introduce support for XDP programs in CPUMAP

Similar to what David Ahern proposed in [1] for DEVMAPs, introduce the capability to attach and run a XDP program to CPUMAP entries.
The idea behind this feature is to add the possibility to define on which CPU run the eBPF program if the underlying hw does not support RSS.
I respin patch 1/6 from a previous series sent by David [2].
The functionality has been tested on Marvell Espressobin, i40e and mlx5. Detailed tests results can be found here:
https://github.com/xdp-project/xdp-project/blob/master/areas/cpumap/cpumap04-map-xdp-prog.org

Changes since v6:
- rebase on top of bpf-next
- move bpf_cpumap_val and bpf_prog in the first bpf_cpu_map_entry cache-line

Changes since v5:
- move bpf_prog_put() in put_cpu_map_entry()
- remove READ_ONCE(rcpu->prog) in cpu_map_bpf_prog_run_xdp
- rely on bpf_prog_get_type() instead of bpf_prog_get_type_dev() in __cpu_map_load_bpf_program()

Changes since v4:
- move xdp_clear_return_frame_no_direct inside rcu section
- update David Ahern's email address

Changes since v3:
- fix typo in commit message
- fix access to ctx->ingress_ifindex in cpumap bpf selftest

Changes since v2:
- improved comments
- fix return value in xdp_convert_buff_to_frame
- added patch 1/9: "cpumap: use non-locked version __ptr_ring_consume_batched"
- do not run kmem_cache_alloc_bulk if all frames have been consumed by the XDP program attached to the CPUMAP entry
- removed bpf_trace_printk in kselftest

Changes since v1:
- added performance test results
- added kselftest support
- fixed memory accounting with page_pool
- extended xdp_redirect_cpu_user.c to load an external program to perform redirect
- reported ifindex to attached eBPF program
- moved bpf_cpumap_val definition to include/uapi/linux/bpf.h

[1] https://patchwork.ozlabs.org/project/netdev/cover/[email protected]/
[2] https://patchwork.ozlabs.org/project/netdev/patch/[email protected]/

David Ahern (1):
net: refactor xdp_convert_buff_to_frame

Jesper Dangaard Brouer (1):
cpumap: use non-locked version __ptr_ring_consume_batched

Lorenzo Bianconi (7):
samples/bpf: xdp_redirect_cpu_user: do not update bpf maps in option
loop
cpumap: formalize map value as a named struct
bpf: cpumap: add the possibility to attach an eBPF program to cpumap
bpf: cpumap: implement XDP_REDIRECT for eBPF programs attached to map
entries
libbpf: add SEC name for xdp programs attached to CPUMAP samples/bpf: xdp_redirect_cpu: load a eBPF program on cpumap
selftest: add tests for XDP programs in CPUMAP entries

 include/linux/bpf.h                           |   6 +
 include/net/xdp.h                             |  41 ++--
 include/trace/events/xdp.h                    |  16 +-
 include/uapi/linux/bpf.h                      |  14 ++
 kernel/bpf/cpumap.c                           | 162 +++++++++++---
 net/core/dev.c                                |   9 +
 samples/bpf/xdp_redirect_cpu_kern.c           |  25 ++-
 samples/bpf/xdp_redirect_cpu_user.c           | 209 ++++++++++++++++--
 tools/include/uapi/linux/bpf.h                |  14 ++
 tools/lib/bpf/libbpf.c                        |   2 +
 .../bpf/prog_tests/xdp_cpumap_attach.c        |  70 ++++++
 .../bpf/progs/test_xdp_with_cpumap_helpers.c  |  36 +++
 12 files changed, 531 insertions(+), 73 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/prog_tests/xdp_cpumap_attach.c
 create mode 100644 tools/testing/selftests/bpf/progs/test_xdp_with_cpumap_helpers.c

2.26.2

From: Jesper Dangaard Brouer <[email protected]>

Commit 77361825bb01 ("bpf: cpumap use ptr_ring_consume_batched") changed away from using single frame ptr_ring dequeue (__ptr_ring_consume) to consume a batched, but it uses a locked version, which as the comment explain isn't needed.

Change to use the non-locked version __ptr_ring_consume_batched.

---
 kernel/bpf/cpumap.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index bd8658055c16..323c91c4fab0 100644
@@ -259,7 +259,7 @@ static int cpu_map_kthread_run(void *data)
 		 * kthread CPU pinned. Lockless access to ptr_ring
 		 * consume side valid as no-resize allowed of queue.
 		 */
-		n = ptr_ring_consume_batched(rcpu->queue, frames, CPUMAP_BATCH);
+		n = __ptr_ring_consume_batched(rcpu->queue, frames, CPUMAP_BATCH);
 
 		for (i = 0; i < n; i++) {
 			void *f = frames[i];

2.26.2

From: David Ahern <[email protected]>

Move the guts of xdp_convert_buff_to_frame to a new helper, xdp_update_frame_from_buff so it can be reused removing code duplication

---
 include/net/xdp.h | 35 ++++++++++++++++++++++-------------
 1 file changed, 22 insertions(+), 13 deletions(-)

index 609f819ed08b..5b383c450858 100644
@@ -121,39 +121,48 @@ void xdp_convert_frame_to_buff(struct xdp_frame *frame, struct xdp_buff *xdp)
 	xdp->frame_sz = frame->frame_sz;
 }
 
-/* Convert xdp_buff to xdp_frame */
 static inline
-struct xdp_frame *xdp_convert_buff_to_frame(struct xdp_buff *xdp)
+int xdp_update_frame_from_buff(struct xdp_buff *xdp,
+			       struct xdp_frame *xdp_frame)
 {
-	struct xdp_frame *xdp_frame;
-	int metasize;
-	int headroom;
-
-	if (xdp->rxq->mem.type == MEM_TYPE_XSK_BUFF_POOL)
-		return xdp_convert_zc_to_xdp_frame(xdp);
+	int metasize, headroom;
 
 	/* Assure headroom is available for storing info */
 	headroom = xdp->data - xdp->data_hard_start;
 	metasize = xdp->data - xdp->data_meta;
 	metasize = metasize > 0 ? metasize : 0;
 	if (unlikely((headroom - metasize) < sizeof(*xdp_frame)))
-		return NULL;
+		return -ENOSPC;
 
 	/* Catch if driver didn't reserve tailroom for skb_shared_info */
 	if (unlikely(xdp->data_end > xdp_data_hard_end(xdp))) {
 		XDP_WARN("Driver BUG: missing reserved tailroom");
-		return NULL;
+		return -ENOSPC;
 	}
 
-	/* Store info in top of packet */
-	xdp_frame = xdp->data_hard_start;
-
 	xdp_frame->data = xdp->data;
 	xdp_frame->len  = xdp->data_end - xdp->data;
 	xdp_frame->headroom = headroom - sizeof(*xdp_frame);
 	xdp_frame->metasize = metasize;
 	xdp_frame->frame_sz = xdp->frame_sz;
 
+	return 0;
+}
+
+/* Convert xdp_buff to xdp_frame */
+static inline
+struct xdp_frame *xdp_convert_buff_to_frame(struct xdp_buff *xdp)
+{
+	struct xdp_frame *xdp_frame;
+
+	if (xdp->rxq->mem.type == MEM_TYPE_XSK_BUFF_POOL)
+		return xdp_convert_zc_to_xdp_frame(xdp);
+
+	/* Store info in top of packet */
+	xdp_frame = xdp->data_hard_start;
+	if (unlikely(xdp_update_frame_from_buff(xdp, xdp_frame) < 0))
+		return NULL;
+
 	/* rxq only valid until napi_schedule ends, convert to xdp_mem_info */
 	xdp_frame->mem = xdp->rxq->mem;
 

2.26.2

Do not update xdp_redirect_cpu maps running while option loop but defer it after all available options have been parsed. This is a preliminary patch to pass the program name we want to attach to the map entries as a user option

---
 samples/bpf/xdp_redirect_cpu_user.c | 36 +++++++++++++++++++++--------
 1 file changed, 27 insertions(+), 9 deletions(-)

index f4e755e0dd73..6bb2d95cb26c 100644
@@ -681,6 +681,7 @@ int main(int argc, char **argv)
 	int add_cpu = -1;
 	int opt, err;
 	int prog_fd;
+	int *cpu, i;
 	__u32 qsize;
 
 	n_cpus = get_nprocs_conf();

@@ -716,6 +717,13 @@ int main(int argc, char **argv)
 	}
 	mark_cpus_unavailable();
 
+	cpu = malloc(n_cpus * sizeof(int));
+	if (!cpu) {
+		fprintf(stderr, "failed to allocate cpu array\n");
+		return EXIT_FAIL;
+	}
+	memset(cpu, 0, n_cpus * sizeof(int));
+
 	/* Parse commands line args */
 	while ((opt = getopt_long(argc, argv, "hSd:s:p:q:c:xzF",
 				  long_options, &longindex)) != -1) {

@@ -760,8 +768,7 @@ int main(int argc, char **argv)
 					errno, strerror(errno));
 				goto error;
 			}
-			create_cpu_entry(add_cpu, qsize, added_cpus, true);
-			added_cpus++;
+			cpu[added_cpus++] = add_cpu;
 			break;
 		case 'q':
 			qsize = atoi(optarg);

@@ -772,6 +779,7 @@ int main(int argc, char **argv)
 		case 'h':
 		error:
 		default:
+			free(cpu);
 			usage(argv, obj);
 			return EXIT_FAIL_OPTION;
 		}

@@ -784,16 +792,21 @@ int main(int argc, char **argv)
 	if (ifindex == -1) {
 		fprintf(stderr, "ERR: required option --dev missing\n");
 		usage(argv, obj);
-		return EXIT_FAIL_OPTION;
+		err = EXIT_FAIL_OPTION;
+		goto out;
 	}
 	/* Required option */
 	if (add_cpu == -1) {
 		fprintf(stderr, "ERR: required option --cpu missing\n");
 		fprintf(stderr, " Specify multiple --cpu option to add more\n");
 		usage(argv, obj);
-		return EXIT_FAIL_OPTION;
+		err = EXIT_FAIL_OPTION;
+		goto out;
 	}
 
