely none of
these files are prog/map/link/btf files so bpf iterator/prog needs
to traverse all these files and not able to return to user space
since there are no seq_file buffer overflow.

This patch fixed the issue in bpf_seq_read() to limit the number
of visited objects. If the maximum number of visited objects is
reached, no more objects will be visited in the current syscall.
If there is nothing written in the seq_file buffer, -EAGAIN will
return to the user so user can try again.

The maximum number of visited objects is set at 1 million.
In our Intel Xeon D-2191 2.3GHZ 18-core server, bpf_seq_read()
visiting 1 million files takes around 0.18 seconds.

We did not use cond_resched() since for some iterators, e.g.,
netlink iterator, where rcu read_lock critical section spans between
consecutive seq_ops->next(), which makes impossible to do cond_resched()
in the key while loop of function bpf_seq_read().

Signed-off-by: Yonghong Song 
Signed-off-by: Alexei Starovoitov 
Cc: Paul E. McKenney 
Link: https://lore.kernel.org/bpf/20200818222309.2181348-1-yhs@fb.com
Signed-off-by: Sasha Levin

bpf: Use get_file_rcu() instead of get_file() for task_file iterator

2020-08-26T09:49:17+00:00

[ Upstream commit cf28f3bbfca097d956f9021cb710dfad56adcc62 ]

With latest `bpftool prog` command, we observed the following kernel
panic.
    BUG: kernel NULL pointer dereference, address: 0000000000000000
    #PF: supervisor instruction fetch in kernel mode
    #PF: error_code(0x0010) - not-present page
    PGD dfe894067 P4D dfe894067 PUD deb663067 PMD 0
    Oops: 0010 [#1] SMP
    CPU: 9 PID: 6023 ...
    RIP: 0010:0x0
    Code: Bad RIP value.
    RSP: 0000:ffffc900002b8f18 EFLAGS: 00010286
    RAX: ffff8883a405f400 RBX: ffff888e46a6bf00 RCX: 000000008020000c
    RDX: 0000000000000000 RSI: 0000000000000001 RDI: ffff8883a405f400
    RBP: ffff888e46a6bf50 R08: 0000000000000000 R09: ffffffff81129600
    R10: ffff8883a405f300 R11: 0000160000000000 R12: 0000000000002710
    R13: 000000e9494b690c R14: 0000000000000202 R15: 0000000000000009
    FS:  00007fd9187fe700(0000) GS:ffff888e46a40000(0000) knlGS:0000000000000000
    CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
    CR2: ffffffffffffffd6 CR3: 0000000de5d33002 CR4: 0000000000360ee0
    DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
    DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
    Call Trace:
     
     rcu_core+0x1a4/0x440
     __do_softirq+0xd3/0x2c8
     irq_exit+0x9d/0xa0
     smp_apic_timer_interrupt+0x68/0x120
     apic_timer_interrupt+0xf/0x20
     
    RIP: 0033:0x47ce80
    Code: Bad RIP value.
    RSP: 002b:00007fd9187fba40 EFLAGS: 00000206 ORIG_RAX: ffffffffffffff13
    RAX: 0000000000000002 RBX: 00007fd931789160 RCX: 000000000000010c
    RDX: 00007fd9308cdfb4 RSI: 00007fd9308cdfb4 RDI: 00007ffedd1ea0a8
    RBP: 00007fd9187fbab0 R08: 000000000000000e R09: 000000000000002a
    R10: 0000000000480210 R11: 00007fd9187fc570 R12: 00007fd9316cc400
    R13: 0000000000000118 R14: 00007fd9308cdfb4 R15: 00007fd9317a9380

After further analysis, the bug is triggered by
Commit eaaacd23910f ("bpf: Add task and task/file iterator targets")
which introduced task_file bpf iterator, which traverses all open file
descriptors for all tasks in the current namespace.
The latest `bpftool prog` calls a task_file bpf program to traverse
all files in the system in order to associate processes with progs/maps, etc.
When traversing files for a given task, rcu read_lock is taken to
access all files in a file_struct. But it used get_file() to grab
a file, which is not right. It is possible file->f_count is 0 and
get_file() will unconditionally increase it.
Later put_file() may cause all kind of issues with the above
as one of sympotoms.

The failure can be reproduced with the following steps in a few seconds:
    $ cat t.c
    #include 
    #include 
    #include 
    #include 
    #include 

    #define N 10000
    int fd[N];
    int main() {
      int i;

      for (i = 0; i < N; i++) {
        fd[i] = open("./note.txt", 'r');
        if (fd[i] < 0) {
           fprintf(stderr, "failed\n");
           return -1;
        }
      }
      for (i = 0; i < N; i++)
        close(fd[i]);

      return 0;
    }
    $ gcc -O2 t.c
    $ cat run.sh
    #/bin/bash
    for i in {1..100}
    do
      while true; do ./a.out; done &
    done
    $ ./run.sh
    $ while true; do bpftool prog >& /dev/null; done

This patch used get_file_rcu() which only grabs a file if the
file->f_count is not zero. This is to ensure the file pointer
is always valid. The above reproducer did not fail for more
than 30 minutes.

Fixes: eaaacd23910f ("bpf: Add task and task/file iterator targets")
Suggested-by: Josef Bacik 
Signed-off-by: Yonghong Song 
Signed-off-by: Alexei Starovoitov 
Reviewed-by: Josef Bacik 
Link: https://lore.kernel.org/bpf/20200817174214.252601-1-yhs@fb.com
Signed-off-by: Sasha Levin