[PATCH net-next] net: tls: Set async_capable for tls zerocopy only if we see EINPROGRESS

[PATCH net-next] net: tls: Set async_capable for tls zerocopy only if we see EINPROGRESS

[Dave Watson]

Currently we don't zerocopy if the crypto framework async bit is set.
However some crypto algorithms (such as x86 AESNI) support async,
but in the context of sendmsg, will never run asynchronously.  Instead,
check for actual EINPROGRESS return code before assuming algorithm is
async.

Signed-off-by: Dave Watson <davejwatson@fb.com>
---
 include/net/tls.h | 1 +
 net/tls/tls_sw.c  | 4 ++--
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/include/net/tls.h b/include/net/tls.h
index 4592606e136a..eb73e62ac8c9 100644
--- a/include/net/tls.h
+++ b/include/net/tls.h
@@ -139,6 +139,7 @@ struct tls_sw_context_tx {
 	struct list_head tx_list;
 	atomic_t encrypt_pending;
 	int async_notify;
+	int async_capable;
 
 #define BIT_TX_SCHEDULED	0
 	unsigned long tx_bitmask;
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index 3f2a6af27e62..2864ad16c2d9 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -673,6 +673,7 @@ static int tls_push_record(struct sock *sk, int flags,
 				tls_merge_open_record(sk, rec, tmp, orig_end);
 			}
 		}
+		ctx->async_capable = 1;
 		return rc;
 	} else if (split) {
 		msg_pl = &tmp->msg_plaintext;
@@ -814,8 +815,7 @@ int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
 	long timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
 	struct tls_context *tls_ctx = tls_get_ctx(sk);
 	struct tls_sw_context_tx *ctx = tls_sw_ctx_tx(tls_ctx);
-	struct crypto_tfm *tfm = crypto_aead_tfm(ctx->aead_send);
-	bool async_capable = tfm->__crt_alg->cra_flags & CRYPTO_ALG_ASYNC;
+	bool async_capable = ctx->async_capable;
 	unsigned char record_type = TLS_RECORD_TYPE_DATA;
 	bool is_kvec = iov_iter_is_kvec(&msg->msg_iter);
 	bool eor = !(msg->msg_flags & MSG_MORE);
-- 
2.17.1

Re: [PATCH net-next] net: tls: Set async_capable for tls zerocopy only if we see EINPROGRESS

[David Miller]

From: Dave Watson <davejwatson@fb.com>
Date: Wed, 30 Jan 2019 22:08:21 +0000

> Currently we don't zerocopy if the crypto framework async bit is set.
> However some crypto algorithms (such as x86 AESNI) support async,
> but in the context of sendmsg, will never run asynchronously.  Instead,
> check for actual EINPROGRESS return code before assuming algorithm is
> async.
> 
> Signed-off-by: Dave Watson <davejwatson@fb.com>

Applied, thanks.