+	for (i = 0; i < added_cpus; i++)
+		create_cpu_entry(cpu[i], qsize, i, true);
+
 	/* Remove XDP program when program is interrupted or killed */
 	signal(SIGINT, int_exit);
 	signal(SIGTERM, int_exit);

@@ -801,27 +814,32 @@ int main(int argc, char **argv)
 	prog = bpf_object__find_program_by_title(obj, prog_name);
 	if (!prog) {
 		fprintf(stderr, "bpf_object__find_program_by_title failed\n");
-		return EXIT_FAIL;
+		err = EXIT_FAIL;
+		goto out;
 	}
 
 	prog_fd = bpf_program__fd(prog);
 	if (prog_fd < 0) {
 		fprintf(stderr, "bpf_program__fd failed\n");
-		return EXIT_FAIL;
+		err = EXIT_FAIL;
+		goto out;
 	}
 
 	if (bpf_set_link_xdp_fd(ifindex, prog_fd, xdp_flags) < 0) {
 		fprintf(stderr, "link set xdp fd failed\n");
-		return EXIT_FAIL_XDP;
+		err = EXIT_FAIL_XDP;
+		goto out;
 	}
 
 	err = bpf_obj_get_info_by_fd(prog_fd, &info, &info_len);
 	if (err) {
 		printf("can't get prog info - %s\n", strerror(errno));
-		return err;
+		goto out;
 	}
 	prog_id = info.id;
 
 	stats_poll(interval, use_separators, prog_name, stress_mode);
-	return EXIT_OK;
+out:
+	free(cpu);
+	return err;
 }

2.26.2

As it has been already done for devmap, introduce 'struct bpf_cpumap_val' to formalize the expected values that can be passed in for a CPUMAP. Update cpumap code to use the struct.

---
 include/uapi/linux/bpf.h       |  9 +++++++++
 kernel/bpf/cpumap.c            | 28 +++++++++++++++-------------
 tools/include/uapi/linux/bpf.h |  9 +++++++++
 3 files changed, 33 insertions(+), 13 deletions(-)

index 5e386389913a..109623527358 100644
@@ -3849,6 +3849,15 @@ struct bpf_devmap_val {
 	} bpf_prog;
 };
 
+/* CPUMAP map-value layout
+ *
+ * The struct data-layout of map-value is a configuration interface.
+ * New members can only be added to the end of this structure.
+ */
+struct bpf_cpumap_val {
+	__u32 qsize;	/* queue size to remote target CPU */
+};
+
 enum sk_action {
 	SK_DROP = 0,
 	SK_PASS,

index 323c91c4fab0..ff48dc00e8d0 100644
@@ -52,7 +52,6 @@ struct xdp_bulk_queue {
 struct bpf_cpu_map_entry {
 	u32 cpu;    /* kthread CPU and map index */
 	int map_id; /* Back reference to map */
-	u32 qsize;  /* Queue size placeholder for map lookup */
 
 	/* XDP can run multiple RX-ring queues, need __percpu enqueue store */
 	struct xdp_bulk_queue __percpu *bulkq;

@@ -62,10 +61,13 @@ struct bpf_cpu_map_entry {
 	/* Queue with potential multi-producers, and single-consumer kthread */
 	struct ptr_ring *queue;
 	struct task_struct *kthread;
-	struct work_struct kthread_stop_wq;
+
+	struct bpf_cpumap_val value;
 
 	atomic_t refcnt; /* Control when this struct can be free'ed */
 	struct rcu_head rcu;
+
+	struct work_struct kthread_stop_wq;
 };
 
 struct bpf_cpu_map {

@@ -307,8 +309,8 @@ static int cpu_map_kthread_run(void *data)
 	return 0;
 }
 
-static struct bpf_cpu_map_entry *__cpu_map_entry_alloc(u32 qsize, u32 cpu,
-						       int map_id)
+static struct bpf_cpu_map_entry *
+__cpu_map_entry_alloc(struct bpf_cpumap_val *value, u32 cpu, int map_id)
 {
 	gfp_t gfp = GFP_KERNEL | __GFP_NOWARN;
 	struct bpf_cpu_map_entry *rcpu;

@@ -338,13 +340,13 @@ static struct bpf_cpu_map_entry *__cpu_map_entry_alloc(u32 qsize, u32 cpu,
 	if (!rcpu->queue)
 		goto free_bulkq;
 
-	err = ptr_ring_init(rcpu->queue, qsize, gfp);
+	err = ptr_ring_init(rcpu->queue, value->qsize, gfp);
 	if (err)
 		goto free_queue;
 
 	rcpu->cpu    = cpu;
 	rcpu->map_id = map_id;
-	rcpu->qsize  = qsize;
+	rcpu->value.qsize  = value->qsize;
 
 	/* Setup kthread */
 	rcpu->kthread = kthread_create_on_node(cpu_map_kthread_run, rcpu, numa,

@@ -437,12 +439,12 @@ static int cpu_map_update_elem(struct bpf_map *map, void *key, void *value,
 			       u64 map_flags)
 {
 	struct bpf_cpu_map *cmap = container_of(map, struct bpf_cpu_map, map);
+	struct bpf_cpumap_val cpumap_value = {};
 	struct bpf_cpu_map_entry *rcpu;
-
 	/* Array index key correspond to CPU number */
 	u32 key_cpu = *(u32 *)key;
-	/* Value is the queue size */
-	u32 qsize = *(u32 *)value;
+
+	memcpy(&cpumap_value, value, map->value_size);
 
 	if (unlikely(map_flags > BPF_EXIST))
 		return -EINVAL;

@@ -450,18 +452,18 @@ static int cpu_map_update_elem(struct bpf_map *map, void *key, void *value,
 		return -E2BIG;
 	if (unlikely(map_flags == BPF_NOEXIST))
 		return -EEXIST;
-	if (unlikely(qsize > 16384)) /* sanity limit on qsize */
+	if (unlikely(cpumap_value.qsize > 16384)) /* sanity limit on qsize */
 		return -EOVERFLOW;
 
 	/* Make sure CPU is a valid possible cpu */
 	if (key_cpu >= nr_cpumask_bits || !cpu_possible(key_cpu))
 		return -ENODEV;
 
-	if (qsize == 0) {
+	if (cpumap_value.qsize == 0) {
 		rcpu = NULL; /* Same as deleting */
 	} else {
 		/* Updating qsize cause re-allocation of bpf_cpu_map_entry */
-		rcpu = __cpu_map_entry_alloc(qsize, key_cpu, map->id);
+		rcpu = __cpu_map_entry_alloc(&cpumap_value, key_cpu, map->id);
 		if (!rcpu)
 			return -ENOMEM;
 		rcpu->cmap = cmap;

@@ -523,7 +525,7 @@ static void *cpu_map_lookup_elem(struct bpf_map *map, void *key)
 	struct bpf_cpu_map_entry *rcpu =
 		__cpu_map_lookup_elem(map, *(u32 *)key);
 
-	return rcpu ? &rcpu->qsize : NULL;
+	return rcpu ? &rcpu->value : NULL;
 }
 
 static int cpu_map_get_next_key(struct bpf_map *map, void *key, void *next_key)

index 5e386389913a..109623527358 100644
@@ -3849,6 +3849,15 @@ struct bpf_devmap_val {
 	} bpf_prog;
 };
 
+/* CPUMAP map-value layout
+ *
+ * The struct data-layout of map-value is a configuration interface.
+ * New members can only be added to the end of this structure.
+ */
+struct bpf_cpumap_val {
+	__u32 qsize;	/* queue size to remote target CPU */
+};
+
 enum sk_action {
 	SK_DROP = 0,
 	SK_PASS,

2.26.2

Introduce the capability to attach an eBPF program to cpumap entries. The idea behind this feature is to add the possibility to define on which CPU run the eBPF program if the underlying hw does not support RSS. Current supported verdicts are XDP_DROP and XDP_PASS.

This patch has been tested on Marvell ESPRESSObin using xdp_redirect_cpu sample available in the kernel tree to identify possible performance regressions. Results show there are no observable differences in packet-per-second:

$./xdp_redirect_cpu --progname xdp_cpu_map0 --dev eth0 --cpu 1
rx: 354.8 Kpps
rx: 356.0 Kpps
rx: 356.8 Kpps
rx: 356.3 Kpps
rx: 356.6 Kpps
rx: 356.6 Kpps
rx: 356.7 Kpps
rx: 355.8 Kpps
rx: 356.8 Kpps
rx: 356.8 Kpps

---
 include/linux/bpf.h            |   6 ++
 include/net/xdp.h              |   5 ++
 include/trace/events/xdp.h     |  14 ++--
 include/uapi/linux/bpf.h       |   5 ++
 kernel/bpf/cpumap.c            | 121 +++++++++++++++++++++++++++++----
 net/core/dev.c                 |   9 +++
 tools/include/uapi/linux/bpf.h |   5 ++
 7 files changed, 148 insertions(+), 17 deletions(-)

index c67c88ad35f8..54ad426dbea1 100644
@@ -1272,6 +1272,7 @@ struct bpf_cpu_map_entry *__cpu_map_lookup_elem(struct bpf_map *map, u32 key);
 void __cpu_map_flush(void);
 int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu, struct xdp_buff *xdp,
 		    struct net_device *dev_rx);
+bool cpu_map_prog_allowed(struct bpf_map *map);
 
 /* Return map's numa specified by userspace */
 static inline int bpf_map_attr_numa_node(const union bpf_attr *attr)

@@ -1432,6 +1433,11 @@ static inline int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu,
 	return 0;
 }
 
+static inline bool cpu_map_prog_allowed(struct bpf_map *map)
+{
+	return false;
+}
+
 static inline struct bpf_prog *bpf_prog_get_type_path(const char *name,
 				enum bpf_prog_type type)
 {

index 5b383c450858..83b9e0142b52 100644
@@ -98,6 +98,11 @@ struct xdp_frame {
 	struct net_device *dev_rx; /* used by cpumap */
 };
 
