Change Details

- LK macro (conditional on SMP for the lock prefix) is unused - SETLK unnecessarily performs xchg. obtained value is never used and the implicit lock prefix adds avoidable cost. there is no correctness issues here either due to ordering guarantees - waiting for blocked_lock to clear first reads the lock word,pauses unconditionally and only then tests for the lock to see if it needs to loop. i don't know if looping is possible with current code, at least in my tests with poudriere there were no instances of such looping occurring. the least which can be done here is move the unlikely (if possible) case out

- LK macro (conditional on SMP for the lock prefix) is unused - SETLK unnecessarily performs xchg. obtained value is never used and the implicit lock prefix adds avoidable cost. there is no correctness issues here either due to ordering guarantees - waiting for blocked_lock to clear first reads the lock word, pauses unconditionally and only then tests for the lock to see if it needs to loop. i don't know if looping is possible with current code, at least in my tests with poudriere there were no instances of such looping occurring. the least which can be done here is move the unlikely (if possible) case out

- LK macro (conditional on SMP for the lock prefix) is unused - SETLK unnecessarily performs xchg. obtained value is never used and the implicit lock prefix adds avoidable cost. there is no correctness issues here either due to ordering guarantees - waiting for blocked_lock to clear first reads the lock word, pauses unconditionally and only then tests for the lock to see if it needs to loop. i don't know if looping is possible with current code, at least in my tests with poudriere there were no instances of such looping occurring. the least which can be done here is move the unlikely (if possible) case out