9a050ae559
Fix function pointer names
...
The ARM function pointers still had the _ARMV8A suffix, not the _ARMV8 suffix
2017-07-31 22:49:35 -04:00
249a5ed869
Consitently use _ARMV8 as Aarch32/Aarch64 function suffix
2017-07-31 18:06:07 -04:00
a8462328f4
Fix Aarch64 build. Cleanup Windows build
2017-07-31 03:07:32 -04:00
a495018af2
Fix GCM under SSSE3 and CLMUL
2017-07-31 00:40:03 -04:00
fd4c7546b5
Update TestScripts/cryptest.sh. Rename X86_SHA256_HashBlocks → SHA256_HashBlocks_SSE2
2017-07-29 15:16:04 -04:00
d5a6d8f81b
Cut-in SHA for Intel and ARMv8a
2017-07-29 14:18:30 -04:00
3e7496803e
Move CRC32 probe code from cpu.cpp to crc-simd.cpp
2017-07-29 03:49:48 -04:00
de1270656c
Avoid extra ByteReverse when using Intel SHA extensions
...
This gains about 0.6 cpb. SHA-1 is down to 1.7 to 1.9 cpb. SHA-256 is not affected
2017-05-26 01:51:44 -04:00
bd7aa155a6
Revert "Avoid extra ByteReverse"
...
This reverts commit 3b56ba118f
2017-05-25 06:46:40 -04:00
3b56ba118f
Avoid extra ByteReverse
...
This gains about 0.6 cpb. SHA-1 is down to 1.9 cpb. SHA-256 is not affected
2017-05-25 06:20:00 -04:00
1df5fc1e21
Revert "Some fixes related to SunCC compiler bugs."
...
This reverts commit 72d8cbe917
2017-04-27 13:50:47 -05:00
b9abd7141e
Fix endian-reversal and loading of MSG0-MSG3
...
Initially we performed a 32-bit word-size ByteReverse() on the entire 64-byte buffer being hashed. Then we performed another fix-up when loading each 16-byte portion of the buffer into the SSE2 registers for SHA processing. The [undesired] consequence was byte swapping and reversals happened twice. Worse, the call to ByteReverse() produced 16 bswaps instead of 1 call pshufb, so it was orders of magnitude slower than it needed to be.
This check-in takes the sane approach to byte reversals and swapping. It performs it once when the message is loaded for SSE processing. The result is SHA1 calculations drop from about 3.0 cpb to about 2.5 cpb.
2017-04-22 12:19:55 -04:00
72d8cbe917
Some fixes related to SunCC compiler bugs.
...
Fixes for "invalid address alignment" errors. See https://groups.google.com/d/topic/cryptopp-users/OYaByDEbSI0/discussion for details.
Changed config.h in accordance to Jeffrey Walton's request;
2017-04-21 15:26:16 -05:00
d9df0961e0
Fix SHA512_SSE2_Transform after whitespace checkin (Issue 365)
2017-01-17 00:49:10 -05:00
3a1a14c885
Update attribution of ARM SHA extensions
...
Also see d3bb0e13de (commitcomment-20482554)
2017-01-15 20:41:54 -05:00
c71803f383
Fix typo
2017-01-14 01:18:31 -05:00
69f812a7af
Add ARM SHA extensions for SHA-224 and SHA-256
2017-01-14 01:16:47 -05:00
d3bb0e13de
Add ARM SHA extensions for SHA1
...
Benchmarking on ARMv8/Aarch64 dev-board shows SHA-1 speeds up by 2.5x
2017-01-13 05:44:19 -05:00
fc306b6474
Whitespace checkin
2017-01-13 04:23:57 -05:00
f197549662
Remove temporary array for SHA1. Whitespace and comments
2016-12-06 11:09:31 -05:00
40230dd3bf
Removed MAINTAIN_BACKWARDS_COMPATIBILITY_562 (Issue 70)
2016-12-03 00:32:07 -05:00
3fff6251aa
Fix anachronism warning for *_HashBlocks functions under Win64
2016-12-02 00:41:13 -05:00
b896c13bcb
Fix missing __fastcall for *_HashBlocks functions under Win32
2016-12-02 00:25:58 -05:00
c8b910aff5
Backed-off automatically setting CRYPTOPP_BOOL_SSE_SHA_INTRINSICS_AVAILABLE due to bad interaction with '-march=x86-64'. Disgorge SSE2 implementation from CXX implementation
2016-12-01 23:35:13 -05:00
50f0c84ea0
Fix "X86_SHA256_HashBlocks was not declared in this scope" when using CRYPTOPP_DISABLE_ASM
2016-12-01 19:00:24 -05:00
cce56d3f79
Add Intel SHA1 extension support (Issue 139)
2016-12-01 15:05:41 -05:00
7ab9b00f90
Add Intel SHA1 extension support
2016-12-01 00:49:59 -05:00
6200029faa
Fix Clang 3.3 and 3.4 compiles (Issue 264)
2016-09-21 13:57:37 -04:00
399a1546de
Add CRYPTOPP_ASSERT (Issue 277, CVE-2016-7420)
...
