kris/boringssl
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
c39be857a8
boringssl/ssl/test/bssl_shim.cc

795 lines
22 KiB
C++
Raw Normal View History

Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
/* Copyright (c) 2014, Google Inc.
*
* Permission to use, copy, modify, and/or distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY
* SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION
* OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN
* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */
Windows build fixes. Windows doesn't have ssize_t, sadly. There's SSIZE_T, but defining an OPENSSL_SSIZE_T seems worse than just using an int. Change-Id: I09bb5aa03f96da78b619e551f92ed52ce24d9f3f Reviewed-on: https://boringssl-review.googlesource.com/1352 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 23:23:51 +01:00
#include <openssl/base.h>
#if !defined(OPENSSL_WINDOWS)
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
#include <arpa/inet.h>
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
#include <netinet/in.h>
#include <signal.h>
#include <sys/socket.h>
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
#include <unistd.h>
Windows build fixes. Windows doesn't have ssize_t, sadly. There's SSIZE_T, but defining an OPENSSL_SSIZE_T seems worse than just using an int. Change-Id: I09bb5aa03f96da78b619e551f92ed52ce24d9f3f Reviewed-on: https://boringssl-review.googlesource.com/1352 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 23:23:51 +01:00
#endif
Remove string.h from base.h. Including string.h in base.h causes any file that includes a BoringSSL header to include string.h. Generally this wouldn't be a problem, although string.h might slow down the compile if it wasn't otherwise needed. However, it also causes problems for ipsec-tools in Android because OpenSSL didn't have this behaviour. This change removes string.h from base.h and, instead, adds it to each .c file that requires it. Change-Id: I5968e50b0e230fd3adf9b72dd2836e6f52d6fb37 Reviewed-on: https://boringssl-review.googlesource.com/3200 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2015-01-31 01:08:37 +00:00
#include <string.h>
Windows build fixes. Windows doesn't have ssize_t, sadly. There's SSIZE_T, but defining an OPENSSL_SSIZE_T seems worse than just using an int. Change-Id: I09bb5aa03f96da78b619e551f92ed52ce24d9f3f Reviewed-on: https://boringssl-review.googlesource.com/1352 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 23:23:51 +01:00
#include <sys/types.h>
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
#include <openssl/bio.h>
Add tests for PSK cipher suites. Only the three plain PSK suites for now. ECDHE_PSK_WITH_AES_128_GCM_SHA256 will be in a follow-up. Change-Id: Iafc116a5b2798c61d90c139b461cf98897ae23b3 Reviewed-on: https://boringssl-review.googlesource.com/2051 Reviewed-by: Adam Langley <agl@google.com>
2014-10-27 05:06:24 +00:00
#include <openssl/buf.h>
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
#include <openssl/bytestring.h>
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
#include <openssl/ssl.h>
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
#include "async_bio.h"
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
#include "packeted_bio.h"
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
#include "test_config.h"
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
static int usage(const char *program) {
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
fprintf(stderr, "Usage: %s [flags...]\n",
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
program);
return 1;
}
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
static int g_ex_data_index = 0;
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
static int g_ex_data_clock_index = 0;
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
Add malloc failure tests. This commit fixes a number of crashes caused by malloc failures. They were found using the -malloc-test=0 option to runner.go which runs tests many times, causing a different allocation call to fail in each case. (This test only works on Linux and only looks for crashes caused by allocation failures, not memory leaks or other errors.) This is not the complete set of crashes! More can be found by collecting core dumps from running with -malloc-test=0. Change-Id: Ia61d19f51e373bccb7bc604642c51e043a74bd83 Reviewed-on: https://boringssl-review.googlesource.com/2320 Reviewed-by: Adam Langley <agl@google.com>
2014-11-18 01:26:55 +00:00
static bool SetConfigPtr(SSL *ssl, const TestConfig *config) {
return SSL_set_ex_data(ssl, g_ex_data_index, (void *)config) == 1;
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
}
static const TestConfig *GetConfigPtr(SSL *ssl) {
return (const TestConfig *)SSL_get_ex_data(ssl, g_ex_data_index);
}
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
static bool SetClockPtr(SSL *ssl, OPENSSL_timeval *clock) {
return SSL_set_ex_data(ssl, g_ex_data_clock_index, (void *)clock) == 1;
}
static OPENSSL_timeval *GetClockPtr(SSL *ssl) {
return (OPENSSL_timeval *)SSL_get_ex_data(ssl, g_ex_data_clock_index);
}
Add basic TLS Channel ID tests. Change-Id: I7ccf2b8282dfa8f3985775e8b67edcf3c2949752 Reviewed-on: https://boringssl-review.googlesource.com/1606 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 06:46:07 +01:00
static EVP_PKEY *LoadPrivateKey(const std::string &file) {
BIO *bio = BIO_new(BIO_s_file());
if (bio == NULL) {
return NULL;
}
if (!BIO_read_filename(bio, file.c_str())) {
BIO_free(bio);
return NULL;
}
EVP_PKEY *pkey = PEM_read_bio_PrivateKey(bio, NULL, NULL, NULL);
BIO_free(bio);
return pkey;
}
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
static int early_callback_called = 0;
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
static int select_certificate_callback(const struct ssl_early_callback_ctx *ctx) {
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
early_callback_called = 1;
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
const TestConfig *config = GetConfigPtr(ctx->ssl);
if (config->expected_server_name.empty()) {
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
return 1;
}
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
const uint8_t *extension_data;
size_t extension_len;
CBS extension, server_name_list, host_name;
uint8_t name_type;
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
if (!SSL_early_callback_ctx_extension_get(ctx, TLSEXT_TYPE_server_name,
&extension_data,
&extension_len)) {
fprintf(stderr, "Could not find server_name extension.\n");
return -1;
}
CBS_init(&extension, extension_data, extension_len);
if (!CBS_get_u16_length_prefixed(&extension, &server_name_list) ||
CBS_len(&extension) != 0 ||
!CBS_get_u8(&server_name_list, &name_type) ||
name_type != TLSEXT_NAMETYPE_host_name ||
!CBS_get_u16_length_prefixed(&server_name_list, &host_name) ||
CBS_len(&server_name_list) != 0) {
fprintf(stderr, "Could not decode server_name extension.\n");
return -1;
