id	summary	reporter	owner	description	type	status	milestone	component	version	severity	resolution	keywords	cc
11798	Implementation of boost::shared_mutex on POSIX is suboptimal	alex@…	viboes	"The current (as of boost 1.59) implementation of boost::shared_mutex for 'pthread' is pretty suboptimal as it's using a heavyweight mutex to guard the internal mutex state. This is more evident when shared locks are used to guard state whose access concurrency is high, due to contention on the mutex state lock (in these cases, the shared mutex is effectively exclusive). In comparison, the 'win32' implementation uses lightweight spinlocks underneath.

There are a couple options to fix this for 'pthread', e.g. using a spinlock or leveraging pthreads_rwlock. I'm happy to provide with an initial patch for this."	Tasks	assigned	To Be Determined	thread	Boost 1.59.0	Optimization		pthread shared_mutex performance concurrency spinlock	Andrey.Semashev@…