Jeffrey Walton
|
0be4110fe5
|
Cleared "signed/unsigned" warning under GCC 4.8 (this may have been a GCC 4.8/Debian-i686 issue because both types were unsigned)
|
2015-07-26 01:09:51 -04:00 |
Jeffrey Walton
|
f36dc3d643
|
Added rules to supress CRYPTOPP_NO_UNALIGNED_DATA_ACCESS warning
|
2015-07-26 00:39:29 -04:00 |
Jeffrey Walton
|
c23c02171d
|
Cleared "signed/unsigned" warning under GCC 4.8
|
2015-07-26 00:18:25 -04:00 |
Jeffrey Walton
|
fe9948e39a
|
Moved IS_MINGW block back under x86
|
2015-07-25 09:27:59 -04:00 |
Jeffrey Walton
|
6ac2555052
|
Removed -Oz from symbolize. Moved MinGW libraries down before recipes
|
2015-07-25 09:17:26 -04:00 |
Jeffrey Walton
|
d215898465
|
Cleared "missing initializer for member '_LARGE_INTEGER::<anonymous struct>::HighPart'" under MinGW
|
2015-07-25 07:23:22 -04:00 |
Jeffrey Walton
|
2c988356a9
|
Revert "Cleared "missing initializer for member '_LARGE_INTEGER::<anonymous struct>::HighPart'" under MinGW".
The commit accidentally committed a number of unneeded files while on Windows during testing.
This reverts commit 576749b4bb.
|
2015-07-25 07:19:41 -04:00 |
Jeffrey Walton
|
576749b4bb
|
Cleared "missing initializer for member '_LARGE_INTEGER::<anonymous struct>::HighPart'" under MinGW
|
2015-07-25 07:16:45 -04:00 |
Jeffrey Walton
|
ff8974f1fd
|
Fixed variable expansion issue on Cygwin (other shells on on OSes were fine, like Bash and C shell)
|
2015-07-25 06:56:38 -04:00 |
Jeffrey Walton
|
e75cb8dd8b
|
Backed out commit 1dc875a44a. MinGW complained "error: 'p0' was not declared in this scope". Thanks to Ilya Bizyaev for reporting
|
2015-07-25 06:15:53 -04:00 |
Jeffrey Walton
|
427b45b4e7
|
Added some sane defaults for -g and -O if they are missing from CXXFLAGS
|
2015-07-25 05:03:13 -04:00 |
Jeffrey Walton
|
6bdcaefccd
|
Added test for CRYPTOPP_NO_UNALIGNED_DATA_ACCESS, added warning if its not defined
|
2015-07-25 04:23:27 -04:00 |
Jeffrey Walton
|
2e37c25acb
|
Removed VC 6.0 dummy function parameter that was also used as a template parametr. Cleared 8 "unused paramter" warnings
|
2015-07-25 02:21:03 -04:00 |
Jeffrey Walton
|
d3b2c03e4e
|
Added missing specialization for i386. Added comment on why we were specializing
|
2015-07-25 02:06:52 -04:00 |
Jeffrey Walton
|
1dc875a44a
|
Cleared "unused variable" warning under GCC 5.1
|
2015-07-25 02:05:38 -04:00 |
Jeffrey Walton
|
ac74f8e4df
|
Added default flag detection for ARFLAGS. Added MACHINE,SYSTEM,RELEASE variables for overrides. Added *.dylib and *.dll to clean rule. Added *.s and *.ii to clean rule
|
2015-07-25 02:04:53 -04:00 |
Jeffrey Walton
|
53eb0f11aa
|
Added check to guard against self-assignment
|
2015-07-25 02:02:41 -04:00 |
Jeffrey Walton
|
ed7566fb8d
|
Fixed mis-detected Fedora 22/i686
|
2015-07-25 01:31:37 -04:00 |
Jeffrey Walton
|
65ee664ae5
|
Cleared "unused variable" warning under GCC 5.1
|
2015-07-25 01:15:54 -04:00 |
Jeffrey Walton
|
3f328922ac
|
Cleared "Base class ‘class ByteQueue’ should be explicitly initialized in the copy constructor" warning under GCC 5.1 with -Wextra
|
2015-07-25 01:13:04 -04:00 |
Jeffrey Walton
|
ed0300b143
|
Cleared "signed/unsigned mismatch" warning under GCC 5.1. The pointer math meant we got a ptrdiff_t back from the expression. Its not clear if we should throw, so an assert was added
|
2015-07-25 00:38:34 -04:00 |
Jeffrey Walton
|
81af925716
|
Cleared "Base class ‘class ByteQueue’ should be explicitly initialized in the copy constructor" warning under GCC 5.1 with -Wextra
|
2015-07-25 00:03:43 -04:00 |
Jeffrey Walton
|
d3e441267b
|
Cleared "unused parameter" warning under GCC 5.1
|
2015-07-24 23:30:32 -04:00 |
Jeffrey Walton
|
d6f5ed03bf
|
Cleaned up and re-organized the makefile. There are now sections, with some comments calling out what various sections do and the work arounds that are in effect
|
2015-07-24 23:08:18 -04:00 |
Jeffrey Walton
|
f4b44dfea8
|
Added trap.h to Visual Studio projects
|
2015-07-24 22:46:38 -04:00 |
Jeffrey Walton
|
1160f4163a
|
Cleared "unused parameter" warning with GCC 5.1 and -Wextra
|
2015-07-23 20:21:17 -04:00 |
Jeffrey Walton
|
6541c7ac72
|
Cleared compile error due to missing include
|
2015-07-23 20:15:51 -04:00 |
Jeffrey Walton
|
20e652abbb
|
Cleared "unused parameter" warning with GCC 5.1 and -Wextra
|
2015-07-23 20:14:15 -04:00 |
Jeffrey Walton
