Rewrite ethread library

Large parts of the ethread library have been rewritten. The
ethread library is an Erlang runtime system internal, portable
thread library used by the runtime system itself.

Most notable improvement is a reader optimized rwlock
implementation which dramatically improve the performance of
read-lock/read-unlock operations on multi processor systems by
avoiding ping-ponging of the rwlock cache lines. The reader
optimized rwlock implementation is used by miscellaneous
rwlocks in the runtime system that are known to be read-locked
frequently, and can be enabled on ETS tables by passing the
`{read_concurrency, true}' option upon table creation. See the
documentation of `ets:new/2' for more information.

The ethread library can now also use the libatomic_ops library
for atomic memory accesses. This makes it possible for the
Erlang runtime system to utilize optimized atomic operations
on more platforms than before. Use the
`--with-libatomic_ops=PATH' configure command line argument
when specifying where the libatomic_ops installation is
located. The libatomic_ops library can be downloaded from:
http://www.hpl.hp.com/research/linux/atomic_ops/

The changed API of the ethread library has also caused
modifications in the Erlang runtime system. Preparations for
the to come "delayed deallocation" feature has also been done
since it depends on the ethread library.

Note: When building for x86, the ethread library will now use
instructions that first appeared on the pentium 4 processor. If
you want the runtime system to be compatible with older
processors (back to 486) you need to pass the
`--enable-ethread-pre-pentium4-compatibility' configure command
line argument when configuring the system.

1 files changed, 18 insertions, 0 deletions

diff --git a/erts/doc/src/erl.xml b/erts/doc/src/erl.xml
index df80142ce1..0e26d62548 100644
--- a/erts/doc/src/erl.xml
+++ b/erts/doc/src/erl.xml
@@ -603,6 +603,24 @@
       <item>
         <p>Force ets memory block to be moved on realloc.</p>
       </item>
+      <tag><marker id="+rg"><c><![CDATA[+rg ReaderGroupsLimit]]></c></marker></tag>
+      <item>
+        <p>Limits the amount of reader groups used by read/write locks
+	   optimized for read operations in the Erlang runtime system. By
+	   default the reader groups limit equals 8.</p>
+	<p>When the amount of schedulers is less than or equal to the reader
+	   groups limit, each scheduler has its own reader group. When the
+	   amount of schedulers is larger than the reader groups limit,
+	   schedulers share reader groups. Shared reader groups degrades
+	   read lock and read unlock performance while a large amount of
+	   reader groups degrades write lock performance, so the limit is a
+	   tradeoff between performance for read operations and performance
+	   for write operations. Each reader group currently consumes 64 byte
+	   in each read/write lock. Also note that a runtime system using
+	   shared reader groups benefits from <seealso marker="#+sbt">binding
+	   schedulers to logical processors</seealso>, since the reader groups
+	   are distributed better between schedulers.</p>
+      </item>
       <tag><marker id="+S"><c><![CDATA[+S Schedulers:SchedulerOnline]]></c></marker></tag>
       <item>
         <p>Sets the amount of scheduler threads to create and scheduler