+struct xdp_cpumap_stats {
+	unsigned int pass;
+	unsigned int drop;
+};
+
 /* Clear kernel pointers in xdp_frame */
 static inline void xdp_scrub_frame(struct xdp_frame *frame)
 {

index b73d3e141323..e2c99f5bee39 100644
@@ -177,9 +177,9 @@ DEFINE_EVENT(xdp_redirect_template, xdp_redirect_map_err,
 TRACE_EVENT(xdp_cpumap_kthread,
 
 	TP_PROTO(int map_id, unsigned int processed,  unsigned int drops,
-		 int sched),
+		 int sched, struct xdp_cpumap_stats *xdp_stats),
 
-	TP_ARGS(map_id, processed, drops, sched),
+	TP_ARGS(map_id, processed, drops, sched, xdp_stats),
 
 	TP_STRUCT__entry(
 		__field(int, map_id)

@@ -188,6 +188,8 @@ TRACE_EVENT(xdp_cpumap_kthread,
 		__field(unsigned int, drops)
 		__field(unsigned int, processed)
 		__field(int, sched)
+		__field(unsigned int, xdp_pass)
+		__field(unsigned int, xdp_drop)
 	),
 
 	TP_fast_assign(

@@ -197,16 +199,20 @@ TRACE_EVENT(xdp_cpumap_kthread,
 		__entry->drops		= drops;
 		__entry->processed	= processed;
 		__entry->sched	= sched;
+		__entry->xdp_pass	= xdp_stats->pass;
+		__entry->xdp_drop	= xdp_stats->drop;
 	),
 
 	TP_printk("kthread"
 		  " cpu=%d map_id=%d action=%s"
 		  " processed=%u drops=%u"
-		  " sched=%d",
+		  " sched=%d"
+		  " xdp_pass=%u xdp_drop=%u",
 		  __entry->cpu, __entry->map_id,
 		  __print_symbolic(__entry->act, __XDP_ACT_SYM_TAB),
 		  __entry->processed, __entry->drops,
-		  __entry->sched)
+		  __entry->sched,
+		  __entry->xdp_pass, __entry->xdp_drop)
 );
 
 TRACE_EVENT(xdp_cpumap_enqueue,

index 109623527358..c010b57fce3f 100644
@@ -227,6 +227,7 @@ enum bpf_attach_type {
 	BPF_CGROUP_INET6_GETSOCKNAME,
 	BPF_XDP_DEVMAP,
 	BPF_CGROUP_INET_SOCK_RELEASE,
+	BPF_XDP_CPUMAP,
 	__MAX_BPF_ATTACH_TYPE
 };
 

@@ -3856,6 +3857,10 @@ struct bpf_devmap_val {
  */
 struct bpf_cpumap_val {
 	__u32 qsize;	/* queue size to remote target CPU */
+	union {
+		int   fd;	/* prog fd on map write */
+		__u32 id;	/* prog id on map read */
+	} bpf_prog;
 };
 
 enum sk_action {

index b61075828358..b820527f0a8d 100644
@@ -5448,6 +5448,8 @@ static int generic_xdp_install(struct net_device *dev, struct netdev_bpf *xdp)
 		for (i = 0; i < new->aux->used_map_cnt; i++) {
 			if (dev_map_can_have_prog(new->aux->used_maps[i]))
 				return -EINVAL;
+			if (cpu_map_prog_allowed(new->aux->used_maps[i]))
+				return -EINVAL;
 		}
 	}
 

@@ -8875,6 +8877,13 @@ int dev_change_xdp_fd(struct net_device *dev, struct netlink_ext_ack *extack,
 			return -EINVAL;
 		}
 
+		if (prog->expected_attach_type == BPF_XDP_CPUMAP) {
+			NL_SET_ERR_MSG(extack,
+				       "BPF_XDP_CPUMAP programs can not be attached to a device");
+			bpf_prog_put(prog);
+			return -EINVAL;
+		}
+
 		/* prog->aux->id may be 0 for orphaned device-bound progs */
 		if (prog->aux->id && prog->aux->id == prog_id) {
 			bpf_prog_put(prog);

index 109623527358..c010b57fce3f 100644
@@ -227,6 +227,7 @@ enum bpf_attach_type {
 	BPF_CGROUP_INET6_GETSOCKNAME,
 	BPF_XDP_DEVMAP,
 	BPF_CGROUP_INET_SOCK_RELEASE,
+	BPF_XDP_CPUMAP,
 	__MAX_BPF_ATTACH_TYPE
 };
 

@@ -3856,6 +3857,10 @@ struct bpf_devmap_val {
  */
 struct bpf_cpumap_val {
 	__u32 qsize;	/* queue size to remote target CPU */
+	union {
+		int   fd;	/* prog fd on map write */
+		__u32 id;	/* prog id on map read */
+	} bpf_prog;
 };
 
 enum sk_action {

2.26.2

Introduce XDP_REDIRECT support for eBPF programs attached to cpumap entries.
This patch has been tested on Marvell ESPRESSObin using a modified version of xdp_redirect_cpu sample in order to attach a XDP program to CPUMAP entries to perform a redirect on the mvneta interface. In particular the following scenario has been tested:

rq (cpu0) --> mvneta - XDP_REDIRECT (cpu0) --> CPUMAP - XDP_REDIRECT (cpu1) --> mvneta

$./xdp_redirect_cpu -p xdp_cpu_map0 -d eth0 -c 1 -e xdp_redirect \
-f xdp_redirect_kern.o -m tx_port -r eth0

tx: 285.2 Kpps rx: 285.2 Kpps

Attaching a simple XDP program on eth0 to perform XDP_TX gives comparable results:

tx: 288.4 Kpps rx: 288.4 Kpps

---
 include/net/xdp.h          |  1 +
 include/trace/events/xdp.h |  6 ++++--
 kernel/bpf/cpumap.c        | 17 +++++++++++++++--
 3 files changed, 20 insertions(+), 4 deletions(-)

index 83b9e0142b52..5be0d4d65b94 100644
@@ -99,6 +99,7 @@ struct xdp_frame {
 };
 
 struct xdp_cpumap_stats {
+	unsigned int redirect;
 	unsigned int pass;
 	unsigned int drop;
 };

index e2c99f5bee39..cd24e8a59529 100644
@@ -190,6 +190,7 @@ TRACE_EVENT(xdp_cpumap_kthread,
 		__field(int, sched)
 		__field(unsigned int, xdp_pass)
 		__field(unsigned int, xdp_drop)
+		__field(unsigned int, xdp_redirect)
 	),
 
 	TP_fast_assign(

@@ -201,18 +202,19 @@ TRACE_EVENT(xdp_cpumap_kthread,
 		__entry->sched	= sched;
 		__entry->xdp_pass	= xdp_stats->pass;
 		__entry->xdp_drop	= xdp_stats->drop;
+		__entry->xdp_redirect	= xdp_stats->redirect;
 	),
 
 	TP_printk("kthread"
 		  " cpu=%d map_id=%d action=%s"
 		  " processed=%u drops=%u"
 		  " sched=%d"
-		  " xdp_pass=%u xdp_drop=%u",
+		  " xdp_pass=%u xdp_drop=%u xdp_redirect=%u",
 		  __entry->cpu, __entry->map_id,
 		  __print_symbolic(__entry->act, __XDP_ACT_SYM_TAB),
 		  __entry->processed, __entry->drops,
 		  __entry->sched,
-		  __entry->xdp_pass, __entry->xdp_drop)
+		  __entry->xdp_pass, __entry->xdp_drop, __entry->xdp_redirect)
 );
 
 TRACE_EVENT(xdp_cpumap_enqueue,

index b3a8aea81ee5..4c95d0615ca2 100644
@@ -239,7 +239,7 @@ static int cpu_map_bpf_prog_run_xdp(struct bpf_cpu_map_entry *rcpu,
 	if (!rcpu->prog)
 		return n;
 
-	rcu_read_lock();
+	rcu_read_lock_bh();
 
 	xdp_set_return_frame_no_direct();
 	xdp.rxq = &rxq;

@@ -267,6 +267,16 @@ static int cpu_map_bpf_prog_run_xdp(struct bpf_cpu_map_entry *rcpu,
 				stats->pass++;
 			}
 			break;
+		case XDP_REDIRECT:
+			err = xdp_do_redirect(xdpf->dev_rx, &xdp,
+					      rcpu->prog);
+			if (unlikely(err)) {
+				xdp_return_frame(xdpf);
+				stats->drop++;
+			} else {
+				stats->redirect++;
+			}
+			break;
 		default:
 			bpf_warn_invalid_xdp_action(act);
 			/* fallthrough */

@@ -277,9 +287,12 @@ static int cpu_map_bpf_prog_run_xdp(struct bpf_cpu_map_entry *rcpu,
 		}
 	}
 
+	if (stats->redirect)
+		xdp_do_flush_map();
+
 	xdp_clear_return_frame_no_direct();
 
-	rcu_read_unlock();
+	rcu_read_unlock_bh(); /* resched point, may call do_softirq() */
 
 	return nframes;
 }

2.26.2

As for DEVMAP, support SEC("xdp_cpumap/") as a short cut for loading the program with type BPF_PROG_TYPE_XDP and expected attach type BPF_XDP_CPUMAP.