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (!CBS_mem_equal(&host_name,
(const uint8_t*)config->expected_server_name.data(),
config->expected_server_name.size())) {
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
fprintf(stderr, "Server name mismatch.\n");
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
}
return 1;
}
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
static int skip_verify(int preverify_ok, X509_STORE_CTX *store_ctx) {
Add tests for the server accepting client certificates. Change-Id: I9acc4363c6b9804d5fe464053393cf16ffb7785c Reviewed-on: https://boringssl-review.googlesource.com/1159 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 20:47:52 +01:00
return 1;
}
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
static int next_protos_advertised_callback(SSL *ssl,
Fix False Start without session tickets. One of the state transitions wasn't rewritten to CR_CHANGE. Add a test to exercise this codepath. Also SSL_cutthrough_complete references the state. Change-Id: Ib2f7ac5ac3f0348864efa93cf13cfd87454572f0 Reviewed-on: https://boringssl-review.googlesource.com/1337 Reviewed-by: Adam Langley <agl@google.com>
2014-07-28 19:52:32 +01:00
const uint8_t **out,
unsigned int *out_len,
void *arg) {
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
const TestConfig *config = GetConfigPtr(ssl);
if (config->advertise_npn.empty())
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
return SSL_TLSEXT_ERR_NOACK;
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
*out = (const uint8_t*)config->advertise_npn.data();
*out_len = config->advertise_npn.size();
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
return SSL_TLSEXT_ERR_OK;
}
Fix False Start without session tickets. One of the state transitions wasn't rewritten to CR_CHANGE. Add a test to exercise this codepath. Also SSL_cutthrough_complete references the state. Change-Id: Ib2f7ac5ac3f0348864efa93cf13cfd87454572f0 Reviewed-on: https://boringssl-review.googlesource.com/1337 Reviewed-by: Adam Langley <agl@google.com>
2014-07-28 19:52:32 +01:00
static int next_proto_select_callback(SSL* ssl,
uint8_t** out,
uint8_t* outlen,
const uint8_t* in,
unsigned inlen,
void* arg) {
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
const TestConfig *config = GetConfigPtr(ssl);
if (config->select_next_proto.empty())
Fix False Start without session tickets. One of the state transitions wasn't rewritten to CR_CHANGE. Add a test to exercise this codepath. Also SSL_cutthrough_complete references the state. Change-Id: Ib2f7ac5ac3f0348864efa93cf13cfd87454572f0 Reviewed-on: https://boringssl-review.googlesource.com/1337 Reviewed-by: Adam Langley <agl@google.com>
2014-07-28 19:52:32 +01:00
return SSL_TLSEXT_ERR_NOACK;
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
*out = (uint8_t*)config->select_next_proto.data();
*outlen = config->select_next_proto.size();
Fix False Start without session tickets. One of the state transitions wasn't rewritten to CR_CHANGE. Add a test to exercise this codepath. Also SSL_cutthrough_complete references the state. Change-Id: Ib2f7ac5ac3f0348864efa93cf13cfd87454572f0 Reviewed-on: https://boringssl-review.googlesource.com/1337 Reviewed-by: Adam Langley <agl@google.com>
2014-07-28 19:52:32 +01:00
return SSL_TLSEXT_ERR_OK;
}
Add tests for ALPN support. Both as client and as server. Also tests that ALPN causes False Start to kick in. Change-Id: Ib570346f3c511834152cd2df2ef29541946d3ab4 Reviewed-on: https://boringssl-review.googlesource.com/1753 Reviewed-by: Adam Langley <agl@google.com>
2014-09-06 17:58:58 +01:00
static int alpn_select_callback(SSL* ssl,
const uint8_t** out,
uint8_t* outlen,
const uint8_t* in,
unsigned inlen,
void* arg) {
const TestConfig *config = GetConfigPtr(ssl);
if (config->select_alpn.empty())
return SSL_TLSEXT_ERR_NOACK;
if (!config->expected_advertised_alpn.empty() &&
(config->expected_advertised_alpn.size() != inlen ||
memcmp(config->expected_advertised_alpn.data(),
in, inlen) != 0)) {
fprintf(stderr, "bad ALPN select callback inputs\n");
exit(1);
}
*out = (const uint8_t*)config->select_alpn.data();
*outlen = config->select_alpn.size();
return SSL_TLSEXT_ERR_OK;
}
Tidy DTLS cookie callback types. const-correctness, unsigned long -> size_t. Change-Id: Ic0c2685a48a0f98396c5753b6077c6c0c3b92326 Reviewed-on: https://boringssl-review.googlesource.com/1540 Reviewed-by: Adam Langley <agl@google.com>
2014-08-15 18:38:15 +01:00
static int cookie_generate_callback(SSL *ssl, uint8_t *cookie, size_t *cookie_len) {
Remove a place where SSL_clear cleans up after client/server confusion. SSL_clear sets s->state and dtls1_clear sets cookie_len on the server. Setting cookie_len on the server seems to serve no purpose but to let the callback know how large the buffer is. This can be done just before calling the callback. It also avoids a bug where the cookie check can be bypassed, should the server not specify an app_verify_cookie_cb, by supplying a cookie of all zeros of the maximum size. (Zero is fine because an empty cookie is rejected.) The goal here is to avoid needing the SSL_clear calls in the handshake functions. They are currently needed to fix the cookie_len setting when using the generic method. (They get set wrong and then flipped back.) Change-Id: I5095891bc0f7df62d83a9c84312fcf0b84826faa Reviewed-on: https://boringssl-review.googlesource.com/2435 Reviewed-by: Adam Langley <agl@google.com>
2014-11-26 05:29:45 +00:00
if (*cookie_len < 32) {
fprintf(stderr, "Insufficient space for cookie\n");
return 0;
}
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
*cookie_len = 32;
memset(cookie, 42, *cookie_len);
return 1;
}
Tidy DTLS cookie callback types. const-correctness, unsigned long -> size_t. Change-Id: Ic0c2685a48a0f98396c5753b6077c6c0c3b92326 Reviewed-on: https://boringssl-review.googlesource.com/1540 Reviewed-by: Adam Langley <agl@google.com>
2014-08-15 18:38:15 +01:00
static int cookie_verify_callback(SSL *ssl, const uint8_t *cookie, size_t cookie_len) {
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
if (cookie_len != 32) {
fprintf(stderr, "Cookie length mismatch.\n");
return 0;
}
for (size_t i = 0; i < cookie_len; i++) {
if (cookie[i] != 42) {
fprintf(stderr, "Cookie mismatch.\n");
return 0;
}
}
return 1;
}
Add tests for PSK cipher suites. Only the three plain PSK suites for now. ECDHE_PSK_WITH_AES_128_GCM_SHA256 will be in a follow-up. Change-Id: Iafc116a5b2798c61d90c139b461cf98897ae23b3 Reviewed-on: https://boringssl-review.googlesource.com/2051 Reviewed-by: Adam Langley <agl@google.com>
2014-10-27 05:06:24 +00:00
static unsigned psk_client_callback(SSL *ssl, const char *hint,
char *out_identity,
unsigned max_identity_len,
uint8_t *out_psk, unsigned max_psk_len) {
const TestConfig *config = GetConfigPtr(ssl);
if (strcmp(hint ? hint : "", config->psk_identity.c_str()) != 0) {
fprintf(stderr, "Server PSK hint did not match.\n");
return 0;
}
// Account for the trailing '\0' for the identity.
