Age | Commit message (Collapse) | Author |
|
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18910)
(cherry picked from commit a9389c0b75e69ebaf74fdc8fee0c983809e45931)
|
|
The patch will process one extra block by scalar in addition to
blocks by SVE/SVE2 in parallel. This is esp. helpful in the
scenario where we only have 128-bit vector length.
The actual uplift to performance is complicated, depending on the
vector length and input data size. SVE/SVE2 implementation don't
always perform better than Neon, but it should prevail in most
cases
On a CPU with 256-bit SVE/SVE2, interleaved processing can
handle 9 blocks in parallel (8 blocks by SVE and 1 by Scalar).
on 128-bit SVE/SVE2 it is 5 blocks. Input size that is a multiple
of 9/5 blocks on respective CPU can be typically handled at
maximum speed.
Here are test data for 256-bit and 128-bit SVE/SVE2 by running
"openssl speed -evp chacha20 -bytes 576" (and other size)
----------------------------------+---------------------------------
256-bit SVE | 128-bit SVE2
----------------------------------|---------------------------------
Input 576 bytes 512 bytes | 320 bytes 256 bytes
----------------------------------|---------------------------------
SVE 1716361.91k 1556699.18k | 1615789.06k 1302864.40k
----------------------------------|---------------------------------
Neon 1262643.44k 1509044.05k | 680075.67k 1060532.31k
----------------------------------+---------------------------------
If the input size gets very large, the advantage of SVE/SVE2 over
Neon will fade out.
Signed-off-by: Daniel Hu <Daniel.Hu@arm.com>
Change-Id: Ieedfcb767b9c08280d7c8c9a8648919c69728fab
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18901)
(cherry picked from commit 3f42f41ad19c631287386fd8d58f9e02466c5e3f)
|
|
The default cross compiler (gcc 9.4.0) for ppc64le on Ubunut 20.04 seems
buggy and causes a seg fault in sslapitest. This doesn't impact any other
CI cross compile platforms and does not seem to impact the gcc 10.3.0 cross
compiler.
We just drop the optimisation level on that platform.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19056)
(cherry picked from commit 200d9521a0d406a7d02778d1c6c5a5230caeecf5)
|
|
Fixes #18820
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19002)
(cherry picked from commit 63df86b041aaafba3e4998b2e3872fa8695a2377)
|
|
In the reference C implementation in bn_asm.c, tp[num + 1] contains the
carry bit for accumulations into tp[num]. tp[num + 1] is only ever
assigned, never itself incremented.
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18883)
(cherry picked from commit 2f1112b22a826dc8854b41b60a422c987f8ddafb)
|
|
Multiplication""
This reverts commit 712d9cc90e355b2c98a959d4e9398610d2269c9e.
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18883)
(cherry picked from commit eae70100fadbc94f18ba7a729bf065cb524a9fc9)
|
|
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/18821)
(cherry picked from commit 45479dcee1672661e4f5b6d8b6c9a50453581e65)
|
|
test/build.info
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/18821)
(cherry picked from commit c02036e1ad759fca228a2201f1c4752670ad59bd)
|
|
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/18821)
(cherry picked from commit 6212fc6814e8a8968bb35239cd454afd22b6a083)
|
|
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18290)
(cherry picked from commit 6136408e6abf10672e399bf95be064868f2f7ca6)
|
|
s390x GHASH assembler implementation assumed it was called from a
gcm128_context structure where the Xi paramter to the ghash function was
embedded in that structure. Since the structure layout resembles the paramter
block required for kimd-GHASH, the assembler code simply assumed the 128 bytes
after Xi are the hash subkey.
This assumption was broken with the introduction of AES-GCM-SIV which uses the
GHASH implementation without a gcm128_context structure. Furthermore, the
bytes following the Xi input parameter to the GHASH function do not contain
the hash subkey. To fix this, we remove the assumption about the calling
context and build the parameter block on the stack. This requires some
copying of data to and from the stack. While this introduces a performance
degradation, new systems anyway use kma for GHASH/AES-GCM.
Finally fixes #18693 for s390x.
