Remove call to cleanup function
Use only one loop to find previous element
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2581)
This fixes the num of fds added/removed returned by ASYNC_WAIT_CTX_get_changed_fds
Previously, the numbers were not consistent with the fds actually written in
the buffers since the fds that have been both added and removed are explicitly
ignored in the loop.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2581)
I don't think this actually affects anything since the cfi_restore
directives aren't strictly needed anyway. (The old values are still in
memory so either will do.)
CLA: trivial
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2582)
EVP_CIPH_FLAG_LENGTH_BITS flag for CFB1 has been broken with the
introduction of the is_partially_overlapping() check that did not take
it into the account (treating number of bits passed as bytes). This
remedies that and allows this flag to work as intended.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1942)
CFI directives annotate instructions that are significant for stack
unwinding procedure. In addition to directives recognized by GNU
assembler this module implements three synthetic ones:
- .cfi_push annotates push instructions in prologue and translates to
.cfi_adjust_cfa_offset (if needed) and .cfi_offset;
- .cfi_pop annotates pop instructions in epilogue and translates to
.cfi_adjust_cfs_offset (if needed) and .cfi_restore;
- .cfi_cfa_expression encodes DW_CFA_def_cfa_expression and passes it
to .cfi_escape as byte vector;
CFA expression syntax is made up mix of DWARF operator suffixes [subset
of] and references to registers with optional bias. Following example
describes offloaded original stack pointer at specific offset from
current stack pointer:
.cfi_cfa_expression %rsp+40,deref,+8
Final +8 has everything to do with the fact that CFA, Canonical Frame
Address, is reference to top of caller's stack, and on x86_64 call to
subroutine pushes 8-byte return address.
Triggered by request from Adam Langley.
Reviewed-by: Rich Salz <rsalz@openssl.org>
or EVP_CTRL_INIT/EVP_CTRL_COPY was not called or failed.
If that happens in EVP_CipherInit_ex/EVP_CIPHER_CTX_copy set cipher = NULL,
aes_gcm_cleanup should check that gctx != NULL before calling OPENSSL_cleanse.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2562)
- harmonize handlers with guidelines and themselves;
- fix some bugs in handlers;
- add missing handlers in chacha and ecp_nistz256 modules;
Reviewed-by: Rich Salz <rsalz@openssl.org>
Fixed a memory leak in ASN1_digest and ASN1_item_digest.
Reworked error handling in asn1_item_embed_new.
Fixed error handling in int_ctx_new and EVP_PKEY_CTX_dup.
Fixed a memory leak in CRYPTO_free_ex_data.
Reworked error handing in x509_name_ex_d2i, x509_name_encode and x509_name_canon.
Check for null pointer in tls_process_cert_verify.
Fixes#2103#2104#2105#2109#2111#2115
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2342)
Updated indentations according project rules, renamed file-local define to the shorter version - USE_RWLOCK, fixed declaration after the if statement in CRYPTO_THREAD_lock_new().
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1981)
Fix compilation on platforms with missing pthread_rwlock_t implementation by replacing it with pthread_mutex_t. An example of such platform can be Android OS 2.0 - 2.1, API level 5 (Eclair), Android NDK platform - android-5 where pthread_rwlock_t is not implemented and is missing in pthread.h.
In case of missing pthread_rwlock_t implementation CRYPTO_RWLOCK will work as exclusive lock in write-only mode of pthread_rwlock_t lock.
The implementation based on pthread_mutex_t must be using PTHREAD_MUTEX_RECURSIVE mode to be compatible with recursive behavior of pthread_rwlock_rdlock.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1981)
The core SipHash supports either 8 or 16-byte output and a configurable
number of rounds.
The default behavior, as added to EVP, is to use 16-byte output and
2,4 rounds, which matches the behavior of most implementations.
There is an EVP_PKEY_CTRL that can control the output size.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2216)
The BIGNUM behaviour is supposed to be "consistent" when going into and
out of APIs, where "consistent" means 'top' is set minimally and that
'neg' (negative) is not set if the BIGNUM is zero (which is iff 'top' is
zero, due to the previous point).
The BN_DEBUG testing (make test) caught the cases that this patch
corrects.
Note, bn_correct_top() could have been used instead, but that is intended
for where 'top' is expected to (sometimes) require adjustment after direct
word-array manipulation, and so is heavier-weight. Here, we are just
catching the negative-zero case, so we test and correct for that
explicitly, in-place.