if (config->psk_identity.size() >= max_identity_len ||
config->psk.size() > max_psk_len) {
fprintf(stderr, "PSK buffers too small\n");
return 0;
}
BUF_strlcpy(out_identity, config->psk_identity.c_str(),
max_identity_len);
memcpy(out_psk, config->psk.data(), config->psk.size());
return config->psk.size();
}
static unsigned psk_server_callback(SSL *ssl, const char *identity,
uint8_t *out_psk, unsigned max_psk_len) {
const TestConfig *config = GetConfigPtr(ssl);
if (strcmp(identity, config->psk_identity.c_str()) != 0) {
fprintf(stderr, "Client PSK identity did not match.\n");
return 0;
}
if (config->psk.size() > max_psk_len) {
fprintf(stderr, "PSK buffers too small\n");
return 0;
}
memcpy(out_psk, config->psk.data(), config->psk.size());
return config->psk.size();
}
Document DTLS timeout API and add current_time_cb hook. This is so the tests needn't be sensitive to the clock. It is, unfortunately, a test-only hook, but the DTLS retransmit/timeout logic more-or-less requires it currently. Use this hook to, for now, freeze the clock at zero. This makes the tests deterministic. It might be worth designing a saner API in the future. The current one, notably, requires that the caller's clock be compatible with the one we internally use. It's also not clear whether the caller needs to call DTLSv1_handle_timeout or can just rely on the state machine doing it internally (as it does do). But mock clocks are relatively tame and WebRTC wants to compile against upstream OpenSSL for now, so we're limited in how much new API we can build. Change-Id: I7aad51570596f69275ed0fc1a8892393e4b7ba13 Reviewed-on: https://boringssl-review.googlesource.com/3210 Reviewed-by: Adam Langley <agl@google.com>
2015-01-26 05:22:12 +00:00
static void current_time_cb(SSL *ssl, OPENSSL_timeval *out_clock) {
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
*out_clock = *GetClockPtr(ssl);
Document DTLS timeout API and add current_time_cb hook. This is so the tests needn't be sensitive to the clock. It is, unfortunately, a test-only hook, but the DTLS retransmit/timeout logic more-or-less requires it currently. Use this hook to, for now, freeze the clock at zero. This makes the tests deterministic. It might be worth designing a saner API in the future. The current one, notably, requires that the caller's clock be compatible with the one we internally use. It's also not clear whether the caller needs to call DTLSv1_handle_timeout or can just rely on the state machine doing it internally (as it does do). But mock clocks are relatively tame and WebRTC wants to compile against upstream OpenSSL for now, so we're limited in how much new API we can build. Change-Id: I7aad51570596f69275ed0fc1a8892393e4b7ba13 Reviewed-on: https://boringssl-review.googlesource.com/3210 Reviewed-by: Adam Langley <agl@google.com>
2015-01-26 05:22:12 +00:00
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
static SSL_CTX *setup_ctx(const TestConfig *config) {
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
SSL_CTX *ssl_ctx = NULL;
runner: Implement DHE-RSA. Use it to test DHE-RSA in BoringSSL. Change-Id: I88f7bfa76507a6f60234d61d494c9f94b7df4e0a Reviewed-on: https://boringssl-review.googlesource.com/1377 Reviewed-by: Adam Langley <agl@google.com>
2014-08-02 22:35:45 +01:00
DH *dh = NULL;
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
Merge SSLv23_method and DTLS_ANY_VERSION. This makes SSLv23_method go through DTLS_ANY_VERSION's version negotiation logic. This allows us to get rid of duplicate ClientHello logic. For compatibility, SSL_METHOD is now split into SSL_PROTOCOL_METHOD and a version. The legacy version-locked methods set min_version and max_version based this version field to emulate the original semantics. As a bonus, we can now handle fragmented ClientHello versions now. Because SSLv23_method is a silly name, deprecate that too and introduce TLS_method. Change-Id: I8b3df2b427ae34c44ecf972f466ad64dc3dbb171
2014-12-12 20:55:27 +00:00
ssl_ctx = SSL_CTX_new(config->is_dtls ? DTLS_method() : TLS_method());
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
if (ssl_ctx == NULL) {
goto err;
}
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
if (config->is_dtls) {
// DTLS needs read-ahead to function on a datagram BIO.
//
// TODO(davidben): this should not be necessary. DTLS code should only
// expect a datagram BIO.
SSL_CTX_set_read_ahead(ssl_ctx, 1);
}
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
if (!SSL_CTX_set_ecdh_auto(ssl_ctx, 1)) {
goto err;
}
if (!SSL_CTX_set_cipher_list(ssl_ctx, "ALL")) {
goto err;
}
runner: Implement DHE-RSA. Use it to test DHE-RSA in BoringSSL. Change-Id: I88f7bfa76507a6f60234d61d494c9f94b7df4e0a Reviewed-on: https://boringssl-review.googlesource.com/1377 Reviewed-by: Adam Langley <agl@google.com>
2014-08-02 22:35:45 +01:00
dh = DH_get_2048_256(NULL);
Add malloc failure tests. This commit fixes a number of crashes caused by malloc failures. They were found using the -malloc-test=0 option to runner.go which runs tests many times, causing a different allocation call to fail in each case. (This test only works on Linux and only looks for crashes caused by allocation failures, not memory leaks or other errors.) This is not the complete set of crashes! More can be found by collecting core dumps from running with -malloc-test=0. Change-Id: Ia61d19f51e373bccb7bc604642c51e043a74bd83 Reviewed-on: https://boringssl-review.googlesource.com/2320 Reviewed-by: Adam Langley <agl@google.com>
2014-11-18 01:26:55 +00:00
if (dh == NULL ||
!SSL_CTX_set_tmp_dh(ssl_ctx, dh)) {
runner: Implement DHE-RSA. Use it to test DHE-RSA in BoringSSL. Change-Id: I88f7bfa76507a6f60234d61d494c9f94b7df4e0a Reviewed-on: https://boringssl-review.googlesource.com/1377 Reviewed-by: Adam Langley <agl@google.com>
2014-08-02 22:35:45 +01:00
goto err;
}
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
SSL_CTX_set_session_cache_mode(ssl_ctx, SSL_SESS_CACHE_BOTH);
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
ssl_ctx->select_certificate_cb = select_certificate_callback;
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
SSL_CTX_set_next_protos_advertised_cb(
ssl_ctx, next_protos_advertised_callback, NULL);
Add tests for ALPN support. Both as client and as server. Also tests that ALPN causes False Start to kick in. Change-Id: Ib570346f3c511834152cd2df2ef29541946d3ab4 Reviewed-on: https://boringssl-review.googlesource.com/1753 Reviewed-by: Adam Langley <agl@google.com>
2014-09-06 17:58:58 +01:00
if (!config->select_next_proto.empty()) {
SSL_CTX_set_next_proto_select_cb(ssl_ctx, next_proto_select_callback, NULL);
}
if (!config->select_alpn.empty()) {
SSL_CTX_set_alpn_select_cb(ssl_ctx, alpn_select_callback, NULL);
}
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
SSL_CTX_set_cookie_generate_cb(ssl_ctx, cookie_generate_callback);
SSL_CTX_set_cookie_verify_cb(ssl_ctx, cookie_verify_callback);
Add basic TLS Channel ID tests. Change-Id: I7ccf2b8282dfa8f3985775e8b67edcf3c2949752 Reviewed-on: https://boringssl-review.googlesource.com/1606 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 06:46:07 +01:00
ssl_ctx->tlsext_channel_id_enabled_new = 1;
Document DTLS timeout API and add current_time_cb hook. This is so the tests needn't be sensitive to the clock. It is, unfortunately, a test-only hook, but the DTLS retransmit/timeout logic more-or-less requires it currently. Use this hook to, for now, freeze the clock at zero. This makes the tests deterministic. It might be worth designing a saner API in the future. The current one, notably, requires that the caller's clock be compatible with the one we internally use. It's also not clear whether the caller needs to call DTLSv1_handle_timeout or can just rely on the state machine doing it internally (as it does do). But mock clocks are relatively tame and WebRTC wants to compile against upstream OpenSSL for now, so we're limited in how much new API we can build. Change-Id: I7aad51570596f69275ed0fc1a8892393e4b7ba13 Reviewed-on: https://boringssl-review.googlesource.com/3210 Reviewed-by: Adam Langley <agl@google.com>
2015-01-26 05:22:12 +00:00
ssl_ctx->current_time_cb = current_time_cb;
runner: Implement DHE-RSA. Use it to test DHE-RSA in BoringSSL. Change-Id: I88f7bfa76507a6f60234d61d494c9f94b7df4e0a Reviewed-on: https://boringssl-review.googlesource.com/1377 Reviewed-by: Adam Langley <agl@google.com>
2014-08-02 22:35:45 +01:00
DH_free(dh);
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
return ssl_ctx;
err:
runner: Implement DHE-RSA. Use it to test DHE-RSA in BoringSSL. Change-Id: I88f7bfa76507a6f60234d61d494c9f94b7df4e0a Reviewed-on: https://boringssl-review.googlesource.com/1377 Reviewed-by: Adam Langley <agl@google.com>
2014-08-02 22:35:45 +01:00
if (dh != NULL) {
DH_free(dh);
}
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
if (ssl_ctx != NULL) {
SSL_CTX_free(ssl_ctx);
}
return NULL;
}
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
static int retry_async(SSL *ssl, int ret, BIO *bio,
OPENSSL_timeval *clock_delta) {
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
// No error; don't retry.
if (ret >= 0) {
return 0;
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
}
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
if (clock_delta->tv_usec != 0 || clock_delta->tv_sec != 0) {
// Process the timeout and retry.