|
db40142116
|
Added -Wno-type-limits for GCC 4.3 and above. Its the GCC equivalent to Clang's -Wno-tautological-compare
|
2015-07-23 20:04:30 -04:00 |
Jeffrey Walton
|
35b213c085
|
Cleared "unused parameter" warning with GCC 5.1 and -Wextra
|
2015-07-23 19:30:14 -04:00 |
Jeffrey Walton
|
e87857f67d
|
Cleared "missing initialization in copy constructor" warning with GCC 5.1 and -Wextra
|
2015-07-23 19:17:07 -04:00 |
Jeffrey Walton
|
5bb0e6fd38
|
Cleared "unused variable" warning under GCC 5.1 with -Wall
|
2015-07-23 18:44:51 -04:00 |
Jeffrey Walton
|
504ba0ea87
|
Cleared "comparison between signed and unsigned integer" warning in SafeConvert by providing template specializations
|
2015-07-23 18:15:29 -04:00 |
Jeffrey Walton
|
3fd7a239f0
|
Backed out Clang integrated assembler changes. LLVM Bug 24232 pushed us over the edge. There is some skeleton code remaining for testing. To activate it, perform "make FORCE_ASM=1"
|
2015-07-23 16:47:03 -04:00 |
Jeffrey Walton
|
07e8319478
|
Cleared "unused function" warning when using GCC with -Wall
|
2015-07-23 14:21:06 -04:00 |
Jeffrey Walton
|
b2d9be1b80
|
Cleared "unused variable" warning when using CRYPTOPP_STATIC_ASSERT under GCC. The "__attribute__ ((unused)" was present in GCC 2.95, so it will be available for the GCC's we support. Non-GCC's will get the non-decorated assert
|
2015-07-23 12:30:13 -04:00 |
Jeffrey Walton
|
929967a042
|
Added CRYTPOPP_ASSERT to cryptest program. The mods also show how to install a debug trap handler such that if CRYPTOPP_ASSERT fires, the program continues
|
2015-07-23 10:48:43 -04:00 |
Jeffrey Walton
|
809a3cbb74
|
Cleared unknown pragma warning under MinGW
|
2015-07-23 09:09:50 -04:00 |
Jeffrey Walton
|
bdb7ea3bf0
|
Cleared unknown pragma warning under MinGW
|
2015-07-23 08:52:08 -04:00 |
Jeffrey Walton
|
5af5c553e3
|
Added default for strip program. Added additional fodder to distclean recipe
|
2015-07-23 08:39:31 -04:00 |
Jeffrey Walton
|
123ca9b76d
|
Added standard makefile targets, including distclean, dist, install-strip. Removed no-lib recipe because there's a proper dependency rule
|
2015-07-23 07:50:55 -04:00 |
Jeffrey Walton
|
7b629dcfa6
|
Cleared UB in cast.cpp due to use of rotVariable. On PowerPC, the specialization that uses inline assembler will activate. The inline assembly uses __rlwnm which takes a mask and does not suffer C/C++ UB
|
2015-07-23 05:10:12 -04:00 |
Jeffrey Walton
|
ac0f94c23a
|
Removed branch from rotlVariable. The mask ensures it stays well defined
|
2015-07-23 04:06:10 -04:00 |
Jeffrey Walton
|
1303f4bdcc
|
Cutover to CRYPTOPP_ASSERT in misc.h. Posix's behavior of "let's crash the program by raising SIGABRT" in Debug builds has got to be one of the most questionable decisions in the history of computing. The new behavior: if there's a SIGTRAP handler somewhere, then the handler will be invoked. Some platforms, like Apple and Microsoft, provide a default SIGTRAP handler. If there's no SIGTRAP handler, then the program will crash just like it raised a SIGABRT. In this case, there's no effective change in behavior
|
2015-07-23 03:39:13 -04:00 |
Jeffrey Walton
|
5daa168a93
|
Cleaned up CRYPTOPP_CLANG_INTEGRATED_ASSEMBLER and WORKAROUND_LLVM_BUG_18916 defines
|
2015-07-23 03:29:21 -04:00 |
Jeffrey Walton
|
cd33142f6c
|
Improved flow of setting variables in makefile
|
2015-07-23 03:17:18 -04:00 |
Jeffrey Walton
|
b3924a7b86
|
Additional changes to support Clang's integrated assembler. They have not been activated though. Uri and other testers will be able to uncomment the TODO line and everything will work as expected
|
2015-07-23 02:18:57 -04:00 |
Jeffrey Walton
|
e6300f52e2
|
Removed template specializations for Clang using preprocessor for rotFixed due to LLVM Bug 24226. Removed asserts from __rlwinm because the mask ensures the operation is well defined (see the comments in ppc_intrinsics.h)
|
2015-07-23 01:53:45 -04:00 |
Jeffrey Walton
|
fcd5f6df72
|
Cleared "invalid operand size" under Clang's integrated assembler. After more testing, this might be applied to other assemblers because Clang is probably correct in this instance
|
2015-07-22 06:14:37 -04:00 |
Jeffrey Walton
|
29c4310aab
|
Cleared remaining issues related to Clang's integarted assembler parsing of the syntax string
|
2015-07-22 03:59:21 -04:00 |