NVIDIA / NVIDIA/cuda-samples

Incorrect padding for the data source of spPostprocess2D in convolutionFFT2D.

Open Beginner friendly
#195 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
C++
Stars
9.6k
Forks
2.4k
Avg merge
53m
Merged PRs (30d)
1

Description

In "convolutionFFT2D.cu":

147 extern "C" void spPostprocess2D(void *d_Dst, void *d_Src, uint DY, uint DX,
148 uint padding, int dir) {
...
162 #if (USE_TEXTURE)
163 cudaTextureObject_t texComplex;
164 cudaResourceDesc texRes;
165 memset(&texRes, 0, sizeof(cudaResourceDesc));
166
167 texRes.resType = cudaResourceTypeLinear;
168 texRes.res.linear.devPtr = d_Src;
169 texRes.res.linear.sizeInBytes = sizeof(fComplex) * DY * (DX + padding);

Here the texture size based on "d_Src" should be "sizeof(fComplex) * DY * DX" because neither the data source passed from "test1()" in "main.cpp" nor the data access in kernel "spPostprocess2D_kernel" has padding.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start in convolutionFFT2D.cu at spPostprocess2D, around lines 147-169, and compare the texture resource size with the source allocation in test1() in main.cpp and accesses in spPostprocess2D_kernel. Verify that the reported linear texture size matches the actual d_Src layout, then run the convolutionFFT2D sample to confirm the postprocessing path still works.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
hpc
Issue type
Bug
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
65/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.