The functions RAND_bytes() and RAND_priv_bytes() are now both based
on a common implementation using RAND_DRBG_bytes() (if the default
OpenSSL rand method is active). This not only simplifies the code
but also has the advantage that additional input from a high precision
timer is added on every generate call if the timer is available.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/5251)
When comparing the implementations of drbg_bytes() and RAND_DRBG_bytes(),
it was noticed that the former split the buffer into chunks when calling
RAND_DRBG_generate() to circumvent the size limitation of the buffer
to outlen <= drb->max_request. This loop was missing in RAND_DRBG_bytes(),
so it was adopted from drbg_bytes().
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/5251)
This check not only prevented the automatic reinstantiation of the
DRBG, which is implemented in RAND_DRBG_generate(), but also prevented
an error message from being generated in the case of failure.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/5251)
BN_from_montgomery_word doesn't have a constant memory access pattern.
Replace the pointer trick with a constant-time select. There is, of
course, still the bn_correct_top leak pervasive in BIGNUM itself.
See also https://boringssl-review.googlesource.com/22904 from BoringSSL.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/5228)
The exponent here is one of d, dmp1, or dmq1 for RSA. This value and its
bit length are both secret. The only public upper bound is the bit width
of the corresponding modulus (RSA n, p, and q, respectively).
Although BN_num_bits is constant-time (sort of; see bn_correct_top notes
in preceding patch), this does not fix the root problem, which is that
the windows are based on the minimal bit width, not the upper bound. We
could use BN_num_bits(m), but BN_mod_exp_mont_consttime is public API
and may be called with larger exponents. Instead, use all top*BN_BITS2
bits in the BIGNUM. This is still sensitive to the long-standing
bn_correct_top leak, but we need to fix that regardless.
This may cause us to do a handful of extra multiplications for RSA keys
which are just above a whole number of words, but that is not a standard
RSA key size.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5154)
(This patch was written by Andy Polyakov. I only wrote the commit
message. Mistakes in the analysis are my fault.)
BN_num_bits, by way of BN_num_bits_word, currently leaks the
most-significant word of its argument via branching and memory access
pattern.
BN_num_bits is called on RSA prime factors in various places. These have
public bit lengths, but all bits beyond the high bit are secret. This
fully resolves those cases.
There are a few places where BN_num_bits is called on an input where the
bit length is also secret. This does *not* fully resolve those cases as
we still only look at the top word. Today, that is guaranteed to be
non-zero, but only because of the long-standing bn_correct_top timing
leak. Once that is fixed, a constant-time BN_num_bits on such inputs
must count bits on each word.
Instead, those cases should not call BN_num_bits at all. In particular,
BN_mod_exp_mont_consttime uses the exponent bit width to pick windows,
but it should be using the maximum bit width. The next patch will fix
this.
Thanks to Dinghao Wu, Danfeng Zhang, Shuai Wang, Pei Wang, and Xiao Liu
for reporting this issue.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5154)
Add SSL_verify_client_post_handshake() for servers to initiate PHA
Add SSL_force_post_handshake_auth() for clients that don't have certificates
initially configured, but use a certificate callback.
Update SSL_CTX_set_verify()/SSL_set_verify() mode:
* Add SSL_VERIFY_POST_HANDSHAKE to postpone client authentication until after
the initial handshake.
* Update SSL_VERIFY_CLIENT_ONCE now only sends out one CertRequest regardless
of when the certificate authentication takes place; either initial handshake,
re-negotiation, or post-handshake authentication.
Add 'RequestPostHandshake' and 'RequirePostHandshake' SSL_CONF options that
add the SSL_VERIFY_POST_HANDSHAKE to the 'Request' and 'Require' options
Add support to s_client:
* Enabled automatically when cert is configured
* Can be forced enabled via -force_pha
Add support to s_server:
* Use 'c' to invoke PHA in s_server
* Remove some dead code
Update documentation
Update unit tests:
* Illegal use of PHA extension
* TLSv1.3 certificate tests
DTLS and TLS behave ever-so-slightly differently. So, when DTLS1.3 is
implemented, it's PHA support state machine may need to be different.
Add a TODO and a #error
Update handshake context to deal with PHA.
The handshake context for TLSv1.3 post-handshake auth is up through the
ClientFinish message, plus the CertificateRequest message. Subsequent
Certificate, CertificateVerify, and Finish messages are based on this
handshake context (not the Certificate message per se, but it's included
after the hash). KeyUpdate, NewSessionTicket, and prior Certificate
Request messages are not included in post-handshake authentication.
After the ClientFinished message is processed, save off the digest state
for future post-handshake authentication. When post-handshake auth occurs,
copy over the saved handshake context into the "main" handshake digest.
This effectively discards the any KeyUpdate or NewSessionTicket messages
and any prior post-handshake authentication.
This, of course, assumes that the ID-22 did not mean to include any
previous post-handshake authentication into the new handshake transcript.
This is implied by section 4.4.1 that lists messages only up to the
first ClientFinished.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4964)
The behavior of resetting the init_lock value to NULL after
freeing it during OPENSSL_cleanup() was added as part of the
global lock commits that were just reverted, but there is desire
to retain this behavior for clarity.
It is unclear that the library would actually remain usable in
any form after OPENSSL_cleanup(), since the required re-initialization
occurs under a CRYPTO_ONCE check that cannot be reset at cleanup time.
That said, a NULL dereference is probably more friendly behavior
in these treacherous waters than using freed memory would be.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/5089)
Conceptually, this is a squashed version of:
Revert "Address feedback"
This reverts commit 75551e07bd.
and
Revert "Add CRYPTO_thread_glock_new"
This reverts commit ed6b2c7938.
But there were some intervening commits that made neither revert apply
cleanly, so instead do it all as one shot.
The crypto global locks were an attempt to cope with the awkward
POSIX semantics for pthread_atfork(); its documentation (the "RATIONALE"
section) indicates that the expected usage is to have the prefork handler
lock all "global" locks, and the parent and child handlers release those
locks, to ensure that forking happens with a consistent (lock) state.
However, the set of functions available in the child process is limited
to async-signal-safe functions, and pthread_mutex_unlock() is not on
the list of async-signal-safe functions! The only synchronization
primitives that are async-signal-safe are the semaphore primitives,
which are not really appropriate for general-purpose usage.
However, the state consistency problem that the global locks were
attempting to solve is not actually a serious problem, particularly for
OpenSSL. That is, we can consider four cases of forking application
that might use OpenSSL:
(1) Single-threaded, does not call into OpenSSL in the child (e.g.,
the child calls exec() immediately)
For this class of process, no locking is needed at all, since there is
only ever a single thread of execution and the only reentrancy is due to
signal handlers (which are themselves limited to async-signal-safe
operation and should not be doing much work at all).
(2) Single-threaded, calls into OpenSSL after fork()
The application must ensure that it does not fork() with an unexpected
lock held (that is, one that would get unlocked in the parent but
accidentally remain locked in the child and cause deadlock). Since
OpenSSL does not expose any of its internal locks to the application
and the application is single-threaded, the OpenSSL internal locks
will be unlocked for the fork(), and the state will be consistent.
(OpenSSL will need to reseed its PRNG in the child, but that is
an orthogonal issue.) If the application makes use of locks from
libcrypto, proper handling for those locks is the responsibility of
the application, as for any other locking primitive that is available
for application programming.
(3) Multi-threaded, does not call into OpenSSL after fork()
As for (1), the OpenSSL state is only relevant in the parent, so
no particular fork()-related handling is needed. The internal locks
are relevant, but there is no interaction with the child to consider.
(4) Multi-threaded, calls into OpenSSL after fork()
This is the case where the pthread_atfork() hooks to ensure that all
global locks are in a known state across fork() would come into play,
per the above discussion. However, these "calls into OpenSSL after
fork()" are still subject to the restriction to async-signal-safe
functions. Since OpenSSL uses all sorts of locking and libc functions
that are not on the list of safe functions (e.g., malloc()), this
case is not currently usable and is unlikely to ever be usable,
independently of the locking situation. So, there is no need to
go through contortions to attempt to support this case in the one small
area of locking interaction with fork().
In light of the above analysis (thanks @davidben and @achernya), go
back to the simpler implementation that does not need to distinguish
"library-global" locks or to have complicated atfork handling for locks.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/5089)
They aren't needed if all they do is set bio->init = 1 and zero other
fields that are already zeroed
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5223)
Without this, every BIO implementation is forced to have a create
method, just to set bio->init = 1.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5223)
Some older glibc versions require the `-lrt` linker option for
resolving the reference to `clock_gettime'. Since it is not desired
to add new library dependencies in version 1.1.1, the call to
clock_gettime() is replaced by a call to gettimeofday() for the
moment. It will be added back in version 1.2.
Signed-off-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/5199)
C preprocessor flags get separated from C flags, which has the
advantage that we don't get loads of macro definitions and inclusion
directory specs when linking shared libraries, DSOs and programs.
This is a step to add support for "make variables" when configuring.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5177)
The new extension is like signature_algorithms, but only for the
signature *on* the certificate we will present to the peer (the
old signature_algorithms extension is still used for signatures that
we *generate*, i.e., those over TLS data structures).
We do not need to generate this extension, since we are the same
implementation as our X.509 stack and can handle the same types
of signatures, but we need to be prepared to receive it, and use the received
information when selecting what certificate to present.
There is a lot of interplay between signature_algorithms_cert and
signature_algorithms, since both affect what certificate we can
use, and thus the resulting signature algorithm used for TLS messages.
So, apply signature_algorithms_cert (if present) as a filter on what
certificates we can consider when choosing a certificate+sigalg
pair.
As part of this addition, we also remove the fallback code that let
keys of type EVP_PKEY_RSA be used to generate RSA-PSS signatures -- the
new rsa_pss_pss_* and rsa_pss_rsae_* signature schemes have pulled
the key type into what is covered by the signature algorithm, so
we should not apply this sort of compatibility workaround.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5068)
Correct error return value in OCSP_basic_sign().
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)
Add a -rsigopt option to the ocsp command that allows signature parameters to be provided for the signing of OCSP responses. The parameters that may be provided to -rsigopt are the same as may be provided to -sigopt in the ca, req, and x509 commands.
