Messages in this thread Patch in this message | | | From | "Paul E. McKenney" <> | Subject | [PATCH tip/core/rcu 22/27] rcu: Make rcu_seq_diff() more exact | Date | Mon, 25 Jun 2018 17:35:08 -0700 |
| |
The current implementatation of rcu_seq_diff() follows tradition in providing a rough-and-ready approximation of the number of elapsed grace periods between the two rcu_seq values. However, this difference is used to flag RCU-failure "near misses", which can be a valuable debugging aid, so more exactitude would be an improvement. This commit therefore improves the accuracy of rcu_seq_diff().
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> --- kernel/rcu/rcu.h | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h index 60f089d08c47..46f12d0fe319 100644 --- a/kernel/rcu/rcu.h +++ b/kernel/rcu/rcu.h @@ -180,7 +180,20 @@ static inline bool rcu_seq_new_gp(unsigned long old, unsigned long new) */ static inline unsigned long rcu_seq_diff(unsigned long new, unsigned long old) { - return (new - old) >> RCU_SEQ_CTR_SHIFT; + unsigned long rnd_diff; + + if (old == new) + return 0; + /* + * Compute the number of grace periods (still shifted up), plus + * one if either of new and old is not an exact grace period. + */ + rnd_diff = (new & ~RCU_SEQ_STATE_MASK) - + ((old + RCU_SEQ_STATE_MASK) & ~RCU_SEQ_STATE_MASK) + + ((new & RCU_SEQ_STATE_MASK) || (old & RCU_SEQ_STATE_MASK)); + if (ULONG_CMP_GE(RCU_SEQ_STATE_MASK, rnd_diff)) + return 1; /* Definitely no grace period has elapsed. */ + return ((rnd_diff - RCU_SEQ_STATE_MASK - 1) >> RCU_SEQ_CTR_SHIFT) + 2; } /* -- 2.17.1
| |