Smaller argparsh binaries
Making argparsh smaller and 3x faster by optimizing binary size!
exploring the world with 1s and 0s
Welcome to my website! I’m a second year PhD Student at UT Austin, advised by Chris Rossbach. My topics of interest are Operating System Design, Systems for ML, and compilers.
All words on this blog my own (and not AI generated) unless specified :)
Interning at Microsoft Research to work on optimizing systems for RL post-training!
Began my PhD at the University of Texas at Austin!
Received an Honorable Mention for my NSF GRFP application
Making argparsh smaller and 3x faster by optimizing binary size!
argparsh allow shell scripts to have better argument parsing. However, it requires many repeated invocations of the argparsh process - how can we reduce the overhead of spawning argparsh processes? In this post we make argparsh 3.46x faster by using static linking and nostd.
Linux lacks a mechanism for injecting a kernel-controlled thread into a userspace process. Recently, I implemented a hacky solution (involving process brain surgery) and wanted to write about it. Implementation code can be found at https://github.com/aneeshdurg/atexec.
see all my posts here
clang and CypherMPI operations from being called during another collective operationls, cat, hexdump and many more!