From: Luis Goncalves <lgoncalv@redhat.com>
To: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: linux-rt-users <linux-rt-users@vger.kernel.org>,
Steven Rostedt <rostedt@goodmis.org>,
stable-rt@vger.kernel.org, Chunyu Hu <chuhu@redhat.com>
Subject: Re: [RFC RT v5.10] [rt] repair usage of raw_v6_hashinfo.lock in raw_seq_start()
Date: Thu, 4 Nov 2021 15:55:50 -0300 [thread overview]
Message-ID: <CAD8J--_nJXgsNVZSFEs4ZVjvANgxt=vU_i3dMcrWq9hC8UyLRw@mail.gmail.com> (raw)
In-Reply-To: <20211027090650.fqpqtze6wfetihej@linutronix.de>
On Wed, Oct 27, 2021 at 6:06 AM Sebastian Andrzej Siewior
<bigeasy@linutronix.de> wrote:
>
> On 2021-10-18 12:51:32 [-0300], Luis Goncalves wrote:
> > Again, the use of tuned in the tests is mostly to reproduce the
> > configuration in use at
> > the moment. I can isolate the specific configuration that makes it
> > possible to trigger the
> > lockdep splat.
>
> I can't reproduce any of those two issues. I tried both and iterated the
> tuned profiles
> tuned-adm profile realtime
> tuned-adm profile throughput-performance
>
> while running the test you mentioned.
> Could you please try if this reproduces with latest v5.15-RT and post
> complete lockdep backtrace?
I couldn't reproduce that on v5.14-rt nor v5.15-rt. But I can easily
reproduce the lockdep splat on v5.10-rt, though.
This is the backtrace for v5.10.73-rt54:
[ 220.768268] run.sh (2905): /proc/2142/oom_adj is deprecated, please
use /proc/2142/oom_score_adj instead.
[ 221.794958] LTP: starting read_all_proc (read_all -d /proc -q -r 3)
[ 222.678607] ICMPv6: process `read_all' is using deprecated sysctl
(syscall) net.ipv6.neigh.default.base_reachable_time - use
net.ipv6.neigh.default.base_reachable_time_ms instead
[ 222.897071] ======================================================
[ 222.897071] WARNING: possible circular locking dependency detected
[ 222.897072] 5.10.73-rt54.lockdep+ #2 Tainted: G S W I
[ 222.897073] ------------------------------------------------------
[ 222.897074] read_all/3946 is trying to acquire lock:
[ 222.897075] ffff9727dee18220
((softirq_ctrl.lock).lock){+.+.}-{2:2}, at:
__local_bh_disable_ip+0x116/0x2f0
[ 222.897084]
but task is already holding lock:
[ 222.897085] ffffffffba2e9720 (raw_v6_hashinfo.lock){++.+}-{0:0},
at: raw_seq_start+0x25/0x60
[ 222.897090]
which lock already depends on the new lock.
[ 222.897091]
the existing dependency chain (in reverse order) is:
[ 222.897092]
-> #1 (raw_v6_hashinfo.lock){++.+}-{0:0}:
[ 222.897093] lock_acquire+0xde/0x380
[ 222.897098] rt_write_lock+0x3a/0x1b0
[ 222.897101] raw_hash_sk+0x3f/0xd0
[ 222.897103] inet6_create+0x2ef/0x690
[ 222.897107] __sock_create+0x1e2/0x430
[ 222.897111] inet_ctl_sock_create+0x33/0x90
[ 222.897113] icmpv6_sk_init+0x71/0x130
[ 222.897118] ops_init+0x3a/0x180
[ 222.897123] register_pernet_operations+0x126/0x220
[ 222.897125] register_pernet_subsys+0x24/0x40
[ 222.897127] icmpv6_init+0x11/0x4c
[ 222.897132] inet6_init+0x112/0x373
[ 222.897133] do_one_initcall+0x74/0x440
[ 222.897137] kernel_init_freeable+0x39d/0x408
[ 222.897140] kernel_init+0xb/0x12e
[ 222.897142] ret_from_fork+0x1f/0x30
[ 222.897144]
-> #0 ((softirq_ctrl.lock).lock){+.+.}-{2:2}:
[ 222.897145] check_prevs_add+0x1bb/0xe50
[ 222.897147] __lock_acquire+0x1187/0x1640
[ 222.897149] lock_acquire+0xde/0x380
[ 222.897151] rt_spin_lock+0x2b/0xd0
[ 222.897152] __local_bh_disable_ip+0x116/0x2f0
[ 222.897155] sock_i_ino+0x22/0x60
[ 222.897157] __ip6_dgram_sock_seq_show+0x3e/0xe0
[ 222.897160] raw6_seq_show+0x28/0x40
[ 222.897162] seq_read_iter+0x27e/0x420
[ 222.897166] seq_read+0x10f/0x160
[ 222.897168] proc_reg_read+0x3e/0x80
[ 222.897171] vfs_read+0xe7/0x1e0
[ 222.897174] ksys_read+0x5f/0xe0
[ 222.897175] do_syscall_64+0x33/0x40
[ 222.897178] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 222.897179]
other info that might help us debug this:
[ 222.897179] Possible unsafe locking scenario:
[ 222.897180] CPU0 CPU1
[ 222.897180] ---- ----
[ 222.897180] lock(raw_v6_hashinfo.lock);
[ 222.897181] lock((softirq_ctrl.lock).lock);
[ 222.897182] lock(raw_v6_hashinfo.lock);
[ 222.897183] lock((softirq_ctrl.lock).lock);
[ 222.897183]
*** DEADLOCK ***
[ 222.897184] 3 locks held by read_all/3946:
[ 222.897185] #0: ffff9720939dfe10 (&p->lock){+.+.}-{0:0}, at:
seq_read_iter+0x56/0x420
[ 222.897188] #1: ffffffffba2e9720
(raw_v6_hashinfo.lock){++.+}-{0:0}, at: raw_seq_start+0x25/0x60
[ 222.897191] #2: ffffffffb9ec8fe0 (rcu_read_lock){....}-{1:2}, at:
rt_read_lock+0x7d/0x1e0
[ 222.897194]
stack backtrace:
[ 222.897195] CPU: 28 PID: 3946 Comm: read_all Tainted: G S W I
5.10.73-rt54.lockdep+ #2
[ 222.897196] Hardware name: Dell Inc. PowerEdge R740/07X9K0, BIOS
2.8.2 08/27/2020
[ 222.897197] Call Trace:
[ 222.897199] dump_stack+0x77/0x97
[ 222.897204] check_noncircular+0xff/0x120
[ 222.897208] check_prevs_add+0x1bb/0xe50
[ 222.897211] __lock_acquire+0x1187/0x1640
[ 222.897215] lock_acquire+0xde/0x380
[ 222.897217] ? __local_bh_disable_ip+0x116/0x2f0
[ 222.897221] ? sock_i_ino+0x5/0x60
[ 222.897223] rt_spin_lock+0x2b/0xd0
[ 222.897224] ? __local_bh_disable_ip+0x116/0x2f0
[ 222.897227] __local_bh_disable_ip+0x116/0x2f0
[ 222.897229] sock_i_ino+0x22/0x60
[ 222.897231] __ip6_dgram_sock_seq_show+0x3e/0xe0
[ 222.897234] raw6_seq_show+0x28/0x40
[ 222.897236] seq_read_iter+0x27e/0x420
[ 222.897239] seq_read+0x10f/0x160
[ 222.897243] proc_reg_read+0x3e/0x80
[ 222.897245] vfs_read+0xe7/0x1e0
[ 222.897247] ksys_read+0x5f/0xe0
[ 222.897249] do_syscall_64+0x33/0x40
[ 222.897251] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 222.897254] RIP: 0033:0x7f71817b99b2
[ 222.897256] Code: 96 20 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb
b6 0f 1f 80 00 00 00 00 f3 0f 1e fa 8b 05 96 da 20 00 85 c0 75 12 31
c0 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 41 54 49 89 d4 55
48 89
[ 222.897258] RSP: 002b:00007ffed3549878 EFLAGS: 00000246 ORIG_RAX:
0000000000000000
[ 222.897259] RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f71817b99b2
[ 222.897261] RDX: 00000000000003ff RSI: 00007ffed3549d20 RDI: 0000000000000003
[ 222.897262] RBP: 00007f7181bec000 R08: 0000000000000020 R09: 0000000000000000
[ 222.897262] R10: 0000000000000000 R11: 0000000000000246 R12: 000000002cab66e1
[ 222.897263] R13: 00007ffed3549921 R14: 0000000000624730 R15: 0000000000000000
[ 232.965010] LTP: starting read_all_sys (read_all -d /sys -q -r 3)
[ 235.930281] rtc_cmos 00:00: Deprecated ABI, please use nvmem
[ 237.200250] WARNING! power/level is deprecated; use power/control instead
prev parent reply other threads:[~2021-11-04 18:56 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-05 1:53 [RFC RT v5.10] [rt] repair usage of raw_v6_hashinfo.lock in raw_seq_start() Luis Claudio R. Goncalves
2021-10-11 16:33 ` Sebastian Andrzej Siewior
2021-10-18 15:51 ` Luis Goncalves
2021-10-27 9:06 ` Sebastian Andrzej Siewior
2021-11-04 18:48 ` Luis Goncalves
2021-11-05 17:18 ` Sebastian Andrzej Siewior
2021-11-05 17:31 ` Sebastian Andrzej Siewior
2021-11-04 18:55 ` Luis Goncalves [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAD8J--_nJXgsNVZSFEs4ZVjvANgxt=vU_i3dMcrWq9hC8UyLRw@mail.gmail.com' \
--to=lgoncalv@redhat.com \
--cc=bigeasy@linutronix.de \
--cc=chuhu@redhat.com \
--cc=linux-rt-users@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=stable-rt@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.