This PR also defines a OCSP_basic_sign_ctx() function, which functions in the same way as OCSP_basic_sign(), except that it accepts a EVP_MD_CTX rather than a key and digest. The OCSP_basic_sign_ctx() function is used to implement the -rsigopt option in the ocsp command.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)
This just adds the various extension functions. More changes will be
required to actually use them.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)
Support added for these two digests, available only via the EVP interface.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5093)
EVP_PKEY_asn1_find_str() would search through standard asn1 methods
first, then those added by the application, which EVP_PKEY_asn1_find()
worked the other way around. Also, EVP_PKEY_asn1_find_str() didn't
handle aliases.
This change brings EVP_PKEY_asn1_find_str() closer to EVP_PKEY_asn1_find().
Fixes#5086
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5137)
Because OPENSSL_SYS_CYGWIN will keep OPENSSL_SYS_UNIX defined, there's
no point having checks of this form:
#if (defined(OPENSSL_SYS_UNIX) || defined(OPENSSL_SYS_CYGWIN))
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5060)
Commit 9f9442918a changed the semantics of BN_copy() to additionally
copy the BN_FLG_CONSTTIME flag if it is set. This turns out to be
ill advised as it has unintended consequences. For example calling
BN_mod_inverse_no_branch() can sometimes return a result with the flag
set and sometimes not as a result. This can lead to later failures if we
go down code branches that do not support constant time, but check for
the presence of the flag.
The original commit was made due to an issue in BN_MOD_CTX_set(). The
original PR fixed the problem in that function, but it was changed in
review to fix it in BN_copy() instead. The solution seems to be to revert
the BN_copy() change and go back to the originally proposed way.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/5080)
Fixes#5076
Since do_rand_drbg_init() allocates three locks, it needs to ensure
that OPENSSL_init_crypto() is called, otherwise these resources are
not cleaned up properly.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/5083)
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4958)
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5000)
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4634)
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4634)
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4634)
SPARC ISA doesn't have provisions to back up 128-bit multiplications
and additions. And so multiplications are done with library calls
and carries with comparisons and conditional moves. As result base
2^51 code is >40% slower...
Reviewed-by: Tim Hudson <tjh@openssl.org>
This is based on RT#3810, which added dedicated modular inversion.
ECDSA verify results improves as well, but not as much.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5001)
The DRGB concept described in NIST SP 800-90A provides for having different
algorithms to generate random output. In fact, the FIPS object module used to
implement three of them, CTR DRBG, HASH DRBG and HMAC DRBG.
When the FIPS code was ported to master in #4019, two of the three algorithms
were dropped, and together with those the entire code that made RAND_DRBG
generic was removed, since only one concrete implementation was left.
This commit restores the original generic implementation of the DRBG, making it
possible again to add additional implementations using different algorithms
(like RAND_DRBG_CHACHA20) in the future.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4998)
The generic part of the FIPS DRBG was implemented in fips_drbg_lib.c and the
algorithm specific parts in fips_drbg_<alg>.c for <alg> in {ctr, hash, hmac}.
Additionally, there was the module fips_drbg_rand.c which contained 'gluing'
code between the RAND_METHOD api and the FIPS DRBG.
When the FIPS code was ported to master in #4019, for some reason the ctr-drbg
implementation from fips_drbg_ctr.c ended up in drbg_rand.c instead of drbg_ctr.c.
This commit renames the module drbg_rand.c back to drbg_ctr.c, thereby restoring
a simple relationship between the original fips modules and the drbg modules
in master:
fips_drbg_lib.c => drbg_lib.c /* generic part of implementation */
fips_drbg_<alg>.c => drbg_<alg>.c /* algorithm specific implementations */
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4998)
Similar to commit 17b602802114d53017ff7894319498934a580b17(
"Remove extra `the` in SSL_SESSION_set1_id.pod"), this commit removes
typos where additional 'the' have been added.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4999)
"Double" is in quotes because improvement coefficient varies
significantly depending on platform and compiler. You're likely
to measure ~2x improvement on popular desktop and server processors,
but not so much on mobile ones, even minor regression on ARM
Cortex series. Latter is because they have rather "weak" umulh
instruction. On low-end x86_64 problem is that contemporary gcc
and clang tend to opt for double-precision shift for >>51, which
can be devastatingly slow on some processors.
Just in case for reference, trick is to use 2^51 radix [currently
only for DH].
Reviewed-by: Rich Salz <rsalz@openssl.org>
Hardware used for benchmarking courtesy of Atos, experiments run by
Romain Dolbeau <romain.dolbeau@atos.net>. Kudos!
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4855)
Previously, the RAND_DRBG_uninstantiate() call was not exactly inverse to
RAND_DRBG_instantiate(), because some important member values of the
drbg->ctr member where cleared. Now these values are restored internally.
Signed-off-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/4402)
Signed-off-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/4402)
Every DRBG now supports automatic reseeding not only after a given
number of generate requests, but also after a specified time interval.
Signed-off-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/4402)
A third shared DRBG is added, the so called master DRBG. Its sole purpose
is to reseed the two other shared DRBGs, the public and the private DRBG.
The randomness for the master DRBG is either pulled from the os entropy
sources, or added by the application using the RAND_add() call.
The master DRBG reseeds itself automatically after a given number of generate
requests, but can also be reseeded using RAND_seed() or RAND_add().
A reseeding of the master DRBG is automatically propagated to the public
and private DRBG. This construction fixes the problem, that up to now
the randomness provided by RAND_add() was added only to the public and
not to the private DRBG.
Signed-off-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/4402)
This commit adds comments to bio_method_st definitions where the
function pointers are defined as NULL. Most of the structs have comments
but some where missing and not all consitent.
CLA: trivial
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4881)
The new ServerHello format is essentially now the same as the old TLSv1.2
one, but it must additionally include supported_versions. The version
field is fixed at TLSv1.2, and the version negotiation happens solely via
supported_versions.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4701)
Reduce RSA_MAX_PRIME_NUM to 5.
Remove no longer used RSA_MIN_PRIME_SIZE.
Make rsa_multip_cap honor RSA_MAX_PRIME_NUM.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4905)
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4916)
If OPENSSL_init_crypto() hasn't been called yet when ERR_get_state()
is called, it need to be called early, so the base initialization is
done. On some platforms (those who support DSO functionality and
don't define OPENSSL_USE_NODELETE), that includes a call of
ERR_set_mark(), which calls this function again.
Furthermore, we know that ossl_init_thread_start(), which is called
later in ERR_get_state(), calls OPENSSL_init_crypto(0, NULL), except
that's too late.
Here's what happens without an early call of OPENSSL_init_crypto():
=> ERR_get_state():
=> CRYPTO_THREAD_get_local():
<= NULL;
# no state is found, so it gets allocated.
=> ossl_init_thread_start():
=> OPENSSL_init_crypto():
# Here, base_inited is set to 1
# before ERR_set_mark() call
=> ERR_set_mark():
=> ERR_get_state():
=> CRYPTO_THREAD_get_local():
<= NULL;
# no state is found, so it gets allocated!!!!!
=> ossl_init_thread_start():
=> OPENSSL_init_crypto():
# base_inited is 1,
# so no more init to be done
<= 1
<=
=> CRYPTO_thread_set_local():
<=
<=
<=
<= 1
<=
=> CRYPTO_thread_set_local() # previous value removed!
<=
Result: double allocation, and we have a leak.
By calling the base OPENSSL_init_crypto() early, we get this instead:
=> ERR_get_state():
=> OPENSSL_init_crypto():
# Here, base_inited is set to 1
# before ERR_set_mark() call
=> ERR_set_mark():
=> ERR_get_state():
=> OPENSSL_init_crypto():
# base_inited is 1,
# so no more init to be done
<= 1
=> CRYPTO_THREAD_get_local():
<= NULL;
# no state is found, so it gets allocated
# let's assume we got 0xDEADBEEF
=> ossl_init_thread_start():
=> OPENSSL_init_crypto():
# base_inited is 1,
# so no more init to be done
<= 1
<= 1
=> CRYPTO_thread_set_local():
<=
<=
<=
<= 1
=> CRYPTO_THREAD_get_local():
<= 0xDEADBEEF
<= 0xDEADBEEF
Result: no leak.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4913)
As per documentation, the RSA keys should not be smaller than 64bit (the
documentation mentions something about a quirk in the prime generation
algorithm). I am adding check into the code which used to be 16 for some
reason.
My primary motivation is to get rid of the last sentence in the
documentation which suggest that typical keys have 1024 bits (instead
updating it to the now default 2048).
I *assume* that keys less than the 2048 bits (say 512) are used for
education purposes.
The 512 bits as the minimum have been suggested by Bernd Edlinger.
Signed-off-by: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4547)
I noticed that some of the BIO_METHOD structs are placing the name on
the same line as the type and some don't. This commit places the name
on a separate line for consistency (which looks like what the majority
do)
CLA: trivial
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4878)
Expression '...' is always true.
The 'b->init' variable is assigned values twice successively
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4753)
Pointer 'o' is set inside a local buffer, so it can't be NULL.
Also fix coding style and add comments
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4754)
256-bit AVX512VL was estimated to deliver ~50% improvement over AVX2
and it did live up to the expectations.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4838)
It was observed that AVX512 code paths can negatively affect overall
Skylake-X system performance. But we are talking specifically about
512-bit code, while AVX512VL, 256-bit variant of AVX512F instructions,
is supposed to fly as smooth as AVX2. Which is why it remains unmasked.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4838)
This line will allow use private keys, which created by Crypto Pro, to
sign with OpenSSL.
CLA: trivial
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4836)
This initial commit is unoptimized reference version that handles
input lengths divisible by 4 blocks.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4830)
Follow up from the conversion to use SSLfatal() in the state machine to
clean things up a bit more.
[extended tests]
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4778)
Sometimes at the top level of the state machine code we know we are
supposed to be in a fatal error condition. This commit adds some sanity
checks to ensure that SSLfatal() has been called.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4778)
The __clang__-guarded #defines cause gas to complain if clang is passed
-fno-integrated-as. Emitting .syntax unified when those are used fixes
this. This matches the change made to ghash-armv4.pl in
6cf412c473.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/3694)
Multi-prime RSA security is not determined by modulus length alone, but
depends even on number of primes. Too many primes render security
inadequate, but there is no common amount of primes or common factors'
length that provide equivalent secuity promise as two-prime for given
modulus length. Maximum amount of permitted primes is determined
according to following table.