OPENSSL_timeval *clock = GetClockPtr(ssl);
clock->tv_usec += clock_delta->tv_usec;
clock->tv_sec += clock->tv_usec / 1000000;
clock->tv_usec %= 1000000;
clock->tv_sec += clock_delta->tv_sec;
memset(clock_delta, 0, sizeof(*clock_delta));
if (DTLSv1_handle_timeout(ssl) < 0) {
printf("Error retransmitting.\n");
return 0;
}
return 1;
}
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
// See if we needed to read or write more. If so, allow one byte through on
// the appropriate end to maximally stress the state machine.
int err = SSL_get_error(ssl, ret);
if (err == SSL_ERROR_WANT_READ) {
async_bio_allow_read(bio, 1);
return 1;
} else if (err == SSL_ERROR_WANT_WRITE) {
async_bio_allow_write(bio, 1);
return 1;
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
}
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
return 0;
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
}
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
static int do_exchange(SSL_SESSION **out_session,
SSL_CTX *ssl_ctx,
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
const TestConfig *config,
bool is_resume,
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
int fd,
SSL_SESSION *session) {
early_callback_called = 0;
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
OPENSSL_timeval clock = {0}, clock_delta = {0};
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
SSL *ssl = SSL_new(ssl_ctx);
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
if (ssl == NULL) {
BIO_print_errors_fp(stdout);
return 1;
}
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
if (!SetConfigPtr(ssl, config) ||
!SetClockPtr(ssl, &clock)) {
Add malloc failure tests. This commit fixes a number of crashes caused by malloc failures. They were found using the -malloc-test=0 option to runner.go which runs tests many times, causing a different allocation call to fail in each case. (This test only works on Linux and only looks for crashes caused by allocation failures, not memory leaks or other errors.) This is not the complete set of crashes! More can be found by collecting core dumps from running with -malloc-test=0. Change-Id: Ia61d19f51e373bccb7bc604642c51e043a74bd83 Reviewed-on: https://boringssl-review.googlesource.com/2320 Reviewed-by: Adam Langley <agl@google.com>
2014-11-18 01:26:55 +00:00
BIO_print_errors_fp(stdout);
return 1;
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (config->fallback_scsv) {
if (!SSL_enable_fallback_scsv(ssl)) {
BIO_print_errors_fp(stdout);
return 1;
}
}
if (!config->key_file.empty()) {
if (!SSL_use_PrivateKey_file(ssl, config->key_file.c_str(),
SSL_FILETYPE_PEM)) {
BIO_print_errors_fp(stdout);
return 1;
}
}
if (!config->cert_file.empty()) {
if (!SSL_use_certificate_file(ssl, config->cert_file.c_str(),
SSL_FILETYPE_PEM)) {
BIO_print_errors_fp(stdout);
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
return 1;
}
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (config->require_any_client_certificate) {
SSL_set_verify(ssl, SSL_VERIFY_PEER|SSL_VERIFY_FAIL_IF_NO_PEER_CERT,
skip_verify);
}
if (config->false_start) {
SSL_set_mode(ssl, SSL_MODE_HANDSHAKE_CUTTHROUGH);
}
if (config->cbc_record_splitting) {
SSL_set_mode(ssl, SSL_MODE_CBC_RECORD_SPLITTING);
}
if (config->partial_write) {
SSL_set_mode(ssl, SSL_MODE_ENABLE_PARTIAL_WRITE);
}
if (config->no_tls12) {
SSL_set_options(ssl, SSL_OP_NO_TLSv1_2);
}
if (config->no_tls11) {
SSL_set_options(ssl, SSL_OP_NO_TLSv1_1);
}
if (config->no_tls1) {
SSL_set_options(ssl, SSL_OP_NO_TLSv1);
}
if (config->no_ssl3) {
SSL_set_options(ssl, SSL_OP_NO_SSLv3);
}
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
if (config->cookie_exchange) {
SSL_set_options(ssl, SSL_OP_COOKIE_EXCHANGE);
}
Add a test for SSL_OP_TLS_D5_BUG. If this is part of SSL_OP_ALL, we should have a test for it. Change-Id: Ia72422beb2da6434726e78e174f3416f90f7c897 Reviewed-on: https://boringssl-review.googlesource.com/1695 Reviewed-by: Adam Langley <agl@google.com>
2014-08-31 05:59:27 +01:00
if (config->tls_d5_bug) {
SSL_set_options(ssl, SSL_OP_TLS_D5_BUG);
}
Add tests for client-initiated renegotiation. These'll get removed once most of renego support is gone, but this is to prove removing the warning alert from the previous commit still prevents legacy renegotiations. Change-Id: I7d9d95e1d4c5d23d3b6d170938a5499a65f2d5ea Reviewed-on: https://boringssl-review.googlesource.com/2236 Reviewed-by: Adam Langley <agl@google.com>
2014-11-08 17:31:52 +00:00
if (config->allow_unsafe_legacy_renegotiation) {
SSL_set_options(ssl, SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION);
}
Add basic TLS Channel ID tests. Change-Id: I7ccf2b8282dfa8f3985775e8b67edcf3c2949752 Reviewed-on: https://boringssl-review.googlesource.com/1606 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 06:46:07 +01:00
if (!config->expected_channel_id.empty()) {
SSL_enable_tls_channel_id(ssl);
}
if (!config->send_channel_id.empty()) {
EVP_PKEY *pkey = LoadPrivateKey(config->send_channel_id);
if (pkey == NULL) {
BIO_print_errors_fp(stdout);
return 1;
}
SSL_enable_tls_channel_id(ssl);
if (!SSL_set1_tls_channel_id(ssl, pkey)) {
EVP_PKEY_free(pkey);
BIO_print_errors_fp(stdout);
return 1;
}
EVP_PKEY_free(pkey);
}
Improve test coverage for server_name extension. Notably, this would have caught ed8270a55c3845abbc85dfeed358597fef059ea9 (although, apart from staring at code coverage, knowing to set resumeSession on the server test isn't exactly obvious). Perhaps we should systematically set it on all extension server tests; ClientHello extension parsing happens after resumption has been determined and is often sensitive to it. Change-Id: Ie83f294a26881a6a41969e9dbd102d0a93cb68b5 Reviewed-on: https://boringssl-review.googlesource.com/1750 Reviewed-by: Adam Langley <agl@google.com>
2014-09-06 17:45:15 +01:00
if (!config->host_name.empty()) {
SSL_set_tlsext_host_name(ssl, config->host_name.c_str());
}
Add tests for ALPN support. Both as client and as server. Also tests that ALPN causes False Start to kick in. Change-Id: Ib570346f3c511834152cd2df2ef29541946d3ab4 Reviewed-on: https://boringssl-review.googlesource.com/1753 Reviewed-by: Adam Langley <agl@google.com>
2014-09-06 17:58:58 +01:00
if (!config->advertise_alpn.empty()) {
SSL_set_alpn_protos(ssl, (const uint8_t *)config->advertise_alpn.data(),
config->advertise_alpn.size());
}
Add tests for PSK cipher suites. Only the three plain PSK suites for now. ECDHE_PSK_WITH_AES_128_GCM_SHA256 will be in a follow-up. Change-Id: Iafc116a5b2798c61d90c139b461cf98897ae23b3 Reviewed-on: https://boringssl-review.googlesource.com/2051 Reviewed-by: Adam Langley <agl@google.com>
2014-10-27 05:06:24 +00:00
if (!config->psk.empty()) {
SSL_set_psk_client_callback(ssl, psk_client_callback);
SSL_set_psk_server_callback(ssl, psk_server_callback);
}
