[bpf-next,v3,3/7] bpf: support epoll from bpf struct_ops links.

Commit Message

Kui-Feng Lee May 10, 2024, 12:29 a.m. UTC

Add epoll support to bpf struct_ops links to trigger EPOLLHUP event upon
detachment.

This patch implements the "poll" of the "struct file_operations" for BPF
links and introduces a new "poll" operator in the "struct bpf_link_ops". By
implementing "poll" of "struct bpf_link_ops" for the links of struct_ops,
the file descriptor of a struct_ops link can be added to an epoll file
descriptor to receive EPOLLHUP events.

Signed-off-by: Kui-Feng Lee <[e-mail]>
---
 include/linux/bpf.h         |  1 +
 kernel/bpf/bpf_struct_ops.c | 17 +++++++++++++++++
 kernel/bpf/syscall.c        | 11 +++++++++++
 3 files changed, 29 insertions(+)

Comments

Martin KaFai Lau May 21, 2024, 1:26 a.m. UTC | #1

On 5/9/24 5:29 PM, Kui-Feng Lee wrote:
> +static __poll_t bpf_struct_ops_map_link_poll(struct file *file,
> +					     struct poll_table_struct *pts)
> +{
> +	struct bpf_struct_ops_link *st_link = file->private_data;
> +
> +	poll_wait(file, &st_link->wait_hup, pts);
> +
> +	return (st_link->map) ? 0 : EPOLLHUP;

nit. It should need a rcu_access_pointer(st_link->map).

Kui-Feng Lee May 21, 2024, 7:31 a.m. UTC | #2

On 5/20/24 18:26, Martin KaFai Lau wrote:
> On 5/9/24 5:29 PM, Kui-Feng Lee wrote:
>> +static __poll_t bpf_struct_ops_map_link_poll(struct file *file,
>> +                         struct poll_table_struct *pts)
>> +{
>> +    struct bpf_struct_ops_link *st_link = file->private_data;
>> +
>> +    poll_wait(file, &st_link->wait_hup, pts);
>> +
>> +    return (st_link->map) ? 0 : EPOLLHUP;
> 
> nit. It should need a rcu_access_pointer(st_link->map).
> 
sure

Patch

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index b600767ebe02..5f7496ef8b7c 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -1612,6 +1612,7 @@  struct bpf_link_ops {
 			      struct bpf_link_info *info);
 	int (*update_map)(struct bpf_link *link, struct bpf_map *new_map,
 			  struct bpf_map *old_map);
+	__poll_t (*poll)(struct file *file, struct poll_table_struct *pts);
 };
 
 struct bpf_tramp_link {
diff --git a/kernel/bpf/bpf_struct_ops.c b/kernel/bpf/bpf_struct_ops.c
index 7e270ee15f6e..497526bd5fec 100644
--- a/kernel/bpf/bpf_struct_ops.c
+++ b/kernel/bpf/bpf_struct_ops.c
@@ -12,6 +12,7 @@ 
 #include <linux/mutex.h>
 #include <linux/btf_ids.h>
 #include <linux/rcupdate_wait.h>
+#include <linux/poll.h>
 
 struct bpf_struct_ops_value {
 	struct bpf_struct_ops_common_value common;
@@ -56,6 +57,7 @@  struct bpf_struct_ops_map {
 struct bpf_struct_ops_link {
 	struct bpf_link link;
 	struct bpf_map __rcu *map;
+	wait_queue_head_t wait_hup;
 };
 
 static DEFINE_MUTEX(update_mutex);
@@ -1167,15 +1169,28 @@  static int bpf_struct_ops_map_link_detach(struct bpf_link *link)
 
 	mutex_unlock(&update_mutex);
 
+	wake_up_interruptible_poll(&st_link->wait_hup, EPOLLHUP);
+
 	return 0;
 }
 
+static __poll_t bpf_struct_ops_map_link_poll(struct file *file,
+					     struct poll_table_struct *pts)
+{
+	struct bpf_struct_ops_link *st_link = file->private_data;
+
+	poll_wait(file, &st_link->wait_hup, pts);
+
+	return (st_link->map) ? 0 : EPOLLHUP;
+}
+
 static const struct bpf_link_ops bpf_struct_ops_map_lops = {
 	.dealloc = bpf_struct_ops_map_link_dealloc,
 	.detach = bpf_struct_ops_map_link_detach,
 	.show_fdinfo = bpf_struct_ops_map_link_show_fdinfo,
 	.fill_link_info = bpf_struct_ops_map_link_fill_link_info,
 	.update_map = bpf_struct_ops_map_link_update,
+	.poll = bpf_struct_ops_map_link_poll,
 };
 
 int bpf_struct_ops_link_create(union bpf_attr *attr)
@@ -1213,6 +1228,8 @@  int bpf_struct_ops_link_create(union bpf_attr *attr)
 	 */
 	RCU_INIT_POINTER(link->map, map);
 
+	init_waitqueue_head(&link->wait_hup);
+
 	err = st_map->st_ops_desc->st_ops->reg(st_map->kvalue.data, &link->link);
 	if (err) {
 		RCU_INIT_POINTER(link->map, NULL);
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 13ad74ecf2cd..ad4f81ed27f0 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -3150,6 +3150,16 @@  static void bpf_link_show_fdinfo(struct seq_file *m, struct file *filp)
 }
 #endif
 
+static __poll_t bpf_link_poll(struct file *file, struct poll_table_struct *pts)
+{
+	struct bpf_link *link = file->private_data;
+
+	if (link->ops->poll)
+		return link->ops->poll(file, pts);
+
+	return 0;
+}
+
 static const struct file_operations bpf_link_fops = {
 #ifdef CONFIG_PROC_FS
 	.show_fdinfo	= bpf_link_show_fdinfo,
@@ -3157,6 +3167,7 @@  static const struct file_operations bpf_link_fops = {
 	.release	= bpf_link_release,
 	.read		= bpf_dummy_read,
 	.write		= bpf_dummy_write,
+	.poll		= bpf_link_poll,
 };
 
 static int bpf_link_alloc_id(struct bpf_link *link)