aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Remove BOTAN_DLL from internal headers, add it to Unexpected_Messagelloyd2011-06-024-14/+14
| | | | exception.
* Change order of arguments to bigint_monty_redc (internal interface,lloyd2011-06-023-46/+43
| | | | | | not exposed to callers) Switch back redc to using the inlined version (accidental change)
* Add monty sqr and multiply routines (they just call karatsuba and thenlloyd2011-06-024-43/+77
| | | | redc, currently)
* Reduce number of temporaries needed in point additionlloyd2011-06-011-22/+14
|
* Multi-expoentiation, using the single bit variation. Using a 2 or 3lloyd2011-06-014-2/+49
| | | | | | bit window may well improve things further. Currently seeing 20-25% improvement in ECDSA signature verification and 25 to 40% in GOST-34.10 verifications.
* Remove amalg on distcleanlloyd2011-06-012-0/+2
|
* PointGFp::get_affine_{x,y} doesn't return a referencelloyd2011-05-311-2/+2
|
* MinGW install commands are the default, so don't specify explicitly.lloyd2011-05-262-5/+2
| | | | | On Solaris, use /usr/ucb/install, as /usr/sbin/install uses some kind of other strange syntax.
* Explicitly specify the types for make_pair when passing a const arraylloyd2011-05-241-2/+10
| | | | | | as the first value. Oddly, Sun Studio's default STL tries to make a pair with the first value a 'const byte[2]', and then subsequently rejects that as invalid. Reported by Jeremy Reed.
* Split the SIMD implementations into their own modules and choose onelloyd2011-05-2412-16/+41
| | | | at config time.
* FD_ZERO on Solaris uses memset, and assumes we included string.hlloyd2011-05-242-0/+2
| | | | already. Reported by Jeremy C. Reed <[email protected]>
* Only call p.sig_words once in CurveGFp constructor.lloyd2011-05-191-4/+2
| | | | Fix build log; copy and paste error.
* Move the curve over GF(p) code back into its own subdirlloyd2011-05-197-7/+22
|
* Reject s == 0 or r == 0 in a ECC signature.lloyd2011-05-192-2/+5
| | | | | | | | | | | | | | | In ECDSA, this cases should all be caught by the later check that R is not zero, so I don't believe there is any security danger. However the GOST 34.10 implementation did not have either check. Fortunately, the function that extracts the affine X coordinate from the Jacobian coordinates will throw an exception if the point is at infinity, so we would not in fact accept invalid signatures, but this is mostly by luck. And still represents a bit of a DoS potential. I checked the history, it looks like not checking for zeros at the start traces back to the original InSiTo code, and I copied the ECDSA code for GOST without thinking about it too much.
* Small cleanup in the Barrett reducer, now doesn't create a temp iflloyd2011-05-182-33/+35
| | | | abs(x) < modulus. Also remove unused member variables.
* Fix the very basic operations for encoding and decoding alloyd2011-05-172-12/+10
| | | | | | | | | certificate policies extension, though it's really not supported at all. Remove test code from secmem.h Fix building the examples
* Modify ECC points to do all math in Montgomery form, rather thanlloyd2011-05-174-156/+144
| | | | | | | | | | | | | converting back and forth. This gives a 10 to 20% speedup on a Core i7. In addition, the CurveGFp no longer contains a Barrett reducer, saving 3 BigInts worth of memory. Add a #if'ed out alternative to point multiplication using the Montgomery ladder technique. It runs in (more or less) constant time, but rather significantly slower than the 4 bit window technique currently used. Tweak the window sizes to match the theoretical optimums.
* Add RSA constructor that does not require a password. Patch submittedlloyd2011-05-161-0/+18
| | | | to the list by William Emmanuel S. Yu <[email protected]>
* Fixes for base64 incremental decoding.lloyd2011-05-164-122/+54
| | | | | | Define the Base64_Decoder filter in terms of base64_decode Don't use locked memory in the hex or base64 filters.
* Always writing to/reading from the carry word makes valgrind unhappy.lloyd2011-05-161-1/+2
|
* The new method of doing comparisons did not work all of the time: iflloyd2011-05-131-1/+3
| | | | | | | | | | | the low bytes were equal, then the saturating subtraction result in that byte would be 0 with the high byte containing a non-zero value. To deal with this, shift and or together the two values into the low byte. Add some new tests which check out the SIMD implementation more carefully, including values that trigger the problem in the earlier version.
