Stripping the .pm file down to just the above entries and running
multiple times gives different results for the produce_decoder.
On multiple runs any iterations over the unordered hash table keys using
foreach my $name (keys %params) results in a different order on multiple
runs. Because of this the mapping from the hash 'value' back to the
'key' will be different.
Note that the code also uses another mechanism in places that uses
"name1" => "value"
"name2" => "*name1"
Rather than fix all the strings the change done was to sort the keys. If
we were to chose to fix the strings then the perl code should be changed
to detect duplicates.
Reviewed-by: Paul Dale <ppzgs1@gmail.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/23688)
slontis [Mon, 4 Mar 2024 02:08:08 +0000 (13:08 +1100)]
Fix BIO_get_new_index() to return an error when it is exhausted.
Fixes #23655
BIO_get_new_index() returns a range of 129..255.
It is set to BIO_TYPE_START (128) initially and is incremented on each
call.
>= 256 is reserved for the class type flags (BIO_TYPE_DESCRIPTOR) so it
should error if it reaches the upper bound.
Reviewed-by: Hugo Landau <hlandau@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23732)
Bernd Edlinger [Wed, 28 Feb 2024 06:14:08 +0000 (07:14 +0100)]
Try to fix intermittent CI failures in sslapitest
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/23705)
Bernd Edlinger [Fri, 23 Feb 2024 11:04:38 +0000 (12:04 +0100)]
Dont run the self-hosted workflows when not available
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23678)
Neil Horman [Tue, 5 Mar 2024 15:56:35 +0000 (10:56 -0500)]
Gate setting of ipi_spec_dst on not building for freebsd
some variants of FreeBSD (notably Dells OneFS) implement IP_PKTINFO
partially, and as such the build breaks for those variants.
specifically, it supports IP_PKTINFO, but the in_pktinfo struct has no
defined ipi_spec_dst field. Work around this by gating the setting of
that variable on not building for FreeBSD
Fixes #23739
Reviewed-by: Hugo Landau <hlandau@openssl.org> Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23753)
Stanislav Zidek [Fri, 1 Mar 2024 14:33:30 +0000 (15:33 +0100)]
interop tests: Fedora 39 config, simplify updates
Imitating Fedora 39 configuration in openssl.cnf with
SECLEVEL lowered to 0 in order to be able to run
TLS 1.3 tests with TLS_AES_128_CCM_8_SHA256.
In order to make updating smoother, check out specific tag rather
than the branch. This way, "old" tests can be fetched until PR
pointing to "new" tests is merged, so backwards-incompatible
changes can be done when needed.
Files specific for openssl upstream moved to separate
directory.
CLA: trivial
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23747)
Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com> Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23695)
Tomas Mraz [Thu, 14 Dec 2023 16:08:56 +0000 (17:08 +0100)]
Allow ignoring unknown sigalgs and groups in the configuration
Related to #20789
Signature algorithms and groups in the configuration that are
preceded with ? character and are unknown to libssl are just ignored.
The handling for them is similar to handling of ciphers.
I.e., there should be a failure only in case the configuration produces
no valid sigalgs or groups.
Also ignore duplicate sigalgs and groups as such confiuration errors
should not be fatal.
apps/cmp: improve -reqin option to read fallback public key from first request message file given
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/21660)
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/21660)
apps/cmp: extend documentation and diagnostics for using -reqin in special situations
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/21660)
apps/cmp: add -reqout_only option for dumping/saving just the initial CMP request message
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/21660)
apps/cmp.c: refactor to fix some coding style nits and more convenient source-level debugging
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/21660)
Neil Horman [Thu, 29 Feb 2024 17:01:31 +0000 (12:01 -0500)]
Dump out qlog json if it is malformed
We're still seeing periodic failures in qlog from malformed json output,
so lets try to catch it.
Modify the verify-qlog.py script to, in the event of an exception in
json.loads, to replay the entire json file to the console, followed by
an exception indicating what line it died trying to parse.
Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23715)
demos/http3: Use `SSL_write_ex2()` together with `SSL_WRITE_FLAG_CONCLUDE`
These calls were introduced by PR #23343.
Change also does a minor tweak to Makefile so CFLAGS and LDFLAGS variables
from the environment are respected.