Signed-off-by: Juergen Christ <jchrist@linux.ibm.com>
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18939)
(cherry picked from commit cd854f225bbef9561fad680e2628dfd55be1b141)
|
|
The rest of the documentation talks about sig, not sigret
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18943)
(cherry picked from commit 2bd8190aace8109a06aff495a3e20c863ef48653)
|
|
Fixes #19673
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19674)
(cherry picked from commit be0161ff100bf10c9549fc09ce4513681011da1c)
|
|
Fixes #18896
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18903)
(cherry picked from commit 186be8ed26f5561faf91d6da3ed14cd9cb6617dd)
|
|
Use hardware acceleration for kmac on s390x. Since klmd does not support
kmac, perform padding of the last block by hand and use kimd. Yields a
performance improvement of between 2x and 3x.
Signed-off-by: Juergen Christ <jchrist@linux.ibm.com>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18863)
(cherry picked from commit affc070aabc930aeaba50f0dd6b3e0b7a2ddc399)
|
|
Rework of GCM code did not include s390x causing NULL pointer dereferences on
GCM operations other than AES-GCM on platforms that support kma. Fix this by
a proper setup of the function pointers.
Fixes: 92c9086e5c2b ("Use separate function to get GCM functions")
Signed-off-by: Juergen Christ <jchrist@linux.ibm.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18862)
(cherry picked from commit 48e35b99bd0071207cfe39da22eb2502db5c09dc)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18835)
(cherry picked from commit d50e0934e5b1537db0ea43986464b8f8f8b4e9fd)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18835)
(cherry picked from commit 95201ef45711220455e8abf1cc6b334393384af2)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18835)
(cherry picked from commit 92c9086e5c2b63606cd28a7f13f09b9ff35a0de3)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18835)
(cherry picked from commit 7da952bcc54604141ea8ed40ec5ed1fd2f74cc25)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18835)
(cherry picked from commit 7b6e19fc4e6cc1a7000f71789ef50636dacdbb85)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18835)
(cherry picked from commit a8b5128fd724bc23f7454d64e401d15129634a01)
|
|
Also add early clobber for two-insn bswap.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18827)
(cherry picked from commit 48b6776678d794406c625dcb5767102b73081962)
|
|
And so clean a few useless includes
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19721)
(cherry picked from commit f2a6f83862be3e20260b708288a8f7d0928e9018)
|
|
Reduce the Miller Rabin counts to the values specified by FIPS 186-5.
The old code was using a fixed value of 64.
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19579)
(cherry picked from commit d2f6e66d2837bff1f5f7636bb2118e3a45c9df61)
|
|
Fixup for e8f1d76b50204d87a0ef7f6879eb1dd507a54368.
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/19675)
|
|
Fixes #17995
Fixes #18578
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/19386)
(cherry picked from commit a88e97fcace01ecf557b207f04328a72df5110df)
|
|
Implement the AES-256-CTR cipher in the dasync engine.
Use that to reproduce the reported problems with the
devcrypto engine in our normal test environment.
See #17995 and #17532 for details.
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19386)
(cherry picked from commit bd363ef32403d58a8b41553b5abd602b30073b10)
|
|
Fixes openssl#14114
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19492)
(cherry picked from commit 858b5d12b85b0639519d21206c9da7e1bb976a00)
|
|
This test had commands that assumes that runner_loop() is used to perform
the tests. These tests still run fine because Unix accepts braces in file
names, but other operating systems might not.
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19731)
|
|
CID 1516955 : Null pointer deref (REVERSE_INULL)
CID 1516954 : Null pointer deref (REVERSE_INULL)
CID 1516953 : RESOURCE_LEAK of child
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19700)
(cherry picked from commit 26cfa4cd85f6b26dd7a48c2ff06bfa4a2cea4764)
|
|
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19690)
(cherry picked from commit 4741c80c0556653c74252ec91425dcb74066b2ec)
|
|
FIPS 186-4 has 5 different algorithms for key generation,
and all of them rely on testing GCD(a,n) == 1 many times.
Cachegrind was showing that during a RSA keygen operation,
the function BN_gcd() was taking a considerable percentage
of the total cycles.
The default provider uses multiprime keygen, which seemed to
be much faster. This is because it uses BN_mod_inverse()
instead.
For a 4096 bit key, the entropy of a key that was taking a
long time to generate was recorded and fed back into subsequent
runs. Roughly 40% of the cycle time was BN_gcd() with most of the
remainder in the prime testing. Changing to use the inverse
resulted in the cycle count being 96% in the prime testing.