Add tests for OCSP stapling and SCT lists. We forgot to add those when we implemented the features. (Also relevant because they will provide test coverage later for configuring features when using the generic method tables rather than *_client_method.) Change-Id: Ie08b27de893095e01a05a7084775676616459807 Reviewed-on: https://boringssl-review.googlesource.com/2410 Reviewed-by: Adam Langley <agl@google.com>
2014-11-25 06:55:35 +00:00
if (!config->psk_identity.empty() &&
!SSL_use_psk_identity_hint(ssl, config->psk_identity.c_str())) {
BIO_print_errors_fp(stdout);
return 1;
Add tests for PSK cipher suites. Only the three plain PSK suites for now. ECDHE_PSK_WITH_AES_128_GCM_SHA256 will be in a follow-up. Change-Id: Iafc116a5b2798c61d90c139b461cf98897ae23b3 Reviewed-on: https://boringssl-review.googlesource.com/2051 Reviewed-by: Adam Langley <agl@google.com>
2014-10-27 05:06:24 +00:00
}
Add tests for OCSP stapling and SCT lists. We forgot to add those when we implemented the features. (Also relevant because they will provide test coverage later for configuring features when using the generic method tables rather than *_client_method.) Change-Id: Ie08b27de893095e01a05a7084775676616459807 Reviewed-on: https://boringssl-review.googlesource.com/2410 Reviewed-by: Adam Langley <agl@google.com>
2014-11-25 06:55:35 +00:00
if (!config->srtp_profiles.empty() &&
!SSL_set_srtp_profiles(ssl, config->srtp_profiles.c_str())) {
BIO_print_errors_fp(stdout);
return 1;
}
if (config->enable_ocsp_stapling &&
!SSL_enable_ocsp_stapling(ssl)) {
BIO_print_errors_fp(stdout);
return 1;
}
if (config->enable_signed_cert_timestamps &&
!SSL_enable_signed_cert_timestamps(ssl)) {
BIO_print_errors_fp(stdout);
return 1;
Add DTLS-SRTP tests. Just the negotiation portion as everything else is external. This feature is used in WebRTC. Change-Id: Iccc3983ea99e7d054b59010182f9a56a8099e116 Reviewed-on: https://boringssl-review.googlesource.com/2310 Reviewed-by: Adam Langley <agl@google.com>
2014-11-16 00:06:08 +00:00
}
ClientHello Padding for Fast Radio Opening in 3G. The ClientHello record is padded to 1024 bytes when fastradio_padding is enabled. As a result, the 3G cellular radio is fast forwarded to DCH (high data rate) state. This mechanism leads to a substantial redunction in terms of TLS handshake latency, and benefits mobile apps that are running on top of TLS. Change-Id: I3d55197b6d601761c94c0f22871774b5a3dad614
2014-11-22 06:47:56 +00:00
SSL_enable_fastradio_padding(ssl, config->fastradio_padding);
Add min_version and max_version APIs. Amend the version negotiation tests to test this new spelling of max_version. min_version will be tested in a follow-up. Change-Id: Ic4bfcd43bc4e5f951140966f64bb5fd3e2472b01 Reviewed-on: https://boringssl-review.googlesource.com/2583 Reviewed-by: Adam Langley <agl@google.com>
2014-12-12 23:17:51 +00:00
if (config->min_version != 0) {
SSL_set_min_version(ssl, (uint16_t)config->min_version);
}
if (config->max_version != 0) {
SSL_set_max_version(ssl, (uint16_t)config->max_version);
}
Add a basic MTU test. The minimum MTU (not consistently enforced) is just under 256, so it's difficult to test everything, but this is a basic test. (E.g., without renego, the only handshake message with encryption is Finished which fits in the MTU.) It tests the server side because the Certificate message is large enough to require fragmentation. Change-Id: Ida11f1057cebae2b800ad13696f98bb3a7fbbc5e Reviewed-on: https://boringssl-review.googlesource.com/2824 Reviewed-by: Adam Langley <agl@google.com>
2015-01-11 21:29:36 +00:00
if (config->mtu != 0) {
SSL_set_options(ssl, SSL_OP_NO_QUERY_MTU);
SSL_set_mtu(ssl, config->mtu);
}
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
BIO *bio = BIO_new_fd(fd, 1 /* take ownership */);
if (bio == NULL) {
BIO_print_errors_fp(stdout);
return 1;
}
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
if (config->is_dtls) {
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
BIO *packeted = packeted_bio_create(&clock_delta);
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
BIO_push(packeted, bio);
bio = packeted;
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (config->async) {
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
BIO *async =
config->is_dtls ? async_bio_create_datagram() : async_bio_create();
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
BIO_push(async, bio);
bio = async;
}
SSL_set_bio(ssl, bio, bio);
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
if (session != NULL) {
if (SSL_set_session(ssl, session) != 1) {
fprintf(stderr, "failed to set session\n");
return 2;
}
}
int ret;
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
do {
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (config->is_server) {
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
ret = SSL_accept(ssl);
} else {
ret = SSL_connect(ssl);
}
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
} while (config->async && retry_async(ssl, ret, bio, &clock_delta));
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
if (ret != 1) {
SSL_free(ssl);
BIO_print_errors_fp(stdout);
return 2;
}
Shush a MSVC bool/int comparison warning. MSVC doesn't like it when you compare the two. Change-Id: I03c5ff2e2668ac2e536de8278e3a7c98a3dfd117 Reviewed-on: https://boringssl-review.googlesource.com/2460 Reviewed-by: Adam Langley <agl@google.com>
2014-12-03 22:56:39 +00:00
if (is_resume && (!!SSL_session_reused(ssl) == config->expect_session_miss)) {
Add tests for client version negotiation and session resumption. BUG=chromium:417134 Change-Id: If5914be98026d899000fde267b2d329861ca3136 Reviewed-on: https://boringssl-review.googlesource.com/1822 Reviewed-by: Adam Langley <agl@google.com>
2014-09-24 20:21:44 +01:00
fprintf(stderr, "session was%s reused\n",
SSL_session_reused(ssl) ? "" : " not");
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
return 2;
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (!config->expected_server_name.empty()) {
Add a test that server_name extensions are parsed correctly. Change-Id: Id4025835df49eb498df9a48fc81061541778569b Reviewed-on: https://boringssl-review.googlesource.com/1092 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 23:37:33 +01:00
const char *server_name =
SSL_get_servername(ssl, TLSEXT_NAMETYPE_host_name);
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (server_name != config->expected_server_name) {
Add a test that server_name extensions are parsed correctly. Change-Id: Id4025835df49eb498df9a48fc81061541778569b Reviewed-on: https://boringssl-review.googlesource.com/1092 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 23:37:33 +01:00
fprintf(stderr, "servername mismatch (got %s; want %s)\n",
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
server_name, config->expected_server_name.c_str());
Add a test that server_name extensions are parsed correctly. Change-Id: Id4025835df49eb498df9a48fc81061541778569b Reviewed-on: https://boringssl-review.googlesource.com/1092 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 23:37:33 +01:00
