Skip to content

Storage for locals #358

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

Jump to bottom
Closed
wants to merge 2 commits into from
Closed

Storage for locals #358

wants to merge 2 commits into from

Conversation

@chfast
Copy link
Collaborator

@chfast chfast commented May 28, 2020

Does not seem to make big difference, more benchmarks to be done soon...

@codecov
Copy link

codecov bot commented May 29, 2020

Codecov Report

Merging #358 into master will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #358   +/-   ##
=======================================
  Coverage   98.78%   98.78%           
=======================================
  Files          38       38           
  Lines       11255    11257    +2     
=======================================
+ Hits        11118    11120    +2     
  Misses        137      137

@gumb0
Copy link
Collaborator

gumb0 commented Jul 2, 2020

I tried to benchmark this (against PR base), here're the results, not much difference

Benchmark                                                            Time             CPU      Time Old      Time New       CPU Old       CPU New
-------------------------------------------------------------------------------------------------------------------------------------------------
fizzy/execute/blake2b/512_bytes_rounds_1_pvalue                    0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/blake2b/512_bytes_rounds_1_mean                     -0.0157         -0.0156           189           186           189           187
fizzy/execute/blake2b/512_bytes_rounds_1_median                   -0.0132         -0.0131           189           186           189           186
fizzy/execute/blake2b/512_bytes_rounds_1_stddev                   -0.5872         -0.5854             2             1             2             1
fizzy/execute/blake2b/512_bytes_rounds_16_pvalue                   0.0036          0.0046      U Test, Repetitions: 10 vs 10
fizzy/execute/blake2b/512_bytes_rounds_16_mean                    -0.0064         -0.0063          2834          2816          2833          2815
fizzy/execute/blake2b/512_bytes_rounds_16_median                  -0.0056         -0.0055          2829          2813          2828          2813
fizzy/execute/blake2b/512_bytes_rounds_16_stddev                  +0.5525         +0.5487            10            16            10            16
fizzy/execute/ecpairing/onepoint_pvalue                            0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/ecpairing/onepoint_mean                             -0.0660         -0.0661       1117536       1043761       1117455       1043634
fizzy/execute/ecpairing/onepoint_median                           -0.0634         -0.0635       1114297       1043633       1114265       1043548
fizzy/execute/ecpairing/onepoint_stddev                           -0.6124         -0.6142          6032          2338          6020          2323
fizzy/execute/keccak256/512_bytes_rounds_1_pvalue                  0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/keccak256/512_bytes_rounds_1_mean                   +0.0536         +0.0534           211           222           211           222
fizzy/execute/keccak256/512_bytes_rounds_1_median                 +0.0502         +0.0500           211           222           211           222
fizzy/execute/keccak256/512_bytes_rounds_1_stddev                 +0.7042         +0.6736             1             2             1             2
fizzy/execute/keccak256/512_bytes_rounds_16_pvalue                 0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/keccak256/512_bytes_rounds_16_mean                  +0.0782         +0.0782          3031          3268          3031          3267
fizzy/execute/keccak256/512_bytes_rounds_16_median                +0.0745         +0.0745          3038          3265          3038          3265
fizzy/execute/keccak256/512_bytes_rounds_16_stddev                +0.8777         +0.8677            26            49            26            49
fizzy/execute/memset/256_bytes_pvalue                              0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/memset/256_bytes_mean                               -0.0174         -0.0176            17            16            17            16
fizzy/execute/memset/256_bytes_median                             -0.0198         -0.0204            17            16            17            16
fizzy/execute/memset/256_bytes_stddev                             +1.0421         +0.9718             0             0             0             0
fizzy/execute/memset/60000_bytes_pvalue                            0.0140          0.0140      U Test, Repetitions: 10 vs 10
fizzy/execute/memset/60000_bytes_mean                             -0.0110         -0.0110          3503          3464          3503          3464
fizzy/execute/memset/60000_bytes_median                           -0.0133         -0.0134          3500          3453          3499          3453
fizzy/execute/memset/60000_bytes_stddev                           +2.0086         +2.0234             9            26             9            26
fizzy/execute/mul256_opt0/input0_pvalue                            0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/mul256_opt0/input0_mean                             -0.0916         -0.0919            65            59            65            59
fizzy/execute/mul256_opt0/input0_median                           -0.0927         -0.0927            65            59            65            59
fizzy/execute/mul256_opt0/input0_stddev                           -0.1719         -0.2832             0             0             0             0
fizzy/execute/mul256_opt0/input1_pvalue                            0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/mul256_opt0/input1_mean                             -0.0887         -0.0887            65            59            65            59
fizzy/execute/mul256_opt0/input1_median                           -0.0877         -0.0878            65            59            65            59
fizzy/execute/mul256_opt0/input1_stddev                           -0.4546         -0.4514             0             0             0             0
fizzy/execute/sha1/512_bytes_rounds_1_pvalue                       0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/sha1/512_bytes_rounds_1_mean                        -0.0107         -0.0104           194           192           194           192
fizzy/execute/sha1/512_bytes_rounds_1_median                      -0.0101         -0.0099           194           192           194           192
fizzy/execute/sha1/512_bytes_rounds_1_stddev                      +0.1994         +0.2311             0             0             0             0
fizzy/execute/sha1/512_bytes_rounds_16_pvalue                      0.0036          0.0058      U Test, Repetitions: 10 vs 10