Change-Id: Iddefbd3c28a13d935648932beebcc765d5b85ae7
Signed-off-by: Geoff Thorpe <geoff@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1672)
Couple of updates to make this code work properly again;
* use OPENSSL_assert() instead of assert() (and #include <assert.h>)
* the circular-dependency-avoidance uses RAND_bytes() (not pseudo)
Change-Id: Iefb5a9dd73f71fd81c1268495c54a64378955354
Signed-off-by: Geoff Thorpe <geoff@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1672)
On error, i2o_SCT_signature() and i2o_SCT() free a pointer that may
have wandered off from the start of the allocated block (not currently
true for i2o_SCT_signature(), but has that potential as the code may
change. To avoid this, save away the start of the allocated block and
free that instead.
Thanks to Guido Vranken for reporting this issue.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2323)
X509_CRL_digest() didn't check if the precomputed sha1 hash was actually
present. This also makes sure there's an appropriate flag to check.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2314)
When the client reads DH parameters from the TLS stream, we only
checked that they all are non-zero. This change updates the check to
use DH_check_params()
DH_check_params() is a new function for light weight checking of the p
and g parameters:
check that p is odd
check that 1 < g < p - 1
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Originally a crash in 32-bit build was reported CHACHA20-POLY1305
cipher. The crash is triggered by truncated packet and is result
of excessive hashing to the edge of accessible memory. Since hash
operation is read-only it is not considered to be exploitable
beyond a DoS condition. Other ciphers were hardened.
Thanks to Robert Święcki for report.
CVE-2017-3731
Reviewed-by: Rich Salz <rsalz@openssl.org>
Originally a crash in 32-bit build was reported CHACHA20-POLY1305
cipher. The crash is triggered by truncated packet and is result
of excessive hashing to the edge of accessible memory (or bogus
MAC value is produced if x86 MD5 assembly module is involved). Since
hash operation is read-only it is not considered to be exploitable
beyond a DoS condition.
Thanks to Robert Święcki for report.
CVE-2017-3731
Reviewed-by: Rich Salz <rsalz@openssl.org>
This function is used to validate application supplied parameters. An
assert should be used to check for an error that is internal to OpenSSL.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2275)
When doing in place encryption the overlapping buffer check can fail
incorrectly where we have done a partial block "Update" operation. This
fixes things to take account of any pending partial blocks.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2275)
If we have previously been passed a partial block in an "Update" call then
make sure we properly increment the output buffer when we use it.
Fixes#2273
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2275)
Lots of references to 16 replaced by AES_BLOCK_SIZE. Also a few other style
tweaks in that function
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2275)
Remove duplicate defines from EVP source files.
Most of them were in evp.h, which is always included.
Add new ones evp_int.h
EVP_CIPH_FLAG_TLS1_1_MULTIBLOCK is now always defined in evp.h, so
remove conditionals on it
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2201)
Add Poly1305 as a "signed" digest.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2128)
ChaCha20 code uses its own custom cipher_data. Add EVP_CIPH_CUSTOM_IV
and EVP_CIPH_ALWAYS_CALL_INIT so that the key and the iv can be set by
different calls of EVP_CipherInit_ex().
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2156)
According to the documentation, the return code should be -1 when
RAND_status does not return 1.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1886)
When setting the digest parameter for DSA parameter generation, the
signature MD was set instead of the parameter generation one.
Fortunately, that's also the one that was used for parameter
generation, but it ultimately meant the parameter generator MD and the
signature MD would always be the same.
Fixes github issue #2016
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2250)
The use of EXFLAG_SET requires the inclusion of openssl/x509v3.h.
openssl/ocsp.h does that, except when OCSP is disabled.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2227)
Still needs to be documented, somehow/somewhere.
The env var OPENSSL_MALLOC_FAILURES controls how often malloc/realloc
should fail. It's a set of fields separated by semicolons. Each field
is a count and optional percentage (separated by @) which defaults to 100.
If count is zero then it lasts "forever." For example: 100;@25 means the
first 100 allocations pass, then the rest have a 25% chance of failing
until the program exits or crashes.
If env var OPENSSL_MALLOC_FD parses as a positive integer, a record
of all malloc "shouldfail" tests is written to that file descriptor.
If a malloc will fail, and OPENSSL_NO_CRYPTO_MDEBUG_BACKTRACE is not set
(platform specific), then a backtrace will be written to the descriptor
when a malloc fails. This can be useful because a malloc may fail but
not be checked, and problems will only occur later.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1252)
That patch also enables support for SHA2 hashes, and
removes support for hashes that were never supported by
cryptodev.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1784)
If we aren't setting public key parameters make EVP_PKEY_CTX available
in SignerInfo so PSS mode and parameters are automatically selected.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
If a key contains any PSS parameter restrictions set them during
sign or verification initialisation. Parameters now become the
default values for sign/verify. Digests are fixed and any attempt
to change them is an error. The salt length can be modified but
must not be less than the minimum value.
