[WIP] mulhigh improvements #2123
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
It looks harder to extend the hardcoded table for sqrhigh efficiently from C, at least if one wants more than a 10% or so improvement. One can try a (n-2)x(n-2) sqrhigh together with a nx1 mul, or perhaps a (n/2) sqr together with a (n/2) mulhigh. In either case the problem is that combining the results is awkward, especially since the mul has to be doubled. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is a work in progress.
Add semi-hardcoded basecases for 10 <= n <= 13 using the n=9 basecase +
mpn_addmul_1+ corrections.Add
_flint_mpn_mulhigh_n_recursivewhich computes exactly the same result as_flint_mpn_mulhigh_basecasebut does so by breaking into smaller muls and mulhighs. With more improvements, I think this approach could obsoletebasecase, but it is not currently strictly faster.Note to Albin: if the assembly
_flint_mpn_mulhigh_basecasewere to assume n >= 10, start by calling the 9x9 subroutine, and update from there, that might obsolete all the above.Change
_flint_mpn_mulhigh_n_mulders_recursiveto add the diagonal corrections so that we compute the more accurate high product. This means that_flint_mpn_mulhigh_n_mulderscan call_flint_mpn_mulhigh_n_mulders_recursivedirectly with lengthn; we no longer need to zero-pad the inputs and work withn + 1.