Mark Wooding [Thu, 27 Mar 2014 03:18:32 +0000 (03:18 +0000)]
math/mpx.c: Eliminate clone-and-hack of shifting primitives.
Replace with some fancy macros.
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c (noise_filter): Use the <mLib/sel.h> machinery.
Makes things a little more verbose but rather simpler.
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c (noise_filter): Use <mLib/mdup.h>.
The code was slightly incorrect previously: if the standard descriptors
are closed to begin with, then the pipe ends up being closed on entry.
As far as I know, this has never actually happened, but I have the
machinery to fix this bug so I might as well use it.
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c (noise_filter): Invoke the shell properly.
This has been buggered forever. :-/
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c: Environment variable to force use of `noise_filter'.
For debugging, mostly. Don't use this in real life.
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c: Order <signal.h> properly.
I try to keep them alphabetical.
Mark Wooding [Wed, 12 Mar 2014 16:20:35 +0000 (16:20 +0000)]
math/gfreduce.c: Fix commentary.
Nothing very serious.
Mark Wooding [Wed, 12 Mar 2014 16:20:06 +0000 (16:20 +0000)]
math/mpreduce.c: Fix some typos in commentary.
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c (noise_timer): Actually update previous time.
A long-standing bug. Oops.
Mark Wooding [Sat, 13 Jul 2013 15:34:40 +0000 (16:34 +0100)]
rand/noise.c (noise_devrandom): Fix reading from the kernel random device.
It was, unfortunately, filled with badgers.
* Don't take short reads for an answer. Loop until the read actually
fails, or we've filled the buffer.
* If we didn't actually read enough to fill the buffer, then don't
return success! The fallback collectors will engage and hopefully
save our bacon.
Mark Wooding [Fri, 6 Sep 2013 09:46:54 +0000 (10:46 +0100)]
gdsa: Generate nonces more securely.
Hash the private key and message, together with some random stuff. This
ought to be hard to guess even if the randomness is bad.
Mark Wooding [Fri, 6 Sep 2013 09:45:22 +0000 (10:45 +0100)]
pub/gkcdsa.c: Fix label in test failure output.
Mark Wooding [Sat, 28 Dec 2013 14:21:51 +0000 (14:21 +0000)]
Release 2.1.6.1.
Mark Wooding [Sat, 28 Dec 2013 13:59:44 +0000 (13:59 +0000)]
symm/multigen: Fix for Python 2.5.
The top-level `next' function was introduced in 2.6, so use `.next()'
and catch `StopIteration'.
Mark Wooding [Sat, 28 Dec 2013 13:36:37 +0000 (13:36 +0000)]
math/gen{primes,wheel}.c: Don't use full header file name in `#include'.
This makes source distributions fail badly.
Mark Wooding [Fri, 27 Dec 2013 14:29:16 +0000 (14:29 +0000)]
Release 2.1.6.
Mark Wooding [Fri, 27 Dec 2013 12:32:12 +0000 (12:32 +0000)]
math/mp-jacobi.c: Fix embarrassing spelling mistake in documentation.
Mark Wooding [Fri, 27 Dec 2013 12:02:26 +0000 (12:02 +0000)]
math/mp-gcd.c: Minor reformatting.
So that it matches the recent fix a little better.
Mark Wooding [Fri, 27 Dec 2013 12:00:53 +0000 (12:00 +0000)]
math/mp-gcd.c: Avoid clobbering constants during the sign fixup.
If the GCD computation was trivial, then the state still contains
references to the constants `MP_ZERO' and `MP_ONE'. It would be Really Bad
to trash their signs -- in particular, this causes comparisons with
`MP_ONE' to fail, leading to an assertion failure from `mp_modinv'.
Mark Wooding [Fri, 27 Dec 2013 11:56:48 +0000 (11:56 +0000)]
Makefile.am: Add some forgotten files to the distribution archive.
Mark Wooding [Sun, 11 Aug 2013 01:07:41 +0000 (02:07 +0100)]
More whitespace fixes.
With a little luck, that'll be the last for a while.
Mark Wooding [Sun, 11 Aug 2013 01:08:00 +0000 (02:08 +0100)]
.gitattributes: New file; Python programs shouldn't have tabs.
Mark Wooding [Sat, 10 Aug 2013 20:19:58 +0000 (21:19 +0100)]
Some more whitespace cleanups.
Mark Wooding [Sat, 10 Aug 2013 08:48:52 +0000 (09:48 +0100)]
math/gfreduce.c, math/mpreduce.c: Remove unused macro from testing code.
Must have been cut-and-paste lossage when the tests were added.
Also remove a spurious blank line.
Mark Wooding [Fri, 9 Aug 2013 12:22:47 +0000 (13:22 +0100)]
math/gfreduce.[ch]: Fix out-of-bounds memory access.
The final pass of the reduction adds a multiple of the extra top bits
from the most significant word; but at this point, the generated
instruction sequence will access a word one beyond the bottom of the
supplied memory vector. While it (probably) won't modify this word, it
will still attempt to read and write it.
This is relatively harmless, since typically the vector will have been
allocated from our custom arena, and therefore there'll be a header word
in this position, but hand-built polynomials may cause trouble.
Fix this bug by keeping track of the first instruction which accesses a
word other than the least significant, and using this alternative entry
point in the final pass. Fortunately, there's an unused slot, `liv', in
the context structure which we can use for this purpose!
(Yes, the previous refactoring was largely for the purpose of fixing
this bug.)
Mark Wooding [Fri, 9 Aug 2013 09:30:48 +0000 (10:30 +0100)]
math/gfreduce.c: Refactor and document.
There was an unfortunate amount of code duplication in the old code,
particularly around the issuing of LSR instruction sequences. Gather
the relevant state into a structure, and split the repeated code into
separate functions.
A casualty of this refactoring work is the work put into calculating the
`liv' slot in the `gfreduce' structure, carefully calculated to be a
pointer immediately after the first `STORE' instruction. Alas, this
information is completely useless.
Also add a bunch of commentary explaining the underlying theory.
Mark Wooding [Fri, 9 Aug 2013 09:13:15 +0000 (10:13 +0100)]
math/gfreduce.h: Missing include.
Mark Wooding [Fri, 9 Aug 2013 09:12:16 +0000 (10:12 +0100)]
math/*.awk: These should have been killed when `mpgen' appeared.
Mark Wooding [Tue, 6 Aug 2013 18:50:02 +0000 (19:50 +0100)]
math/t/mpreduce: Add a slew more randomized tests.
With a little luck, that's the main edge cases in the algorithm
tested (as well as merely proven correct).
Mark Wooding [Tue, 6 Aug 2013 18:49:08 +0000 (19:49 +0100)]
math/mpreduce.[ch]: Extend the domain to all positive integers.
Integers of the form (100...)_2 are now acceptable, at last, now that
I've got a grip on the underlying theory. (It's somewhat embarrassing
that it's taken so long, given that the algorithm was my own work to
begin with, but it was all rather trial-and-error.)
Negative numbers still don't work, and probably never will.
Mark Wooding [Tue, 6 Aug 2013 18:37:34 +0000 (19:37 +0100)]
utils/naf.c: Wrap up the state machine correctly.
The old code just assumed a particular resulution (in particular, the
`X' case), which is far from universally applicable. Now we simulate up
to two more steps, feeding in zero bits.
Mark Wooding [Tue, 6 Aug 2013 17:47:50 +0000 (18:47 +0100)]
math/mpgen: Remove spurious newlines in `mplimits.c'.
Mark Wooding [Mon, 5 Aug 2013 20:16:53 +0000 (21:16 +0100)]
utils/naf.c: Non-adjacent form calculator, from math/mpreduce.c.
Rescue the plain old state machine, in case it's interesting later.
Mark Wooding [Mon, 5 Aug 2013 20:15:45 +0000 (21:15 +0100)]
math/mpreduce.c: Remove ancient debugging code.
Mark Wooding [Mon, 5 Aug 2013 20:14:06 +0000 (21:14 +0100)]
math/gfreduce.c: Fix misformatting in the commentary.