* Add specialization for i486 that uses -mtune=generic, as Debianlloyd2011-05-132-13/+12
| | | | | | | | | | | | invokes the build with --cpu=i486 on x86-32. Add -momit-leaf-frame-pointer to x86 specializations. Use -march=atom for atom32 as well. Use -mcpu=cell for Cell PPU - it's not documented, but GCC 4.6 has it. Remove ancient/untested rs64a support from ppc64
* I can't spelllloyd2011-05-134-9/+9
|
* Fix the problem that prevented the SSE2 IDEA implementation fromlloyd2011-05-122-6/+1
| | | | | | | | | | working correctly under Clang - the technique for emulating unsigned compare relied on signed overflow. The new method does not, and works under GCC, ICC, and Clang. Even better, the compare takes only 2 instructions instead of 4. Prevent using any of the asm implementations under Clang on x86-32. All of them crash under Clang 2.9, unclear why.
* Add base64_decode. Seems to work OK, though incremental decoding islloyd2011-05-124-10/+222
| | | | not well tested.
* Renaming the ppc target to ppc32 prevented -mcpu flags from being setlloyd2011-05-123-2/+8
| | | | | | | by GCC. Add Niagra targets for sparc, and extend/fix the Sun Studio flags for SPARC64.
* Add support for targetting the PowerPCSPE (an embedded PPC with anlloyd2011-05-123-5/+16
| | | | | experimental Debian port), and add aliases for ARM to match the Debian architectures.
* Modify soname to match Unix conventions at suggestion of Debianlloyd2011-05-124-11/+13
| | | | | | | | | | maintainer. Namespace the headers in /usr/local/include/botan-1.9/botan, so multiple versions can be installed in parallel. Down version to 1.9.18 since there are a good number of changes in this release already.
* Add back BigInt::to_u32bitlloyd2011-05-103-18/+46
| | | | | | | Fix BigInt::get_substring when length is equal to 32 - an overflow would cause the mask to be equal to 0 thus producing nothing at all. Disable CVC by default, it's not ready for prime time in any sense.
* Make different targets for 32 and 64 bit Atom processors, since mostlloyd2011-05-093-6/+7
| | | | consumer/desktop level Atoms are actually 32 bit.
* dword was set to a size_t if we were using a 16 bit limb. Caused by anlloyd2011-05-091-1/+1
| | | | overzealous perl script...
* Disable CMS by default, since it's pretty untested and likely unusablelloyd2011-05-061-0/+2
| | | | | | | | currently. Include the source directory name in object file names. Add some more DES test vectors.
* Search for the delimiter bytes in OAEP using a loop that doesn't havelloyd2011-05-051-25/+34
| | | | | | | conditionals to help avoid timing anylsis. Unfortunately GCC is too smart for us and compiles it to jumps anyway; probably would need to put the delim search into its own function and pass variables by volatile pointers to force the compiler to do what we want.
* When building for GCC, always get the version number, and turn off TR1lloyd2011-05-037-11/+14
| | | | | | | | | | | and -fvisibility support if the version is too old. You can also turn them off explicitly with the (hidden) option --without-visibility. We get the version number from the binary specified with --cc-bin, if the user set that, rather than from plain 'g++'. Fix Solaris install - apparently 'install' cmd is broken/dumb. Fix Ekopath flags for submodels.
* Tweak settings for SuperH setup, based on Debian bug 594159. Since aslloyd2011-05-023-8/+9
| | | | | far as I know this is the first and only real use, or even test, of botan on the SH, I'll let them pick the names and flags...
* Calling &str[str.size()] is only valid if str is const; otherwise thelloyd2011-04-292-5/+12
| | | | | | | | | | | | | | results are undefined. This happens to work under GCC and most other compilers, but does not under Visual C++ 2010. This broke hex_encode when encoding an empty input, and this subsequently broke SSL handshaking. 2010 includes a TR1 that works fine for SSL, but it puts the headers in the main header space rather than under tr1/, so account for that. Hack the socket header into working under WinSock Tick version to 1.10.0
* Default sun4u to sparc32-v9 rather than sparc64lloyd2011-04-292-4/+2
|
* Flatten out the mp directory since no alternative asm implemenationslloyd2011-04-256-13/+3
| | | | | | | for Montgomery or multiply/square currently exist and almost ceratainly won't be added during 1.10 Fix the name for Sun Studio in mp_asm64
* Rename AES_XXX_Intel to AES_XXX_NI, since AMD's Bulldozer will alsolloyd2011-04-254-33/+33
| | | | support AES-NI.
