rcu: Increase rcutorture test coverage
authorPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Wed, 12 Jun 2013 22:12:21 +0000 (15:12 -0700)
committerPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Tue, 20 Aug 2013 18:38:41 +0000 (11:38 -0700)
commit2ec1f2d98752293f4831ce7d7bdbc3fc36bdd114
treec1dc25a48d0bfac1381879587986578bc5100d0b
parentd2818df168b2c80c7449e47bd349094c308fa323
rcu: Increase rcutorture test coverage

Currently, rcutorture has separate torture_types to test synchronous,
asynchronous, and expedited grace-period primitives.  This has
two disadvantages: (1) Three times the number of runs to cover the
combinations and (2) Little testing of concurrent combinations of the
three options.  This commit therefore adds a pair of module parameters
that control normal and expedited state, with the default being both
types, randomly selected, by the fakewriter processes, thus reducing
source-code size and increasing test coverage.  In addtion, the writer
task switches between asynchronous-normal and expedited grace-period
primitives driven by the same pair of module parameters.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Documentation/RCU/torture.txt
kernel/rcutorture.c