[libmath-prime-util-perl] annotated tag upstream/0.65 created (now 78dbf88)
gregor herrmann
gregoa at debian.org
Fri Aug 4 21:58:40 UTC 2017
This is an automated email from the git hooks/post-receive script.
gregoa pushed a change to annotated tag upstream/0.65
in repository libmath-prime-util-perl.
at 78dbf88 (tag)
tagging 797a8337dece69ef19fe56e60f85ac26bd8e1e1b (commit)
replaces upstream/0.60
tagged by gregor herrmann
on Fri Aug 4 17:45:36 2017 -0400
- Log -----------------------------------------------------------------
Upstream version 0.65
Dana Jacobsen (170):
Handle is_power with small negative arguments in XS
Fix PP is_power with -1
Change PP Bernoulli method to Seidel from Brent-Harvey
Small redo of PP LambertW code
Merge pull request #17 from trizen/patch-1
Style update for PP bernoulli
Merge pull request #18 from trizen/patch-2
Rewrite of LambertW code, both C and PP
Code style; docs for LambertW
LambertW accuracy near branch point
Fix RiemannZeta PP post-process of GMP result
Use GMP LambertW if possible
Tune AKS params for 33-64 bit inputs
Add clang support for uint128
Use Montgomery math for sqrtmod, is_primitive_root, and p-1 factor
Make sure we don't use Mont math on even modulus
Make primes.pl use twin_primes function
Handle some chinese() cases that overflowed before
Switch to Bytes::Random::Secure::Tiny
Add is_pillai
chinese goes directly to GMP
Tests for zero modulus on chinese
Sort chinese moduli with Shell sort
Export inverse_li including PP, docs, tests, etc.
Merge branch 'master' of https://github.com/danaj/Math-Prime-Util
Add is_semiprime
is_semiprime in GMP, minor typo in test output
Use MPU::GMP random primes functions
Use GMP version number to select random prime package
Remove unused variables
is_mersenne_prime got pushed out of XS case
Use GMP random strong and maurer primes if possible
Small speedup for is_carmichael
Timing on new AWS machine
Add tests for is_pillai function
Comment about Axler 2017 lower prime count bound
PP powmod and mulmod will try GMP
Update for new GMP random prime functionality
Update version number
Merge branch 'master' of https://github.com/danaj/Math-Prime-Util
Fixes for portability and PP
Back out PP nth_prime_approx change to use inverse_li
Better Lucky primes and clusters
Slightly faster is_pillai
Comment change
Update for release
Comment updates
Add internal double-word to string function
Use our own 128-bit return for vecsum
Add faster prime sum
Uninitialized variable in 128-bit string return
Remove warnings for unused variables in conditional code
Speed up is_pillai for composites (early 0 detection)
Add prime restriction to forpart
Update ppport
Add PP prime sum, fix result of both versions
Speed up cluster sieve a bit
Formatting
Update Mersenne number milestones
is_primitive_root needs two args
Merge branch 'master' of https://github.com/danaj/Math-Prime-Util
Update large prime count bounds with Axler 2017
Ramanujan primes, finer nth upper limit
Faster ramanujan prime count
Ramanujan prime loops use segmented sieve for big memory savings
ramanujan prime count updates
Move some thread logic to its own include file
Skip a big semiprime test if not using GMP
Add random functions. Massive changes.
Move random_nbit_prime and random_ndigit_prime to XS
mismatched ifdef
Make tests pass on 32-bit
Perl_seed didn't exist in 5.6
Remember thread context for seed
standalone compilation was skipping the test for uint64_t
Remove debugging comments
Move random_proven_prime and random_strong_prime into XS
Call urandomm from PP code if possible
Move remaining random primes functions to XS
Remove RandomPrimesGMP -- not needed any more
cleanup release tests
Update timings in RandomPrimes comments
Latest Ramanujan prime limits
ChaCha20 csprng in C and Perl
More options to have uint64_t
memtile is a little more generic
Ramanujan prime updates
Fix for odd substr behavior
Add but don't use new ChaCha core
Switch to new simple ChaCha20
Update docs for ChaCha20
Fixes for 32-bit
More 32-bit fixes
Updates for PP ChaCha and drand
Consistent ChaCha20 for C and Perl
ChaCha20 PP random_bytes fix
Consistent srand on 32-bit machines
Consistency for ISAAC, more float changes
Add support for quadmath to drand.