<1024 | >=1024 | >=4096 | >=8192
------+--------+--------+-------
2 | 3 | 4 | 5
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4791)
Only chacha_internal_test is affected, since this path is not used
from EVP.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4758)
Convert AVX512F+VL+BW code path to pure AVX512F, so that it can be
executed even on Knights Landing. Trigger for modification was
observation that AVX512 code paths can negatively affect overall
Skylake-X system performance. Since we are likely to suppress
AVX512F capability flag [at least on Skylake-X], conversion serves
as kind of "investment protection".
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4758)
This avoids taking quadratic time to pretty-print certificates with
excessively large integer fields. Very large integers aren't any more
readable in decimal than hexadecimal anyway, and the i2s_* functions
will parse either form.
Found by libFuzzer.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4790)
Switch to make it return an uint32_t instead of the various different
types it returns now.
Fixes: #3125
Reviewed-by: Andy Polyakov <appro@openssl.org>
GH: #4757
Originally it was thought that it's possible to use AVX512VL+BW
instructions with XMM and YMM registers without kernel enabling
ZMM support, but it turned to be wrong assumption.
Reviewed-by: Rich Salz <rsalz@openssl.org>
* Introduce RSA_generate_multi_prime_key to generate multi-prime
RSA private key. As well as the following functions:
RSA_get_multi_prime_extra_count
RSA_get0_multi_prime_factors
RSA_get0_multi_prime_crt_params
RSA_set0_multi_prime_params
RSA_get_version
* Support EVP operations for multi-prime RSA
* Support ASN.1 operations for multi-prime RSA
* Support multi-prime check in RSA_check_key_ex
* Support multi-prime RSA in apps/genrsa and apps/speed
* Support multi-prime RSA manipulation functions
* Test cases and documentation are added
* CHANGES is updated
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/4241)
EVP_PKEY_public_check() and EVP_PKEY_param_check()
Doc and test cases are added
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4647)
All exponentiation subroutines but BN_mod_exp_mont_consttime produce
non-negative result for negative input, which is confusing for fuzzer.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/4676)
Performance regression was reported for EC key generation between
1.0.2 and 1.1.x [in GH#2891]. It naturally depends on platform,
values between 6 and 9% were observed.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4743)
It's argued that /WX allows to keep better focus on new code, which
motivates its comeback...
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4721)
Even though |Blen| is declared uint64_t it was casted implicitly to int.
[Caught by VC warning subsytem.]
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4721)
|flags| argument to do_esc_char was apparently truncated by implicit
cast. [Caught by VC warning subsytem.]
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4721)
Add a new function OCSP_resp_get0_signer() that looks in the
certs bundled with the response as well as in additional certificates
provided as a function argument, returning the certificate that signed
the given response (if present).
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4573)
Around 138 distinct errors found and fixed; thanks!
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3459)
* addressing", Proc. 6th Conference on Very Large Databases: 212–223
^
The EN DASH ('–') in this line is one UTF-8 character (hex: e2 80 93).
Under some code page setting (e.g. 936), Visual Studio may report C4819
warning: The file contains a character that cannot be represented in the
current code page.
Replace this character with the ASCII char '-' (Hex Code: 2D).
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4691)
There were 4 macros in ocsp.h that have not worked since 1.1.0 because
they attempt to access the internals of an opaque structure.
For OCSP_REQUEST_sign() applications should use OCSP_request_sign() instead.
For OCSP_BASICRESP_sign() applications should use OCSP_basic_sign() instead.
For OCSP_REQUEST_verify() applications should use OCSP_request_verify()
instead.
For OCSP_BASICRESP_verify() applications should use OCSP_basic_verify()
instead.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4635)
SM3 is a secure hash function which is part of the Chinese
"Commercial Cryptography" suite of algorithms which use is
required for certain commercial applications in China.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4616)
It's not clear if it's a feature or bug, but binutils-2.29[.1]
interprets 'adr' instruction with Thumb2 code reference differently,
in a way that affects calculation of addresses of constants' tables.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/4669)
information about the length of the scalar used in ECDSA operations
from a large number (2^32) of signatures.
This doesn't rate as a CVE because:
* For the non-constant time code, there are easier ways to extract
more information.
* For the constant time code, it requires a significant number of signatures
to leak a small amount of information.
Thanks to Neals Fournaise, Eliane Jaulmes and Jean-Rene Reinhard for
reporting this issue.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4576)
information about the length of a value used in DSA operations from
a large number of signatures.
This doesn't rate as a CVE because:
* For the non-constant time code, there are easier ways to extract
more information.
* For the constant time code, it requires a significant number of signatures
to leak a small amount of information.
Thanks to Neals Fournaise, Eliane Jaulmes and Jean-Rene Reinhard for
reporting this issue.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4576)
No two public key ASN.1 methods with the same pkey_id can be
registered at the same time.
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/4596)
If the list of fds contains only (one or more) entries marked
as deleted prior to the entry currently being deleted, and the
entry currently being deleted was only just added, the 'prev'
pointer would never be updated from its initial NULL value, and
we would dereference NULL while trying to remove the entry from
the linked list.
Reported by Coverity.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/4602)
Extend the s390x capability vector to store the longer facility list
available from z13 onwards. The bits indicating the vector extensions
are set to zero, if the kernel does not enable the vector facility.
Also add capability bits returned by the crypto instructions' query
functions.
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4542)
Use the newly introduced sk_TYPE_new_reserve API to simplify the
reservation of stack as creating it.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4592)
The functions strcasecmp() and strncasecmp() will use locale specific rules
when performing comparison. This could cause some problems in certain
locales. For example in the Turkish locale an 'I' character is not the
uppercase version of 'i'. However IA5 strings should not use locale specific
rules, i.e. for an IA5 string 'I' is uppercase 'i' even if using the
Turkish locale.
This fixes a bug in name constraints checking reported by Thomas Pornin
(NCCGroup).
This is not considered a security issue because it would require both a
Turkish locale (or other locale with similar issues) and malfeasance by
a trusted name-constrained CA for a certificate to pass name constraints
in error. The constraints also have to be for excluded sub-trees which are
extremely rare. Failure to match permitted subtrees is a bug, not a
vulnerability.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4569)
This is a combination of sk_new and sk_reserve, to make it more
convenient to allocate a new stack with reserved memory and comaprison
function (if any).
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4559)
asn1_item_embed_free() will try unlocking and fail in this case, and
since the new item was just allocated on the heap, free it directly
with OPENSSL_free() instead.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/4579)
The previous change with this intention didn't quite do it. An
embedded item must not be freed itself, but might potentially contain
non-embedded elements, which must be freed.
So instead of calling ASN1_item_ex_free(), where we can't pass the
embed flag, we call asn1_item_embed_free() directly.
This changes asn1_item_embed_free() from being a static function to
being a private non-static function.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/4579)
The lhash expand() function can fail if realloc fails. The previous
implementation made changes to the structure and then attempted to do a
realloc. If the realloc failed then it attempted to undo the changes it
had just made. Unfortunately changes to lh->p were not undone correctly,
ultimately causing subsequent expand() calls to increment num_nodes to a
value higher than num_alloc_nodes, which can cause out-of-bounds reads/
writes. This is not considered a security issue because an attacker cannot
cause realloc to fail.
This commit moves the realloc call to near the beginning of the function
before any other changes are made to the lhash structure. That way if a
failure occurs we can immediately fail without having to undo anything.
Thanks to Pavel Kopyl (Samsung) for reporting this issue.
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4550)
An embedded item wasn't allocated separately on the heap, so don't
free it as if it was.
Issue discovered by Pavel Kopyl
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4572)
The function BN_security_bits() uses the values from SP800-57 to assign
security bit values for different FF key sizes. However the value for 192
security bits is wrong. SP800-57 has it as 7680 but the code had it as
7690.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4546)
Since return is inconsistent, I removed unnecessary parentheses and
unified them.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4541)
The drbg's lock must be held across calls to RAND_DRBG_generate()
to prevent simultaneous modification of internal state.
This was observed in practice with simultaneous SSL_new() calls attempting
to seed the (separate) per-SSL RAND_DRBG instances from the global
rand_drbg instance; this eventually led to simultaneous calls to
ctr_BCC_update() attempting to increment drbg->bltmp_pos for their
respective partial final block, violating the invariant that bltmp_pos < 16.
The AES operations performed in ctr_BCC_blocks() makes the race window
quite easy to trigger. A value of bltmp_pos greater than 16 induces
catastrophic failure in ctr_BCC_final(), with subtraction overflowing
and leading to an attempt to memset() to zero a very large range,
which eventually reaches an unmapped page and segfaults.
Provide the needed locking in get_entropy_from_parent(), as well as
fixing a similar issue in RAND_priv_bytes(). There is also an
unlocked call to RAND_DRBG_generate() in ssl_randbytes(), but the
requisite serialization is already guaranteed by the requirements on
the application's usage of SSL objects, and no further locking is
needed for correct behavior. In that case, leave a comment noting
the apparent discrepancy and the reason for its safety (at present).
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4328)
The DRBG_RESEED state plays an analogue role to the |reseed_required_flag| in
Appendix B.3.4 of [NIST SP 800-90A Rev. 1]. The latter is a local variable,
the scope of which is limited to the RAND_DRBG_generate() function. Hence there
is no need for a DRBG_RESEED state outside of the generate function. This state
was removed and replaced by a local variable |reseed_required|.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4328)
Reseeding is handled very differently by the classic RAND_METHOD API
and the new RAND_DRBG api. These differences led to some problems when
the new RAND_DRBG was made the default OpenSSL RNG. In particular,
RAND_add() did not work as expected anymore. These issues are discussed
on the thread '[openssl-dev] Plea for a new public OpenSSL RNG API'
and in Pull Request #4328. This commit fixes the mentioned issues,
introducing the following changes:
- Replace the fixed size RAND_BYTES_BUFFER by a new RAND_POOL API which
facilitates collecting entropy by the get_entropy() callback.
- Don't use RAND_poll()/RAND_add() for collecting entropy from the
get_entropy() callback anymore. Instead, replace RAND_poll() by
RAND_POOL_acquire_entropy().
- Add a new function rand_drbg_restart() which tries to get the DRBG
in an instantiated state by all means, regardless of the current
state (uninstantiated, error, ...) the DRBG is in. If the caller
provides entropy or additional input, it will be used for reseeding.
