commit 962b375bcb7de6422c2bc5c4ef6e527d8f74c74d
author David Benjamin <davidben@google.com> Mon May 10 15:17:18 2021 -0400
committer CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org> Thu May 13 14:35:32 2021 +0000
tree a66198bbe77c19f452889c5a1f607118c76802a3
parent 8349dfc87e46d5914b0fefbb33241a95a9eef07d

Move session ID assignment out of ssl_get_new_session.

It's kind of weird that we assign a session ID, based on whether we
detect the handshake wants stateful resumption, and then erase it
afterwards.

Also remove the is_server parameter, which we can get from hs.

Change-Id: I94ac817c63abb08a457e0e0c29f5c2d2b60aa498
Reviewed-on: https://boringssl-review.googlesource.com/c/boringssl/+/47444
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>

ssl/internal.h

diff --git a/ssl/internal.h b/ssl/internal.h
index 084e9c3..a753329 100644
--- a/ssl/internal.h
+++ b/ssl/internal.h
@@ -2966,7 +2966,7 @@
 bool ssl_compare_public_and_private_key(const EVP_PKEY *pubkey,
                                        const EVP_PKEY *privkey);
 bool ssl_cert_check_private_key(const CERT *cert, const EVP_PKEY *privkey);
-int ssl_get_new_session(SSL_HANDSHAKE *hs, int is_server);
+bool ssl_get_new_session(SSL_HANDSHAKE *hs);
 int ssl_encrypt_ticket(SSL_HANDSHAKE *hs, CBB *out, const SSL_SESSION *session);
 int ssl_ctx_rotate_ticket_encryption_key(SSL_CTX *ctx);