dtls 1.3: allow rtx interval to be less than a second #9623
Conversation
|
🛟 Devin Lifeguard found 1 likely issues in this PR
@julek-wolfssl |
julek-wolfssl
force-pushed
the
dtls-1.3-ms-interval
branch
from
d8b2992 to
8c416f2
Compare
|
Retest this please RequestAbortedException |
| /* Save the message count to make sure no new messages are sent */ | ||
| ExpectIntGE(c_msg_count = test_ctx.c_msg_count, 2); |
There was a problem hiding this comment.
NIT: two instruction to make it more readable
| test_memio_clear_buffer(&test_ctx, 1); | ||
|
|
||
| /* First timeout. This one should trigger a retransmission */ | ||
| if (wolfSSL_dtls13_use_quick_timeout(ssl_s)) |
There was a problem hiding this comment.
wrap the result of the wolfSSL_dtls13_use_quick_timeout in an Expect assert
| #define DTLS13_MIN_CIPHERTEXT 16 | ||
| #define DTLS13_MIN_RTX_INTERVAL 1 | ||
| #ifndef DTLS13_MIN_RTX_INTERVAL | ||
| #define DTLS13_MIN_RTX_INTERVAL 1000 |
There was a problem hiding this comment.
what do you think of using DTLS_TIMEOUT_INIT as default here?
| * wolfSSL_dtls_got_timeout() to simulate timeouts and verify that | ||
| * retransmissions are spaced at least DTLS13_MIN_RTX_INTERVAL apart. | ||
| */ | ||
| int test_dtls13_min_rtx_interval(void) |
There was a problem hiding this comment.
This test relies on timing and may introduce flakiness. I don't see an easy way to test without timing dependency. If we want to keep it as it is, we need at least a clear comment about the risk imo.
2 participants
No description provided.