- Restore the original documented behaviour of RAND_add() and RAND_poll()
(namely to reseed the DRBG immediately) by a new implementation based
on rand_drbg_restart().
- Add automatic error recovery from temporary failures of the entropy
source to RAND_DRBG_generate() using the rand_drbg_restart() function.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4328)
An instruction's QUERY function is executed at initialization, iff the required
MSA level is installed. Therefore, it is sufficient to check the bits returned
by the QUERY functions. The MSA level does not have to be checked at every
function call.
crypto/aes/asm/aes-s390x.pl: The AES key schedule must be computed if the
required KM or KMC function codes are not available. Formally, the availability
of a KMC function code does not imply the availability of the corresponding KM
function code.
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4501)
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4501)
Thanks to David Benjamin for spotting this.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4532)
Names were not removed.
Some comments were updated.
Replace Andy's address with openssl.org
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/4516)
Use atomic operations for the counters
Rename malloc_lock to memdbg_lock
Also fix some style errors in mem_dbg.c
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/4359)
If q is non-NULL but p is indeed a safe prime, a modified copy
of p could be leaked.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4525)
Add functions to return DH parameters using NID and to return the
NID if parameters match a named set. Currently this supports only
RFC7919 parameters but could be expanded in future.
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4485)
If BN_FLG_STATIC_DATA is set don't cleanse a->d as it will reside
in read only memory. If BN_FLG_MALLOCED is not set don't modify the
BIGNUM at all.
This change applies to BN_clear_free() and BN_free(). Now the BIGNUM
structure is opaque applications cannot create a BIGNUM structure
without BN_FLG_MALLOCED being set so they are unaffected.
Update internal DH routines so they only copy pointers for read only
parameters.
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4485)
Add an ENGINE to EVP_PKEY structure which can be used for cryptographic
operations: this will typically be used by an HSM key to redirect calls
to a custom EVP_PKEY_METHOD.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4503)
If we are passed an ENGINE to use in int_ctx_new e.g. via EVP_PKEY_CTX_new()
use it instead of the default.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4503)
RSA_setup_blinding() calls BN_BLINDING_create_param() which later calls
BN_mod_exp() as follows:
BN_mod_exp(ret->A, ret->A, ret->e, ret->mod, ctx)
ret->mod will have BN_FLG_CONSTTIME set, but ret->e does not. In
BN_mod_exp() we only test the third param for the existence of this flag.
We should test all the inputs.
Thanks to Samuel Weiser (samuel.weiser@iaik.tugraz.at) for reporting this
issue.
This typically only happens once at key load, so this is unlikely to be
exploitable in any real scenario.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4477)
indicate the level of locking required for various operations.
Remove the lock and atomics from the lhash code. These we're not complete
or adequate.
Refer to #4418 and #4427 for details.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4429)
Rewrite the -req-nodes flag from CA.pl (idea from Andy)
Rewrite ERR_string_error_n
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4478)
When an SSL's context is swtiched from a ticket-enabled context to
a ticket-disabled context in the servername callback, no session-id
is generated, so the session can't be resumed.
If a servername callback changes the SSL_OP_NO_TICKET option, check
to see if it's changed to disable, and whether a session ticket is
expected (i.e. the client indicated ticket support and the SSL had
tickets enabled at the time), and whether we already have a previous
session (i.e. s->hit is set).
In this case, clear the ticket-expected flag, remove any ticket data
and generate a session-id in the session.
If the SSL hit (resumed) and switched to a ticket-disabled context,
assume that the resumption was via session-id, and don't bother to
update the session.
Before this fix, the updated unit-tests in 06-sni-ticket.conf would
fail test #4 (server1 = SNI, server2 = no SNI).
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/1529)
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4379)
insert() or an explicit call to OPENSSL_sk_reserve
Factorise STACK item deletion code
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4379)
This change adds a comment to the exceptional case in point_add that
handles the case of a doubling, which explains when this case may occur
during normal processing.
Thanks go to Antonio Sanso for noting this.
Reviewed-by: Emilia Käsper <emilia@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4424)
Change argument type of xxxelem_is_zero_int to const void*
to avoid the need of type casts.
Fixes#4413
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4450)
Remove all stack headers from some includes that don't use them.
Avoid a genearic untyped stack use.
Update stack POD file to include the OPENSSL_sk_ API functions in the notes
section. They were mentioned in the name section but not defined anywhere.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4430)
This allows the caller to guarantee that there is sufficient space for a
number of insertions without reallocation.
The expansion ratio when reallocating the array is reduced to 1.5 rather than 2.
Change bounds testing to use a single size rather than both INT_MAX and
SIZE_MAX. This simplifies some of the tests.
Switch the stack pointers to data from char * to void *
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4386)
The pub_key field for DH isn't actually used in DH_compute_key at all.
(Note the peer public key is passed in as as BIGNUM.) It's mostly there
so the caller may extract it from DH_generate_key. It doesn't
particularly need to be present if filling in a DH from external
parameters.
The check in DH_set0_key conflicts with adding OpenSSL 1.1.0 to Node.
