v4.0.0

Breaking Change

• Types have been renamed to int128 and uint128 from int128_t and uint128_t respectively

What's Changed

• Add SYCL support by @mborland in #419
• Fix and test module support by @mborland in #421
• Use extern c++ in the module and add GCC + MSVC runners by @mborland in #422
• Fix consistency of const parameters in the docs by @mborland in #426
• Add support for the entire library to be shipped as a single header by @mborland in #431
• Add support for digit separators in literals by @mborland in #425
• Fix signed negation in add/sub operators causing UBSAN to flag operation by @mborland in #429
• Add Boost.ContainerHash integration and example by @mborland in #446
• Fix for 452. by @MarekKnapek in #454
• Fix for incorrect signed natvis. by @MarekKnapek in #456
• Fix 2x parsing problems with literals (prefix and oversize value) by @mborland in #445
• Update division operation behavior and docs by @mborland in #451
• Add to_wstring function by @mborland in #453
• Rename functions from op_sat to saturating_op by @mborland in #457
• Improve floating point conversions by @mborland in #458
• Remove shift guardrails so builtin delegation and synthetic match behavior by @mborland in #459
• Add C++26 integer comparison functions by @mborland in #460
• Improve performance of widening muls by @mborland in #461
• Add support for C++26 constexpr std::format by @mborland in #462
• Debugger pretty printer for MSVC. by @MarekKnapek in #464
• Replace int128_t high word with uint64_t by @mborland in #469
• Add Additional GPU examples by @mborland in #470
• Address collected doc review comments by @mborland in #471
• Add the P3724 division functions by @mborland in #472
• Use more builtin bit/byte functions when available (and more performant) by @mborland in #474
• Drop the _t from the type names by @mborland in #475
• Address bugs found in #476 by @mborland in #488
• Add support for floating point and integral compound operators by @mborland in #490
• Separate {fmt} and examples by @mborland in #494
• Add endian transformation functions by @mborland in #495
• Add absl benchmarks for Windows platforms by @mborland in #497
• Rename examples to example by @mborland in #498

New Contributors

• @MarekKnapek made their first contribution in #454

Full Changelog: v3.1.0...v4.0.0

v3.1.0

What's Changed

• Use edit_url instead of branch by @mborland in #413
• Reduce noise in the examples by @mborland in #414
• Review documentations and fix formatting behaviors by @mborland in #416

Full Changelog: v3.0.0...v3.1.0

v3.0.0

Breaking Changes

The library behavior now matches exactly the behavior of (unsigned) __int128 instead of erroring on sign compare or conversion.

What's Changed

• MSVC added constexpr support to their 128-bit ints by @mborland in #386
• Explain why this is different from Boost.MP by @mborland in #387
• Remove the sign conversions and any differentiated handling by @mborland in #390
• Fix consteval fallthrough by @mborland in #393
• Fix possible overflow in mini_from_chars by @mborland in #392
• Update numeric literals and documentation by @mborland in #395
• Add modular exponentiation function by @mborland in #399
• Make cross type conversions implicit by @mborland in #401
• Add support for std::hash by @mborland in #404
• Add ipow function by @mborland in #405
• Add sqrt function by @mborland in #406
• Bump the all-dependencies group in /doc with 3 updates by @dependabot[bot] in #407
• Add C23 style checked arithmetic by @mborland in #409
• Optimize further mul and div algorithms by @mborland in #411

Full Changelog: v2.0.0...v3.0.0

v2.0.0

What's Changed

• Improve Overview and Readme by @mborland in #336
• Add additional modules testing by @mborland in #339
• Allow explicit operations only between int128_t and uint128_t by @mborland in #340
• Add int128_t support for Boost.Random by @mborland in #344
• Add CUDA support for fundamental operations by @mborland in #346
• Add CUDA testing of bitwise operations by @mborland in #350
• Add cuda testing of functions by @mborland in #355
• Add CUDA testing of cstdlib functions by @mborland in #356
• Add CUDA testing of all comparison operators by @mborland in #359
• Add CUDA testing of conversion functions by @mborland in #360
• Add CUDA testing of functions by @mborland in #358
• Add and test CUDA support to charconv functions by @mborland in #361
• Add testing of literals with CUDA devices by @mborland in #362
• Fix CUDA selection macro by @mborland in #364
• Disable std::is_constant_evaluated on device and expand CUDA testing by @mborland in #366
• Replace deprecated find CUDA module by @mborland in #368
• Fix std::showbase vs std::noshowbase by @mborland in #371
• Add CUDA usage example by @mborland in #372
• Fix bug in division of minimum value by @mborland in #378
• Fix useless cast from decimal by @mborland in #380

New Contributors

• @dependabot[bot] made their first contribution in #341

Full Changelog: v1.6.1...v2.0.0

V1.6.1

What's Changed

• Avoid using strlen to get return size by @pps83 in #326
• Allow unaligned load with SSE on win x32 by @mborland in #328

New Contributors

• @pps83 made their first contribution in #326

Full Changelog: v1.6.0...v1.6.1

v1.6.0

What's Changed

• Move docs to Antora by @mborland in #291
• Add lunr extension for doc search by @mborland in #294
• Improve the docs of the individual types by @mborland in #296
• Improve charconv performance with num_digits specialization by @mborland in #299
• Expand Examples by @mborland in #300
• Add {fmt} and alignment support by @mborland in #301
• Apply Sam's Antora recommendations from decimal by @mborland in #302
• Doc updates by @mborland in #303
• Fix for implementation defined behavior causing failures by @mborland in #305
• Fix formatting of padding by @mborland in #306
• Add spaceship operator support for both int128_t and uint128_t by @mborland in #309
• Mark unreachable paths by @mborland in #310
• Coverage Improvements by @mborland in #312
• Add example of rollover behavior by @mborland in #314
• Make operator bool() explicit by @mborland in #323
• Add to_string functions by @mborland in #324
• Adjust alignment on 32-bit systems by @mborland in #322
• Add example of new to_string method by @mborland in #325

Full Changelog: v1.5.1...v1.6.0

v1.5.1

What's Changed

• Fixes for formatting headers by @mborland in #288

Full Changelog: v1.5.0...v1.5.1

v1.5.0

What's Changed

• Add LLDB pretty printer module by @mborland in #282
• Add GDB printer by @mborland in #283
• Fix strict aliasing violation by @mborland in #285
• Merge to master for v1.5.0 by @mborland in #287

Full Changelog: v1.4.0...v1.5.0

v1.4.0

What's Changed

• Fix behavior of prefix increment/decrement by @mborland in #278
• Fix ODR violation in numeric limits pre-C++17 by @mborland in #279

Full Changelog: v1.3.0...v1.4.0

v1.3.0

• Implementation of std::div like functionality
• Fix overconsumption of characters with istream