Mark Wooding [Mon, 5 Aug 2013 20:13:48 +0000 (21:13 +0100)]
math/mpreduce.c: Add extensive commentary.
The behaviour of this code must have been something of a mystery. It's
not arbitrary, but it is a little subtle in places. Add a full
explanation of the whole thing.
Mark Wooding [Mon, 5 Aug 2013 20:12:32 +0000 (21:12 +0100)]
math/mpreduce.h: Missing include files.
This all seems rather unfinished, really.
Mark Wooding [Mon, 5 Aug 2013 01:52:41 +0000 (02:52 +0100)]
math/mp-jacobi.c: Improve the presentation.
Make the commentary a bit more useful, and move the code about to fit
in.
Mark Wooding [Mon, 5 Aug 2013 01:52:20 +0000 (02:52 +0100)]
math/ec-prime.c: Fix commentary typo.
Mark Wooding [Mon, 5 Aug 2013 01:41:12 +0000 (02:41 +0100)]
math/ec.c (ec_stdsub): Remove redundant `EC_FIX'.
This was left over from the old days, when the right input to `add'
was guaranteed to be in canonical form. It's not been necessary since
391faf4...
Mark Wooding [Sat, 3 Aug 2013 11:28:45 +0000 (12:28 +0100)]
math/mp-modsqrt.c: Reformat and add commentary.
Mark Wooding [Sat, 3 Aug 2013 11:28:13 +0000 (12:28 +0100)]
math/mp.h, math/mp-jacobi.c: Whitespace fix.
Mark Wooding [Sat, 3 Aug 2013 10:51:37 +0000 (11:51 +0100)]
vars.am: Remove extra prerequisite from the `.to.t' rule.
It doesn't do what one expects.
Mark Wooding [Thu, 11 Jul 2013 09:09:14 +0000 (10:09 +0100)]
progs/catsign.1: Fix the warning about `verify' output.
Obviously copied from `catcrypt decrypt', which does only output
verified data because of its chunking; but `catsign' doesn't work like
this by default.
Mark Wooding [Thu, 11 Jul 2013 09:05:59 +0000 (10:05 +0100)]
progs/catsign.c: Remove old debugging code.
Commented out, but still annoying.
Mark Wooding [Wed, 10 Jul 2013 23:09:05 +0000 (00:09 +0100)]
progs/pixie.c: Use GNU capital-letter convention in HELP output.
Mark Wooding [Wed, 10 Jul 2013 23:07:43 +0000 (00:07 +0100)]
progs/pixie.[c1]: Don't replace the existing pixie by default.
It's too easy to screw things up by replacing the running pixie, by
calling the pixie with no arguments. The old behaviour is available via
the shiny new `--replace' option.
Mark Wooding [Sun, 30 Jun 2013 00:17:18 +0000 (01:17 +0100)]
progs/cookie.c: Constant-time MAC tag checking.
Mark Wooding [Fri, 28 Jun 2013 23:42:04 +0000 (00:42 +0100)]
Release 2.1.5.
Mark Wooding [Fri, 28 Jun 2013 23:23:07 +0000 (00:23 +0100)]
Mollify various warnings which occur in 64-bit builds.
Mark Wooding [Fri, 28 Jun 2013 23:22:28 +0000 (00:22 +0100)]
pixie.c: Use `socklen_t' rather than `size_t'.
Mark Wooding [Sat, 22 Jun 2013 15:23:56 +0000 (16:23 +0100)]
key/pixie-common.c, progs/pixie.c: Handle error returns better.
This was rather bad, really. It's still not perfect, by a long way.
Mark Wooding [Sat, 22 Jun 2013 15:24:52 +0000 (16:24 +0100)]
progs/pixie.c: Fix format security bugs.
Really not very impressive.
Mark Wooding [Sat, 22 Jun 2013 14:04:44 +0000 (15:04 +0100)]
Use the shiny new `mLib' warning-control macros.
Mark Wooding [Sat, 22 Jun 2013 14:03:33 +0000 (15:03 +0100)]
progs: Fix a number of format-related errors.
Picked up by the new `mLib' attribute machinery.
