Fix overflow in type cast for GUPS benchmark #58
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Changed size_t n = (size_t)(1 << logn) to size_t n = ((size_t)1) << logn to prevent integer overflow when logn >= 31. The original code would perform the left shift on an int (1) before casting to size_t, causing overflow. Now the cast happens first, allowing proper handling of large values like n = 2^31. Fixes NVIDIA-developer-blog#57
harrism
approved these changes
Sep 17, 2025
posts/gups/gups.cu
Outdated
| CUDA_RT_CALL(cudaFree(d_t)); | ||
| return 0; | ||
| } | ||
| } |
Member
There was a problem hiding this comment.
💡 suggestion: Add a newline to the end of the file.
Member
There was a problem hiding this comment.
Please fix the EOF, then we can merge.
Using unsigned long literal (1UL) is cleaner and more idiomatic C++ than C-style cast. This achieves the same result of avoiding integer overflow while being more readable.
harrism
reviewed
Sep 18, 2025
- Remove redundant cast on line 462 where n is already size_t - Fix similar pattern on line 393 using 1UL instead of C-style cast
dmikushin
force-pushed
the
fix-gups-overflow
branch
from
be368dd to
f0f1e14
Compare
Contributor
Author
|
@harrism Hm, this is a tricky one 😅 The newline behavior is set by my git somewhere internally. Does it look fine for you if I add a newline manually? |
Member
|
You fixed it. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
This PR fixes an integer overflow bug in the GUPS benchmark that prevents running with problem size n = 2^31.
Changes
size_t n = (size_t)(1 << logn)tosize_t n = ((size_t)1) << logninposts/gups/gups.cu:4611 << lognperforms left shift on anint(1) before casting tosize_t, causing overflow whenlogn >= 31size_thappens first, allowing proper handling of large valuesTesting
This fix is required to benchmark size n = 2^31 as mentioned in issue #57.
Fixes #57