4 ms·Finesse: Using LD_PRELOAD for high performance kernel bypass FUSE [pdf]16 points by a-dub 6y ago