---
 tools/lib/bpf/libbpf.c | 2 ++
 1 file changed, 2 insertions(+)

index 4489f95f1d1a..f55fd8a5c008 100644
@@ -6912,6 +6912,8 @@ static const struct bpf_sec_def section_defs[] = {
 		.attach_fn = attach_iter),
 	BPF_EAPROG_SEC("xdp_devmap/",		BPF_PROG_TYPE_XDP,
 						BPF_XDP_DEVMAP),
+	BPF_EAPROG_SEC("xdp_cpumap/",		BPF_PROG_TYPE_XDP,
+						BPF_XDP_CPUMAP),
 	BPF_PROG_SEC("xdp",			BPF_PROG_TYPE_XDP),
 	BPF_PROG_SEC("perf_event",		BPF_PROG_TYPE_PERF_EVENT),
 	BPF_PROG_SEC("lwt_in",			BPF_PROG_TYPE_LWT_IN),

2.26.2

Extend xdp_redirect_cpu_{usr,kern}.c adding the possibility to load a XDP program on cpumap entries. The following options have been added:
- mprog-name: cpumap entry program name
- mprog-filename: cpumap entry program filename
- redirect-device: output interface if the cpumap program performs a XDP_REDIRECT to an egress interface
- redirect-map: bpf map used to perform XDP_REDIRECT to an egress interface
- mprog-disable: disable loading XDP program on cpumap entries

Add xdp_pass, xdp_drop, xdp_redirect stats accounting

---
 samples/bpf/xdp_redirect_cpu_kern.c |  25 ++--
 samples/bpf/xdp_redirect_cpu_user.c | 175 +++++++++++++++++++++++++---
 2 files changed, 178 insertions(+), 22 deletions(-)

index 2baf8db1f7e7..8255025dea97 100644
@@ -21,7 +21,7 @@
 struct {
 	__uint(type, BPF_MAP_TYPE_CPUMAP);
 	__uint(key_size, sizeof(u32));
-	__uint(value_size, sizeof(u32));
+	__uint(value_size, sizeof(struct bpf_cpumap_val));
 	__uint(max_entries, MAX_CPUS);
 } cpu_map SEC(".maps");
 

@@ -30,6 +30,9 @@ struct datarec {
 	__u64 processed;
 	__u64 dropped;
 	__u64 issue;
+	__u64 xdp_pass;
+	__u64 xdp_drop;
+	__u64 xdp_redirect;
 };
 
 /* Count RX packets, as XDP bpf_prog doesn't get direct TX-success

@@ -692,13 +695,16 @@ int trace_xdp_cpumap_enqueue(struct cpumap_enqueue_ctx *ctx)
  * Code in:         kernel/include/trace/events/xdp.h
  */
 struct cpumap_kthread_ctx {
-	u64 __pad;		// First 8 bytes are not accessible by bpf code
-	int map_id;		//	offset:8;  size:4; signed:1;
-	u32 act;		//	offset:12; size:4; signed:0;
-	int cpu;		//	offset:16; size:4; signed:1;
-	unsigned int drops;	//	offset:20; size:4; signed:0;
-	unsigned int processed;	//	offset:24; size:4; signed:0;
-	int sched;		//	offset:28; size:4; signed:1;
+	u64 __pad;			// First 8 bytes are not accessible
+	int map_id;			//	offset:8;  size:4; signed:1;
+	u32 act;			//	offset:12; size:4; signed:0;
+	int cpu;			//	offset:16; size:4; signed:1;
+	unsigned int drops;		//	offset:20; size:4; signed:0;
+	unsigned int processed;		//	offset:24; size:4; signed:0;
+	int sched;			//	offset:28; size:4; signed:1;
+	unsigned int xdp_pass;		//	offset:32; size:4; signed:0;
+	unsigned int xdp_drop;		//	offset:36; size:4; signed:0;
+	unsigned int xdp_redirect;	//	offset:40; size:4; signed:0;
 };
 
 SEC("tracepoint/xdp/xdp_cpumap_kthread")

@@ -712,6 +718,9 @@ int trace_xdp_cpumap_kthread(struct cpumap_kthread_ctx *ctx)
 		return 0;
 	rec->processed += ctx->processed;
 	rec->dropped   += ctx->drops;
+	rec->xdp_pass  += ctx->xdp_pass;
+	rec->xdp_drop  += ctx->xdp_drop;
+	rec->xdp_redirect  += ctx->xdp_redirect;
 
 	/* Count times kthread yielded CPU via schedule call */
 	if (ctx->sched)

2.26.2

Similar to what have been done for DEVMAP, introduce tests to verify ability to add a XDP program to an entry in a CPUMAP. Verify CPUMAP programs can not be attached to devices as a normal XDP program, and only programs with BPF_XDP_CPUMAP attach type can be loaded in a CPUMAP.

---
 .../bpf/prog_tests/xdp_cpumap_attach.c        | 70 +++++++++++++++++++
 .../bpf/progs/test_xdp_with_cpumap_helpers.c  | 36 ++++++++++
 2 files changed, 106 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/prog_tests/xdp_cpumap_attach.c
 create mode 100644 tools/testing/selftests/bpf/progs/test_xdp_with_cpumap_helpers.c

new file mode 100644
index 000000000000..0176573fe4e7
@@ -0,0 +1,70 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <uapi/linux/bpf.h>
+#include <linux/if_link.h>
+#include <test_progs.h>
+
+#include "test_xdp_with_cpumap_helpers.skel.h"
+
+#define IFINDEX_LO	1
+
+void test_xdp_with_cpumap_helpers(void)
+{
+	struct test_xdp_with_cpumap_helpers *skel;
+	struct bpf_prog_info info = {};
+	struct bpf_cpumap_val val = {
+		.qsize = 192,
+	};
+	__u32 duration = 0, idx = 0;
+	__u32 len = sizeof(info);
+	int err, prog_fd, map_fd;
+
+	skel = test_xdp_with_cpumap_helpers__open_and_load();
+	if (CHECK_FAIL(!skel)) {
+		perror("test_xdp_with_cpumap_helpers__open_and_load");
+		return;
+	}
+
+	/* can not attach program with cpumaps that allow programs
+	 * as xdp generic
+	 */
+	prog_fd = bpf_program__fd(skel->progs.xdp_redir_prog);
+	err = bpf_set_link_xdp_fd(IFINDEX_LO, prog_fd, XDP_FLAGS_SKB_MODE);
+	CHECK(err == 0, "Generic attach of program with 8-byte CPUMAP",
+	      "should have failed\n");
+
+	prog_fd = bpf_program__fd(skel->progs.xdp_dummy_cm);
+	map_fd = bpf_map__fd(skel->maps.cpu_map);
+	err = bpf_obj_get_info_by_fd(prog_fd, &info, &len);
+	if (CHECK_FAIL(err))
+		goto out_close;
+
+	val.bpf_prog.fd = prog_fd;
+	err = bpf_map_update_elem(map_fd, &idx, &val, 0);
+	CHECK(err, "Add program to cpumap entry", "err %d errno %d\n",
+	      err, errno);
+
+	err = bpf_map_lookup_elem(map_fd, &idx, &val);
+	CHECK(err, "Read cpumap entry", "err %d errno %d\n", err, errno);
+	CHECK(info.id != val.bpf_prog.id, "Expected program id in cpumap entry",
+	      "expected %u read %u\n", info.id, val.bpf_prog.id);
+
+	/* can not attach BPF_XDP_CPUMAP program to a device */
+	err = bpf_set_link_xdp_fd(IFINDEX_LO, prog_fd, XDP_FLAGS_SKB_MODE);
+	CHECK(err == 0, "Attach of BPF_XDP_CPUMAP program",
+	      "should have failed\n");
+
+	val.qsize = 192;
+	val.bpf_prog.fd = bpf_program__fd(skel->progs.xdp_dummy_prog);
+	err = bpf_map_update_elem(map_fd, &idx, &val, 0);
+	CHECK(err == 0, "Add non-BPF_XDP_CPUMAP program to cpumap entry",
+	      "should have failed\n");
+
+out_close:
+	test_xdp_with_cpumap_helpers__destroy(skel);
+}
+
+void test_xdp_cpumap_attach(void)
+{
+	if (test__start_subtest("cpumap_with_progs"))
+		test_xdp_with_cpumap_helpers();
+}

new file mode 100644
index 000000000000..59ee4f182ff8
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/bpf.h>
+#include <bpf/bpf_helpers.h>
+
+#define IFINDEX_LO	1
+
+struct {
+	__uint(type, BPF_MAP_TYPE_CPUMAP);
+	__uint(key_size, sizeof(__u32));
+	__uint(value_size, sizeof(struct bpf_cpumap_val));
+	__uint(max_entries, 4);
+} cpu_map SEC(".maps");
+
+SEC("xdp_redir")
+int xdp_redir_prog(struct xdp_md *ctx)
+{
+	return bpf_redirect_map(&cpu_map, 1, 0);
+}
+
+SEC("xdp_dummy")
+int xdp_dummy_prog(struct xdp_md *ctx)
+{
+	return XDP_PASS;
+}
+
+SEC("xdp_cpumap/dummy_cm")
+int xdp_dummy_cm(struct xdp_md *ctx)
+{
+	if (ctx->ingress_ifindex == IFINDEX_LO)
+		return XDP_DROP;
+
+	return XDP_PASS;
+}
+
+char _license[] SEC("license") = "GPL";

2.26.2

fyi. I'm waiting on Daniel to do one more look, since he commented in the past.

ack, thx. I have just figured out today that v6 is not applying anymore.

Regards,
Lorenzo

LGTM, I've applied it, thanks!

This started showing up with when running ./test_progs from recent bpf-next (bfdfa51702de). Any chance it is related?

[ 2950.440613] =============================================

[ 3073.281578] INFO: task cpumap/0/map:26:536 blocked for more than 860 seconds.
[ 3073.285492]       Tainted: G        W         5.8.0-rc4-01471-g15d51f3a516b #814
[ 3073.289177] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 3073.293021] cpumap/0/map:26 D    0   536      2 0x00004000
[ 3073.295755] Call Trace:
[ 3073.297143]  __schedule+0x5ad/0xf10
[ 3073.299032]  ? pci_mmcfg_check_reserved+0xd0/0xd0
[ 3073.301416]  ? static_obj+0x31/0x80
[ 3073.303277]  ? mark_held_locks+0x24/0x90
[ 3073.305313]  ? cpu_map_update_elem+0x6d0/0x6d0
[ 3073.307544]  schedule+0x6f/0x160
[ 3073.309282]  schedule_preempt_disabled+0x14/0x20
[ 3073.311593]  kthread+0x175/0x240
[ 3073.313299]  ? kthread_create_on_node+0xd0/0xd0
[ 3073.315106]  ret_from_fork+0x1f/0x30
[ 3073.316365]
               Showing all locks held in the system:
[ 3073.318423] 1 lock held by khungtaskd/33:
[ 3073.319642]  #0: ffffffff82d246a0 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x28/0x1c3

[ 3073.322249] =============================================

'test_maps' looks sad :-( Not sure if related.

bash-5.0# ./test_maps
Fork 1024 tasks to 'test_update_delete' Fork 1024 tasks to 'test_update_delete' Fork 100 tasks to 'test_hashmap' Fork 100 tasks to 'test_hashmap_percpu' Fork 100 tasks to 'test_hashmap_sizes'