Their public API is a thin wrapper over the old OpenSSL one:
https://nodejs.org/api/crypto.html#crypto_class_diffiehellman
They have separate setPrivateKey and setPublicKey methods, so the public
key may be set last or not at all. In 1.0.2, either worked fine since
operations on DH objects generally didn't use the public key. (Like
with OpenSSL, Node's setPublicKey method is also largely a no-op, but so
it goes.) In 1.1.0, DH_set0_key prevents create a private-key-only DH
object.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/4384)
This guards against the name constraints check consuming large amounts
of CPU time when certificates in the presented chain contain an
excessive number of names (specifically subject email names or subject
alternative DNS names) and/or name constraints.
Name constraints checking compares the names presented in a certificate
against the name constraints included in a certificate higher up in the
chain using two nested for loops.
Move the name constraints check so that it happens after signature
verification so peers cannot exploit this using a chain with invalid
signatures. Also impose a hard limit on the number of name constraints
check loop iterations to further mitigate the issue.
Thanks to NCC for finding this issue. Fix written by Martin Kreichgauer.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4393)
c2i_ASN1_BIT_STRING takes length as a long but uses it as an int. Check
bounds before doing so. Previously, excessively large inputs to the
function could write a single byte outside the target buffer. (This is
unreachable as asn1_ex_c2i already uses int for the length.)
Thanks to NCC for finding this issue. Fix written by Martin Kreichgauer.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4385)
This reverts commit cc9c568946 for the file
pbe_scrypt.c instead of scrypt.c
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4357)
A new method is added to EVP_PKEY_METH as:
int (*check) (EVP_PKEY_CTX *ctx);
and to EVP_PKEY_ASN1_METHOD as:
int (*pkey_check) (EVP_PKEY_CTX *ctx);
This is used to check the validity of a specific key.
The order of calls is:
EVP_PKEY_check -> pmeth.check -> ameth.pkey_check.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4337)
... script data load.
On related note an attempt was made to merge rotations with logical
operations. I mean as we know, ARM ISA has merged rotate-n-logical
instructions which can be used here. And they were used to improve
keccak1600-armv4 performance. But not here. Even though this approach
resulted in improvement on Cortex-A53 proportional to reduction of
amount of instructions, ~8%, it didn't exactly worked out on
non-Cortex cores. Presumably because they break merged instructions
to separate μ-ops, which results in higher *operations* count. X-Gene
and Denver went ~20% slower and Apple A7 - 40%. The optimization was
therefore dismissed.
Reviewed-by: Rich Salz <rsalz@openssl.org>
This allows completely arbitrary passphrases to be entered, including
NUL bytes.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/3821)
This quiets down complaints about the use of uninitialised memory
[extended tests]
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4340)
OPENSSL_ia32cap.pod discusses possibility to disable operations on
XMM register bank. This formally means that this flag has to be checked
in combination with other flags. But it customarily isn't. But instead
of chasing all the cases we can flip more bits together with FXSR one.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4303)
Move struct timeval includes into e_os.h (where the Windows ones were).
Enaure that the include is guarded canonically.
Refer #4271
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4312)
SNI and ALPN must be set to be consistent with the PSK. Otherwise this is
an error.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/3926)
cryptilib.h is the second.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4188)
Removed e_os.h from all bar three headers (apps/apps.h crypto/bio/bio_lcl.h and
ssl/ssl_locl.h).
Added e_os.h into the files that need it now.
Directly reference internal/nelem.h when required.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4188)
The one creating the DRBG should instantiate it, it's there that we
know which parameters we should use to instantiate it.
This splits the rand init in two parts to avoid a deadlock
because when the global drbg is created it wands to call
rand_add on the global rand method.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
GH: #4268
The DRBG callbacks 'get_entropy()' and 'cleanup_entropy()' are designed
in such a way that the randomness buffer does not have to be allocated
by the calling function. It receives the address of a dynamically
allocated buffer from get_entropy() and returns this address to
cleanup_entropy(), where it is freed. If these two calls are properly
paired, the address can be stored in a stack local variable of the
calling function, so there is no need for having a 'randomness' member
(and a 'filled' member) in 'RAND_DRBG'.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4266)
With the introduction of RAND_poll_ex(), the `RAND_add()` calls were
replaced by meaningless cb(...). This commit changes the 'cb(...)'
calls back to 'rand_add(...)' calls by changing the signature as follows:
-int RAND_poll_ex(RAND_poll_fn cb, void *arg);
+int RAND_poll_ex(RAND_poll_cb rand_add, void *arg);
Changed the function typedef name to 'RAND_poll_cb' to emphasize the fact
that the function type represents a callback function.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4266)
Unlike the NIST DRBG standard, entropy counts are in bits and
buffer lengths are in bytes. This has lead to some confusion and
errors in the past, see my comment on PR 3789.
To clarify the destinction between entropy counts and buffer lengths,
a 'len' suffix has been added to all member names of RAND_DRBG which
represent buffer lengths:
- {min,max}_{entropy,adin,nonce,pers}
+ {min,max}_{entropy,adin,nonce,pers}len
This change makes naming also more consistent, as can be seen in the
diffs, for example:
- else if (adinlen > drbg->max_adin) {
+ else if (adinlen > drbg->max_adinlen) {
Also replaced all 'ent's by 'entropy's, following a suggestion of Paul Dale.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4266)
