weiji14
cog3pio
BlogDocsChangelog

Performance History

Latest Results

:beers: Symlink nvtiff.h to targets/x86_64-linux/lib and set RUSTFLAGS Brute force symlinking to get nvtiff-sys to compile with conda-forge's libnvtiff that is under $CONDA_PREFIX/include/ instead of $CONDA_PREFIX/targets/x86_64-linux/include/ where most other header files are. One key part is to use RUSTFLAGS instead of LD_LIBRARY_PATH to actually get rustc to search the correct lib/ folder for the .so files.
dlpack_to_cupy
5 days ago
:construction_worker: Track Rust performance benchmarks on CI (#61) * :arrow_up: Bump criterion from 0.5.1 to 0.7.0 Bumps [criterion](https://github.com/bheisler/criterion.rs) from 0.5.1 to 0.7.0. - [Changelog](https://github.com/bheisler/criterion.rs/blob/master/CHANGELOG.md) - [Commits](https://github.com/bheisler/criterion.rs/compare/0.5.1...0.7.0) No need to use the criterion.rs fork set at f07ae721237c5ab107d0b5429f8fa349cd16a072/#27, because the original error `assertion failed: slice.len() > 1 && slice.iter().all(|x| !x.is_nan())` can been fixed by setting nresamples to 2, yeah I know right! * :construction_worker: Run criterion benchmarks on CI with cargo-codspeed Use codspeed-criterion-compat crate inplace of criterion, and setup CI job to run the Rust benchmarks! * :green_heart: Run on ghcr.io/osgeo/gdal:ubuntu-small-3.11.5 container And bump to georust/gdal version with GDAL 3.11 bindings. * :rotating_light: Set git config core.longpaths to true Fix `failed to update submodule `gdal-src/source` ... Caused by: path too long: 'C:/Users/runneradmin/.cargo/git/checkouts/gdal-3b51095beae5b4fc/3df2c28/gdal-src/source/...`. * :green_heart: Cargo remove gdal-src and gdal-sys in benchmark tests Fix `undefined reference to `GDALGetRasterBand'` and various other linker errors. Just rely on existing GDAL installation within the docker container, rather than the pre-built bindings. * :alembic: Benchmark with LiberTIFF driver Benchmark reading with GDAL 3.11's LiberTIFF driver, using 4 threads. Also renumbered the benchmark IDs to start from 0. * :egg: Do host to device copy when 'cuda' flag is enabled Ensure fair comparison between nvTIFF which reads to CUDA memory (even though it's the fastest already), and the GDAL LiberTIFF and image-tiff backend readers.
main
6 days ago

Active Branches

:sparkles: Python bindings for CudaCogReader
last run
5 days ago
#58
CodSpeed Performance Gauge
0%
© 2025 CodSpeed Technology
Home Terms Privacy Docs