memory-model/Documentation/ordering.txt

12 	all of the CPU's prior operations against some or all of its
16 	against some or all of the CPU's prior accesses or some or all
68 accesses against all subsequent accesses from the viewpoint of all CPUs.
119 typically instead used to provide ordering against RCU read-side critical
263 end in _release.  These operations order their own store against all
287 against "y".  On x86, the version using smp_store_release() might compile
307 	Note that release ordering is guaranteed only against the
308 	memory-store portion of the RMW operation, and not against the
322 operations order their own load against all of the CPU's subsequent
333 improves upon smp_rmb() by ordering against subsequent stores as well
334 as against subsequent loads.
344 	Note that acquire ordering is guaranteed only against the
345 	memory-load portion of the RMW operation, and not against the
390 srcu_dereference(), order their load (which must be a pointer) against any