Replace calls to time(NULL) that occur on the order of once per read, one per write, or once per cell with calls to a function that looks at a cached value of time. This is tricksy to benchmark, since it will only help on systems where time() is a syscall and syscalls are relatively slow.

svn:r17690

1 files changed, 3 insertions, 0 deletions

diff --git a/ChangeLog b/ChangeLog
index 933ed102d5..56efc89102 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -43,6 +43,9 @@ Changes in version 0.2.1.9-alpha - 2008-12-2?
     - On platforms with a maximum syslog string length, truncate syslog
       messages to that length ourselves, rather than relying on the
       system to do it for us.
+    - Optimize out calls to time(NULL) that occur for every IO operation, or
+      for every cell. On systems where time() is a slow syscalls, this will
+      be slightly helpful.
 
   o Minor features (controller):
     - New CONSENSUS_ARRIVED event to note when a new consensus has