trap.h and CRYPTOPP_ASSERT has existed for over a year in Master. We deferred on the cut-over waiting for a minor version bump (5.7). We have to use it now due to CVE-2016-7420
2016-09-16 11:27:15 -04:00
db768200ab
Uncomment existing SHA256::Transform, enable when __OPTIMIZE_SIZE__ is defined
...
This is Wei's original code. It was commented out, but it might be useful to some since IoT gadgets are becoming more popular
2016-09-13 21:06:42 -04:00
ff67abdec5
Add virtual dtor for IteratedHash and ClonableImpl due to non-trivial data members
...
Solaris is showing unusual signs with SunCC 5.13 and 5.14. One user is experiencing a SIGBUS in SHA512::Transform due to data alignment of 'data', which was only 2-byte aligned. The project experienced an exception "Coneable not implemented" during the hashing test after building with Cmake. Its not clear how much Cmake influenced the project's results.
2016-09-13 20:44:14 -04:00
894874fe75
Whitespace checkin
2016-09-10 04:57:48 -04:00
db42a4ac2f
Fixed compile under Clang x86
2016-02-01 01:12:44 -05:00
0e3ee98ff6
Fixed Clang 3.4 compiler error on Linux
2015-12-25 08:00:05 -05:00
885b94707c
Fixed Clang 3.3 integrated assembler crash on Linux
2015-12-25 03:37:14 -05:00
6ac1e46a1f
Cleared issues 11,12,13 (Clang integrated assembler), 58 (RC rollup), 66 (Coverity rollup)
2015-11-18 15:32:28 -05:00
48809d4e85
CRYPTOPP 5.6.3 RC6 checkin
2015-11-05 01:59:46 -05:00
29c4310aab
Cleared remaining issues related to Clang's integarted assembler parsing of the syntax string
2015-07-22 03:59:21 -04:00
6d0dd41eb0
Specialized defines to work with/around Clang integrated assembler. The two defines are WORKAROUND_LLVM_BUG_18916 and CRYPTOPP_USING_CLANG_INTEGRATED_ASSEMBLER. Once the 18916 bug is fixed, the other issues around the integrated assembler will remain
2015-07-22 03:48:10 -04:00
3c56063c3f
Added CRYPTOPP_CLANG_NO_INTEGRATED_AS as a guard in config.h (disabled by default since integrated-as is default behavior and the bug is active). GNU_ATT_SYNTAX → GNU_AS_ATT_SYNTAX and GNUINTEL_SYNTAX → GNU_AS_INTEL_SYNTAX
2015-07-21 18:05:48 -04:00
444d1f1a3b
Cut-in to utilize GNU_ATT_SYNTAX and GNU_INTEL_SYNTAX macros. Also see LLVM Bug 18916 ( https://llvm.org/bugs/show_bug.cgi?id=18916 )
2015-07-21 14:37:55 -04:00
d7399e0fd7
Cleared GCC/Clang warning on unknown pragma
2015-06-09 11:33:32 -04:00
9a759e4492
fix incorrect SHA-256 hash on x64 when compiled with GCC with optimizations enabled
2010-06-29 00:46:23 +00:00
9d1843e8fc
fix for http://sourceforge.net/apps/trac/cryptopp/ticket/2 (Ken Raeburn)
2010-05-13 16:58:21 +00:00
5221b19822
fix incorrect SHA-256 computation on non-SSE2 x86 machines on non-aligned input (reported by
...
pycryptopp project)
2009-07-05 18:10:10 +00:00
13c3aa2f4e
fix compile on MSVC 6 and MSVC 2003
2009-03-15 03:53:58 +00:00
3202bf2809
fix compile on OpenSolaris 8.11
2009-03-13 11:15:21 +00:00
d6b4e54448
fix compile on MSVC 6
2009-03-13 02:55:23 +00:00
2779fc6050
- add EAX mode, XSalsa20
...
- speed up GCM key setup
- wipe stack in AES assembly code
- speed up CFB mode
2009-03-12 11:24:12 +00:00
8ca6148ad5
add x86/x64 assembly for SHA-256,
...
add DEFAULT_CHANNEL and AAD_CHANNEL,
fix macChannel for AuthenticatedEncryptionFilter
2009-03-10 02:56:19 +00:00
Jeffrey Walton
Anton Gorev
weidai