libsmctrl.git - Library to enable intra-context SM/TPC partitioning on NVIDIA GPUs

Branch	Commit message	Author	Age
ecrts25-ae	Support installation on Linux4Tegra	Joshua Bakita	8 weeks
master	De-duplicate CUDA version checks and omit when building with CUDA > 6.5	Joshua Bakita	4 months


Age	Commit message	Author
2025-07-03	Support installation on Linux4Tegraecrts25-ae	Joshua Bakita
2025-06-25	Speed up nvtaskset by skipping CUDA context creation if possible	Joshua Bakita
2025-06-17	Remove dependency on patchelf and use /usr/lib instead of /lib for install	Joshua Bakita
2025-06-17	Rewrite nvtaskset and implementation of partitioning for unmodified tasks	Joshua Bakita
2025-05-20	Correct how path of libsmctrl.so is specified in LD_PRELOAD by nvtaskset	Joshua Bakita
2025-05-09	Major update for ECRTS'25: fix TPC to GPU mapping and add a "supreme" mask	Joshua Bakita
2025-05-05	De-duplicate CUDA version checks and omit when building with CUDA > 6.5HEAD master	Joshua Bakita
2025-04-07	Support stream masking on CUDA 12.7 (x86) and 12.8 (x86)	Joshua Bakita
2025-03-20	Makefile improvements	Joshua Bakita
2024-12-23	Correctly pass arguments to Python's ctypes find_library() function	Joshua Bakita
[...]

Clone
http://rtsrv.cs.unc.edu/cgit/cgit.cgi/libsmctrl.git
or
ssh://rtsrv.cs.unc.edu/public/libsmctrl.git