Incorrect output for CUDA version? about sw4lite HOT 3 CLOSED

geodynamics commented on June 2, 2024

Incorrect output for CUDA version?

from sw4lite.

Comments (3)

andersp commented on June 2, 2024

That result indicates that the code is broken. /Anders From: René van Oostrum <[email protected]> Reply-To: geodynamics/sw4lite <[email protected]> Date: Thursday, June 7, 2018 at 6:56 AM To: geodynamics/sw4lite <[email protected]> Cc: Subscribed <[email protected]> Subject: [geodynamics/sw4lite] Incorrect output for CUDA version? (#12) When building the latest CUDA version, and running it with the "pointsource.in" example input, as described in README.md, the expected output is Errors at time 0.6 Linf = 0.569416 L2 = 0.0245361 norm of solution = 3.7439 Instead, I'm getting Errors at time 0.6 Linf = 3.7439 L2 = 1.01524 norm of solution = 3.7439 (I.e., Linf and L2 are different, norm of solution is identical). I interpret the statement in README.md "When modifying the code, it is important to verify that these numbers have not changed" to say that all three values should match, and they don't. Does this indicate an error in the code? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#12>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AIgih_tRCmhkxhuxIMSPJpR5dxHcSHgsks5t6TDrgaJpZM4UedXI>.

from sw4lite.

rwvo commented on June 2, 2024

I manage to correctly build the CPU version from the latest commit from 2017-12-17 (commit 5ab806), and with a minor fix, the GPU version too. Both give correct output for pointsource.in.

Yesterday's commit (merging the pull request) broke both the GPU version (incorrect output for pointsource.in), and the CPU version, which no longer builds; CPU code is calling GPU functions that are undefined in the CPU version:

../src/EW.C: In member function ‘void EW::timesteploop(std::vector<Sarray>&, std::vector<Sarray>&)’:                   
../src/EW.C:2602:43: error: ‘pack_HaloArrayCU_X’ was not declared in this scope                                        
            pack_HaloArrayCU_X (Up[g], g, 0);
                                           ^
../src/EW.C:2604:45: error: ‘unpack_HaloArrayCU_X’ was not declared in this scope                                      
            unpack_HaloArrayCU_X (Up[g], g, 0);
                                             ^
../src/EW.C:2605:43: error: ‘pack_HaloArrayCU_Y’ was not declared in this scope                                        
            pack_HaloArrayCU_Y (Up[g], g, 0);
                                           ^
../src/EW.C:2607:45: error: ‘unpack_HaloArrayCU_Y’ was not declared in this scope                                      
            unpack_HaloArrayCU_Y (Up[g], g, 0);
                                             ^
../src/EW.C:2609:32: error: ‘cudaDeviceSynchronize’ was not declared in this scope                                     
          cudaDeviceSynchronize();
                                ^
../src/EW.C:2645:32: error: ‘cudaDeviceSynchronize’ was not declared in this scope                                     
          cudaDeviceSynchronize();
                                ^
../src/EW.C:2730:43: error: ‘pack_HaloArrayCU_X’ was not declared in this scope                                        
            pack_HaloArrayCU_X (Up[g], g, 0);
                                           ^
../src/EW.C:2732:45: error: ‘unpack_HaloArrayCU_X’ was not declared in this scope                                      
            unpack_HaloArrayCU_X (Up[g], g, 0);
                                             ^
../src/EW.C:2733:43: error: ‘pack_HaloArrayCU_Y’ was not declared in this scope                                        
            pack_HaloArrayCU_Y (Up[g], g, 0);
                                           ^
../src/EW.C:2735:45: error: ‘unpack_HaloArrayCU_Y’ was not declared in this scope                                      
            unpack_HaloArrayCU_Y (Up[g], g, 0);
                                             ^

from sw4lite.

rwvo commented on June 2, 2024

Fixed in commit c52471, Thu Jun 7 20:38:56 2018 -0700:
Both CPU and GPU versions compile, and run with correct output for test input pointsource.in.

from sw4lite.

Incorrect output for CUDA version? about sw4lite HOT 3 CLOSED

Comments (3)

Related Issues (5)

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent