On Mon, Feb 06, 2023 at 11:15:21PM -0800, Jakub Kicinski wrote:
>
> > aead_request_set_callback(aead_req, CRYPTO_TFM_REQ_MAY_BACKLOG,
> > - tls_encrypt_done, sk);
> > + tls_encrypt_done, aead_req);
>
> ... let's just pass rec instead of aead_req here, then?
Good point. Could we do this as a follow-up patch? Reposting
the whole series would disturb a lot of people. Of course if
other major issues crop up I can fold this into the existing
patch.
Thanks!
---8<---
The function tls_encrypt_done only uses aead_req to get ahold of
the tls_rec object. So we could pass that in instead of aead_req
to simplify the code.
Suggested-by: Jakub Kicinski <kuba@xxxxxxxxxx>
Signed-off-by: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index 0515cda32fe2..6dfec2e8fdfa 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -430,18 +430,16 @@ int tls_tx_records(struct sock *sk, int flags)
static void tls_encrypt_done(void *data, int err)
{
- struct aead_request *aead_req = data;
struct tls_sw_context_tx *ctx;
struct tls_context *tls_ctx;
struct tls_prot_info *prot;
+ struct tls_rec *rec = data;
struct scatterlist *sge;
struct sk_msg *msg_en;
- struct tls_rec *rec;
bool ready = false;
struct sock *sk;
int pending;
- rec = container_of(aead_req, struct tls_rec, aead_req);
msg_en = &rec->msg_encrypted;
sk = rec->sk;
@@ -536,7 +534,7 @@ static int tls_do_encryption(struct sock *sk,
data_len, rec->iv_data);
aead_request_set_callback(aead_req, CRYPTO_TFM_REQ_MAY_BACKLOG,
- tls_encrypt_done, aead_req);
+ tls_encrypt_done, rec);
/* Add the record in tx_list */
list_add_tail((struct list_head *)&rec->list, &ctx->tx_list);
--
Email: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
