Skip to content

Add more multiplication primitives #107

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 25 commits into
base: master
Choose a base branch
from
Open

Add more multiplication primitives #107

Changes from 1 commit
Commits
Show all changes
25 commits
Select commit Hold shift + click to select a range
b4db29e
add main logic
jamierpond Jan 13, 2025
1261006
add even lane mask
jamierpond Jan 13, 2025
f976ae4
even/odd lane mask
jamierpond Jan 13, 2025
f0720bd
clean up a little
jamierpond Jan 13, 2025
f928811
clean some more
jamierpond Jan 13, 2025
51f2987
just return overflow
jamierpond Jan 13, 2025
35fffa7
wow seems to be working
jamierpond Jan 14, 2025
b87b408
tidy a little
jamierpond Jan 14, 2025
07be9f9
rm spurious tests
jamierpond Jan 14, 2025
93e4bac
rename
jamierpond Jan 14, 2025
05468a2
start test refactor
jamierpond Jan 15, 2025
ac45f1b
tidy tests
jamierpond Jan 16, 2025
2302504
rm tests
jamierpond Jan 16, 2025
2214ac8
add consume msb
jamierpond Jan 16, 2025
61a7506
rename lower/upper
jamierpond Jan 16, 2025
7b41db0
make doubling multi nicer
jamierpond Jan 16, 2025
5cf88df
consolidate exponentation and make naming consistent
jamierpond Jan 17, 2025
3a65ed2
works
jamierpond Jan 17, 2025
2b613ee
mv tests
jamierpond Feb 25, 2025
fa0667b
oops
jamierpond Feb 25, 2025
dea354a
tidy
jamierpond Feb 25, 2025
81237a6
tidy
jamierpond Feb 25, 2025
b792e30
tidy
jamierpond Feb 25, 2025
5d41262
make pair for generatlity
jamierpond Feb 25, 2025
cfb1072
tidy
jamierpond Feb 25, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 6 additions & 6 deletions inc/zoo/swar/associative_iteration.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -474,8 +474,8 @@ constexpr auto halvePrecision(SWAR<NB, T> even, SWAR<NB, T> odd) {


template <int NB, typename T> struct MultiplicationResult {
SWAR<NB, T> result;
SWAR<NB, T> overflow;
SWAR<NB, T> lower;
SWAR<NB, T> upper;
Comment on lines +482 to +483
Copy link
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

merge

};

template <int NB, typename T>
Expand All @@ -497,11 +497,11 @@ wideningMultiplication(SWAR<NB, T> multiplicand, SWAR<NB, T> multiplier) {
constexpr auto
HalfLane = S::NBits,
UpperHalfOfLanes = SWAR<S::NBits, T>::oddLaneMask().value();
auto [res_even, res_odd] = doublingMultiplication(multiplicand, multiplier);
auto result = halvePrecision(res_even, res_odd);
auto [lower, upper] = doublingMultiplication(multiplicand, multiplier);
auto result = halvePrecision(lower, upper);
auto
over_even = D{(res_even.value() & UpperHalfOfLanes) >> HalfLane},
over_odd = D{(res_odd.value() & UpperHalfOfLanes) >> HalfLane};
over_even = D{(lower.value() & UpperHalfOfLanes) >> HalfLane},
over_odd = D{(upper.value() & UpperHalfOfLanes) >> HalfLane};
Copy link
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

shift intra lane allows you to provide the mask.
Please use those primitives instead of deploying the pick-axe

auto upper_lanes_overflow = halvePrecision(over_even, over_odd);
return {result, upper_lanes_overflow};
}
Expand Down
Loading