Spaces:

ejschwartz
/

cuobjdump

Running

App Files Files Community

cuobjdump / README.md

Edward J. Schwartz

Route old-format CUDA ELF to nvdisasm 12.9 so CFGs work

efc2358 22 days ago

preview code

Raw History Blame Contribute Delete

686 Bytes

A newer version of the Gradio SDK is available: 6.31.0

Upgrade

metadata

title: cuobjdump / nvdisasm
emoji: 🔍
colorFrom: green
colorTo: gray
sdk: gradio
sdk_version: 6.29.1
app_file: app.py
pinned: false
short_description: Disassemble CUDA cubins with cuobjdump and nvdisasm

Upload a CUDA cubin, fatbin or ELF object and run NVIDIA's cuobjdump or nvdisasm on it. Both tools are installed from PyPI (nvidia-cuda-cuobjdump, nvidia-cuda-nvdisasm); no GPU required. nvdisasm 13.x cannot build control flow graphs for old-format CUDA ELF (pre-CUDA 13, OSABI 0x33), so those files are routed to nvdisasm 12.9, downloaded from NVIDIA's redist site at startup.

examples/ holds two cubins assembled with ptxas from the included PTX sources.