search

flatironinstitute / FMM3D

Flatiron Institute Fast Multipole Libraries --- This codebase is a set of libraries to compute N-body interactions governed by the Laplace and Helmholtz equations, to a specified precision, in three dimensions, on a multi-core shared-memory machine.
https://fmm3d.readthedocs.io
Other
100 stars 38 forks source link

integer8 update #19

Closed lu1and10 closed 2 years ago

lu1and10 commented 3 years ago

Change 4-bytes integer to 8-bytes integer for potential large problems.

Passing all the tests except helm3d_mps(Requested precision = 0.50000E-09, Result rel_err = 0.82247E-06).

Successfully completed 5 out of 5 tests in helmrouts3d testing suite Successfully completed 18 out of 18 tests in hfmm3d testing suite Successfully completed 6 out of 6 in hfmm3d scale testing suite Successfully completed 18 out of 18 tests in hfmm3d vec testing suite

Successfully completed 0 out of 1 tests in helm3d_mps testing suite

Successfully completed 5 out of 5 tests in laprouts3d testing suite Successfully completed 27 out of 27 tests in lfmm3d testing suite Successfully completed 2 out of 2 in lfmm3d scale testing suite Successfully completed 27 out of 27 tests in lfmm3d testing suite

Successfully completed 4 out of 4 tests in stokkernels testing suite

Successfully completed 6 out of 6 tests in emfmm3d testing suite

ahbarnett commented 3 years ago

looks like a fun use of sed :) (I had a similar thing with setting up BIGINT in FINUFFT - I was worried that int8 was slower than in4, which it did used to be, but not in recent years...)

On Mon, Dec 14, 2020 at 12:06 PM Libin Lu notifications@github.com wrote:

Change 4-bytes integer to 8-bytes integer for potential large problems.

Passing all the tests except helm3d_mps(Requested precision = 0.50000E-09, Result rel_err = 0.82247E-06).

Successfully completed 5 out of 5 tests in helmrouts3d testing suite Successfully completed 18 out of 18 tests in hfmm3d testing suite Successfully completed 6 out of 6 in hfmm3d scale testing suite Successfully completed 18 out of 18 tests in hfmm3d vec testing suite

Successfully completed 0 out of 1 tests in helm3d_mps testing suite

Successfully completed 5 out of 5 tests in laprouts3d testing suite Successfully completed 27 out of 27 tests in lfmm3d testing suite Successfully completed 2 out of 2 in lfmm3d scale testing suite Successfully completed 27 out of 27 tests in lfmm3d testing suite

Successfully completed 4 out of 4 tests in stokkernels testing suite

Successfully completed 6 out of 6 tests in emfmm3d testing suite

You can view, comment on, or merge this pull request online at:

https://github.com/flatironinstitute/FMM3D/pull/19 Commit Summary

  • integer8 update

File Changes

Patch Links:

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/flatironinstitute/FMM3D/pull/19, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACNZRSWTBRTARLD66KO733DSUZAX7ANCNFSM4U3BKVOA .

-- *---------------------------------------------------------------------~^`^~._.~' |\ Alex H. Barnett Center for Computational Mathematics, Flatiron Institute | \ http://users.flatironinstitute.org/~ahb 646-876-5942

zgimbutas commented 3 years ago

You can also try to use -finteger-4-integer-8 flag to debug your conversion. This flag has been created specifically for this task.

On Dec 14, 2020, at 10:16 AM, Alex Barnett notifications@github.com wrote:

looks like a fun use of sed :) (I had a similar thing with setting up BIGINT in FINUFFT - I was worried that int8 was slower than in4, which it did used to be, but not in recent years...)

On Mon, Dec 14, 2020 at 12:06 PM Libin Lu notifications@github.com wrote:

Change 4-bytes integer to 8-bytes integer for potential large problems.

Passing all the tests except helm3d_mps(Requested precision = 0.50000E-09, Result rel_err = 0.82247E-06).

Successfully completed 5 out of 5 tests in helmrouts3d testing suite Successfully completed 18 out of 18 tests in hfmm3d testing suite Successfully completed 6 out of 6 in hfmm3d scale testing suite Successfully completed 18 out of 18 tests in hfmm3d vec testing suite

Successfully completed 0 out of 1 tests in helm3d_mps testing suite

Successfully completed 5 out of 5 tests in laprouts3d testing suite Successfully completed 27 out of 27 tests in lfmm3d testing suite Successfully completed 2 out of 2 in lfmm3d scale testing suite Successfully completed 27 out of 27 tests in lfmm3d testing suite

Successfully completed 4 out of 4 tests in stokkernels testing suite

Successfully completed 6 out of 6 tests in emfmm3d testing suite

You can view, comment on, or merge this pull request online at:

https://github.com/flatironinstitute/FMM3D/pull/19 Commit Summary

  • integer8 update

File Changes

Patch Links:

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/flatironinstitute/FMM3D/pull/19, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACNZRSWTBRTARLD66KO733DSUZAX7ANCNFSM4U3BKVOA .

-- *---------------------------------------------------------------------~^`^~._.~' |\ Alex H. Barnett Center for Computational Mathematics, Flatiron Institute | \ http://users.flatironinstitute.org/~ahb 646-876-5942 — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/flatironinstitute/FMM3D/pull/19#issuecomment-744584586, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWJ3GR2BCKDDPEGDCBWRWTSUZB5XANCNFSM4U3BKVOA.

lu1and10 commented 2 years ago

close legacy PR, new PR for integer 8 coming.