If the key parameters are invalid then verification or signing
initialisation returns an error.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
New function rsa_pss_get_param to extract and sanity check PSS parameters.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
Pad mode setting returns an error if the mode is anything other then PSS.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
New function EVP_PKEY_CTX_md() which takes a string and passes a digest
to a ctrl.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
Add support for common operations in PSS by adding a new function
RSA_pkey_ctx_ctrl() which calls EVP_PKEY_CTX_ctrl if the key type
is RSA or PSS.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
Print out RSA-PSS key parameters if present.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
For RSA PSS keys encode and decode parameters when handling public
and private keys.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
Split PSS parameter creation. This adds a new function rsa_pss_params_create
which creates PSS parameters from digest and salt values. This will be
used for PSS key generation.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
Make RSA method more flexible by using the key type from the
method instead of hard coding EVP_PKEY_RSA: by doing this the
same code supports both RSA and RSA-PSS.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
Store hash algorithm used for MGF1 masks in PSS and OAEP modes in PSS and
OAEP parameter structure: this avoids the need to decode part of the ASN.1
structure every time it is used.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2177)
More importantly, port CRL test from boringSSL crypto/x509/x509_test.cc
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1775)
$1<<32>>32 worked fine with either 32- or 64-bit perl for a good while,
relying on quirk that [pure] 32-bit perl performed it as $1<<0>>0. But
this apparently changed in some version past minimally required 5.10,
and operation result became 0. Yet, it went unnoticed for another while,
because most perl package providers configure their packages with
-Duse64bitint option.
Reviewed-by: Rich Salz <rsalz@openssl.org>
On all platforms, if the controlling tty isn't an actual tty, this is
flagged by setting is_a_tty to zero... except on VMS, where this was
treated as an error. Change this to behave like the other platforms.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2063)
There is code that retries calling RAND_bytes() until it gets something
other than 0, which just hangs if we always return 0.
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #2041
If on a non-tty stdin, TTY_get() will fail with errno == ENODEV.
We didn't catch that.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2039)
TTY_get() sometimes surprises us with new errno values to determine if
we have a controling terminal or not. This generated error is a
helpful tool to figure out that this was what happened and what the
unknown value is.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2043)
Since there are many parts of UI_process() that can go wrong, it isn't
very helpful to only return -1 with no further explanation. With this
change, the error message will at least show which part went wrong.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2037)
GH issue #1916 affects only big-endian platforms. TLS is not affected,
because TLS fragment is never big enough.
Reviewed-by: Matt Caswell <matt@openssl.org>
The bug was introduced in 80d27cdb84,
one too many instructions was removed. It went unnoticed, because
new subroutine introduced in previous commit is called in real-life
RSA/DSA/DH cases, while original code is called only in rare tests.
The bug was caught in test_fuzz.
Reviewed-by: Rich Salz <rsalz@openssl.org>
It's called with 0 when it's already locked, with 1 when it's not.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1500
In non-__KERNEL__ context 32-bit-style __ARMEB__/__ARMEL__ macros were
set in arm_arch.h, which is shared between 32- and 64-bit builds. Since
it's not included in __KERNEL__ case, we have to adhere to official
64-bit pre-defines, __AARCH64EB__/__AARCH64EL__.
[If we are to share more code, it would need similar adjustment.]
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Both strdup or malloc failure should raise à err.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1905)
This reflects its position in include/openssl/ct.h.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1548)
Otherwise, |dec| gets moved past the end of the signature by
o2i_SCT_signature and then can't be correctly freed afterwards.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1548)
Factorise multiple bn_get_top(group->field) calls
Add missing checks on some conditional BN_copy return value
Add missing checks on some BN_copy return value
Add missing checks on a few bn_wexpand return value
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1626)
Do not call the time "current", as a different time can be provided.
For example, a time slightly in the future, to provide tolerance for
CT logs with a clock that is running fast.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1554)
Under certain circumstances, the libcrypto init code would loop,
causing a deadlock. This would typically happen if something in
ossl_init_base() caused an OpenSSL error, and the error stack routines
would recurse into the init code before the flag that ossl_init_base()
had been run was checked.
This change makes sure ossl_init_base isn't run once more of the base
is initiated.
Thanks to Dmitry Kostjuchenko for the idea.