return 2;
}
Port early callback support to CBS. Resolve one of the TODOs since it's quick. Adjust the -expect-server-name test to assert it both in the normal codepath and in the early callback, to provide test coverage for SSL_early_callback_ctx_extension_get. Change-Id: I4d71158b9fd2f4fbb54d3e51184bd25d117bdc91 Reviewed-on: https://boringssl-review.googlesource.com/1120 Reviewed-by: Adam Langley <agl@google.com>
2014-07-09 14:30:38 +01:00
if (!early_callback_called) {
fprintf(stderr, "early callback not called\n");
return 2;
}
Add a test that server_name extensions are parsed correctly. Change-Id: Id4025835df49eb498df9a48fc81061541778569b Reviewed-on: https://boringssl-review.googlesource.com/1092 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 23:37:33 +01:00
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (!config->expected_certificate_types.empty()) {
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
uint8_t *certificate_types;
int num_certificate_types =
Add basic TLS Channel ID tests. Change-Id: I7ccf2b8282dfa8f3985775e8b67edcf3c2949752 Reviewed-on: https://boringssl-review.googlesource.com/1606 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 06:46:07 +01:00
SSL_get0_certificate_types(ssl, &certificate_types);
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (num_certificate_types !=
(int)config->expected_certificate_types.size() ||
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
memcmp(certificate_types,
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
config->expected_certificate_types.data(),
Add a test for certificate types parsing. Change-Id: Icddd39ae183f981f78a65427a4dda34449ca389a Reviewed-on: https://boringssl-review.googlesource.com/1111 Reviewed-by: Adam Langley <agl@google.com>
2014-07-08 22:30:11 +01:00
num_certificate_types) != 0) {
fprintf(stderr, "certificate types mismatch\n");
return 2;
}
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (!config->expected_next_proto.empty()) {
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
const uint8_t *next_proto;
unsigned next_proto_len;
SSL_get0_next_proto_negotiated(ssl, &next_proto, &next_proto_len);
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (next_proto_len != config->expected_next_proto.size() ||
memcmp(next_proto, config->expected_next_proto.data(),
next_proto_len) != 0) {
Add a server NPN test. Change-Id: Ib34a24e86bb5de117ecf5609918e130c1ff9532e Reviewed-on: https://boringssl-review.googlesource.com/1161 Reviewed-by: Adam Langley <agl@google.com>
2014-07-12 21:18:02 +01:00
fprintf(stderr, "negotiated next proto mismatch\n");
return 2;
}
}
Add tests for ALPN support. Both as client and as server. Also tests that ALPN causes False Start to kick in. Change-Id: Ib570346f3c511834152cd2df2ef29541946d3ab4 Reviewed-on: https://boringssl-review.googlesource.com/1753 Reviewed-by: Adam Langley <agl@google.com>
2014-09-06 17:58:58 +01:00
if (!config->expected_alpn.empty()) {
const uint8_t *alpn_proto;
unsigned alpn_proto_len;
SSL_get0_alpn_selected(ssl, &alpn_proto, &alpn_proto_len);
if (alpn_proto_len != config->expected_alpn.size() ||
memcmp(alpn_proto, config->expected_alpn.data(),
alpn_proto_len) != 0) {
fprintf(stderr, "negotiated alpn proto mismatch\n");
return 2;
}
}
Add basic TLS Channel ID tests. Change-Id: I7ccf2b8282dfa8f3985775e8b67edcf3c2949752 Reviewed-on: https://boringssl-review.googlesource.com/1606 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 06:46:07 +01:00
if (!config->expected_channel_id.empty()) {
uint8_t channel_id[64];
if (!SSL_get_tls_channel_id(ssl, channel_id, sizeof(channel_id))) {
fprintf(stderr, "no channel id negotiated\n");
return 2;
}
if (config->expected_channel_id.size() != 64 ||
memcmp(config->expected_channel_id.data(),
channel_id, 64) != 0) {
fprintf(stderr, "channel id mismatch\n");
return 2;
}
}
Extended master secret support. This change implements support for the extended master secret. See https://tools.ietf.org/html/draft-ietf-tls-session-hash-01 https://secure-resumption.com/ Change-Id: Ifc7327763149ab0894b4f1d48cdc35e0f1093b93 Reviewed-on: https://boringssl-review.googlesource.com/1930 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-10-11 00:23:43 +01:00
if (config->expect_extended_master_secret) {
if (!ssl->session->extended_master_secret) {
fprintf(stderr, "No EMS for session when expected");
return 2;
}
}
Add tests for OCSP stapling and SCT lists. We forgot to add those when we implemented the features. (Also relevant because they will provide test coverage later for configuring features when using the generic method tables rather than *_client_method.) Change-Id: Ie08b27de893095e01a05a7084775676616459807 Reviewed-on: https://boringssl-review.googlesource.com/2410 Reviewed-by: Adam Langley <agl@google.com>
2014-11-25 06:55:35 +00:00
if (!config->expected_ocsp_response.empty()) {
const uint8_t *data;
size_t len;
SSL_get0_ocsp_response(ssl, &data, &len);
if (config->expected_ocsp_response.size() != len ||
memcmp(config->expected_ocsp_response.data(), data, len) != 0) {
fprintf(stderr, "OCSP response mismatch\n");
return 2;
}
}
if (!config->expected_signed_cert_timestamps.empty()) {
const uint8_t *data;
size_t len;
SSL_get0_signed_cert_timestamp_list(ssl, &data, &len);
if (config->expected_signed_cert_timestamps.size() != len ||
memcmp(config->expected_signed_cert_timestamps.data(),
data, len) != 0) {
fprintf(stderr, "SCT list mismatch\n");
return 2;
}
}
Test server-side renegotiation. This change adds support to the Go code for renegotiation as a client, meaning that we can test BoringSSL's renegotiation as a server. Change-Id: Iaa9fb1a6022c51023bce36c47d4ef7abee74344b Reviewed-on: https://boringssl-review.googlesource.com/2082 Reviewed-by: Adam Langley <agl@google.com>
2014-10-29 00:29:33 +00:00
if (config->renegotiate) {
if (config->async) {
fprintf(stderr, "--renegotiate is not supported with --async.\n");
return 2;
}
SSL_renegotiate(ssl);
ret = SSL_do_handshake(ssl);
if (ret != 1) {
SSL_free(ssl);
BIO_print_errors_fp(stdout);
return 2;
}
SSL_set_state(ssl, SSL_ST_ACCEPT);
ret = SSL_do_handshake(ssl);
if (ret != 1) {
SSL_free(ssl);
BIO_print_errors_fp(stdout);
return 2;
}
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (config->write_different_record_sizes) {
Add initial DTLS tests. Change-Id: I7407261bdb2d788c879f2e67e617a615d9ff8f8b Reviewed-on: https://boringssl-review.googlesource.com/1505 Reviewed-by: Adam Langley <agl@google.com>
2014-08-11 23:43:38 +01:00
if (config->is_dtls) {
fprintf(stderr, "write_different_record_sizes not supported for DTLS\n");
return 6;
}
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
// This mode writes a number of different record sizes in an attempt to
// trip up the CBC record splitting code.