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23602)
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23404)
Job Snijders [Wed, 21 Feb 2024 21:26:50 +0000 (21:26 +0000)]
Add appropriate lower bound checks for GeneralizedTime and UTCTime
ITU-T X.690 / ISO/IEC 8825-1 section 11.7 and section 11.8
impose specific constraints on how GeneralizedTime and UTCTime
can be encoded in BER/CER/DER. Following from these constraints
a minimum length can be derived.
Checking the length in this context can potentially help prevent
applications from interpreting an invalid GeneralizedTime as a
valid UTCTime.
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23483)
Hamilton Chapman [Wed, 21 Feb 2024 13:47:19 +0000 (13:47 +0000)]
Ensure `$(MAKE)` commands and `CFLAGS` are appropriately quoted in the Makefile.
If a user's `make` command came from a path that contained a space then both the
`$(MAKE)` variable (and parts of the generated `CFLAGS`, when building for iOS)
would not be properly quoted and the build would fail.
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23663)
Reviewed-by: Hugo Landau <hlandau@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23651)
响马 [Sat, 17 Feb 2024 22:57:57 +0000 (06:57 +0800)]
chachap10-ppc.pl: Fix truncated relocation
Fix error: relocation truncated to fit: R_PPC64_REL14 (stub)
against symbol `ChaCha20_ctr32_vsx_8x'
CLA: trivial
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23618)
Michael Baentsch [Mon, 19 Feb 2024 05:41:35 +0000 (06:41 +0100)]
SSL_set1_groups_list(): Fix memory corruption with 40 groups and more
Fixes #23624
The calculation of the size for gid_arr reallocation was wrong.
A multiplication by gid_arr array item size was missing.
Testcase is added.
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23625)
MrRurikov [Wed, 21 Feb 2024 08:11:34 +0000 (11:11 +0300)]
s_cb.c: Add missing return value checks
Return value of function 'SSL_CTX_ctrl', that is called from
SSL_CTX_set1_verify_cert_store() and SSL_CTX_set1_chain_cert_store(),
is not checked, but it is usually checked for this function.
CLA: trivial
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23647)
Matt Caswell [Tue, 20 Feb 2024 15:11:26 +0000 (15:11 +0000)]
Don't print excessively long ASN1 items in fuzzer
Prevent spurious fuzzer timeouts by not printing ASN1 which is excessively
long.
This fixes a false positive encountered by OSS-Fuzz.
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/23640)
Ijtaba Hussain [Fri, 9 Jun 2023 06:04:53 +0000 (11:04 +0500)]
Extended SSL_SESSION functions using time_t
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/21206)
Hugo Landau [Wed, 14 Feb 2024 10:06:45 +0000 (10:06 +0000)]
Minor updates
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Mon, 12 Feb 2024 11:31:18 +0000 (11:31 +0000)]
Fix warning
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Mon, 12 Feb 2024 10:06:23 +0000 (10:06 +0000)]
JSON_ENC: Ensure ossl_json_flush() really flushes the BIO
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Mon, 12 Feb 2024 09:50:59 +0000 (09:50 +0000)]
QUIC CHANNEL: Defer QLOG instantiation until first event
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Mon, 12 Feb 2024 09:50:42 +0000 (09:50 +0000)]
QUIC TXP: Allow QLOG instance retrieval via callback
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Mon, 12 Feb 2024 09:49:58 +0000 (09:49 +0000)]
QUIC QTX: Allow QLOG instance retrieval via callback
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Mon, 12 Feb 2024 09:49:32 +0000 (09:49 +0000)]
QUIC FIFD: Allow QLOG instance retrieval via callback
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Fri, 9 Feb 2024 14:22:51 +0000 (14:22 +0000)]
Add entry to CHANGES.md
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Fri, 9 Feb 2024 14:08:52 +0000 (14:08 +0000)]
QUIC MULTISTREAM TEST: Test explicit event handling mode
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Fri, 9 Feb 2024 12:52:49 +0000 (12:52 +0000)]
QUIC APL: Add implementation of SSL_VALUE_EVENT_HANDLING_MODE
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Fri, 9 Feb 2024 12:52:33 +0000 (12:52 +0000)]
QUIC: Add API for SSL_VALUE_EVENT_HANDLING_MODE
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Hugo Landau [Fri, 9 Feb 2024 12:52:09 +0000 (12:52 +0000)]
QUIC: Add docs for SSL_VALUE_EVENT_HANDLING_MODE
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23535)
Shakti Shah [Tue, 13 Feb 2024 19:03:19 +0000 (00:33 +0530)]
OpenSSL License is applied for some source files, change to Apache 2
The following files
include/openssl/hpke.h
crypto/hpke/hpke.c
crypto/ec/asm/ecp_sm2p256-armv8.pl
crypto/chacha/asm/chacha-loongarch64.pl
still seem to be released under the OpenSSL License instead of the Apache 2 license.