Fixes Github issue #1899
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1922)
prio openssl 1.1.0 seed_len < q was accepted and the seed argument was
then ignored. Now DSA_generate_parameters_ex() returns an error in such
a case but no error string.
Signed-off-by: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1657)
This reverts commit 349d1cfddc.
The proposed fix is incorrect. It marks the "run_once" code as having
finished before it has. The intended semantics of run_once is that no
threads should proceed until the code has run exactly once. With this
change the "second" thread will think the run_once code has already been
run and will continue, even though it is still in progress. This could
result in a crash or other incorrect behaviour.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Fixed deadlock in CRYPTO_THREAD_run_once() if call to init() is causing
a recursive call to CRYPTO_THREAD_run_once() again that is causing a hot
deadloop inside do { } while (result == ONCE_ININIT); section.
CLA: trivial
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1913)
llvm's ubsan reported:
runtime error: negation of -9223372036854775808 cannot be represented in
type 'int64_t' (aka 'long'); cast to an unsigned type to negate this
value to itself
Found using libfuzzer
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1908
In order to minimize dependency on assembler version a number of
post-SSE2 instructions are encoded manually. But in order to simplify
the procedure only register operands are considered. Non-register
operands are passed down to assembler. Module in question uses pshufb
with memory operands, and old [GNU] assembler can't handle it.
Fortunately in this case it's possible skip just the problematic
segment without skipping SSSE3 support altogether.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
When configured no-dso, there are no DSO_{whatever} macros defined.
Therefore, before checking those, you have to check if OPENSSL_NO_DSO
is defined.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1902)
Now that we can link specifically with static libraries, the immediate
need to split ppccap.c (and eventually other *cap.c files) is no more.
This reverts commit e3fb4d3d52.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Don't set choice selector on parse failure: this can pass unexpected
values to the choice callback. Instead free up partial structure
directly.
CVE-2016-7053
Thanks to Tyler Nighswander of ForAllSecure for reporting this issue.
Reviewed-by: Richard Levitte <levitte@openssl.org>
The offset to the memory to clear was incorrect, causing a heap buffer
overflow.
CVE-2016-7054
Thanks to Robert Święcki for reporting this
Reviewed-by: Rich Salz <rsalz@openssl.org>
Some of stone-age assembler can't cope with r0 in address. It's actually
sensible thing to do, because r0 is shunted to 0 in address arithmetic
and by refusing r0 assembler effectively makes you understand that.
Reviewed-by: Rich Salz <rsalz@openssl.org>
At the moment you can only do an HKDF Extract and Expand in one go. For
TLS1.3 we need to be able to do an Extract first, and the subsequently do
a number of Expand steps on the same PRK.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Having that code in one central object file turned out to cause
trouble when building test/modes_internal_test.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1883)
BN_RECP_CTX_new direclty use bn_init to avoid twice memset calls
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1879)
RFC 3447, section 8.2.2, steps 3 and 4 states that verifiers must encode
the DigestInfo struct and then compare the result against the public key
operation result. This implies that one and only one encoding is legal.
OpenSSL instead parses with crypto/asn1, then checks that the encoding
round-trips, and allows some variations for the parameter. Sufficient
laxness in this area can allow signature forgeries, as described in
https://www.imperialviolet.org/2014/09/26/pkcs1.html
Although there aren't known attacks against OpenSSL's current scheme,
this change makes OpenSSL implement the algorithm as specified. This
avoids the uncertainty and, more importantly, helps grow a healthy
ecosystem. Laxness beyond the spec, particularly in implementations
which enjoy wide use, risks harm to the ecosystem for all. A signature
producer which only tests against OpenSSL may not notice bugs and
accidentally become widely deployed. Thus implementations have a
responsibility to honor the specification as tightly as is practical.
In some cases, the damage is permanent and the spec deviation and
security risk becomes a tax all implementors must forever pay, but not
here. Both BoringSSL and Go successfully implemented and deployed
RSASSA-PKCS1-v1_5 as specified since their respective beginnings, so
this change should be compatible enough to pin down in future OpenSSL
releases.
See also https://tools.ietf.org/html/draft-thomson-postel-was-wrong-00
As a bonus, by not having to deal with sign/verify differences, this
version is also somewhat clearer. It also more consistently enforces
digest lengths in the verify_recover codepath. The NID_md5_sha1 codepath
wasn't quite doing this right.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1474
All of these don't compile cleanly any more, probably haven't for quite
some time
Reviewed-by: Emilia Käsper <emilia@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1789)
