Changes between Version 3 and Version 4 of ConcurrentHashTable

Timestamp:

2012-08-08T11:43:00Z (12 years ago)

Author:

Adam Hraska

Comment:

rcu update overhead benchmark description

ConcurrentHashTable

@@ -127 +127 @@
 chtbench 4 4 0 -w
 }}}
+Then build with Podzimek-RCU and rerun:
+{{{
+chtbench 3 1 0 -w
+chtbench 3 2 0 -w
+chtbench 3 3 0 -w
+chtbench 3 4 0 -w
+}}}
 
+=== RCU write side overhead ===
 [[Image(r1589-list-upd.png)]]
 [[Image(r1589-list-upd-trim.png)]]
+
+Second, we examine the overhead of rcu writers (ie invoking {{{rcu_call()}}} and processing the callback in the background) 
+compared to acquiring a spinlock. The figure above shows the number of traversals 
+or updates of a five element list with an increasing percentage of updates. 
+The benchmark ran in 4 threads/CPUs. In each iteration a thread selected at random
+whether to walk the entire list or to replace an item in the list (ie to update the 
+list). All items were preallocated. More is better ;-).
+- //ideal// - the list was accessed without any synchronization whatsoever on a single cpu;
+  and the result multiplied by the number of cpus (ie 4)
+- //a-rcu + spinlock// - each list traversal and update was protected by A-RCU; concurrent updates
+  were synchronized by means of a spinlock
+- //podzimek-rcu + spinlock// - same as a-rcu but protected by preemptible version of Podzimek's RCU
+- //spinlock// - guarded by an ordinary preemption disabling spinlock
+
+To reproduce these results, switch to the kernel console and run:
+{{{
+chtbench 6 1 0 -w
+chtbench 7 4 0 -w
+chtbench 7 4 5 -w
+chtbench 7 4 10 -w
+chtbench 7 4 20 -w
+chtbench 7 4 30 -w
+chtbench 7 4 40 -w
+chtbench 7 4 60 -w
+chtbench 7 4 100 -w
+chtbench 8 4 0 -w
+chtbench 8 4 5 -w
+chtbench 8 4 10 -w
+chtbench 8 4 20 -w
+chtbench 8 4 30 -w
+chtbench 8 4 40 -w
+chtbench 8 4 60 -w
+chtbench 8 4 100 -w
+}}}
+Then rebuild with Podzimek-RCU and rerun:
+{{{
+chtbench 7 4 0 -w
+chtbench 7 4 5 -w
+chtbench 7 4 10 -w
+chtbench 7 4 20 -w
+chtbench 7 4 30 -w
+chtbench 7 4 40 -w
+chtbench 7 4 60 -w
+chtbench 7 4 100 -w
+}}}
+
+
+=== Hash table lookup scalability ===
 [[Image(r1589-ht-lookup.png)]]
+
+=== Hash table update overhead ===
 [[Image(r1589-ht-upd.png)]]