Mark Wooding [Wed, 19 Jun 2013 02:09:46 +0000 (03:09 +0100)]
Muffle GCC warnings in various ways.
Some don't require much in the way of contortion to muffle without using
GCC-specific tricks. Others are hard or impossible to avoid -- because
they're to do with macro definitions, or even actually GCC bugs (e.g.,
the array-bounds warning in `square.c') -- without just muffling the
warnings explicitly, by name.
Mark Wooding [Wed, 19 Jun 2013 00:43:57 +0000 (01:43 +0100)]
progs/pixie.c: Rewrite list hacking to avoid strict-aliasing badness.
The circular list stuff was quite pretty but involved some really
unpleasant casting which modern GCC (quite properly) complains about
vociferously.
Replace it with more traditional doubly-linked-list hacking with
null-pointer sentinels, with the slightly nasty pointer swizzling tucked
away in useful macros. Some of the uses of these macros (e.g.,
unlinking the first or last item in a list) could be made more efficient
by using special-case versions, but it doesn't seem worthwhile.
Mark Wooding [Wed, 19 Jun 2013 00:39:03 +0000 (01:39 +0100)]
progs/pixie.c: Rename `log' function to `pxlog'.
Shuts up an annoying compiler warning.
Mark Wooding [Fri, 14 Jun 2013 19:43:15 +0000 (20:43 +0100)]
debian: Use Debhelper 8 rather than the awful custom script.
Mark Wooding [Fri, 14 Jun 2013 19:43:15 +0000 (20:43 +0100)]
debian/control: Remove irrelevant blather about `xpixie'.
The script was never shipped.
Mark Wooding [Wed, 5 Jun 2013 16:14:30 +0000 (17:14 +0100)]
Overhaul `math' representation machinery.
Collect type information from the C compiler at configuration
time (using a rather complicated hack so that it works with
cross-compilers). Read this from a Python script `mpgen' which is now
responsible for knowing all of the `mp' representation details.
Since `mpgen' generates all of the constant tables directly, we no
longer have any need for the programs `genlimits' or `mpdump' -- or the
random collection of `awk' scripts for turning `mumbletab.in' files into
`mumbletab.c' files. And this means that we can kill `libmpbase.la'.
With this change, Catacomb is finally safe for cross-compilation.
Mark Wooding [Sun, 2 Jun 2013 20:37:47 +0000 (21:37 +0100)]
Generate precomputed tables as sources in `precomps/'.
Rather than header files in the build tree. The precomputations are
distributed, and not built in cross builds.
Mark Wooding [Sun, 2 Jun 2013 11:54:11 +0000 (12:54 +0100)]
Rearrange the file tree.
It's actually, like, a tree now. Testing is a bit wobbly: you really do
have to make the library before the tests will build. I don't like
this, but the pieces of the library are all rather intertwined.
One small piece of unintertwining: `strongprime.c' no longer includes
`rand.h', which detaches the mathematical code from the symmetric-crypto
build system disaster.
Test files have been moved into `t/' directories as is now established
practice.
Also take the opportunity to eradicate the CVS `$Id...$' droppings.
Mark Wooding [Sat, 1 Jun 2013 22:39:51 +0000 (23:39 +0100)]
configure.ac: Replace with a new version.
Mark Wooding [Sat, 1 Jun 2013 19:35:54 +0000 (20:35 +0100)]
blowfish-mktab.c: Remove the eye-candy progress meter.
It really makes a mess of parallel builds.
Mark Wooding [Sat, 1 Jun 2013 19:28:26 +0000 (20:28 +0100)]
Abandon the `m4'-based build system. And there was great rejoicing.
The makefile has been entirely rewritten using old-fashioned Automake.
I've used the undocumented `foo_OBJECTS' variables in a couple of places
to make parallel builds work, but those will disappear again in the
course of future rearrangements of the source tree.
A couple of the scungier programs for building boilerplate source files
have been replaced by the marvellous new template-substitution program
`multigen'.
The `qcc' script is a casualty of the rearrangements. Supporting it is
easy, but probably pointless. So it's gone.