uint8_t buf[32769];
memset(buf, 0x42, sizeof(buf));
static const size_t kRecordSizes[] = {
0, 1, 255, 256, 257, 16383, 16384, 16385, 32767, 32768, 32769};
for (size_t i = 0; i < sizeof(kRecordSizes) / sizeof(kRecordSizes[0]);
i++) {
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
int w;
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
const size_t len = kRecordSizes[i];
size_t off = 0;
if (len > sizeof(buf)) {
fprintf(stderr, "Bad kRecordSizes value.\n");
return 5;
}
Test state machine asynchronous behavior. Add a framework for testing the asynchronous codepath. Move some handshake state machine coverage tests to cover a range of record-layer and handshake-layer asynchronicity. This adds tests for the previous two async bugs fixed. Change-Id: I422ef33ba3eeb0ad04766871ed8bc59b677b169e Reviewed-on: https://boringssl-review.googlesource.com/1410 Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 07:28:57 +01:00
do {
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
w = SSL_write(ssl, buf + off, len - off);
if (w > 0) {
off += (size_t) w;
}
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
} while ((config->async && retry_async(ssl, w, bio, &clock_delta)) ||
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
(w > 0 && off < len));
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
if (w < 0 || off != len) {
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
SSL_free(ssl);
BIO_print_errors_fp(stdout);
return 4;
}
}
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
} else {
Add a test to ensure False Start occurs. This adds the missing test coverage for 7e3305eebd7fb06d57e7f25b3bbf9c10d526f7d5. Change-Id: I8c9f1dc998afa9bb1f6fb2a7872a651037bb4844 Reviewed-on: https://boringssl-review.googlesource.com/1610 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 08:47:07 +01:00
if (config->shim_writes_first) {
int w;
do {
w = SSL_write(ssl, "hello", 5);
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
} while (config->async && retry_async(ssl, w, bio, &clock_delta));
Add a test to ensure False Start occurs. This adds the missing test coverage for 7e3305eebd7fb06d57e7f25b3bbf9c10d526f7d5. Change-Id: I8c9f1dc998afa9bb1f6fb2a7872a651037bb4844 Reviewed-on: https://boringssl-review.googlesource.com/1610 Reviewed-by: Adam Langley <agl@google.com>
2014-08-24 08:47:07 +01:00
}
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
for (;;) {
uint8_t buf[512];
int n;
do {
n = SSL_read(ssl, buf, sizeof(buf));
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
} while (config->async && retry_async(ssl, n, bio, &clock_delta));
Return SSL_ERROR_SYSCALL on unclean EOF. This regressed in fcf25833bc4d057377da1ea755f57b2f5126e555. 0 return code on unclean shutdown means the underlying BIO returned EOF, didn't push any error code, but we haven't seen close_notify yet. The intent seems to be that you go check errno or some BIO-specific equivalent if you care about close_notify. Make sure test code routes all SSL_read return codes through SSL_get_error since that's supposed to work in all cases. (Note that rv == 0 can still give SSL_ERROR_SSL if the error queue is not empty.) Change-Id: I45bf9614573f876d93419ce169a4e0d9ceea9052 Reviewed-on: https://boringssl-review.googlesource.com/2981 Reviewed-by: Adam Langley <agl@google.com>
2015-01-22 21:06:11 +00:00
int err = SSL_get_error(ssl, n);
if (err == SSL_ERROR_ZERO_RETURN ||
(n == 0 && err == SSL_ERROR_SYSCALL)) {
if (n != 0) {
fprintf(stderr, "Invalid SSL_get_error output\n");
return 3;
}
/* Accept shutdowns with or without close_notify.
* TODO(davidben): Write tests which distinguish these two cases. */
break;
} else if (err != SSL_ERROR_NONE) {
if (n > 0) {
fprintf(stderr, "Invalid SSL_get_error output\n");
return 3;
}
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
SSL_free(ssl);
BIO_print_errors_fp(stdout);
return 3;
Return SSL_ERROR_SYSCALL on unclean EOF. This regressed in fcf25833bc4d057377da1ea755f57b2f5126e555. 0 return code on unclean shutdown means the underlying BIO returned EOF, didn't push any error code, but we haven't seen close_notify yet. The intent seems to be that you go check errno or some BIO-specific equivalent if you care about close_notify. Make sure test code routes all SSL_read return codes through SSL_get_error since that's supposed to work in all cases. (Note that rv == 0 can still give SSL_ERROR_SSL if the error queue is not empty.) Change-Id: I45bf9614573f876d93419ce169a4e0d9ceea9052 Reviewed-on: https://boringssl-review.googlesource.com/2981 Reviewed-by: Adam Langley <agl@google.com>
2015-01-22 21:06:11 +00:00
}
/* Successfully read data. */
if (n <= 0) {
fprintf(stderr, "Invalid SSL_get_error output\n");
return 3;
}
for (int i = 0; i < n; i++) {
buf[i] ^= 0xff;
}
int w;
do {
w = SSL_write(ssl, buf, n);
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
} while (config->async && retry_async(ssl, w, bio, &clock_delta));
Return SSL_ERROR_SYSCALL on unclean EOF. This regressed in fcf25833bc4d057377da1ea755f57b2f5126e555. 0 return code on unclean shutdown means the underlying BIO returned EOF, didn't push any error code, but we haven't seen close_notify yet. The intent seems to be that you go check errno or some BIO-specific equivalent if you care about close_notify. Make sure test code routes all SSL_read return codes through SSL_get_error since that's supposed to work in all cases. (Note that rv == 0 can still give SSL_ERROR_SSL if the error queue is not empty.) Change-Id: I45bf9614573f876d93419ce169a4e0d9ceea9052 Reviewed-on: https://boringssl-review.googlesource.com/2981 Reviewed-by: Adam Langley <agl@google.com>
2015-01-22 21:06:11 +00:00
if (w != n) {
SSL_free(ssl);
BIO_print_errors_fp(stdout);
return 4;
Retry sending record split fragment when SSL write fails. When the write size was exactly SSL3_RT_MAX_PLAIN_LENGTH+1 and record splitting is needed, an extra byte would be added to the max size of the message to be written. This would cause the requested size to not exceed the max. If the SSL_WANT_WRITE error were returned, the next packet would not get the extra byte added to the max packet size since record_split_done is set. Since a different set of arguments (SSL3_RT_MAX_PLAIN_LENGTH+1 vs SSL3_RT_MAX_PLAIN_LENGTH) would be passed to do_ssl3_write, it would return an "SSL3_WRITE_PENDING:bad write retry" error. To avoid a failure in the opposite direction, the max variable increment is removed as well. This can happen when SSL_MODE_ENABLE_PARTIAL_WRITE is not enabled and the call to ssl3_write_bytes contains, e.g., a buffer of 2*SSL3_RT_MAX_PLAIN_LENGTH, where the first call into do_ssl3_write succeeds writing the first SSL3_RT_MAX_PLAIN_LENGTH bytes, but writing the second SSL3_RT_MAX_PLAIN_LENGTH bytes fails. This means the first time the the second section of SSL3_RT_MAX_PLAIN_LENGTH bytes has called do_ssl3_write with "max" bytes, but next call to ssl3_write_bytes in turn calls into do_ssl3_write with "max+1" bytes. Change-Id: Icf8453195c1145a54d31b8e8146801118207df03 Reviewed-on: https://boringssl-review.googlesource.com/1420 Reviewed-by: Kenny Root <kroot@google.com> Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-05 23:23:37 +01:00
}
}
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
}
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
