summaryrefslogtreecommitdiff
path: root/src/lib/libc/stdlib/reallocarray.c (unfollow)
Commit message (Collapse)AuthorFilesLines
2020-10-14Free peekaboo pbuf at end of s_client_main()tb1-1/+2
Otherwise each run of the s_client leaks 16k of memory. This hurts in interactive mode. ok inoguchi jsing
2020-10-13Make sure an OCSP query sends a host headertb1-4/+13
While OCSP uses HTTP/1.0 where a host header is optional, some widely used OCSP responders will return 400 bad request if it is missing. Add such a header unless it's already provided in the user's custom headers. OpenSSL did something similar in ff4a9394a23 and 76e0cd12f68 (both commits are under the old license) ok inoguchi
2020-10-12make fixed-sized fixed-value mib[] arrays be constderaadt3-12/+7
ok guenther tb millert
2020-10-12avoid two overlong lines and drop an unused printf argumenttb1-4/+4
2020-10-12drop references to the SSL protocol because support was removed long ago;schwarze1-6/+6
suggested by tb@
2020-10-12List and describe the recommended methods first and relegate theschwarze1-27/+26
deprecated methods to a separate table. Simplify and shorten the surrounding verbiage. Joint work with tb@.
2020-10-11In ssl_methods.c rev. 1.18, jsing@ deprecated *_server_method(3)schwarze1-11/+18
and *_client_method(3). Adjust the documentation. While here, delete most of the verbiage regarding the deprecated functions SSLv23_*(3) and add the missing entry to RETURN VALUES. OK tb@
2020-10-11SSL3_ENC_METHOD is just a flag word; merge it into SSL_METHOD_INTERNALguenther5-42/+28
with #defines for the per-version initializers instead of extern globals. Add SSL_USE_SHA256_PRF() to complete the abstraction. ok tb@ jsing@
2020-10-11Fix an assert conditioned on DTLS1_VERSION.jsing1-9/+4
This condition previously existed for DTLS BAD_VER, which has long been removed. Furthermore, conditioning on DTLS1_VERSION means this is broken for any newer DTLS version. While here roll up two assertions into one. ok tb@
2020-10-11Grow init_buf before stashing a handshake message for the legacy stack.jsing1-1/+3
When transitioning from the TLSv1.3 stack to the legacy stack, grow init_buf before stashing the handshake message. The TLSv1.3 stack has already received the handshake message (potentially from multiple TLS records) and validated its size, however the default allocation is only for a single plaintext record, which can result in the handshake message failing to fit in certain cases. Issue noted by tb@ via tlsfuzzer. ok tb@
2020-10-11No longer need to cast away const in srtp_find_profile_by_name()tb1-2/+2
2020-10-11Make profile_name const in srtp_find_profile_by_name()tb2-7/+5
There is no reason (and there never was any) for profile_name to be non-const, it was always just passed to strncmp(). Changing this allows removing an ugly instance of casting away const. ok guenther jsing
2020-10-11Update regress for method changes.jsing1-4/+4
2020-10-11Condense and simplify TLS methods.jsing7-442/+73
Historically, OpenSSL has had client and server specific methods - the only difference between these is that the .ssl_connect or .ssl_accept function pointer is set to ssl_undefined_function, with the intention of reducing code size for a statically linked binary that was only a client or server. These days the difference is minimal or non-existant in many cases and we can reduce the amount of code and complexity by having single method. Internally remove all of the client and server specific methods, simplifying code in the process. The external client/server specific API remain, however these now return the same thing as TLS_method() does. ok tb@
2020-10-11Unbreak regress following SRTP_PROTECTION_PROFILE const change.jsing1-2/+2
2020-10-11Constipate srtp_known_profiles, pushing it into .data.rel.roguenther3-17/+18
ok tb@ jsing@
2020-10-11Constipate ssl3_ciphers and tls1[23]_sigalgs*, pushing them intoguenther6-32/+32
.data.rel.ro and .rodata respectively. ok tb@ jsing@
2020-10-10Don't leak bundle_file and cert_file paths at the end.tb1-1/+3
2020-10-09Fix leak or double free with OCSP_request_add0_id()tb1-8/+9
On success, OCSP_request_add0_id() transfers ownership of cid to either 'one' or 'req' depending on whether the latter is NULL or not. On failure, the caller can't tell whether OCSP_ONEREQ_new() failed (in which case cid needs to be freed) or whether it was a failure to allocate memory in sk_insert() (in which case cid must not be freed). The caller is thus faced with the choice of leaving either a leak or a potential double free. Fix this by transferring ownership only at the end of the function. Found while reviewing an upcoming diff by beck. ok jsing
2020-10-09As noted by tb@ previous commit only removed an unused fucntion.otto1-4/+9
So redo previous commit properly: Use random value for canary bytes; ok tb@.
2020-10-08set SO_REUSEADDR on the server socket when we play oscp server.beck1-1/+2
ok inoguchi@ tb@ deraadt@
2020-10-08A void function has no return value, so zap RETURN VALUES sectiontb1-5/+2
documenting that SSL_set_bio(3) cannot fail. A similar commit was made by schwarze a while ago for a few functions in libcrypto.
