This is taken from upstream, although it originally came from us. This
will only take effect on 64-bit systems (x86-64 and aarch64).
Before:
Did 1496 ECDH P-256 operations in 1038743us (1440.2 ops/sec)
Did 2783 ECDSA P-256 signing operations in 1081006us (2574.5 ops/sec)
Did 2400 ECDSA P-256 verify operations in 1059508us (2265.2 ops/sec)
After:
Did 4147 ECDH P-256 operations in 1061723us (3905.9 ops/sec)
Did 9372 ECDSA P-256 signing operations in 1040589us (9006.4 ops/sec)
Did 4114 ECDSA P-256 verify operations in 1063478us (3868.4 ops/sec)
Change-Id: I11fabb03239cc3a7c4a97325ed4e4c97421f91a9
This change adds a benchmark for the RNG and also allows specific speed
tests to be selected via a command-line argument, since the full speed
suite is getting quite lengthy now and is only going to get longer in
the future.
Change-Id: If62c69177d58d3eb945d6108524c144ea0044137
Reviewed-on: https://boringssl-review.googlesource.com/4326
Reviewed-by: Adam Langley <agl@google.com>
Beyond generally eliminating unnecessary includes, eliminate as many
includes of headers that declare/define particularly error-prone
functionality like strlen, malloc, and free. crypto/err/internal.h was
added to remove the dependency on openssl/thread.h from the public
openssl/err.h header. The include of <stdlib.h> in openssl/mem.h was
retained since it defines OPENSSL_malloc and friends as macros around
the stdlib.h functions. The public x509.h, x509v3.h, and ssl.h headers
were not changed in order to minimize breakage of source compatibility
with external code.
Change-Id: I0d264b73ad0a720587774430b2ab8f8275960329
Reviewed-on: https://boringssl-review.googlesource.com/4220
Reviewed-by: Adam Langley <agl@google.com>
A previous change in BoringSSL renamed ERR_print_errors_fp to
BIO_print_errors_fp as part of refactoring the code to improve the
layering of modules within BoringSSL. Rename it back for better
compatibility with code that was using the function under the original
name. Move its definition back to crypto/err using an implementation
that avoids depending on crypto/bio.
Change-Id: Iee7703bb1eb4a3d640aff6485712bea71d7c1052
Reviewed-on: https://boringssl-review.googlesource.com/4310
Reviewed-by: Adam Langley <agl@google.com>
MinGW on Linux needs lowercase include files. On Windows this doesn't
matter since the filesystems are case-insensitive, but building
BoringSSL on Linux with MinGW has case-sensitive filesystems.
Change-Id: Id9c120d819071b041341fbb978352812d6d073bc
Reviewed-on: https://boringssl-review.googlesource.com/4090
Reviewed-by: Adam Langley <agl@google.com>
The TLS AEADs require EVP_AEAD_CTX_init_with_direction now. The speed command
was never updated to account for this.
BUG=467762
Change-Id: I8c7d4566b51d54d91d17318b8f4813ad8f19c777
Reviewed-on: https://boringssl-review.googlesource.com/3971
Reviewed-by: Adam Langley <agl@google.com>
Windows x64 uses the IL32P64 data model, which means that unsigned int
is 32 bits and size_t is 64 bits. Previously, the expression
|~(alignment - 1)| resulted in the 32-bit value 0xFFFFFFF0, which was
then extended to the 64-bit value 0x00000000FFFFFFF0 when promoted to
size_t. When the input pointer was masked with this value, the result
was a pointer that was usually way outside the boundaries of the array.
The new code casts |alignment| to size_t first prior to the bitwise
negation, resulting in the correct mask value of 0xFFFFFFFFFFFFFFF0.
Change-Id: I04754aa9e1ce7a615c2b4c74051cfcca38dbb52f
Reviewed-on: https://boringssl-review.googlesource.com/3961
Reviewed-by: Adam Langley <agl@google.com>
Including string.h in base.h causes any file that includes a BoringSSL
header to include string.h. Generally this wouldn't be a problem,
although string.h might slow down the compile if it wasn't otherwise
needed. However, it also causes problems for ipsec-tools in Android
because OpenSSL didn't have this behaviour.