if (out_session) {
*out_session = SSL_get1_session(ssl);
}
SSL_shutdown(ssl);
Add some basic server tests to runner.go. client_shim.cc and runner.go are generalized to handle both ends. Plumb a bit through the test case to control which and add server versions of all the cipher suite tests. Change-Id: Iab2640b390f7ed7160c9a9bf6bb34b8bec761b2e Reviewed-on: https://boringssl-review.googlesource.com/1091 Reviewed-by: Adam Langley <agl@google.com>
2014-07-02 00:53:04 +01:00
SSL_free(ssl);
return 0;
}
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
int main(int argc, char **argv) {
Windows build fixes. Windows doesn't have ssize_t, sadly. There's SSIZE_T, but defining an OPENSSL_SSIZE_T seems worse than just using an int. Change-Id: I09bb5aa03f96da78b619e551f92ed52ce24d9f3f Reviewed-on: https://boringssl-review.googlesource.com/1352 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 23:23:51 +01:00
#if !defined(OPENSSL_WINDOWS)
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
signal(SIGPIPE, SIG_IGN);
Windows build fixes. Windows doesn't have ssize_t, sadly. There's SSIZE_T, but defining an OPENSSL_SSIZE_T seems worse than just using an int. Change-Id: I09bb5aa03f96da78b619e551f92ed52ce24d9f3f Reviewed-on: https://boringssl-review.googlesource.com/1352 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 23:23:51 +01:00
#endif
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (!SSL_library_init()) {
return 1;
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
g_ex_data_index = SSL_get_ex_new_index(0, NULL, NULL, NULL, NULL);
Add DTLS timeout and retransmit tests. This extends the packet adaptor protocol to send three commands: type command = | Packet of []byte | Timeout of time.Duration | TimeoutAck When the shim processes a Timeout in BIO_read, it sends TimeoutAck, fails the BIO_read, returns out of the SSL stack, advances the clock, calls DTLSv1_handle_timeout, and continues. If the Go side sends Timeout right between sending handshake flight N and reading flight N+1, the shim won't read the Timeout until it has sent flight N+1 (it only processes packet commands in BIO_read), so the TimeoutAck comes after N+1. Go then drops all packets before the TimeoutAck, thus dropping one transmit of flight N+1 without having to actually process the packets to determine the end of the flight. The shim then sees the updated clock, calls DTLSv1_handle_timeout, and re-sends flight N+1 for Go to process for real. When dropping packets, Go checks the epoch and increments sequence numbers so that we can continue to be strict here. This requires tracking the initial sequence number of the next epoch. The final Finished message takes an additional special-case to test. DTLS triggers retransmits on either a timeout or seeing a stale flight. OpenSSL only implements the former which should be sufficient (and is necessary) EXCEPT for the final Finished message. If the peer's final Finished message is lost, it won't be waiting for a message from us, so it won't time out anything. That retransmit must be triggered on stale message, so we retransmit the Finished message in Go. Change-Id: I3ffbdb1de525beb2ee831b304670a3387877634c Reviewed-on: https://boringssl-review.googlesource.com/3212 Reviewed-by: Adam Langley <agl@google.com>
2015-01-27 06:09:43 +00:00
g_ex_data_clock_index = SSL_get_ex_new_index(0, NULL, NULL, NULL, NULL);
if (g_ex_data_index < 0 || g_ex_data_clock_index < 0) {
Fix a couple more malloc test crashes. The ex_data index may fail to be allocated. Also don't leave a dangling pointer in handshake_dgst if EVP_DigestInit_ex fails and check a few more init function failures. Change-Id: I2e99a89b2171c9d73ccc925a2f35651af34ac5fb Reviewed-on: https://boringssl-review.googlesource.com/2342 Reviewed-by: Adam Langley <agl@google.com>
2014-11-19 02:52:26 +00:00
return 1;
}
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
TestConfig config;
if (!ParseConfig(argc - 1, argv + 1, &config)) {
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
return usage(argv[0]);
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
SSL_CTX *ssl_ctx = setup_ctx(&config);
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
if (ssl_ctx == NULL) {
BIO_print_errors_fp(stdout);
return 1;
}
Fix (harmless) memory leak in the test harness. Change-Id: Ia0daaaaf464cfa0e9d563d7f376ce2bb2e338685 Reviewed-on: https://boringssl-review.googlesource.com/1560 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-19 19:36:45 +01:00
SSL_SESSION *session = NULL;
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
int ret = do_exchange(&session,
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
ssl_ctx, &config,
false /* is_resume */,
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
3 /* fd */, NULL /* session */);
if (ret != 0) {
Fix (harmless) memory leak in the test harness. Change-Id: Ia0daaaaf464cfa0e9d563d7f376ce2bb2e338685 Reviewed-on: https://boringssl-review.googlesource.com/1560 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-19 19:36:45 +01:00
goto out;
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
}
Move configuration into a dedicated TestConfig struct. This removes some duplicate code in parsing command-line flags and, more importantly, makes configuration available when constructing the SSL_CTX and avoids a number of globals. Change-Id: I26e2d2285b732f855a2c82752bc8e0db480c3b30 Reviewed-on: https://boringssl-review.googlesource.com/1502 Reviewed-by: Adam Langley <agl@google.com>
2014-08-12 00:51:50 +01:00
if (config.resume) {
Fix (harmless) memory leak in the test harness. Change-Id: Ia0daaaaf464cfa0e9d563d7f376ce2bb2e338685 Reviewed-on: https://boringssl-review.googlesource.com/1560 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-19 19:36:45 +01:00
ret = do_exchange(NULL,
ssl_ctx, &config,
true /* is_resume */,
4 /* fd */,
config.is_server ? NULL : session);
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
if (ret != 0) {
Fix (harmless) memory leak in the test harness. Change-Id: Ia0daaaaf464cfa0e9d563d7f376ce2bb2e338685 Reviewed-on: https://boringssl-review.googlesource.com/1560 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-19 19:36:45 +01:00
goto out;
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
}
}
Fix (harmless) memory leak in the test harness. Change-Id: Ia0daaaaf464cfa0e9d563d7f376ce2bb2e338685 Reviewed-on: https://boringssl-review.googlesource.com/1560 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-19 19:36:45 +01:00
ret = 0;
out:
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
SSL_SESSION_free(session);
SSL_CTX_free(ssl_ctx);
Fix (harmless) memory leak in the test harness. Change-Id: Ia0daaaaf464cfa0e9d563d7f376ce2bb2e338685 Reviewed-on: https://boringssl-review.googlesource.com/1560 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2014-08-19 19:36:45 +01:00
return ret;
Add test coverage for session resumption with tickets. The shim is now passed two file descriptors. In a session resumption test, the second is used in an abbreviated handshake immediately after the first. Change-Id: I1f348c05f1a8ff2881fb46fc9e869696f74071c6 Reviewed-on: https://boringssl-review.googlesource.com/1291 Reviewed-by: Adam Langley <agl@google.com>
2014-07-23 00:20:02 +01:00
}
Reference in New Issue Copy Permalink