2020-10-08grammar fix: if/when a renegotiation takeS placetb2-6/+6
2020-10-08Read cert.pem once and reuse it instead of reading it twice per test certtb1-18/+10
chain. It only takes a few dozens of ms to read it, but doing this 7290 times adds up to a few minutes run time. This way, the test completes in a handful of seconds. Diagnosed by jsing, ok beck
2020-10-07zap trailing whitespacetb1-21/+21
2020-10-07add arguments to skip a bunch of x448 teststb1-3/+18
2020-10-07fix line wrappingtb1-3/+2
2020-10-07Hook up tlslegacy regress.jsing1-1/+2
2020-10-07Mop up various things that are now unused with the new record layer.jsing3-32/+6
ok inoguchi@ tb@
2020-10-07Add regress tests for switching from the new TLS stack to the legacy stack.jsing2-0/+642
This includes a test where the server response includes multiple handhshake messages in the single TLS plaintext record (which would have caught the bug just fixed in tls13_legacy.c).
2020-10-07Include a TLS record header when switching to the legacy stack.jsing1-9/+27
When switching to the legacy TLS stack we previously copied any remaining handshake messages into the receive buffer, but do not include any TLS record header (largely due to the fact that we've already processed part of the TLS record that we actually received - that part is placed into the init_buf). This worked fine with the old record layer implementation, however the new record layer expects to find the TLS record header. This means that if we switch from the new stack to the legacy stack (i.e. the remote side does not support TLSv1.3) and there is more than one handshake message in the TLS plaintext record (which Microsoft's TLS stack is known to do), we now read a TLS record of zero bytes instead of getting the correct length. Fix this by generating a pseudo-TLS record header when switching from the new TLS stack to the legacy stack. Found the hard way by guenther@. Thanks to tb@ for coming up with a reproducible test case and doing much of the debugging. ok inoguchi@ tb@
2020-10-06Use std=gnu99 to be able to build the tests with both base clang and gcc.mpi1-1/+3
2020-10-06Use random value for canary bytes; ok tb@otto1-23/+1
2020-10-03Merge d1_{clnt,srvr}.c into ssl_{clnt,srvr}.cjsing6-359/+103
The d1_{clnt,srvr}.c contain a single function each - merge these into the ssl_{clnt,srvr}.c, renaming them with an ssl3_ prefix for consistency. ok beck@ tb@
2020-10-03Use TLSv1_1_enc_data instead of DTLSv1_enc_data.jsing3-11/+6
DTLSv1 is TLSv1.1 over datagrams - there is no need for a separate SSL3_ENC_METHOD struct, just use TLSv1_1_enc_data and remove DTLSv1_enc_data entirely. ok tb@
2020-10-03Reimplement the TLSv1.2 record handling for the read side.jsing6-605/+426
This is the next step in replacing the TLSv1.2 record layer. The existing record handling code does decryption and processing in place, which is not ideal for various reasons, however it is retained for now as other code depends on this behaviour. Additionally, CBC requires special handling to avoid timing oracles - for now the existing timing safe code is largely retained. ok beck@ inoguchi@ tb@
2020-10-03Rename tls13_record_layer_alert() to tls13_record_layer_enqueue_alert()jsing1-3/+4
This avoids naming confusion with an upcoming TLSv1.2 record layer change and is more descriptive of this function. Requested by tb@
2020-10-03Make dtls1_copy_record() take a DTLS1_RECORD_DATA_INTERNAL *.jsing1-10/+4
This removes the need for extra variables and casts. ok inoguchi@ tb@
2020-10-03Inline two macros that are only used in one place each.jsing1-16/+6
This improves readability - while here also add a missing return value check (although it cannot currently fail). ok inoguchi@ tb@
2020-10-03KNFtb1-12/+11
2020-10-03typotb1-1/+1
2020-10-02spelling, punctuation, whitespacetb2-6/+6
2020-10-02Add -[46] nops for compatibility with the OpenSSL s_servertb1-1/+9
On OpenBSD it's necessary to use the eopenssl11 s_server with either -4 or -6 to choose an address family. I often want to try something with an OpenSSL server and then test the same thing with LibreSSL or vice versa. Adding and removing -4s on top of editing the command is annoying and distracting. This commits teaches our s_server to ignore -4 and -6 and thus makes commands that work with eopenssl11 more likely to work with openssl(1). These options are deliberately undocumented and don't show up in help listings. ok bcook inoguchi jsing
2020-10-02Add possibility to link and run this test against the OpenSSL 1.1.1 packagetb1-1/+7
using the make variable EOPENSSL11. Suggested by jsing
2020-10-02Use += and ?= and tidy up whitespacetb1-8/+8
2020-10-02Tidy up: no need to link statically against libcrypto and no needtb1-4/+4
to look at its private headers either.
2020-10-02Make this test compile against OpenSSL 1.1tb1-11/+21
The X509_STORE_CTX struct is opaque in OpenSSL 1.1. To avoid reaching inside it, reuse the trusted certificate store that was just assigned to it and use X509_STORE_CTX_get0_param(3) to access the verification parameters.
2020-09-26Ensure leaf is set up on X509_STORE_CTX before verification.jsing1-9/+7
Previously the leaf certificate was only being set up on the X509_STORE_CTX after two verification steps were performed, however at least one of those steps could result in the verification callback being triggered and existing code breaking. Issue noticed by Raf Czlonka when attempting to connect to talk.google.com using profanity (which does not set SNI and ends up receiving an invalid certificate). ok beck@ deraadt@ tb@
2020-09-26Call dtls1_hm_fragment_free() from dtls1_drain_fragments()jsing2-8/+7
Currently dtls1_drain_fragments() has a incomplete handrolled version of dtls1_hm_fragment_free(), which has the potential to leak memory. Replace the handrolled free with a call to dtls1_hm_fragment_free(). ok inoguchi@ tb@
2020-09-26Have dtls1_new() call dtls1_free() on failure.jsing1-36/+22
Allocate into the appropriate structures and call dtls1_free() on failure, rather than allocating into local variables and then remembering to free various things on failure. ok tb@
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-05-17 17:08:05 +0000