Skip to content

power uses Float64 exponents for integers#53967

Merged
oscardssmith merged 25 commits intoJuliaLang:masterfrom
KlausC:krc/pow-float-path
Apr 22, 2024
Merged

power uses Float64 exponents for integers#53967
oscardssmith merged 25 commits intoJuliaLang:masterfrom
KlausC:krc/pow-float-path

Conversation

@KlausC
Copy link
Copy Markdown
Contributor

@KlausC KlausC commented Apr 5, 2024 •
edited
Loading

Improve performance of ^(::Float64, n::Integer) in the case of abs(n) > 2^13.

While pow_body is unreliable for abs(n) > 2^25 this implementation provides errors of a few ULPs, while runtime is capped to that of the Float64 implementation.

Fixes #53881
See also #53886.

@dkarrasch dkarrasch added the maths Mathematical functions label Apr 5, 2024
@KlausC
Copy link
Copy Markdown
Contributor Author

KlausC commented Apr 6, 2024

@oscardssmith I would be interested in your opinion!

@oscardssmith
Copy link
Copy Markdown
Member

oscardssmith commented Apr 6, 2024

this looks really good to me. I need to spend a little time benchmarking to make sure all the details look good, but I'm theory, this looks great!

@KlausC KlausC marked this pull request as draft April 9, 2024 09:05
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
oscardssmith
oscardssmith reviewed Apr 12, 2024
View reviewed changes
@KlausC
Copy link
Copy Markdown
Contributor Author

KlausC commented Apr 13, 2024 •
edited
Loading

Results of benchmark tests:
The abscissa is sign(n) * log2(abs(n) for integer exponents n.
The ordinate is in time per x^n for x close to but > 1.0.
The first graph show execution times in ns for ´x^n´ power_by_squaring, the second one for n -> exp(log(x)*n).
Hardware: Intel® Core™ i7-3610QM × 8

grafik

The result of the merged version with parameters nmin = -2^12, nmax = 3*2^13:

grafik

@KlausC

This comment was marked as off-topic.

Sign in to view
@KlausC KlausC marked this pull request as ready for review April 13, 2024 11:44
@oscardssmith oscardssmith mentioned this pull request Feb 19, 2025
Closed
@oscardssmith oscardssmith added bugfix This change fixes an existing bug backport 1.10 Change should be backported to the 1.10 release backport 1.11 Change should be backported to release-1.11 labels Feb 19, 2025
@oscardssmith oscardssmith mentioned this pull request Feb 21, 2025
Closed
KristofferC pushed a commit that referenced this pull request Mar 11, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
71ec30c 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
This was referenced Mar 11, 2025
Merged
Merged
KristofferC pushed a commit that referenced this pull request Mar 25, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
237ad1a 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
@KristofferC
Copy link
Copy Markdown
Member

KristofferC commented Apr 10, 2025

Causes a test error on 1.11, won't backport.

@KristofferC KristofferC removed the backport 1.11 Change should be backported to release-1.11 label Apr 10, 2025
KristofferC added a commit that referenced this pull request Apr 10, 2025
@KristofferC
Revert "power uses Float64 exponents for integers (#53967)"
8063884 
This reverts commit 237ad1a.
@KlausC KlausC restored the krc/pow-float-path branch April 10, 2025 13:03
@KlausC
Copy link
Copy Markdown
Contributor Author

KlausC commented Apr 10, 2025

Causes a test error on 1.11, won't backport.

@KristofferC is there a chance to look at the failed test output? I could not verify that locally.

@KlausC KlausC mentioned this pull request Apr 11, 2025
Open
@KristofferC
Copy link
Copy Markdown
Member

KristofferC commented Apr 11, 2025

It is only on 32-bit. #58062 might fix it. But it is too late for 1.11.5, it can go into 1.11.6 if it is fixed with the earlier-mentioned PR.

The error was https://buildkite.com/julialang/julia-release-1-dot-11/builds/451#019619be-f75a-400d-b019-9d047c239bd3


Test Failed at /cache/build/tester-amdci5-15/julialang/julia-release-1-dot-11/julia-88568a8784/share/julia/test/math.jl:1600
--
  | Expression: Core.Compiler.is_foldable(effects)
  | Context: effects = (+c,+e,+n,!t,+s,+m,+u,+o,+r)
  | T = Float64

KristofferC pushed a commit that referenced this pull request Jun 5, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
b9c867c 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
KristofferC pushed a commit that referenced this pull request Jun 5, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
560797e 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
KristofferC pushed a commit that referenced this pull request Jun 5, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
fe073c4 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
KristofferC added a commit that referenced this pull request Jun 5, 2025
@KristofferC
Revert "power uses Float64 exponents for integers (#53967)"
cbc56b0 
This reverts commit fe073c4.
KristofferC pushed a commit that referenced this pull request Jun 5, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
4b0ea7b 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
KristofferC added a commit that referenced this pull request Jun 5, 2025
@KristofferC
Revert "power uses Float64 exponents for integers (#53967)"
86a6f68 
This reverts commit fe073c4.
KristofferC pushed a commit that referenced this pull request Jun 5, 2025
@KlausC @KristofferC
power uses Float64 exponents for integers (#53967)
ee8dd0f 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
KristofferC added a commit that referenced this pull request Jun 5, 2025
@KristofferC
Revert "power uses Float64 exponents for integers (#53967)"
7f6ffc7 
This reverts commit fe073c4.
KristofferC pushed a commit that referenced this pull request Jul 3, 2025
@KlausC
power uses Float64 exponents for integers (#53967)
7d55587 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

(cherry picked from commit fe49d56)
KristofferC added a commit that referenced this pull request Jul 3, 2025
@KristofferC
Revert "power uses Float64 exponents for integers (#53967)"
4cf9373 
This reverts commit fe073c4.
@DilumAluthge DilumAluthge mentioned this pull request Sep 13, 2025
Merged
71 tasks
@github-actions github-actions bot removed the backport 1.10 Change should be backported to the 1.10 release label Jan 18, 2026
@adienes adienes mentioned this pull request Jan 28, 2026
Merged
DilumAluthge pushed a commit that referenced this pull request Jan 28, 2026
@KlausC @DilumAluthge
power uses Float64 exponents for integers (#53967)
5982ccd 
Improve performance of `^(::Float64, n::Integer)` in the case of `abs(n)
> 2^13`.

While `pow_body` is unreliable for `abs(n) > 2^25` this implementation
provides errors of a few ULPs, while runtime is capped to that of the
`Float64` implementation.

Fixes  #53881
See also #53886.

Note: This was originally backported to 1.10 as
ee8dd0f,
but then reverted in 7f6ffc7

(cherry picked from commit fe49d56)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@oscardssmith oscardssmith oscardssmith approved these changes

Assignees

No one assigned

Labels

bugfix This change fixes an existing bug maths Mathematical functions

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

BUG: ^(::Float64, ::Union{Int, Float64}) incorrect for very large negative exponents

4 participants

@KlausC @oscardssmith @KristofferC @dkarrasch