fizzy/execute/sha1/512_bytes_rounds_16_mean                       -0.0038         -0.0036          2660          2650          2660          2650
fizzy/execute/sha1/512_bytes_rounds_16_median                     -0.0054         -0.0052          2661          2646          2660          2647
fizzy/execute/sha1/512_bytes_rounds_16_stddev                     +0.9248         +0.9149             7            14             7            14
fizzy/execute/sha256/512_bytes_rounds_1_pvalue                     0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/sha256/512_bytes_rounds_1_mean                      -0.0806         -0.0804           197           181           197           181
fizzy/execute/sha256/512_bytes_rounds_1_median                    -0.0804         -0.0802           197           181           197           181
fizzy/execute/sha256/512_bytes_rounds_1_stddev                    -0.5585         -0.5576             0             0             0             0
fizzy/execute/sha256/512_bytes_rounds_16_pvalue                    0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/sha256/512_bytes_rounds_16_mean                     -0.0851         -0.0850          2667          2440          2667          2440
fizzy/execute/sha256/512_bytes_rounds_16_median                   -0.0851         -0.0850          2668          2441          2667          2441
fizzy/execute/sha256/512_bytes_rounds_16_stddev                   -0.2784         -0.2762             4             3             4             3
fizzy/execute/micro/factorial/10_pvalue                            0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/factorial/10_mean                             +0.0175         +0.0178             3             3             3             3
fizzy/execute/micro/factorial/10_median                           +0.0194         +0.0199             3             3             3             3
fizzy/execute/micro/factorial/10_stddev                           -0.7293         -0.7382             0             0             0             0
fizzy/execute/micro/factorial/20_pvalue                            0.3847          0.1859      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/factorial/20_mean                             +0.0001         +0.0008             5             5             5             5
fizzy/execute/micro/factorial/20_median                           +0.0009         +0.0016             5             5             5             5
fizzy/execute/micro/factorial/20_stddev                           -0.8678         -0.8655             0             0             0             0
fizzy/execute/micro/fibonacci/24_pvalue                            0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/fibonacci/24_mean                             +0.0101         +0.0104         35247         35604         35237         35604
fizzy/execute/micro/fibonacci/24_median                           +0.0093         +0.0095         35232         35558         35224         35558
fizzy/execute/micro/fibonacci/24_stddev                           -0.1711         -0.1403           104            86           100            86
fizzy/execute/micro/host_adler32/1_pvalue                          0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/host_adler32/1_mean                           -0.0695         -0.0691             1             1             1             1
fizzy/execute/micro/host_adler32/1_median                         -0.0711         -0.0707             1             1             1             1
fizzy/execute/micro/host_adler32/1_stddev                         -0.5073         -0.5047             0             0             0             0
fizzy/execute/micro/host_adler32/100_pvalue                        0.6232          0.6232      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/host_adler32/100_mean                         -0.0017         -0.0017            15            15            15            15
fizzy/execute/micro/host_adler32/100_median                       -0.0028         -0.0028            15            15            15            15
fizzy/execute/micro/host_adler32/100_stddev                       -0.0183         -0.0193             0             0             0             0
fizzy/execute/micro/host_adler32/1000_pvalue                       0.9698          0.9698      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/host_adler32/1000_mean                        -0.0014         -0.0014           149           148           149           148
fizzy/execute/micro/host_adler32/1000_median                      -0.0059         -0.0060           149           148           149           148
fizzy/execute/micro/host_adler32/1000_stddev                      -0.2565         -0.2571             2             1             2             1
fizzy/execute/micro/spinner/1_pvalue                               0.0002          0.0004      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/spinner/1_mean                                +0.0086         +0.0084             1             1             1             1
fizzy/execute/micro/spinner/1_median                              +0.0090         +0.0089             1             1             1             1
fizzy/execute/micro/spinner/1_stddev                              -0.5344         -0.5380             0             0             0             0
fizzy/execute/micro/spinner/1000_pvalue                            0.0002          0.0002      U Test, Repetitions: 10 vs 10
fizzy/execute/micro/spinner/1000_mean                             +0.0281         +0.0275            24            24            24            24
fizzy/execute/micro/spinner/1000_median                           +0.0306         +0.0298            24            24            24            24
fizzy/execute/micro/spinner/1000_stddev                           -0.8765         -0.8854             0             0             0             0

@chfast
Copy link
Collaborator Author

chfast commented Jul 6, 2020

I didn't have spectacular results either.

@axic axic mentioned this pull request Jul 16, 2020
Merged
@axic
Copy link
Member

axic commented Jul 16, 2020

What is the story with this?

@axic axic closed this Jul 16, 2020
@axic axic reopened this Jul 16, 2020
@axic
Copy link
Member

axic commented Jul 17, 2020

Wouldn't this be needed if we want to move execute to not have exceptions? I changed it to not throw, but std::vector could still imho.

@axic axic mentioned this pull request Sep 3, 2020
Closed
@chfast
Copy link
Collaborator Author

chfast commented Sep 30, 2020

Similar optimization done in #382.

@chfast chfast closed this Sep 30, 2020
@chfast chfast deleted the locals_storage branch September 30, 2020 13:47
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@chfast @gumb0 @axic