[   66.961150] test_maps invoked oom-killer: gfp_mask=0x100cca(GFP_HIGHUSER_MOVABLE), order=0, oom_score_adj=0
[   66.962490] CPU: 3 PID: 3263 Comm: test_maps Not tainted 5.8.0-rc4-01471-g15d51f3a516b #814
[   66.963617] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.fedoraproject.org-3.fc31 04/01/2014
[   66.965958] Call Trace:
[   66.966404]  dump_stack+0x9e/0xe0
[   66.966978]  dump_header+0x89/0x49a
[   66.967624]  oom_kill_process.cold+0xb/0x10
[   66.968379]  out_of_memory+0x1b1/0x820
[   66.969060]  ? oom_killer_disable+0x210/0x210
[   66.969833]  __alloc_pages_slowpath.constprop.0+0x125f/0x1460
[   66.970852]  ? warn_alloc+0x120/0x120
[   66.971508]  ? __alloc_pages_nodemask+0x30f/0x5c0
[   66.972237]  __alloc_pages_nodemask+0x4fd/0x5c0
[   66.972863]  ? __alloc_pages_slowpath.constprop.0+0x1460/0x1460
[   66.973681]  ? find_held_lock+0x85/0xa0
[   66.974223]  ? lock_downgrade+0x360/0x360
[   66.974760]  ? policy_nodemask+0x19/0x90
[   66.975266]  ? policy_node+0x56/0x60
[   66.975719]  pagecache_get_page+0xf7/0x360
[   66.976259]  filemap_fault+0xe4a/0xfe0
[   66.976743]  ? read_cache_page_gfp+0x20/0x20
[   66.977274]  ? find_held_lock+0x85/0xa0
[   66.977792]  ? filemap_page_mkwrite+0x140/0x140
[   66.978383]  __do_fault+0x6a/0x1e0
[   66.978789]  handle_mm_fault+0x16eb/0x1fb0
[   66.979319]  ? copy_page_range+0xf80/0xf80
[   66.979755]  ? vmacache_find+0xba/0x100
[   66.980166]  do_user_addr_fault+0x2ce/0x5ed
[   66.980625]  exc_page_fault+0x5e/0xc0
[   66.981005]  ? asm_exc_page_fault+0x8/0x30
[   66.981445]  asm_exc_page_fault+0x1e/0x30
[   66.981860] RIP: 0033:0x7f33d40901bd
[   66.982248] Code: Bad RIP value.
[   66.982579] RSP: 002b:00007ffe84e1dd18 EFLAGS: 00010202
[   66.983143] RAX: fffffffffffffff4 RBX: 0000000000000002 RCX: 00007f33d40901bd
[   66.984038] RDX: 0000000000000078 RSI: 00007ffe84e1dd50 RDI: 0000000000000000
[   66.984831] RBP: 00007ffe84e1dd30 R08: 00007ffe84e1dd50 R09: 00007ffe84e1dd50
[   66.985619] R10: 0000000000000000 R11: 0000000000000202 R12: 0000000000020000
[   66.986399] R13: 0000000000000004 R14: 0000000000000000 R15: 0000000000000060
[   66.987656] Mem-Info:
[   66.987924] active_anon:1424 inactive_anon:126 isolated_anon:0
[   66.987924]  active_file:53 inactive_file:33 isolated_file:0
[   66.987924]  unevictable:0 dirty:0 writeback:0
[   66.987924]  slab_reclaimable:11120 slab_unreclaimable:98031
[   66.987924]  mapped:88 shmem:137 pagetables:120 bounce:0
[   66.987924]  free:21175 free_pcp:725 free_cma:0
[   66.993308] Node 0 active_anon:5696kB inactive_anon:504kB active_file:212kB inactive_file:132kB unevictable:0kB isolated(anon):0kB isolated(file):0kB mapped:352kB dirty:0kB writeback:0kB shmem:548kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 0kB writeback_tmp:0kB all_unreclaimable? yes
[   66.997665] Node 0 DMA free:13708kB min:308kB low:384kB high:460kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:15992kB managed:15908kB mlocked:0kB kernel_stack:96kB pagetables:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB
[   67.002330] lowmem_reserve[]: 0 2925 3354 3354
[   67.003090] Node 0 DMA32 free:60008kB min:58668kB low:73332kB high:87996kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:3129212kB managed:3001752kB mlocked:0kB kernel_stack:25536kB pagetables:292kB bounce:0kB free_pcp:844kB local_pcp:248kB free_cma:0kB
[   67.008157] lowmem_reserve[]: 0 0 428 428
[   67.008843] Node 0 Normal free:9528kB min:8600kB low:10748kB high:12896kB reserved_highatomic:2048KB active_anon:5404kB inactive_anon:504kB active_file:336kB inactive_file:0kB unevictable:0kB writepending:0kB present:1048576kB managed:439260kB mlocked:0kB kernel_stack:5824kB pagetables:128kB bounce:0kB free_pcp:1596kB local_pcp:660kB free_cma:0kB
[   67.014007] lowmem_reserve[]: 0 0 0 0
[   67.014673] Node 0 DMA: 1*4kB (U) 1*8kB (U) 2*16kB (UE) 1*32kB (U) 1*64kB (E) 2*128kB (UE) 2*256kB (UE) 1*512kB (E) 2*1024kB (UE) 3*2048kB (UME) 1*4096kB (M) = 13708kB
[   67.016793] Node 0 DMA32: 8*4kB (UM) 5*8kB (UM) 4*16kB (UME) 8*32kB (UME) 7*64kB (ME) 8*128kB (UME) 6*256kB (UM) 8*512kB (ME) 3*1024kB (M) 2*2048kB (ME) 11*4096kB (M) = 59720kB
[   67.018915] Node 0 Normal: 256*4kB (UME) 149*8kB (UM) 89*16kB (UM) 41*32kB (UM) 20*64kB (UMH) 8*128kB (UM) 2*256kB (MH) 1*512kB (H) 1*1024kB (H) 0*2048kB 0*4096kB = 9304kB
[   67.020673] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB
[   67.021811] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB
[   67.022768] 184 total pagecache pages
[   67.023418] 0 pages in swap cache
[   67.023912] Swap cache stats: add 0, delete 0, find 0/0
[   67.024726] Free swap  = 0kB
[   67.025150] Total swap = 0kB
[   67.025530] 1048445 pages RAM
[   67.025865] 0 pages HighMem/MovableOnly
[   67.026510] 184215 pages reserved
[   67.027002] 0 pages cma reserved
[   67.027542] 0 pages hwpoisoned
[   67.027928] Unreclaimable slab info:
[   67.028500] Name                      Used          Total
[   67.029057] 9p-fcall-cache           297KB        445KB
[   67.029793] 9p-fcall-cache            49KB         49KB
[   67.030375] 9p-fcall-cache           123KB        272KB
[   67.030897] 9p-fcall-cache           346KB        495KB
[   67.031539] p9_req_t                  16KB         16KB
[   67.032326] fib6_nodes                 4KB          4KB
[   67.032950] RAWv6                     31KB         31KB
[   67.033960] mqueue_inode_cache         31KB         31KB
[   67.034708] ext4_bio_post_read_ctx         15KB         15KB
[   67.035647] bio-2                      7KB          7KB
[   67.036471] UNIX                     372KB        372KB
[   67.037769] tcp_bind_bucket            4KB          4KB
[   67.038701] ip_fib_trie                4KB          4KB
[   67.039496] ip_fib_alias               3KB          3KB
[   67.040414] ip_dst_cache               4KB          4KB
[   67.041365] RAW                       31KB         31KB
[   67.042341] UDP                      121KB        121KB
[   67.043540] tw_sock_TCP                7KB          7KB
[   67.044491] request_sock_TCP           7KB          7KB
[   67.045447] TCP                       58KB         58KB
[   67.046445] hugetlbfs_inode_cache         31KB         31KB
[   67.047443] bio-1                     15KB         15KB
[   67.048375] eventpoll_pwq             23KB         23KB
[   67.049388] eventpoll_epi             35KB         35KB
[   67.050361] inotify_inode_mark          3KB          3KB
[   67.051399] request_queue             62KB         62KB
[   67.052563] blkdev_ioc                 7KB          7KB
[   67.053623] bio-0                     20KB         20KB
[   67.054603] biovec-max               327KB        327KB
[   67.055593] skbuff_fclone_cache         15KB         15KB
[   67.056627] skbuff_head_cache        281KB        312KB
[   67.057868] file_lock_cache           31KB         31KB
[   67.058878] file_lock_ctx             15KB         15KB
[   67.059874] fsnotify_mark_connector          4KB          4KB
[   67.060957] task_delay_info          451KB        451KB
[   67.061644] proc_dir_entry           125KB        125KB
[   67.062404] pde_opener                59KB         59KB
[   67.063696] seq_file                 210KB        241KB
[   67.064759] sigqueue                  19KB         19KB
[   67.065763] shmem_inode_cache        795KB        795KB
[   67.066778] kernfs_node_cache       2565KB       2565KB
[   67.067762] mnt_cache                 31KB         31KB
[   67.068798] filp                   11820KB      11820KB
[   67.069807] names_cache              327KB        476KB
[   67.070795] key_jar                   15KB         15KB
[   67.071787] nsproxy                    3KB          3KB
[   67.072802] vm_area_struct          3236KB       3400KB
[   67.073773] mm_struct                870KB       1154KB
[   67.074787] fs_cache                 232KB        264KB
[   67.076008] files_cache              860KB        972KB
[   67.076969] signal_cache            2626KB       2686KB
[   67.077911] sighand_cache           2370KB       2405KB
[   67.078963] task_struct             8564KB       8564KB
[   67.080472] cred_jar                 392KB        420KB
[   67.081767] anon_vma_chain          1191KB       1413KB
[   67.082736] anon_vma                 200KB        220KB
[   67.083805] pid                      593KB        640KB
[   67.084733] Acpi-Operand             265KB        308KB
[   67.085598] Acpi-ParseExt             47KB         47KB
[   67.086504] Acpi-Parse               189KB        205KB
[   67.091509] Acpi-State               200KB        216KB
[   67.092511] Acpi-Namespace            24KB         24KB
[   67.093388] numa_policy                3KB          3KB
[   67.094249] trace_event_file         151KB        151KB
[   67.095076] ftrace_event_field        196KB        196KB
[   67.099230] pool_workqueue            16KB         16KB
[   67.100094] vmap_area                567KB        567KB
[   67.105224] page->ptl                385KB        433KB
[   67.106079] kmemleak_scan_area        286KB        286KB
[   67.107190] kmemleak_object       163919KB     168558KB
[   67.108081] kmalloc-8k             21840KB      21840KB
[   67.108932] kmalloc-4k             24592KB      24592KB
[   67.114003] kmalloc-2k             15684KB      15684KB
[   67.114896] kmalloc-1k             44936KB      44936KB
[   67.121018] kmalloc-512             4192KB       4192KB
[   67.122212] kmalloc-256             1128KB       1128KB
[   67.123089] kmalloc-192             5008KB       5008KB
[   67.128978] kmalloc-128              356KB        356KB
[   67.129877] kmalloc-96               145KB        156KB
[   67.132262] kmalloc-64              1032KB       1032KB
[   67.133085] kmalloc-32               233KB        244KB
[   67.138993] kmalloc-16               108KB        108KB
[   67.139847] kmalloc-8                440KB        533KB
[   67.142221] kmem_cache_node           43KB         43KB
[   67.143038] kmem_cache               140KB        140KB
[   67.149977] Tasks state (memory values in pages):
[   67.150743] [  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj name
[   67.153263] [    103]     0   103     7766     1042    81920        0         -1000 systemd-udevd
[   67.157980] [    194]     0   194     1032      136    40960        0             0 bash
[   67.161986] [    195]     0   195      758       40    45056        0             0 test_maps
[   67.167074] [   3178]     0  3178      758       39    45056        0             0 test_maps
[   67.170572] [   3186]     0  3186      758       39    45056        0             0 test_maps
[   67.172052] [   3205]     0  3205      758       39    45056        0             0 test_maps
[   67.177580] [   3213]     0  3213      758       39    45056        0             0 test_maps
[   67.179026] [   3221]     0  3221      758       39    45056        0             0 test_maps
[   67.184143] [   3222]     0  3222      758       39    45056        0             0 test_maps
[   67.187571] [   3230]     0  3230      758       39    45056        0             0 test_maps
[   67.189053] [   3241]     0  3241      758       39    45056        0             0 test_maps
[   67.194580] [   3243]     0  3243      758       39    45056        0             0 test_maps
[   67.196058] [   3250]     0  3250      758       39    45056        0             0 test_maps
[   67.201124] [   3263]     0  3263      758       39    45056        0             0 test_maps
[   67.206581] [   3298]     0  3298      758       39    45056        0             0 test_maps
[   67.210563] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/,task=bash,pid=194,uid=0
[   67.214666] Out of memory: Killed process 194 (bash) total-vm:4128kB, anon-rss:544kB, file-rss:0kB, shmem-rss:0kB, UID:0 pgtables:40kB oom_score_adj:0
[   67.226601] oom_reaper: reaped process 3298 (test_maps), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[   67.228355] oom_reaper: reaped process 3178 (test_maps), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[   67.234148] oom_reaper: reaped process 3222 (test_maps), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[   67.236873] oom_reaper: reaped process 3213 (test_maps), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[   67.239028] oom_reaper: reaped process 3186 (test_maps), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[   67.268146] oom_reaper: reaped process 3221 (test_maps), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[   67.275959] kthreadd invoked oom-killer: gfp_mask=0x40cc0(GFP_KERNEL|__GFP_COMP), order=1, oom_score_adj=0
[   67.277227] CPU: 0 PID: 2 Comm: kthreadd Not tainted 5.8.0-rc4-01471-g15d51f3a516b #814
[   67.278278] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.fedoraproject.org-3.fc31 04/01/2014