* Use STLport with Sun Studio by default on Linux. It is probably thelloyd2011-04-251-1/+1
| | | | | | | | | right thing for most modern distros, and if someone is using Sun CC on Linux they probably know if they don't want that. Record successful Sun Studio 5.10 build, also note that Clang 2.9 also seems to miscompile SSE2 IDEA. The Clang 2.9 build on FreeBSD did not have this failure, wonder what happened there.
* Rename all references of ia32 to x86-32 and amd64 to x86-64.lloyd2011-04-2248-193/+193
| | | | | | Back the reported version from 1.10.0 to 1.9.17 for the time being. Still on the fence if this will be 1.10.0 or another release candidate instead.
* Intel C++ doesn't seem to have an option for setting the soname, so itlloyd2011-04-221-1/+1
| | | | | | | | | | | | will default to the full output file name. That actually worked as expected when the so was always written as libbotan-1.8.2, but doesn't anymore since soname does not match the written file. This probably won't work right on Windows, but I don't have access to Intel C++ on Windows. Note good Intel C++ 11.1 build in build log Note that CPython 2.7 works for configuring the build
* Enable unix_procs for FreeBSD. It was disabled in 2006 to worklloyd2011-04-219-26/+24
| | | | | | | | | | | | | | | | | | | | | | | | around a bug in FreeBSD 6.1, which is long EOL. If we can't figure out the CPU in configure.py, if running verbosely dump the entire list of CPUs we know about. Some doc cleanups. Rename the 'beos' target to 'haiku', since testing shows that botan can't compile under the old BeOS GCC 2.95 anyway. Remove the call to idle_time in the stats entropy source - it causes a crash on Haiku R1-alpha2 somewhere inside a system DLL. I didn't bother debugging it beyond looking at the backtrace. Add a 'bepc' alias for i386 as that is what Haiku reports its processor as. Fix the install dirs to match Haiku R1, though apparently they will change in R2 anyway when they add package management. Enable use of gmtime_r on Haiku.
* It's likely that other FPE methods will be desirable once they arelloyd2011-04-203-9/+17
| | | | | | | standardized by NIST; the FPE currently included is just a random one that was relatively easy to implement. Move the header to fpe_fe1.h, and rename the function. Update the example and add some documentation for it.
* Add a --no-autoload examplelloyd2011-04-191-0/+4
| | | | | | Add missing dependency in the SIMD engine that would cause build failures when using --no-autoload and explicitly requesting a SIMD algorithm like aes_ssse3 using --enable-modules.
* Maintainer mode warning cleanups, mostly for C style casts which Illoyd2011-04-1815-58/+77
| | | | added to the flags here.
* My workaround for gcc pr 44174 caused cpuid support to be silentlylloyd2011-04-183-6/+12
| | | | | | | disabled under gcc. Hide asm labels in the x86-32 code; some did not begin with a .L so they were being exported.
* Integrate building docs using Sphinx and Doxygen into the makefilelloyd2011-04-184-86/+45
| | | | | | | | | | | | | proper. Enabled using --use-sphinx and --use-doxygen options, both of which default to false. If Sphinx isn't enabled, the ReST sources are copied directly (a case where having a very readable source format comes in handy...) The reference manual (either Sphinx processed into HTML, or the raw source) and the Doxygen output (if enabled) are copied into the doc directory upon install on Unix. Currently not done on Windows, the install target is fairly bogus there currently, and hasn't been tested in some time.
* Fix location of online docs in readmelloyd2011-04-153-23/+15
| | | | | | | | | | Some fixes for the Windows installer config Remove the SIMD rotate overloads; VC 2010 does not like passing a __m128i by value, which is required to match the template overload for the regular rotates. Could change it to a const reference, but I would worry this would inhibit compiler optimizations. Only used in one place (Noekeon), so just use the long expressions there.
* One minor fix for QNX 6.4.1 (Python identifies CPU as 'x86pc'), andlloyd2011-04-153-2/+8
| | | | | add support for features that it has that apparently the last version I tested did not, including dynamic loading.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-22 18:02:41 +0000