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19578)
(cherry picked from commit dd1d7bcb69994d81662e709b0ad838880b943870)
|
|
dtlstest.c needs some adjusting to handle no-dtls1_2 since commit
7bf2e4d7f0c banned DTLSv1 at the default security level - causing the
test to fail.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18848)
(cherry picked from commit a6843e6ae8ae0551aae8555783f06dab7951f112)
|
|
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18480)
(cherry picked from commit 2885b2ca4eee5586baa50208e41a1ca54532eb3a)
|
|
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18480)
(cherry picked from commit 2b11a8ecc8ed1355b99a6d88b8e7e7a75a67bd0a)
|
|
Longer private key sizes unnecessarily raise the cycles needed to
compute the shared secret without any increase of the real security.
We use minimum key sizes as defined in RFC7919.
For arbitrary parameters we cannot know whether they are safe
primes (we could test but that would be too inefficient) we have
to keep generating large keys.
However we now set a small dh->length when we are generating safe prime
parameters because we know it is safe to use small keys with them.
That means users need to regenerate the parameters if they
want to take the performance advantage of small private key.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18480)
(cherry picked from commit ddb13b283be84d771deba1e964610b1670641f03)
|
|
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18289)
(cherry picked from commit ca6286c382a7eb527fac9aba2a018354acb27b16)
|
|
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18256)
(cherry picked from commit 081f3484593cdd3be2b7fdd8818c3f928ce729bc)
|
|
Test that if we receive a packet from the next epoch, we can buffer it
and still use it.
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18601)
(cherry picked from commit e1c153d31d4f913ebe2202a4bc20305919274d1f)
|
|
The way it was implemented didn't play well with perl's join(), so it's
reimplemented a bit differently.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19347)
(cherry picked from commit 1ec0acf264652bd981e95842723e5414d634cd93)
|
|
VMS linking complains a lot about multiply defined symbols unless told
otherwise, especially when shareable images are involved. For example, this
involves the legacy provider, where there are overriding implementations of
certain ERR functions.
To quiet the linker down, we need to say that symbols should be searched
selectively in shareable images.
However, that's not quite enough. The order in which the VMS linker
processes files isn't necessarily top to bottom as given on the command line
or the option file(s), which may result in some symbols appearing undefined,
even though they are. To remedy that, it's necessary to explicitly include
all object files and object libraries into a cluster, thus ensuring that
they will be processed first. This allows the search for remaining symbol
references to be done in the as desired in the shareable images that follow.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19327)
(cherry picked from commit c62a9cd720eccdbb388890ee4a36801d01315be4)
|
|
And move ROTATE inline asm to header.
Now this benefits SM3, SHA (when with Zbb only and no Zknh)
and other hash functions
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18287)
(cherry picked from commit eea820f3e239a4c11d618741fd5d00a6bc877347)
|
|
This works for both RV32 and RV64
Signed-off-by: Hongren (Zenithal) Zheng <i@zenithal.me>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18287)
(cherry picked from commit 7ae2bc9df6e0916a8f16183f07dfa1815dd4b66d)
|
|
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17937)
(cherry picked from commit 5317b6ee1fc3db20de5976fbb46cc49a45c0768a)
|
|
This patch improves existing chacha20 SVE patch by using SVE2,
which is an optional architecture feature of aarch64, with XAR
instruction that can improve the performance of chacha20.
Signed-off-by: Daniel Hu <Daniel.Hu@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18522)
(cherry picked from commit bcb52bcc9f9c36a85d037976676fd5ca52f307cd)
|
|
Test makes sure, that both time stamping certificate according to rfc3161 (no
requirements for keyUsage extension) and according to CAB forum (keyUsage
extension must be digitalSignature and be set critical) are accepted. Misuse
cases as stated in CAB forum are rejected, only exeption is a missing
"critial" flag on keyUsage.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18597)
(cherry picked from commit 386ab7f1fefdd77521e670d9593e9894e2774be0)
|
|
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18571)
(cherry picked from commit 909d590fe7a0935e7856ec618afd652ae03a9260)
|
|
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18518)
(cherry picked from commit a644cb7c1c19c78e2ca393c8ca36989e7ca61715)
|
|
Signed-off-by: Hongren (Zenithal) Zheng <i@zenithal.me>
Tested-by: Jiatai He <jiatai2021@iscas.ac.cn>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18197)
(cherry picked from commit ee11118deb65d2b22b94721125a5649d05591e7b)
|