[PATCH bpf] bpf: add schedule point in htab_init_buckets()

[PATCH bpf] bpf: add schedule point in htab_init_buckets()

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

We noticed that with a LOCKDEP enabled kernel,
allocating a hash table with 65536 buckets would
use more than 60ms.

htab_init_buckets() runs from process context,
it is safe to schedule to avoid latency spikes.

Fixes: c50eb518e262 ("bpf: Use separate lockdep class for each hashtab")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-By: John Sperbeck <jsperbeck@google.com>
Cc: Song Liu <songliubraving@fb.com>
---
 kernel/bpf/hashtab.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index 7e848200cd268a0f9ed063f0b641d3c355787013..c1ac7f964bc997925fd427f5192168829d812e5d 100644
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -152,6 +152,7 @@ static void htab_init_buckets(struct bpf_htab *htab)
 			lockdep_set_class(&htab->buckets[i].lock,
 					  &htab->lockdep_key);
 		}
+		cond_resched();
 	}
 }
 
-- 
2.29.2.729.g45daf8777d-goog

Re: [PATCH bpf] bpf: add schedule point in htab_init_buckets()

[Song Liu]

> On Dec 21, 2020, at 11:25 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> 
> From: Eric Dumazet <edumazet@google.com>
> 
> We noticed that with a LOCKDEP enabled kernel,
> allocating a hash table with 65536 buckets would
> use more than 60ms.
> 
> htab_init_buckets() runs from process context,
> it is safe to schedule to avoid latency spikes.
> 
> Fixes: c50eb518e262 ("bpf: Use separate lockdep class for each hashtab")
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Reported-By: John Sperbeck <jsperbeck@google.com>
> Cc: Song Liu <songliubraving@fb.com>

Acked-by: Song Liu <songliubraving@fb.com>

Thanks for the fix!

> ---
> kernel/bpf/hashtab.c | 1 +
> 1 file changed, 1 insertion(+)
> 
> diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
> index 7e848200cd268a0f9ed063f0b641d3c355787013..c1ac7f964bc997925fd427f5192168829d812e5d 100644
> --- a/kernel/bpf/hashtab.c
> +++ b/kernel/bpf/hashtab.c
> @@ -152,6 +152,7 @@ static void htab_init_buckets(struct bpf_htab *htab)
> 			lockdep_set_class(&htab->buckets[i].lock,
> 					  &htab->lockdep_key);
> 		}
> +		cond_resched();
> 	}
> }
> 
> -- 
> 2.29.2.729.g45daf8777d-goog
> 

Re: [PATCH bpf] bpf: add schedule point in htab_init_buckets()

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to bpf/bpf.git (refs/heads/master):

On Mon, 21 Dec 2020 11:25:06 -0800 you wrote:
> From: Eric Dumazet <edumazet@google.com>
> 
> We noticed that with a LOCKDEP enabled kernel,
> allocating a hash table with 65536 buckets would
> use more than 60ms.
> 
> htab_init_buckets() runs from process context,
> it is safe to schedule to avoid latency spikes.
> 
> [...]

Here is the summary with links:
  - [bpf] bpf: add schedule point in htab_init_buckets()
    https://git.kernel.org/bpf/bpf/c/e7e518053c26

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html