Currently most of the generated code is left in the build tree (rather
than the source tree) and not distributed. This wants to be fixed
later, but the exact machinery will be rather different.
This is only the start of a programme of overhauling Catacomb's ageing
and rather nasty build system. The objective is to end up with modern-
looking build scripts, and a pleasant and sensible directory tree. Not
much of the actual code will change during this, though lots of it will
move around. In particular, the testing machinery is likely to be
overhauled quite a lot.
Mark Wooding [Sat, 1 Jun 2013 22:38:03 +0000 (23:38 +0100)]
.links: Drop obsolete `lib-config.in' file.
Mark Wooding [Sat, 1 Jun 2013 19:34:48 +0000 (20:34 +0100)]
key-flags.c, key-pack.c, key-pass.c: Don't use the `key.h' machinery.
These are logically part of the `key-data.h' layer, and don't need
anything beyond that, so remove the header-file inclusion.
Mark Wooding [Mon, 27 May 2013 21:34:38 +0000 (22:34 +0100)]
Release 2.1.4.
Mark Wooding [Mon, 27 May 2013 21:23:58 +0000 (22:23 +0100)]
oaep.c, pkcs1.c: Use official constant-time operations.
The logic is a bit more contorted in places, but the security is better.
Mark Wooding [Mon, 27 May 2013 21:23:35 +0000 (22:23 +0100)]
ct.c, ct.h: New constant-time operations.
Mark Wooding [Mon, 27 May 2013 22:05:44 +0000 (23:05 +0100)]
Revert "group.h: Fix the struct tag for `group'."
I'm a moron. The trailing `_' was there to distinguish my groups from
`struct group' in <grp.h>. Put it back.
This reverts commit
c6c823084467cc7a60808cdee8015529115b8b91.
Mark Wooding [Mon, 27 May 2013 21:28:12 +0000 (22:28 +0100)]
Merge branch 'master' of git.distorted.org.uk:~mdw/publish/public-git/catacomb
* 'master' of git.distorted.org.uk:~mdw/publish/public-git/catacomb:
mpmul.[ch]: Move internal `HWM' and `LWM' constants to implementation.
group.h: Fix the struct tag for `group'.
Mark Wooding [Tue, 30 Apr 2013 22:00:34 +0000 (23:00 +0100)]
mpmul.[ch]: Move internal `HWM' and `LWM' constants to implementation.
No idea what I was thinking when I put them in the public header file.
Mark Wooding [Thu, 11 Apr 2013 10:37:10 +0000 (11:37 +0100)]
Release 2.1.3.
Mark Wooding [Thu, 11 Apr 2013 11:02:21 +0000 (12:02 +0100)]
The pixie no longer needs to be setuid-root.
So turn off the option by default, and downgrade the question. Also
make the documentation more useful and up-to-date.
Mark Wooding [Wed, 10 Apr 2013 19:40:36 +0000 (20:40 +0100)]
group.h: Fix the struct tag for `group'.
Maybe I was going to give it a longer name once upon a time. It's stuck
now, though.
Mark Wooding [Mon, 8 Apr 2013 16:10:50 +0000 (17:10 +0100)]
mkphrase.c: Allow a range for phrase entropy.
Mark Wooding [Mon, 8 Apr 2013 16:09:49 +0000 (17:09 +0100)]
mkphrase.c: Better error checking on the length range parameter.
Mark Wooding [Sun, 7 Apr 2013 22:44:34 +0000 (23:44 +0100)]
New function and example program computes Fibonacci numbers fairly fast.
Mark Wooding [Thu, 28 Feb 2013 16:55:41 +0000 (16:55 +0000)]
hashsum.c: Document `--progress' in the `--help' display.
Release 2.1.2.1.
Mark Wooding [Thu, 28 Feb 2013 17:36:22 +0000 (17:36 +0000)]
.gitignore: Ignore `auto-version' script.
Mark Wooding [Wed, 9 Jan 2013 03:26:52 +0000 (03:26 +0000)]
Release 2.1.2.
Mark Wooding [Wed, 9 Jan 2013 03:29:48 +0000 (03:29 +0000)]
Cleanups.
