[PATCH] bpf: Fix infinite loop in pcpu_freelist push with one possible CPU
From: Hui Su
Date: Thu Aug 06 2026 - 13:58:53 EST
__pcpu_freelist_push() can loop forever when only one CPU is possible
and an NMI re-enters pcpu_freelist_push() while the interrupted context
holds that CPU's freelist lock.
After the current-CPU fast path fails, the fallback loop walks
cpu_possible_mask while skipping the current CPU. With CONFIG_SMP=n, or
when an SMP kernel is limited to one possible CPU with nr_cpus=1 or
possible_cpus=1, there are no other possible CPUs to examine. The loop
therefore makes no lock acquisition attempt and can never make progress.
The following stack was observed on a UP system:
NMI context:
pcpu_freelist_push
free_htab_elem
htab_map_delete_elem
[perf-event BPF program]
__perf_event_overflow
perf_event_nmi_handler
exc_nmi
Interrupted context:
__pcpu_freelist_push
pcpu_freelist_push
free_htab_elem
htab_map_delete_elem
[raw_tp/sys_enter BPF program]
__bpf_trace_sys_enter
do_syscall_64
raw_res_spin_lock() detects the same-CPU recursive acquisition and
returns -EDEADLK, but the subsequent fallback loop has no candidate head
on a system with one possible CPU.
Restore the extra fallback head that existed before the rqspinlock
conversion. Keep the current-CPU fast path, then try the other possible
CPUs and finally the extra head. The additional head lets a push, which
cannot fail without losing a preallocated element, make progress when the
only per-CPU head is held by the interrupted context.
Also check the extra head from the pop path so that nodes placed there
can be reused.
Fixes: f2ac0e5d1c4d ("bpf: Convert percpu_freelist.c to rqspinlock")
Cc: stable@xxxxxxxxxxxxxxx
Signed-off-by: Hui Su <sh_def@xxxxxxx>
---
kernel/bpf/percpu_freelist.c | 35 +++++++++++++++++++++++++++--------
kernel/bpf/percpu_freelist.h | 1 +
2 files changed, 28 insertions(+), 8 deletions(-)
diff --git a/kernel/bpf/percpu_freelist.c b/kernel/bpf/percpu_freelist.c
index 632762b57299..06ce588d13a3 100644
--- a/kernel/bpf/percpu_freelist.c
+++ b/kernel/bpf/percpu_freelist.c
@@ -17,6 +17,8 @@ int pcpu_freelist_init(struct pcpu_freelist *s)
raw_res_spin_lock_init(&head->lock);
head->first = NULL;
}
+ raw_res_spin_lock_init(&s->extralist.lock);
+ s->extralist.first = NULL;
return 0;
}
@@ -46,22 +48,28 @@ void __pcpu_freelist_push(struct pcpu_freelist *s,
struct pcpu_freelist_node *node)
{
struct pcpu_freelist_head *head;
- int cpu;
+ int cpu, this_cpu;
if (___pcpu_freelist_push(this_cpu_ptr(s->freelist), node))
return;
+ this_cpu = raw_smp_processor_id();
while (true) {
- for_each_cpu_wrap(cpu, cpu_possible_mask, raw_smp_processor_id()) {
- if (cpu == raw_smp_processor_id())
+ for_each_cpu_wrap(cpu, cpu_possible_mask, this_cpu) {
+ if (cpu == this_cpu)
continue;
+
head = per_cpu_ptr(s->freelist, cpu);
- if (raw_res_spin_lock(&head->lock))
- continue;
- pcpu_freelist_push_node(head, node);
- raw_res_spin_unlock(&head->lock);
- return;
+ if (___pcpu_freelist_push(head, node))
+ return;
}
+
+ /*
+ * Push cannot fail. Use the extra list when none of the
+ * per-CPU freelists can accept the node.
+ */
+ if (___pcpu_freelist_push(&s->extralist, node))
+ return;
}
}
@@ -117,6 +125,17 @@ static struct pcpu_freelist_node *___pcpu_freelist_pop(struct pcpu_freelist *s)
}
raw_res_spin_unlock(&head->lock);
}
+
+ /* Per-CPU lists are empty or unavailable, try the extra list. */
+ head = &s->extralist;
+ if (!READ_ONCE(head->first))
+ return NULL;
+ if (raw_res_spin_lock(&head->lock))
+ return NULL;
+ node = head->first;
+ if (node)
+ WRITE_ONCE(head->first, node->next);
+ raw_res_spin_unlock(&head->lock);
return node;
}
diff --git a/kernel/bpf/percpu_freelist.h b/kernel/bpf/percpu_freelist.h
index 914798b74967..980cf2884fd2 100644
--- a/kernel/bpf/percpu_freelist.h
+++ b/kernel/bpf/percpu_freelist.h
@@ -14,6 +14,7 @@ struct pcpu_freelist_head {
struct pcpu_freelist {
struct pcpu_freelist_head __percpu *freelist;
+ struct pcpu_freelist_head extralist;
};
struct pcpu_freelist_node {
--
2.43.0