summaryrefslogtreecommitdiff
path: root/include/linux/compiler.h
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2015-06-22 14:01:01 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2015-06-22 14:01:01 -0700
commitfc934d40178ad4e551a17e2733241d9f29fddd70 (patch)
tree43e3f3c9019915cc41befec4dad5d1328a4cae63 /include/linux/compiler.h
parent052b398a43a7de8c68c13e7fa05d6b3d16ce6801 (diff)
parent085c789783f5f95c253740d3d0a1659ec5ff32a8 (diff)
Merge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull RCU updates from Ingo Molnar: - Continued initialization/Kconfig updates: hide most Kconfig options from unsuspecting users. There's now a single high level configuration option: * * RCU Subsystem * Make expert-level adjustments to RCU configuration (RCU_EXPERT) [N/y/?] (NEW) Which if answered in the negative, leaves us with a single interactive configuration option: Offload RCU callback processing from boot-selected CPUs (RCU_NOCB_CPU) [N/y/?] (NEW) All the rest of the RCU options are configured automatically. Later on we'll remove this single leftover configuration option as well. - Remove all uses of RCU-protected array indexes: replace the rcu_[access|dereference]_index_check() APIs with READ_ONCE() and rcu_lockdep_assert() - RCU CPU-hotplug cleanups - Updates to Tiny RCU: a race fix and further code shrinkage. - RCU torture-testing updates: fixes, speedups, cleanups and documentation updates. - Miscellaneous fixes - Documentation updates * 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (60 commits) rcutorture: Allow repetition factors in Kconfig-fragment lists rcutorture: Display "make oldconfig" errors rcutorture: Update TREE_RCU-kconfig.txt rcutorture: Make rcutorture scripts force RCU_EXPERT rcutorture: Update configuration fragments for rcutree.rcu_fanout_exact rcutorture: TASKS_RCU set directly, so don't explicitly set it rcutorture: Test SRCU cleanup code path rcutorture: Replace barriers with smp_store_release() and smp_load_acquire() locktorture: Change longdelay_us to longdelay_ms rcutorture: Allow negative values of nreaders to oversubscribe rcutorture: Exchange TREE03 and TREE08 NR_CPUS, speed up CPU hotplug rcutorture: Exchange TREE03 and TREE04 geometries locktorture: fix deadlock in 'rw_lock_irq' type rcu: Correctly handle non-empty Tiny RCU callback list with none ready rcutorture: Test both RCU-sched and RCU-bh for Tiny RCU rcu: Further shrink Tiny RCU by making empty functions static inlines rcu: Conditionally compile RCU's eqs warnings rcu: Remove prompt for RCU implementation rcu: Make RCU able to tolerate undefined CONFIG_RCU_KTHREAD_PRIO rcu: Make RCU able to tolerate undefined CONFIG_RCU_FANOUT_LEAF ...
Diffstat (limited to 'include/linux/compiler.h')
-rw-r--r--include/linux/compiler.h16
1 files changed, 16 insertions, 0 deletions
diff --git a/include/linux/compiler.h b/include/linux/compiler.h
index 867722591be2..5d66777914db 100644
--- a/include/linux/compiler.h
+++ b/include/linux/compiler.h
@@ -252,6 +252,22 @@ static __always_inline void __write_once_size(volatile void *p, void *res, int s
#define WRITE_ONCE(x, val) \
({ typeof(x) __val = (val); __write_once_size(&(x), &__val, sizeof(__val)); __val; })
+/**
+ * READ_ONCE_CTRL - Read a value heading a control dependency
+ * @x: The value to be read, heading the control dependency
+ *
+ * Control dependencies are tricky. See Documentation/memory-barriers.txt
+ * for important information on how to use them. Note that in many cases,
+ * use of smp_load_acquire() will be much simpler. Control dependencies
+ * should be avoided except on the hottest of hotpaths.
+ */
+#define READ_ONCE_CTRL(x) \
+({ \
+ typeof(x) __val = READ_ONCE(x); \
+ smp_read_barrier_depends(); /* Enforce control dependency. */ \
+ __val; \
+})
+
#endif /* __KERNEL__ */
#endif /* __ASSEMBLY__ */