bpf: Emit bpf_timer in vmlinux BTF
authorYonghong Song <yhs@fb.com>
Fri, 11 Feb 2022 19:49:48 +0000 (11:49 -0800)
committerAlexei Starovoitov <ast@kernel.org>
Fri, 11 Feb 2022 21:21:47 +0000 (13:21 -0800)
Currently the following code in check_and_init_map_value()
  *(struct bpf_timer *)(dst + map->timer_off) =
      (struct bpf_timer){};
can help generate bpf_timer definition in vmlinuxBTF.
But the code above may not zero the whole structure
due to anonymour members and that code will be replaced
by memset in the subsequent patch and
bpf_timer definition will disappear from vmlinuxBTF.
Let us emit the type explicitly so bpf program can continue
to use it from vmlinux.h.

Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20220211194948.3141529-1-yhs@fb.com
kernel/bpf/helpers.c

index 01cfdf4..55c0842 100644 (file)
@@ -2,6 +2,7 @@
 /* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
  */
 #include <linux/bpf.h>
+#include <linux/btf.h>
 #include <linux/bpf-cgroup.h>
 #include <linux/rcupdate.h>
 #include <linux/random.h>
@@ -1075,6 +1076,7 @@ static enum hrtimer_restart bpf_timer_cb(struct hrtimer *hrtimer)
        void *key;
        u32 idx;
 
+       BTF_TYPE_EMIT(struct bpf_timer);
        callback_fn = rcu_dereference_check(t->callback_fn, rcu_read_lock_bh_held());
        if (!callback_fn)
                goto out;