Previous PP drand change broken arguments. Fix
Instead of NV size and USE_QUADMATH, use NVMANTBITS for drand behavior
Look for PERL_VERSION to decide if we're compiling as part of Perl or not
Simplify ChaCha irand/irand64/random_bytes code
Revert STANDALONE change
Stateful testing for rand
More drand updates
Endian update for PP ChaCha20
Speedup PP testing on super slow machines
Optimizations for PP ChaCha
Update benchmarks and PPFE selection of RNG
fix old seed_csprng to csrand. Enable rand tag test.
Remove old comments in rand tag test
Fix seeding alignment differences on 32-bit and 64-bit
Release testing and add 6 new ramanujan prime functions
Bump version number
Move RNGSeed to Entropy
Add entropy_bytes
Split util.c into 3 pieces
32-bit issues with some of the PP functions
Bump requested GMP version
math.h defines fro functions used
Merge branch 'master' of https://github.com/danaj/Math-Prime-Util
Change RNGSeed to Entropy in MANIFEST
Perl.h's U8TO32_LE doesn't wrap ptr in parens
Redo how we find math.h functions
Speed up PP is_semiprime
is_semi test faster but still slow on 32-bit PP, so trim
PP next_prime prev_prime, inline bigint primality test
Fix some warnings
Redo small Pi to help quadmath not depend on sprintf
Comments about drand
Move spigot Pi algorithm to util. Use it in Pi().
Compatibility for Perl 5.6 and old GMP code
Add entropy_bytes to release tests
Run t/34 core rand detection on 32-bit
miller_rabin_random broken out and better catching of negative number of bases
Bump version number
We should be fine with MPU::GMP 0.44
Change inverse_li to Halley
Don't call GMP miller_rabin_random for older versions
Bump version
If GMP returns undef, we pass it through unchanged
Rewrite of LogarithmicIntegral and inverse_li
Support Math::BigInt::Lite
New version of inverse R used in nth_prime_approx
Use our rand calls, especially urandomm, for our xt and bench files
Use inverse_li instead of Cipolla for PP nth_prime_approx
Faster convergence for inverse_li
Internal: remove _XS_ prefix from Ei,Li,RiemannR
Make sure tests pass for all non-bigint object inputs
Merge branch 'master' of https://github.com/danaj/Math-Prime-Util
Internal: Rename _XS_prime_count to segment_prime_count, add prime_count
Internal: Rename _XS_*_pi functions
Rename private XS interface functions for prime count algorithms
inverse Riemann R: slightly better accuracy
Internal: Change _XS_is_prime to is_prime
Build and test fixes from test reports for v0.64
Use binary search for trimming ramanujan_primes
Adjust segment / LMO tradeoff in prime_count
Work around Math::BigInt::Lite issue
Lock-free prefix for cache optimization in next_prime etc.
Make sure we turn on dynamic config
Fix sum_primes, tune nth_prime_upper, more Ramanujan bounds
Deprecate irand and primeinc
Update some benchmarks
Update last double-checked Mersenne
Updates to large Ramanujan prime bounds
Bump version number
Make PP inverse_R function private
Prep for release
Daniel Șuteu (2):
_bernoulli_seidel() minor improvements
minor precedence fix in _bernoulli_seidel()
gregor herrmann (1):
New upstream version 0.65
-----------------------------------------------------------------------
No new revisions were added by this update.
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/pkg-perl/packages/libmath-prime-util-perl.git
More information about the Pkg-perl-cvs-commits
mailing list