bemanproject / big_int / 31041432589

92%

push

github

Iterative steps school mul optimize (#305)

* Iterative steps school mul optimize

* Next iteration separate header low-level mul

* Run pre-commit locally

* Try again de-mangled optimized name

* Rearrange params

* Next architectural iteration

* Minor repairs runtime path isolation

* Repair an overlooked typo

* Linting and intuitive file naming

* Complete skeleton for multiply_long_runtime

* Improve guidance in comments

* Further improve guiding comments

* ASM and macro switch to/from ASM/C++

* Update include/beman/big_int/detail/multiply_long_runtime.hpp

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

* Get x86_64 ASM running try CI

* School mul opt (#306)

* removed a buggy instruction (neg r12)
and added a missing instruction (test r14, r14) which now no longer causes segfaults

* added an extra macro check for x86-64 cpus so as to allow build using clang on linux which was previously failing. details in the comment.

* switched the macro to the assembly function

* Get x86_64 ASM running try CI

---------

Co-authored-by: ckormanyos <ckormanyos@yahoo.com>

* Simplify syntax to be closer to original

* Further optimization by reducing 3-cycle dependency chain

* Handle reviews and syntax

* Update docs and fix benches

* Minor syntax improvements

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: The Blessed Machine <121212552+EpsilonNought117@users.noreply.github.com>
Co-authored-by: Matt Borland <matt@mattborland.com>

5 of 5 new or added lines in 2 files covered. (100.0%)

5700 of 5913 relevant lines covered (96.4%)

3312803.55 hits per line