Fixes #23570
Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23576)
Hugo Landau [Wed, 14 Feb 2024 08:08:01 +0000 (08:08 +0000)]
QUIC: Test that SSL_ctrl, SSL_set_mode are routed correctly on QSSOs
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23581)
Hugo Landau [Wed, 14 Feb 2024 08:04:12 +0000 (08:04 +0000)]
QUIC: Fix SSL_ctrl operation for QSSOs
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23581)
Neil Horman [Sat, 16 Dec 2023 20:32:48 +0000 (15:32 -0500)]
Check for NULL cleanup function before using it in encoder_process
encoder_process assumes a cleanup function has been set in the currently
in-use encoder during processing, which can lead to segfaults if said
function hasn't been set
Add a NULL check for this condition, returning -1 if it is not set
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23069)
Neil Horman [Thu, 8 Feb 2024 14:17:44 +0000 (09:17 -0500)]
Skip shlibload tests if no-atexit is configured
the shared library load tests fail if no-atexit is configured. The
entire test suite relies on atexit handling to indicate an at exit
handler has run, by producing a file that the test recipe then reads.
With no-atexit that never happens, and the test fails
If no-atexit is specified, skip all the tests
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23523)
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23545)
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23545)
test_cmp_http: decrease risk of timeouts due to delays caused by the underlying system running tests
Fixes #22870
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/23502)
Tomas Mraz [Mon, 12 Feb 2024 09:59:27 +0000 (10:59 +0100)]
os-zoo.yml: Do not add enable-unstable-qlog as this CI tests all branches
The enable-unstable-qlog is enabled in windows.yml,
which is sufficient for testing it on Windows.
Reviewed-by: Hugo Landau <hlandau@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23550)
Remove unused function arguments from tls_int_new_record_layer
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23514)
Hugo Landau [Thu, 8 Feb 2024 10:27:56 +0000 (10:27 +0000)]
THREADING: Make CRYPTO_MUTEX and CRYPTO_CONDVAR typesafe
There was really no need for this to be void and it made bugs very easy
to introduce accidentally, especially given that the free functions
needed to be passed a pointer to the pointer.
Also fix some bugs in the QUIC code detected immediately by this change.
.
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23519)
Amend CHANGES.md/NEWS.md to reflect DN output changes
Fixes #23492
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23507)
Hugo Landau [Thu, 8 Feb 2024 10:36:50 +0000 (10:36 +0000)]
Windows: Add _dclass to the allowed symbols list
We use isnan() and isinf() in JSON_ENC now, which is translated to a
call to Microsoft's standard library function _dclass.
.
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23517)
Hugo Landau [Thu, 8 Feb 2024 10:12:45 +0000 (10:12 +0000)]
QUIC QLOG: Fix use of size_t and uint64_t
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23517)
Hugo Landau [Thu, 8 Feb 2024 09:55:24 +0000 (09:55 +0000)]
QUIC QLOG: Retrieve PID correctly on Windows
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23517)
MSVC cannot handle this, presumably because this technically involves a
load from a variable to determine that script_name equals "xxx" and it
is unable to do this during evaluation of a constant initializer list.
Resolve this by changing script_name and script_title to be arrays
instead, allowing the correct pointer values to be filled into
script_info as symbol addresses/relocations rather than dereferences.
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23517)
Hugo Landau [Thu, 8 Feb 2024 09:11:46 +0000 (09:11 +0000)]
QUIC qlog: Enable qlog in Windows CI
Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23517)
Tomas Mraz [Wed, 7 Feb 2024 09:27:50 +0000 (10:27 +0100)]
Fix memory leaks on error cases during drbg initializations
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/23503)
Angel Baez [Wed, 7 Feb 2024 15:34:48 +0000 (10:34 -0500)]
Rearrange terms in gf_mul to prevent segfault
CLA: trivial
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23512)