Skip to content

Performance improvements to fannkuchredux-fast #36

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 6 commits into
base: master
Choose a base branch
from
Open

Performance improvements to fannkuchredux-fast #36

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
dcb04f1
make fannkuchredux-fast slighly faster
hycakir Nov 22, 2018
796712c
fix type issues
hycakir Dec 27, 2018
75bce92
fix type issues
hycakir Dec 27, 2018
88ae8e6
Merge branch 'master' of https://github.com/hycakir/BenchmarksGame.jl
hycakir Dec 28, 2018
89b498c
Merge branch 'master' of https://github.com/KristofferC/BenchmarksGam…
hycakir Feb 1, 2019
8bdec7e
Performance improvements
hycakir Feb 1, 2019
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 fannkuchredux/fannkuchredux-fast.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
# https://salsa.debian.org/benchmarksgame-team/benchmarksgame/

# based on Oleg Mazurov's Java Implementation and Jeremy Zerfas' C implementation
# transliterated by Hamza Yusuf Çakır
# transliterated and modified by Hamza Yusuf Çakır

global const preferred_num_blocks = 24

Expand Down Expand Up @@ -42,7 +42,7 @@ Base.@propagate_inbounds @inline function first_permutation(perm::Perm, idx)
pp = perm.pp

for i = 2:length(p)
p[i] = i - 1
p[i] = (i - 1) % Int8
end

for i = length(p):-1:2
Expand Down Expand Up @@ -102,7 +102,7 @@ Base.@propagate_inbounds @inline function count_flips(perm::Perm)
while true
flips += one(flips)
new_first = pp[first]
pp[first] = first - 1
pp[first] = (first - 1) % Int8

if first > 3
lo = 2; hi = first - 1
Expand Down Expand Up @@ -134,7 +134,7 @@ Base.@propagate_inbounds function run_task(f::Fannkuch, perm::Perm, idxmin, idxm
while true
if perm.p[1] != 0
flips = count_flips(perm)
maxflips = max(maxflips, flips)
(flips > maxflips) && (maxflips = flips)
chksum += iseven(i) ? flips : -flips
end
i != idxmax || break
Expand All @@ -143,7 +143,7 @@ Base.@propagate_inbounds function run_task(f::Fannkuch, perm::Perm, idxmin, idxm
end

id = Threads.threadid()
f.maxflips[id] = max(f.maxflips[id], maxflips)
(maxflips > f.maxflips[id]) && (f.maxflips[id] = maxflips)
f.chksums[id] += chksum
nothing
end
Expand Down Expand Up @@ -172,4 +172,4 @@ function fannkuchredux(n)
end

n = parse(Int, ARGS[1])
fannkuchredux(n)
fannkuchredux(n)