* Remove bogus old `catacomb-config.in' file from the pre-pkgconfig
days.
* Fix library version in Makefile.
Mark Wooding [Wed, 9 Jan 2013 03:56:20 +0000 (03:56 +0000)]
cc-hash.c (fhash): The FILE name may be null.
So pass a dummy string on for the progress indicator.
Mark Wooding [Sat, 25 Feb 2012 23:35:18 +0000 (23:35 +0000)]
Hash utilities: Check for and report on junk files.
That is, when verifying a list of hashes, we optionally detect and
report files which are present in the filesystem, alongside files which
we're checking, but which aren't in our list.
Thanks to Patrick Gosling for the idea.
Mark Wooding [Mon, 7 Jan 2013 23:27:08 +0000 (23:27 +0000)]
Hash utilities: maintain a hash state object, not a bundle of arguments.
This makes a number of things a bit cleaner, except that we (currently
pointlessly) remember to free up this state object when we're finished
with it.
Mark Wooding [Thu, 3 May 2012 12:48:59 +0000 (13:48 +0100)]
Use auto-version machinery for building.
Mark Wooding [Thu, 3 May 2012 12:35:47 +0000 (13:35 +0100)]
Makefile.m4: Remove mplimits.[ch] on clean.
Mark Wooding [Thu, 3 May 2012 12:28:24 +0000 (13:28 +0100)]
configure.in: Put the auxiliary scripts in a useful place.
For some reason, modern Autoconf finds the wrong copy otherwise and
all ell breaks loose.
Mark Wooding [Thu, 3 May 2012 09:54:16 +0000 (10:54 +0100)]
mpint.c (touint): Compare unsigned with unsigned, not unsigned long.
Otherwise the tests fail on LP64 platforms.
Mark Wooding [Thu, 3 May 2012 09:44:20 +0000 (10:44 +0100)]
tests/Makefile.m4: Distribute the converted AES test-vector files.
Otherwise later versions of distcheck fail.
Mark Wooding [Thu, 3 May 2012 09:35:19 +0000 (10:35 +0100)]
Makefile.m4: Don't build pgroups.kr: it introduces circular dependency.
Mark Wooding [Sat, 25 Feb 2012 23:34:16 +0000 (23:34 +0000)]
hashsum.c: Return nonzero from `checkhash' on errors.
A serious bug: `checkhash' carefully maintained the `rc' variable -- and
then ignored it and always returned zero anyway.
Mark Wooding [Sat, 25 Feb 2012 13:41:19 +0000 (13:41 +0000)]
cc.h: Fix FHF_MASK.
Must have been a typo. Nobody uses this for anything anyway.
Mark Wooding [Mon, 23 Jan 2012 02:46:53 +0000 (02:46 +0000)]
key-data.[ch]: Fix trivial typo.
`key_mewmp'. Miaow.
Mark Wooding [Mon, 23 Jan 2012 02:37:40 +0000 (02:37 +0000)]
key-data.c (key_struct{set,steal}): Assert no other references.
Otherwise I predict serious trouble when someone gets the reference
counting wrong.
Mark Wooding [Sun, 22 Jan 2012 13:12:15 +0000 (13:12 +0000)]
dsig.c: Allow precomputed hashes to be read from a file.
This lets you convert a hashsum(1) file or similar into a dsig(1)
signature file.
Mark Wooding [Sun, 22 Jan 2012 13:12:15 +0000 (13:12 +0000)]
cc-hash.c, hashsum.c: Move hash-file parsing stuff to `cc-hash.c'.
This is a bit trickier than just slinging existing functions about and
tarting them up a bit: it introduces a proper interface to parsing hash
files, which previously was interleaved with actually verifying the
hashes.
Also moved a couple of auxiliary functions which are needed by the moved
code.
Mark Wooding [Sun, 22 Jan 2012 13:12:14 +0000 (13:12 +0000)]
cc-hash.c: New file containing hash-related code from hashsum and dsig.
There's a fair amount of duplication already, most notably the
`getstring'/`putstring' functions, and `fhash'. The encoding stuff
isn't common yet, but will be needed in a later change.