yeah, it was fun! it felt a bit like a sidequest, and to be honest, i'm happer at the fact that i have the publication under my belt than i was happy at the experience of doing the work that resulted in it.
i dunno if i told you about it, but you might actually be interested (it's about determinism, but not formal logic) -- here's the paper; my summary is that we (the Columbia folks) built a determinizing userspace thread scheduler for pthread programs, and added annotations to it which let you sacrifice determinism in small regions to get some performance back, and then we (Jiri and me) hooked up a systematic tester to verify the small nondeterministic sections that resulted
( ... )
Comments 4
(Also, why do people pronounce each individual letter in SOSP? It's totally a pronounceable word!)
Reply
i dunno if i told you about it, but you might actually be interested (it's about determinism, but not formal logic) -- here's the paper; my summary is that we (the Columbia folks) built a determinizing userspace thread scheduler for pthread programs, and added annotations to it which let you sacrifice determinism in small regions to get some performance back, and then we (Jiri and me) hooked up a systematic tester to verify the small nondeterministic sections that resulted ( ... )
Reply
Reply
Reply
Leave a comment