[   67.279895] Call Trace:
[   67.280250]  dump_stack+0x9e/0xe0
[   67.280674]  dump_header+0x89/0x49a
[   67.281113]  out_of_memory.cold+0xa/0xbb
[   67.281608]  ? oom_killer_disable+0x210/0x210
[   67.282157]  __alloc_pages_slowpath.constprop.0+0x125f/0x1460
[   67.282894]  ? warn_alloc+0x120/0x120
[   67.283377]  ? __alloc_pages_nodemask+0x30f/0x5c0
[   67.283925]  __alloc_pages_nodemask+0x4fd/0x5c0
[   67.284446]  ? __alloc_pages_slowpath.constprop.0+0x1460/0x1460
[   67.285146]  alloc_slab_page+0x2e/0x7a0
[   67.285590]  ? new_slab+0x22e/0x2b0
[   67.285999]  new_slab+0x276/0x2b0
[   67.286402]  ___slab_alloc+0x4ba/0x6d0
[   67.286892]  ? copy_process+0x256d/0x2f80
[   67.287391]  ? lock_downgrade+0x360/0x360
[   67.287908]  ? copy_process+0x256d/0x2f80
[   67.288482]  ? __slab_alloc.isra.0+0x4b/0x90
[   67.289207]  __slab_alloc.isra.0+0x4b/0x90
[   67.289874]  ? copy_process+0x256d/0x2f80
[   67.290565]  kmem_cache_alloc_node+0xb7/0x330
[   67.291163]  ? trace_hardirqs_on+0x1e/0x130
[   67.291699]  copy_process+0x256d/0x2f80
[   67.292231]  ? mark_lock+0x13f/0xc30
[   67.292746]  ? find_held_lock+0x85/0xa0
[   67.293314]  ? __cleanup_sighand+0x60/0x60
[   67.293889]  _do_fork+0xcf/0x840
[   67.294354]  ? copy_init_mm+0x20/0x20
[   67.294856]  ? lockdep_hardirqs_on_prepare+0x14c/0x240
[   67.295530]  ? _raw_spin_unlock_irq+0x24/0x50
[   67.296036]  ? trace_hardirqs_on+0x1e/0x130
[   67.296529]  ? preempt_count_sub+0x14/0xc0
[   67.297010]  ? lock_acquire+0x133/0x4e0
[   67.297467]  kernel_thread+0xa8/0xe0
[   67.297886]  ? legacy_clone_args_valid+0x30/0x30
[   67.298429]  ? kthread_create_on_node+0xd0/0xd0
[   67.298966]  ? do_raw_spin_unlock+0xa3/0x130
[   67.299458]  ? preempt_count_sub+0x14/0xc0
[   67.299907]  kthreadd+0x2be/0x340
[   67.300290]  ? kthread_create_on_cpu+0x120/0x120
[   67.300775]  ? lockdep_hardirqs_on_prepare+0x14c/0x240
[   67.301324]  ? _raw_spin_unlock_irq+0x24/0x50
[   67.301771]  ? trace_hardirqs_on+0x1e/0x130
[   67.302206]  ? kthread_create_on_cpu+0x120/0x120
[   67.302685]  ret_from_fork+0x1f/0x30
[   67.303234] Mem-Info:
[   67.303485] active_anon:1210 inactive_anon:126 isolated_anon:0
[   67.303485]  active_file:5 inactive_file:12 isolated_file:0
[   67.303485]  unevictable:0 dirty:0 writeback:0
[   67.303485]  slab_reclaimable:11187 slab_unreclaimable:98423
[   67.303485]  mapped:12 shmem:137 pagetables:70 bounce:0
[   67.303485]  free:20722 free_pcp:1 free_cma:0
[   67.307090] Node 0 active_anon:4840kB inactive_anon:504kB active_file:20kB inactive_file:48kB unevictable:0kB isolated(anon):0kB isolated(file):0kB mapped:48kB dirty:0kB writeback:0kB shmem:548kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 0kB writeback_tmp:0kB all_unreclaimable? yes
[   67.309863] Node 0 DMA free:13708kB min:308kB low:384kB high:460kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:15992kB managed:15908kB mlocked:0kB kernel_stack:96kB pagetables:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB
[   67.312641] lowmem_reserve[]: 0 2925 3354 3354
[   67.313081] Node 0 DMA32 free:60868kB min:58668kB low:73332kB high:87996kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:3129212kB managed:3001752kB mlocked:0kB kernel_stack:25824kB pagetables:240kB bounce:0kB free_pcp:248kB local_pcp:0kB free_cma:0kB
[   67.316022] lowmem_reserve[]: 0 0 428 428
[   67.316560] Node 0 Normal free:8312kB min:8600kB low:10748kB high:12896kB reserved_highatomic:0KB active_anon:4916kB inactive_anon:504kB active_file:308kB inactive_file:32kB unevictable:0kB writepending:0kB present:1048576kB managed:439260kB mlocked:0kB kernel_stack:8192kB pagetables:40kB bounce:0kB free_pcp:392kB local_pcp:0kB free_cma:0kB
[   67.319987] lowmem_reserve[]: 0 0 0 0
[   67.320421] Node 0 DMA: 1*4kB (U) 1*8kB (U) 2*16kB (UE) 1*32kB (U) 1*64kB (E) 2*128kB (UE) 2*256kB (UE) 1*512kB (E) 2*1024kB (UE) 3*2048kB (UME) 1*4096kB (M) = 13708kB
[   67.321913] Node 0 DMA32: 6*4kB (M) 26*8kB (UM) 29*16kB (UM) 11*32kB (UME) 9*64kB (UME) 10*128kB (UME) 6*256kB (UM) 7*512kB (ME) 4*1024kB (M) 2*2048kB (ME) 11*4096kB (M) = 61272kB
[   67.323562] Node 0 Normal: 488*4kB (UME) 251*8kB (UM) 86*16kB (UM) 37*32kB (UM) 18*64kB (UM) 6*128kB (M) 4*256kB (UM) 0*512kB 0*1024kB 0*2048kB 0*4096kB = 9464kB
[   67.325773] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB
[   67.326999] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB
[   67.328287] 155 total pagecache pages
[   67.328836] 0 pages in swap cache
[   67.329365] Swap cache stats: add 0, delete 0, find 0/0
[   67.330140] Free swap  = 0kB
[   67.330596] Total swap = 0kB
[   67.331043] 1048445 pages RAM
[   67.331478] 0 pages HighMem/MovableOnly
[   67.332080] 184215 pages reserved
[   67.332617] 0 pages cma reserved
[   67.333114] 0 pages hwpoisoned
[   67.333816] Unreclaimable slab info:
[   67.334439] Name                      Used          Total
[   67.335297] 9p-fcall-cache           297KB        445KB
[   67.336094] 9p-fcall-cache            49KB         49KB
[   67.336933] 9p-fcall-cache           123KB        272KB
[   67.337751] 9p-fcall-cache           445KB        495KB
[   67.338572] p9_req_t                  16KB         16KB
[   67.339391] fib6_nodes                 4KB          4KB
[   67.340229] RAWv6                     31KB         31KB
[   67.341033] mqueue_inode_cache         31KB         31KB
[   67.341997] ext4_bio_post_read_ctx         15KB         15KB
[   67.342835] bio-2                      7KB          7KB
[   67.343695] UNIX                     372KB        372KB
[   67.344560] tcp_bind_bucket            4KB          4KB
[   67.345368] ip_fib_trie                4KB          4KB
[   67.346190] ip_fib_alias               3KB          3KB
[   67.346997] ip_dst_cache               4KB          4KB
[   67.347819] RAW                       31KB         31KB
[   67.348668] UDP                      121KB        121KB
[   67.349856] tw_sock_TCP                7KB          7KB
[   67.350723] request_sock_TCP           7KB          7KB
[   67.351567] TCP                       58KB         58KB
[   67.352412] hugetlbfs_inode_cache         31KB         31KB
[   67.353319] bio-1                     15KB         15KB
[   67.354116] eventpoll_pwq             23KB         23KB
[   67.354919] eventpoll_epi             35KB         35KB
[   67.355735] inotify_inode_mark          3KB          3KB
[   67.356618] request_queue             62KB         62KB
[   67.357680] blkdev_ioc                 7KB          7KB
[   67.358569] bio-0                     20KB         20KB
[   67.359520] biovec-max               327KB        327KB
[   67.360396] skbuff_fclone_cache         15KB         15KB
[   67.361277] skbuff_head_cache        281KB        312KB
[   67.362116] file_lock_cache           31KB         31KB
[   67.362967] file_lock_ctx             15KB         15KB
[   67.363597] fsnotify_mark_connector          4KB          4KB
[   67.364279] task_delay_info          455KB        455KB
[   67.364832] proc_dir_entry           125KB        125KB
[   67.365663] pde_opener                59KB         59KB
[   67.366265] seq_file                 210KB        241KB
[   67.366805] sigqueue                  19KB         19KB
[   67.367386] shmem_inode_cache        795KB        795KB
[   67.367909] kernfs_node_cache       2565KB       2565KB
[   67.368489] mnt_cache                 31KB         31KB
[   67.369014] filp                   11820KB      11820KB
[   67.369585] names_cache              327KB        476KB
[   67.370113] key_jar                   15KB         15KB
[   67.370664] nsproxy                    3KB          3KB
[   67.371230] vm_area_struct          2870KB       3100KB
[   67.371754] mm_struct                831KB       1154KB
[   67.372337] fs_cache                 176KB        212KB
[   67.372862] files_cache              843KB        956KB
[   67.373625] signal_cache            2809KB       2809KB
[   67.374210] sighand_cache           2615KB       2615KB
[   67.374750] task_struct             9207KB       9207KB
[   67.375335] cred_jar                 392KB        420KB
[   67.375856] anon_vma_chain          1058KB       1291KB
[   67.376437] anon_vma                 144KB        160KB
[   67.376952] pid                      609KB        640KB
[   67.377519] Acpi-Operand             265KB        308KB
[   67.378063] Acpi-ParseExt             47KB         47KB
[   67.378645] Acpi-Parse               189KB        205KB
[   67.379216] Acpi-State               200KB        216KB
[   67.379747] Acpi-Namespace            24KB         24KB
[   67.380345] numa_policy                3KB          3KB
[   67.380870] trace_event_file         151KB        151KB
[   67.381649] ftrace_event_field        196KB        196KB
[   67.382199] pool_workqueue            16KB         16KB
[   67.382722] vmap_area                567KB        567KB
[   67.383261] page->ptl                384KB        429KB
[   67.383781] kmemleak_scan_area        286KB        286KB
[   67.384527] kmemleak_object       162939KB     167943KB
[   67.385080] kmalloc-8k             21840KB      21840KB
[   67.385618] kmalloc-4k             25736KB      25736KB
[   67.386158] kmalloc-2k             15684KB      15684KB
[   67.386679] kmalloc-1k             44936KB      44936KB
[   67.387208] kmalloc-512             4192KB       4192KB
[   67.387826] kmalloc-256             1184KB       1184KB
[   67.388514] kmalloc-192             5032KB       5032KB
[   67.389397] kmalloc-128              356KB        356KB
[   67.390100] kmalloc-96               145KB        156KB
[   67.390775] kmalloc-64              1052KB       1052KB
[   67.391489] kmalloc-32               233KB        244KB
[   67.392261] kmalloc-16               108KB        108KB
[   67.392964] kmalloc-8                440KB        533KB
[   67.393657] kmem_cache_node           43KB         43KB
[   67.394424] kmem_cache               140KB        140KB
[   67.395296] Tasks state (memory values in pages):
[   67.395889] [  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj name
[   67.397031] [    103]     0   103     7766     1042    81920        0         -1000 systemd-udevd
[   67.398411] Out of memory and no killable processes...
[   67.399203] Kernel panic - not syncing: System is deadlocked on memory
[   67.400146] CPU: 0 PID: 2 Comm: kthreadd Not tainted 5.8.0-rc4-01471-g15d51f3a516b #814
[   67.401254] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.fedoraproject.org-3.fc31 04/01/2014
[   67.403154] Call Trace:
[   67.403536]  dump_stack+0x9e/0xe0
[   67.404021]  panic+0x1ab/0x3ae
[   67.404497]  ? __warn_printk+0xf3/0xf3
[   67.405075]  ? __rcu_read_unlock+0x58/0x250
[   67.405705]  ? out_of_memory.cold+0x2d/0xbb
[   67.406272]  ? out_of_memory.cold+0x1f/0xbb
[   67.406910]  out_of_memory.cold+0x45/0xbb
[   67.407540]  ? oom_killer_disable+0x210/0x210
[   67.408225]  __alloc_pages_slowpath.constprop.0+0x125f/0x1460
[   67.409178]  ? warn_alloc+0x120/0x120
[   67.409849]  ? __alloc_pages_nodemask+0x30f/0x5c0
[   67.410722]  __alloc_pages_nodemask+0x4fd/0x5c0
[   67.411533]  ? __alloc_pages_slowpath.constprop.0+0x1460/0x1460
[   67.412641]  alloc_slab_page+0x2e/0x7a0
[   67.413345]  ? new_slab+0x22e/0x2b0
[   67.413883]  new_slab+0x276/0x2b0
[   67.414534]  ___slab_alloc+0x4ba/0x6d0
[   67.415221]  ? copy_process+0x256d/0x2f80
[   67.415930]  ? lock_downgrade+0x360/0x360
[   67.416659]  ? copy_process+0x256d/0x2f80
[   67.417390]  ? __slab_alloc.isra.0+0x4b/0x90
[   67.418171]  __slab_alloc.isra.0+0x4b/0x90
[   67.418903]  ? copy_process+0x256d/0x2f80
[   67.419600]  kmem_cache_alloc_node+0xb7/0x330
[   67.420339]  ? trace_hardirqs_on+0x1e/0x130
[   67.421078]  copy_process+0x256d/0x2f80
[   67.421722]  ? mark_lock+0x13f/0xc30
[   67.422284]  ? find_held_lock+0x85/0xa0
[   67.422838]  ? __cleanup_sighand+0x60/0x60
[   67.423515]  _do_fork+0xcf/0x840
[   67.423978]  ? copy_init_mm+0x20/0x20
[   67.424496]  ? lockdep_hardirqs_on_prepare+0x14c/0x240
[   67.425321]  ? _raw_spin_unlock_irq+0x24/0x50
[   67.426018]  ? trace_hardirqs_on+0x1e/0x130
[   67.426702]  ? preempt_count_sub+0x14/0xc0
[   67.427328]  ? lock_acquire+0x133/0x4e0
[   67.427938]  kernel_thread+0xa8/0xe0
[   67.428523]  ? legacy_clone_args_valid+0x30/0x30
[   67.429283]  ? kthread_create_on_node+0xd0/0xd0
[   67.430026]  ? do_raw_spin_unlock+0xa3/0x130
[   67.430708]  ? preempt_count_sub+0x14/0xc0
[   67.431402]  kthreadd+0x2be/0x340
[   67.431958]  ? kthread_create_on_cpu+0x120/0x120
[   67.432726]  ? lockdep_hardirqs_on_prepare+0x14c/0x240
[   67.433575]  ? _raw_spin_unlock_irq+0x24/0x50
[   67.434288]  ? trace_hardirqs_on+0x1e/0x130
[   67.434976]  ? kthread_create_on_cpu+0x120/0x120
[   67.435746]  ret_from_fork+0x1f/0x30
[   67.436548] Kernel Offset: disabled

[...]

I trigger this issue even switching back to commit:

commit 59632b220f2d61df274ed3a14a204e941051fdad

Author: Randy Dunlap <[email protected]>
Date:   Wed Jul 15 09:42:46 2020 -0700

net: ipv6: drop duplicate word in comment

[  244.327081] [   8645]     0  8645      760       31    40960        0             0 test_maps
[  244.327148] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),global_oom,task_memcg=/system.slice/polkit.service,task=polkitd,pid=211,uid=997
[  244.327264] Out of memory: Killed process 211 (polkitd) total-vm:1541300kB, anon-rss:7516kB, file-rss:0kB, shmem-rss:0kB, UID:997 pgtables:180kB oom_score_adj:0
[  244.333067] test_maps invoked oom-killer: gfp_mask=0x140cc0(GFP_USER|__GFP_COMP), order=0, oom_score_adj=0
[  244.333141] CPU: 0 PID: 8632 Comm: test_maps Not tainted 5.8.0-rc4-kvm+ #196
[  244.333196] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014
[  244.333270] Call Trace:
[  244.333301]  dump_stack+0x57/0x70
[  244.333334]  dump_header+0x4a/0x1d7
[  244.333367]  oom_kill_process.cold+0xb/0x10
[  244.333398]  out_of_memory+0x1ca/0x280
[  244.333428]  __alloc_pages_slowpath.constprop.0+0x946/0xc00
[  244.333489]  __alloc_pages_nodemask+0x1f7/0x230
[  244.333529]  alloc_slab_page+0x157/0x2b0
[  244.333556]  allocate_slab+0x2da/0x320
[  244.333585]  ___slab_alloc.constprop.0+0x283/0x4f0
[  244.333622]  ? update_load_avg+0x5b/0x520
[  244.333649]  ? htab_map_alloc+0x3a/0x480
[  244.333677]  ? set_next_entity+0x60/0x80
[  244.333705]  ? pick_next_task_fair+0x25f/0x2c0
[  244.333742]  ? kvm_sched_clock_read+0xd/0x20
[  244.333776]  kmem_cache_alloc_trace+0x1ca/0x1e0
[  244.333811]  ? htab_map_alloc+0x3a/0x480
[  244.333839]  htab_map_alloc+0x3a/0x480
[  244.333869]  __do_sys_bpf+0x2a4/0x1c30
[  244.333897]  ? hrtimer_nanosleep+0xb8/0x190
[  244.333926]  do_syscall_64+0x45/0x240
[  244.333953]  ? exc_page_fault+0x1b4/0x4a0
[  244.333993]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[  244.334037] RIP: 0033:0x7fe678a5b43d
[  244.334072] Code: Bad RIP value.
[  244.334103] RSP: 002b:00007ffed0c408a8 EFLAGS: 00000202 ORIG_RAX: 0000000000000141
[  244.334166] RAX: ffffffffffffffda RBX: 0000000000000013 RCX: 00007fe678a5b43d
[  244.334219] RDX: 0000000000000078 RSI: 00007ffed0c408e0 RDI: 0000000000000000
[  244.334268] RBP: 00007ffed0c408c0 R08: 00007ffed0c408e0 R09: 00007ffed0c408e0
[  244.334326] R10: 0000000000000000 R11: 0000000000000202 R12: 0000000000000001
[  244.334393] R13: 0000000000000002 R14: 0000000000000054 R15: 0000000000000000
[  244.334457] Mem-Info:
[  244.334488] active_anon:13749 inactive_anon:81 isolated_anon:0
                active_file:0 inactive_file:2 isolated_file:0
                unevictable:0 dirty:0 writeback:0
                slab_reclaimable:4723 slab_unreclaimable:25886
                mapped:38 shmem:125 pagetables:1432 bounce:0
                free:25932 free_pcp:676 free_cma:0
[  244.334684] Node 0 active_anon:54996kB inactive_anon:324kB active_file:0kB inactive_file:8kB unevictable:0kB isolated(anon):0kB isolated(file):0kB mapped:152kB dirty:0kB writeback:0kB shmem:500kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 4096kB writeback_tmp:0kB all_unreclaimable? no
[  244.334839] DMA free:8528kB min:704kB low:880kB high:1056kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:15992kB managed:15908kB mlocked:0kB kernel_stack:16kB pagetables:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB
[  244.334991] lowmem_reserve[]: 0 1972 1972 1972
[  244.335029] DMA32 free:95200kB min:97596kB low:119944kB high:142292kB reserved_highatomic:4096KB active_anon:54996kB inactive_anon:324kB active_file:100kB inactive_file:492kB unevictable:0kB writepending:0kB present:2080624kB managed:2024508kB mlocked:0kB kernel_stack:6496kB pagetables:5728kB bounce:0kB free_pcp:2704kB local_pcp:1468kB free_cma:0kB
[  244.335202] lowmem_reserve[]: 0 0 0 0
[  244.335230] DMA: 6*4kB (U) 1*8kB (U) 1*16kB (U) 1*32kB (U) 0*64kB 0*128kB 1*256kB (U) 0*512kB 0*1024kB 0*2048kB 2*4096kB (ME) = 8528kB
[  244.335313] DMA32: 1607*4kB (MEH) 1383*8kB (UMEH) 913*16kB (UMH) 556*32kB (UMEH) 368*64kB (MEH) 82*128kB (MH) 27*256kB (MH) 0*512kB 0*1024kB 2*2048kB (UM) 0*4096kB = 94948kB
[  244.335415] 159 total pagecache pages
[  244.335444] 0 pages in swap cache
[  244.335472] Swap cache stats: add 0, delete 0, find 0/0
[  244.335503] Free swap  = 0kB
[  244.335530] Total swap = 0kB
[  244.335559] 524154 pages RAM
[  244.335589] 0 pages HighMem/MovableOnly
[  244.335618] 14050 pages reserved
[  244.335645] Unreclaimable slab info:
[  244.335672] Name                      Used          Total
[  244.335760] 9p-fcall-cache            32KB         32KB
[  244.335808] p9_req_t                   3KB          3KB
[  244.335842] PINGv6                    31KB         31KB
[  244.335880] RAWv6                     78KB         78KB
[  244.335915] UDPv6                     30KB         30KB
[  244.335951] tw_sock_TCPv6              3KB          3KB
[  244.335987] request_sock_TCPv6          3KB          3KB
[  244.336022] TCPv6                     30KB         30KB

Increasing vm memory from 2G to 4G fixes the issue for me. Can you please double check?

Regards,
Lorenzo

[...]

Hi Jakub,

can you please provide more info? can you please identify the test that trigger the issue? I run test_progs with bpf-next master branch and it works fine for me. I run the tests in a vm with 4 vcpus and 4G of memory.

Regards,
Lorenzo

Was able to trigger it running the newly added selftest:

virtme-init: console is ttyS0
bash-5.0# ./test_progs -n 100
#100/1 cpumap_with_progs:OK
#100 xdp_cpumap_attach:OK
Summary: 1/1 PASSED, 0 SKIPPED, 0 FAILED
bash-5.0# [ 247.177168] INFO: task cpumap/0/map:3:198 blocked for more than 122 seconds.

[  247.181306]       Not tainted 5.8.0-rc4-01456-gbfdfa51702de #815
[  247.184487] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  247.188876] cpumap/0/map:3  D    0   198      2 0x00004000
[  247.192624] Call Trace:
[  247.194327]  __schedule+0x5ad/0xf10
[  247.196860]  ? pci_mmcfg_check_reserved+0xd0/0xd0
[  247.199853]  ? static_obj+0x31/0x80
[  247.201917]  ? mark_held_locks+0x24/0x90
[  247.204398]  ? cpu_map_update_elem+0x6d0/0x6d0
[  247.207098]  schedule+0x6f/0x160
[  247.209079]  schedule_preempt_disabled+0x14/0x20
[  247.211863]  kthread+0x175/0x240
[  247.213698]  ? kthread_create_on_node+0xd0/0xd0
[  247.216054]  ret_from_fork+0x1f/0x30
[  247.218363]
[  247.218363] Showing all locks held in the system:
[  247.222150] 1 lock held by khungtaskd/33:
[  247.224894]  #0: ffffffff82d246a0 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x28/0x1c3
[  247.231113]
[  247.232335] =============================================
[  247.232335]

qemu running with 4 vCPUs, 4 GB of memory. .config uploaded at
https://paste.centos.org/view/0c14663d

HTH,
-jkbs

[...]

ack, thx Jakub. I will look at it.

Regards,
Lorenzo

[...]

Hi Jakub,

can you please test the patch below when you have some free cycles? It fixes the issue in my setup.

Regards,
Lorenzo

index 4c95d0615ca2..f1c46529929b 100644
@@ -453,24 +453,27 @@ __cpu_map_entry_alloc(struct bpf_cpumap_val *value, u32 cpu, int map_id)
 	rcpu->map_id = map_id;
 	rcpu->value.qsize  = value->qsize;
 
+	if (fd > 0 && __cpu_map_load_bpf_program(rcpu, fd))
+		goto free_ptr_ring;
+
 	/* Setup kthread */
 	rcpu->kthread = kthread_create_on_node(cpu_map_kthread_run, rcpu, numa,
 					       "cpumap/%d/map:%d", cpu, map_id);
 	if (IS_ERR(rcpu->kthread))
-		goto free_ptr_ring;
+		goto free_prog;
 
 	get_cpu_map_entry(rcpu); /* 1-refcnt for being in cmap->cpu_map[] */
 	get_cpu_map_entry(rcpu); /* 1-refcnt for kthread */
 
-	if (fd > 0 && __cpu_map_load_bpf_program(rcpu, fd))
-		goto free_ptr_ring;
-
 	/* Make sure kthread runs on a single CPU */
 	kthread_bind(rcpu->kthread, cpu);
 	wake_up_process(rcpu->kthread);
 
 	return rcpu;
 
+free_prog:
+	if (rcpu->prog)
+		bpf_prog_put(rcpu->prog);
 free_ptr_ring:
 	ptr_ring_cleanup(rcpu->queue, NULL);
 free_queue: