[syzbot] [mm?] INFO: rcu detected stall in sys_clone (8)

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: f99c5f563c17 Merge tag 'nf-24-03-21' of git://git.kernel.o..
git tree: net
console+strace: https://syzkaller.appspot.com/x/log.txt?x=12163769180000
kernel config: https://syzkaller.appspot.com/x/.config?x=6fb1be60a193d440
dashboard link: https://syzkaller.appspot.com/bug?extid=c4c6c3dc10cc96bcf723
compiler: Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=125d023a180000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=14a54006180000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/65d3f3eb786e/disk-f99c5f56.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/799cf7f28ff8/vmlinux-f99c5f56.xz
kernel image: https://storage.googleapis.com/syzbot-assets/ab26c60c3845/bzImage-f99c5f56.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+c4c6c3...@syzkaller.appspotmail.com

rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
rcu: 1-...!: (1 ticks this GP) idle=f744/1/0x4000000000000000 softirq=6822/6822 fqs=0
rcu: (detected by 0, t=10503 jiffies, g=8373, q=658 ncpus=2)
Sending NMI from CPU 0 to CPUs 1:
NMI backtrace for cpu 1
CPU: 1 PID: 4736 Comm: dhcpcd Not tainted 6.8.0-syzkaller-05271-gf99c5f563c17 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 02/29/2024
RIP: 0010:lockdep_enabled kernel/locking/lockdep.c:122 [inline]
RIP: 0010:lock_release+0x124/0x9d0 kernel/locking/lockdep.c:5767
Code: 00 65 48 8b 04 25 80 ce 03 00 48 89 44 24 18 48 8d 98 d4 0a 00 00 48 89 d8 48 c1 e8 03 42 0f b6 04 38 84 c0 0f 85 c3 05 00 00 <83> 3b 00 0f 85 f1 04 00 00 4c 8d b4 24 b0 00 00 00 4c 89 f3 48 c1
RSP: 0018:ffffc90000a08a60 EFLAGS: 00000046
RAX: 0000000000000000 RBX: ffff888018308ad4 RCX: ffffffff8171c080
RDX: 0000000000000000 RSI: ffffffff8bfec640 RDI: ffffffff8bfec600
RBP: ffffc90000a08b90 R08: ffffffff8f86ae6f R09: 1ffffffff1f0d5cd
R10: dffffc0000000000 R11: fffffbfff1f0d5ce R12: 1ffff92000141158
R13: ffffffff84ac5555 R14: ffffc90000a08bc0 R15: dffffc0000000000
FS: 00007f8e49f9b740(0000) GS:ffff8880b9500000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f8e4a140b10 CR3: 0000000028508000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<NMI>
</NMI>
<IRQ>
__raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:149 [inline]
_raw_spin_unlock_irqrestore+0x79/0x140 kernel/locking/spinlock.c:194
debug_object_deactivate+0x2d5/0x390 lib/debugobjects.c:778
debug_hrtimer_deactivate kernel/time/hrtimer.c:428 [inline]
debug_deactivate+0x1b/0x200 kernel/time/hrtimer.c:484
__run_hrtimer kernel/time/hrtimer.c:1660 [inline]
__hrtimer_run_queues+0x30f/0xd00 kernel/time/hrtimer.c:1756
hrtimer_interrupt+0x396/0x990 kernel/time/hrtimer.c:1818
local_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1032 [inline]
__sysvec_apic_timer_interrupt+0x107/0x3a0 arch/x86/kernel/apic/apic.c:1049
instr_sysvec_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1043 [inline]
sysvec_apic_timer_interrupt+0xa1/0xc0 arch/x86/kernel/apic/apic.c:1043
</IRQ>
<TASK>
asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:702
RIP: 0010:check_kcov_mode kernel/kcov.c:184 [inline]
RIP: 0010:write_comp_data kernel/kcov.c:236 [inline]
RIP: 0010:__sanitizer_cov_trace_cmp8+0x35/0x90 kernel/kcov.c:284
Code: 0c 25 80 ce 03 00 65 8b 05 a0 db 6e 7e a9 00 01 ff 00 74 10 a9 00 01 00 00 74 57 83 b9 14 16 00 00 00 74 4e 8b 81 f0 15 00 00 <83> f8 03 75 43 48 8b 91 f8 15 00 00 44 8b 89 f4 15 00 00 49 c1 e1
RSP: 0018:ffffc900035ff500 EFLAGS: 00000246
RAX: 0000000000000000 RBX: ffffffff81000000 RCX: ffff888018308000
RDX: ffffc900035ff615 RSI: ffffffff81000000 RDI: ffffffff81ecc4e1
RBP: ffffffff81ecc4e1 R08: ffffffff81409149 R09: ffffc900035ff6d0
R10: 0000000000000003 R11: ffffffff8180dbc0 R12: ffffc900035ff5e0
R13: ffffc900035ff630 R14: dffffc0000000000 R15: ffffffff81ecc4e2
orc_find arch/x86/kernel/unwind_orc.c:206 [inline]
unwind_next_frame+0x1d9/0x2a00 arch/x86/kernel/unwind_orc.c:494
arch_stack_walk+0x151/0x1b0 arch/x86/kernel/stacktrace.c:25
stack_trace_save+0x118/0x1d0 kernel/stacktrace.c:122
kasan_save_stack mm/kasan/common.c:47 [inline]
kasan_save_track+0x3f/0x80 mm/kasan/common.c:68
unpoison_slab_object mm/kasan/common.c:312 [inline]
__kasan_slab_alloc+0x66/0x80 mm/kasan/common.c:338
kasan_slab_alloc include/linux/kasan.h:201 [inline]
slab_post_alloc_hook mm/slub.c:3813 [inline]
slab_alloc_node mm/slub.c:3860 [inline]
kmem_cache_alloc_node+0x192/0x380 mm/slub.c:3903
alloc_task_struct_node kernel/fork.c:176 [inline]
dup_task_struct+0x57/0x7d0 kernel/fork.c:1106
copy_process+0x5d1/0x3df0 kernel/fork.c:2219
kernel_clone+0x21e/0x8d0 kernel/fork.c:2796
__do_sys_clone kernel/fork.c:2939 [inline]
__se_sys_clone kernel/fork.c:2923 [inline]
__x64_sys_clone+0x258/0x2a0 kernel/fork.c:2923
do_syscall_64+0xfb/0x240
entry_SYSCALL_64_after_hwframe+0x6d/0x75
RIP: 0033:0x7f8e4a04ba12
Code: 41 5d 41 5e 41 5f c3 64 48 8b 04 25 10 00 00 00 45 31 c0 31 d2 31 f6 bf 11 00 20 01 4c 8d 90 d0 02 00 00 b8 38 00 00 00 0f 05 <48> 3d 00 f0 ff ff 76 10 48 8b 15 e7 43 0f 00 f7 d8 64 89 02 48 83
RSP: 002b:00007fff3d31e228 EFLAGS: 00000246 ORIG_RAX: 0000000000000038
RAX: ffffffffffffffda RBX: 00005624b41b1601 RCX: 00007f8e4a04ba12
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000001200011
RBP: 00007fff3d33e798 R08: 0000000000000000 R09: 00005624b41b15d0
R10: 00007f8e49f9ba10 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 00005624b41b1604
</TASK>
rcu: rcu_preempt kthread timer wakeup didn't happen for 10502 jiffies! g8373 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402
rcu: Possible timer handling issue on cpu=1 timer-softirq=4283
rcu: rcu_preempt kthread starved for 10503 jiffies! g8373 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402 ->cpu=1
rcu: Unless rcu_preempt kthread gets sufficient CPU time, OOM is now expected behavior.
rcu: RCU grace-period kthread stack dump:
task:rcu_preempt state:I stack:24656 pid:16 tgid:16 ppid:2 flags:0x00004000
Call Trace:
<TASK>
context_switch kernel/sched/core.c:5409 [inline]
__schedule+0x17d3/0x4a20 kernel/sched/core.c:6736
__schedule_loop kernel/sched/core.c:6813 [inline]
schedule+0x14b/0x320 kernel/sched/core.c:6828
schedule_timeout+0x1be/0x310 kernel/time/timer.c:2572
rcu_gp_fqs_loop+0x2df/0x1370 kernel/rcu/tree.c:1663
rcu_gp_kthread+0xa7/0x3b0 kernel/rcu/tree.c:1862
kthread+0x2f0/0x390 kernel/kthread.c:388
ret_from_fork+0x4b/0x80 arch/x86/kernel/process.c:147
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:243
</TASK>


---
This report 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 syzk...@googlegroups.com.

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

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want syzbot to run the reproducer, reply with:
#syz test: git://repo/address.git branch-or-commit-hash
If you attach or paste a git patch, syzbot will apply it before testing.

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

[syzbot]

For archival purposes, forwarding an incoming command email to
linux-...@vger.kernel.org, syzkall...@googlegroups.com.

***

Subject: Test
Author: radoslaw...@gmail.com

#syz test https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 71b1543c83d65af8215d7558d70fc2ecbee77dcf

---
net/sched/sch_taprio.c | 26 ++++++++++++++++++++++++++
1 file changed, 26 insertions(+)

diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
index a0d54b422186..41add4a1d407 100644
--- a/net/sched/sch_taprio.c
+++ b/net/sched/sch_taprio.c
@@ -44,6 +44,8 @@ static struct static_key_false taprio_have_working_mqprio;
(TCA_TAPRIO_ATTR_FLAG_TXTIME_ASSIST | TCA_TAPRIO_ATTR_FLAG_FULL_OFFLOAD)
#define TAPRIO_FLAGS_INVALID U32_MAX

+#define TAPRIO_MAX_NUM_SCHEDULES_BEFORE_NOW 10
+
struct sched_entry {
/* Durations between this GCL entry and the GCL entry where the
* respective traffic class gate closes
@@ -62,6 +64,8 @@ struct sched_entry {
u32 gate_mask;
u32 interval;
u8 command;
+ /* Used to skip some cycles to prevent other tasks from starving */
+ u8 num_schedules_before_now;
};

struct sched_gate_list {
@@ -111,6 +115,9 @@ struct __tc_taprio_qopt_offload {
struct tc_taprio_qopt_offload offload;
};

+static void setup_txtime(struct taprio_sched *q,
+ struct sched_gate_list *sched, ktime_t base);
+
static void taprio_calculate_gate_durations(struct taprio_sched *q,
struct sched_gate_list *sched)
{
@@ -919,6 +926,7 @@ static enum hrtimer_restart advance_sched(struct hrtimer *timer)
struct sched_entry *entry, *next;
struct Qdisc *sch = q->root;
ktime_t end_time;
+ ktime_t basenow;
int tc;

spin_lock(&q->current_entry_lock);
@@ -977,6 +985,23 @@ static enum hrtimer_restart advance_sched(struct hrtimer *timer)
taprio_set_budgets(q, oper, next);

first_run:
+ basenow = ktime_get();
+ if (ktime_before(end_time, basenow))
+ {
+ next->num_schedules_before_now++;
+ if (next->num_schedules_before_now >= TAPRIO_MAX_NUM_SCHEDULES_BEFORE_NOW)
+ {
+ end_time = ktime_add_ns(basenow, oper->cycle_time);
+ next->end_time = end_time;
+ next->num_schedules_before_now = 0;
+ setup_txtime(q, oper, basenow);
+ }
+ }
+ else
+ {
+ next->num_schedules_before_now = 0;
+ }
+
rcu_assign_pointer(q->current_entry, next);
spin_unlock(&q->current_entry_lock);

@@ -1074,6 +1099,7 @@ static int parse_sched_entry(struct taprio_sched *q, struct nlattr *n,
}

entry->index = index;
+ entry->num_schedules_before_now = 0;

return fill_sched_entry(q, tb, entry, extack);
}
--
2.43.0

[syzbot]

For archival purposes, forwarding an incoming command email to
linux-...@vger.kernel.org, syzkall...@googlegroups.com.

***

Subject: Test
Author: radoslaw...@gmail.com

#syz test https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git fe46a7dd189e25604716c03576d05ac8a5209743

---
include/linux/sched.h | 7 +++++++
kernel/sched/core.c | 1 +
kernel/sched/rt.c | 26 +++++++++++++++++++++++---
kernel/signal.c | 9 +++++++++
4 files changed, 40 insertions(+), 3 deletions(-)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 17cb0761ff65..123bc16ad3d0 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1121,6 +1121,13 @@ struct task_struct {
size_t sas_ss_size;
unsigned int sas_ss_flags;

+ /*
+ * Number of signals received by an RT task between scheduling ticks.
+ * This counter is used to throttle RT tasks when too many signals
+ * (e.g., POSIX timers) are sent to the task, which can cause an RCU stall.
+ */
+ atomic_t rt_signals_recv_count; /* used outside of the rq lock */
+
struct callback_head *task_works;

#ifdef CONFIG_AUDIT
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index d44efa0d0611..9def826bd35f 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -4779,6 +4779,7 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p)
p->policy = SCHED_NORMAL;
p->static_prio = NICE_TO_PRIO(0);
p->rt_priority = 0;
+ atomic_set(&p->rt_signals_recv_count, 0);
} else if (PRIO_TO_NICE(p->static_prio) < 0)
p->static_prio = NICE_TO_PRIO(0);

diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
index 3261b067b67e..9b22d67d1746 100644
--- a/kernel/sched/rt.c
+++ b/kernel/sched/rt.c
@@ -24,6 +24,15 @@ int sysctl_sched_rt_period = 1000000;
*/
int sysctl_sched_rt_runtime = 950000;

+/*
+ * To avoid an RCU stall due to a large number of signals received by RT tasks
+ * (e.g., POSIX timers), the RT task needs to be throttled.
+ * When the number of signals received by an RT task during a scheduling
+ * tick period exceeds the threshold, the RT task will be throttled.
+ * The value of 100 has not been thoroughly tested and may need adjustment.
+ */
+#define RT_RECV_SGINAL_THROTTLE_THRESHOLD 100
+
#ifdef CONFIG_SYSCTL
static int sysctl_sched_rr_timeslice = (MSEC_PER_SEC * RR_TIMESLICE) / HZ;
static int sched_rt_handler(struct ctl_table *table, int write, void *buffer,
@@ -951,7 +960,7 @@ static inline int rt_se_prio(struct sched_rt_entity *rt_se)
return rt_task_of(rt_se)->prio;
}

-static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq)
+static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq, int rt_signal_recv)
{
u64 runtime = sched_rt_runtime(rt_rq);

@@ -966,7 +975,15 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq)
if (runtime == RUNTIME_INF)
return 0;

- if (rt_rq->rt_time > runtime) {
+ /*
+ * When a large number of signals are sent to this task (e.g., POSIX timers)
+ * the delta time deviates significantly from real time due to the overhead
+ * of handling signals. For RT tasks, this can cause an RCU stall.
+ * To avoid this, throttle the task when the number of signals received
+ * exceeds a certain threshold.
+ */
+ if (rt_rq->rt_time > runtime ||
+ rt_signal_recv >= RT_RECV_SGINAL_THROTTLE_THRESHOLD) {
struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);

/*
@@ -1021,7 +1038,9 @@ static void update_curr_rt(struct rq *rq)
if (sched_rt_runtime(rt_rq) != RUNTIME_INF) {
raw_spin_lock(&rt_rq->rt_runtime_lock);
rt_rq->rt_time += delta_exec;
- exceeded = sched_rt_runtime_exceeded(rt_rq);
+ exceeded = sched_rt_runtime_exceeded(
+ rt_rq,
+ atomic_read(&curr->rt_signals_recv_count));
if (exceeded)
resched_curr(rq);
raw_spin_unlock(&rt_rq->rt_runtime_lock);
@@ -1029,6 +1048,7 @@ static void update_curr_rt(struct rq *rq)
do_start_rt_bandwidth(sched_rt_bandwidth(rt_rq));
}
}
+ atomic_set(&curr->rt_signals_recv_count, 0);
}

static void
diff --git a/kernel/signal.c b/kernel/signal.c
index bdca529f0f7b..d58e0ba9336c 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -629,6 +629,15 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask,
bool resched_timer = false;
int signr;

+ /*
+ * To prevent an RCU stall due to receiving too many signals by RT tasks,
+ * count all signals regardless of their type.
+ * Based on this counter, the RT scheduler will decide whether the task
+ * should be throttled or not.
+ */
+ if (tsk->policy == SCHED_FIFO || tsk->policy == SCHED_RR)
+ atomic_inc(&tsk->rt_signals_recv_count);
+
/* We only dequeue private signals from ourselves, we don't let
* signalfd steal them
*/
--
2.43.0