This change removes string.h from base.h and, instead, adds it to each
.c file that requires it.
Change-Id: I5968e50b0e230fd3adf9b72dd2836e6f52d6fb37
Reviewed-on: https://boringssl-review.googlesource.com/3200
Reviewed-by: David Benjamin <davidben@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
This avoids having Windows be different and is also easier for testing
because it's a simple matter to unalign the pointer if needed.
Change-Id: I32cfa5834e3fe4f16304a25092b9c71946d4744d
Reviewed-on: https://boringssl-review.googlesource.com/3131
Reviewed-by: Adam Langley <agl@google.com>
This avoids a conflict with the Chromium build system, which
defines WIN32_LEAN_AND_MEAN with a different value.
BUG=crbug.com/453196
Change-Id: Ia15ec7c20325c1961af4f32e5208266e5f846f35
Reviewed-on: https://boringssl-review.googlesource.com/3150
Reviewed-by: David Benjamin <davidben@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
Define WIN32_LEAN_AND_MEAN before including Windows Platform SDK
headers to preempt naming conflicts and to make the build faster. Avoid
including those headers in BoringSSL headers. Document that Platform
SDK 8.1 or later is required on Windows.
Change-Id: I907ada21dc722527ea37e839c71c5157455a7003
Reviewed-on: https://boringssl-review.googlesource.com/3100
Reviewed-by: Adam Langley <agl@google.com>
This change adds the AES-128-CBC-SHA and AES-256-CBC-SHA AEADs to the
speed test. These AEADs need an 11 byte additional data so the test is
extended to be able to provide that.
Change-Id: I9a57c2321a979a68ab0df9faf1bb26b44a3009c4
Reviewed-on: https://boringssl-review.googlesource.com/2922
Reviewed-by: Adam Langley <agl@google.com>
This eliminates a source of variability from the benchmarks.
Change-Id: I8ce07bd68e7591f8c5545040b02b96d21609a0e5
Reviewed-on: https://boringssl-review.googlesource.com/2920
Reviewed-by: David Benjamin <davidben@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
This change adds the stitched RC4-MD5 code from upstream OpenSSL but
exposes it as an AEAD. It's not a normal AEAD (it's stateful thus
doesn't take an nonce) but forcing pre-AEAD cipher suites in the AEAD
interface is less painful than forcing AEADs into the EVP_CIPHER
interface. Over time, more and more cipher suites will be exposed as
TLS-specific AEADs and then ssl/ can drop support for EVP_CIPHER.
See original code from upstream:
https://github.com/openssl/openssl/blob/master/crypto/evp/e_rc4_hmac_md5.c
Change-Id: Ia9267b224747f02be6b934ea0b2b50e1f529fab9
Reviewed-on: https://boringssl-review.googlesource.com/1043
Reviewed-by: Adam Langley <agl@google.com>
Apart from the obvious little issues, this also works around a
(seeming) libtool/linker:
a.c defines a symbol:
int kFoo;
b.c uses it:
extern int kFoo;
int f() {
return kFoo;
}
compile them:
$ gcc -c a.c
$ gcc -c b.c
and create a dummy main in order to run it, main.c:
int f();
int main() {
return f();
}
this works as expected:
$ gcc main.c a.o b.o
but, if we make an archive:
$ ar q lib.a a.o b.o
and use that:
$ gcc main.c lib.a
Undefined symbols for architecture x86_64
"_kFoo", referenced from:
_f in lib.a(b.o)
(It doesn't matter what order the .o files are put into the .a)
Linux and Windows don't seem to have this problem.
nm on a.o shows that the symbol is of type "C", which is a "common symbol"[1].
Basically the linker will merge multiple common symbol definitions together.
If ones makes a.c read:
int kFoo = 0;
Then one gets a type "D" symbol - a "data section symbol" and everything works
just fine.
This might actually be a libtool bug instead of an ld bug: Looking at `xxd
lib.a | less`, the __.SYMDEF SORTED index at the beginning of the archive
doesn't contain an entry for kFoo unless initialised.
Change-Id: I4cdad9ba46e9919221c3cbd79637508959359427
