From 300b419486c1ca88e33938f182d5d5a8b90fb73f Mon Sep 17 00:00:00 2001 From: Rickard Green Date: Thu, 17 Jun 2010 10:23:50 +0200 Subject: 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. --- lib/stdlib/doc/src/ets.xml | 34 +++++++++++++++++++++++++++-- lib/stdlib/test/ets_SUITE.erl | 50 ++++++++++++++++++++++++++++++++++++++----- 2 files changed, 77 insertions(+), 7 deletions(-) (limited to 'lib/stdlib') diff --git a/lib/stdlib/doc/src/ets.xml b/lib/stdlib/doc/src/ets.xml index ee1befc882..5df60a92e5 100644 --- a/lib/stdlib/doc/src/ets.xml +++ b/lib/stdlib/doc/src/ets.xml @@ -1039,15 +1039,22 @@ ets:select(Table,MatchSpec), the owner terminates.

+

{write_concurrency,bool()} - Performance tuning. Default is false, which means that the table - is optimized towards concurrent read access. An operation that + Performance tuning. Default is false. An operation that mutates (writes to) the table will obtain exclusive access, blocking any concurrent access of the same table until finished. If set to true, the table is optimized towards concurrent write access. Different objects of the same table can be mutated (and read) by concurrent processes. This is achieved to some degree at the expense of single access and concurrent reader performance. + The write_concurrency option can be combined with the + read_concurrency + option. You typically want to combine these when large concurrent + read bursts and large concurrent write bursts are common (see the + documentation of the + read_concurrency + option for more information). Note that this option does not change any guarantees about atomicy and isolation. Functions that makes such promises over several objects (like @@ -1055,6 +1062,29 @@ ets:select(Table,MatchSpec),

Table type ordered_set is not affected by this option in current implementation.

 + + +

{read_concurrency,bool()} + Performance tuning. Default is false. When set to + true, the table is optimized for concurrent read + operations. When this option is enabled on a runtime system with + SMP support, read operations become much cheaper; especially on + systems with multiple physical processors. However, switching + between read and write operations becomes more expensive. You + typically want to enable this option when concurrent read + operations are much more frequent than write operations, or when + concurrent reads and writes comes in large read and write + bursts (i.e., lots of reads not interrupted by writes, and lots + of writes not interrupted by reads). You typically do + not want to enable this option when the common access + pattern is a few read operations interleaved with a few write + operations repeatedly. In this case you will get a performance + degradation by enabling this option. The read_concurrency + option can be combined with the + write_concurrency + option. You typically want to combine these when large concurrent + read bursts and large concurrent write bursts are common.

+ diff --git a/lib/stdlib/test/ets_SUITE.erl b/lib/stdlib/test/ets_SUITE.erl index 13c87ca005..1f68c6b8c4 100644 --- a/lib/stdlib/test/ets_SUITE.erl +++ b/lib/stdlib/test/ets_SUITE.erl @@ -33,7 +33,7 @@ -export([misc/1, dups/1, misc1/1, safe_fixtable/1, info/1, tab2list/1]). -export([files/1, tab2file/1, tab2file2/1, tab2file3/1, tabfile_ext1/1, tabfile_ext2/1, tabfile_ext3/1, tabfile_ext4/1]). --export([heavy/1, heavy_lookup/1, heavy_lookup_element/1]). +-export([heavy/1, heavy_lookup/1, heavy_lookup_element/1, heavy_concurrent/1]). -export([lookup_element/1, lookup_element_mult/1]). -export([fold/1]). -export([foldl_ordered/1, foldr_ordered/1, foldl/1, foldr/1, fold_empty/1]). @@ -89,7 +89,8 @@ match_delete_do/1, match_delete3_do/1, firstnext_do/1, slot_do/1, match1_do/1, match2_do/1, match_object_do/1, match_object2_do/1, misc1_do/1, safe_fixtable_do/1, info_do/1, dups_do/1, heavy_lookup_do/1, - heavy_lookup_element_do/1, member_do/1, otp_5340_do/1, otp_7665_do/1, meta_wb_do/1 + heavy_lookup_element_do/1, member_do/1, otp_5340_do/1, otp_7665_do/1, meta_wb_do/1, + do_heavy_concurrent/1 ]). -include("test_server.hrl"). @@ -3877,7 +3878,7 @@ make_sub_binary(List, Num) when is_list(List) -> {_,B} = split_binary(Bin, N+1), B. -heavy(suite) -> [heavy_lookup, heavy_lookup_element]. +heavy(suite) -> [heavy_lookup, heavy_lookup_element, heavy_concurrent]. %% Lookup stuff like crazy... heavy_lookup(doc) -> ["Performs multiple lookups for every key ", @@ -3940,6 +3941,44 @@ do_lookup_element(Tab, N, M) -> end. +heavy_concurrent(Config) -> + repeat_for_opts(do_heavy_concurrent). + +do_heavy_concurrent(Opts) -> + ?line Size = 20000, + ?line EtsMem = etsmem(), + ?line Tab = ets:new(blupp, [set, public, {keypos, 2} | Opts]), + ?line ok = fill_tab2(Tab, 0, Size), + ?line Procs = lists:map( + fun (N) -> + spawn_link( + fun () -> + do_heavy_concurrent_proc(Tab, Size, N) + end) + end, + lists:seq(1, 500)), + ?line lists:foreach(fun (P) -> + M = erlang:monitor(process, P), + receive + {'DOWN', Mon, process, P, _} -> + ok + end + end, + Procs), + ?line true = ets:delete(Tab), + ?line verify_etsmem(EtsMem). + +do_heavy_concurrent_proc(_Tab, 0, _Offs) -> + done; +do_heavy_concurrent_proc(Tab, N, Offs) when (N+Offs) rem 100 == 0 -> + Data = {"here", are, "S O M E ", data, "toooooooooooooooooo", insert, + make_ref(), make_ref(), make_ref()}, + true=ets:insert(Tab, {{self(),Data}, N}), + do_heavy_concurrent_proc(Tab, N-1, Offs); +do_heavy_concurrent_proc(Tab, N, Offs) -> + _ = ets:lookup(Tab, N), + do_heavy_concurrent_proc(Tab, N-1, Offs). + fold(suite) -> [foldl_ordered, foldr_ordered, foldl, foldr, fold_empty]. @@ -5336,7 +5375,7 @@ only_if_smp(Schedulers, Func) -> %% Repeat test function with different combination of table options %% repeat_for_opts(F) -> - repeat_for_opts(F, [write_concurrency]). + repeat_for_opts(F, [write_concurrency, read_concurrency]). repeat_for_opts(F, OptGenList) when is_atom(F) -> repeat_for_opts(fun(Opts) -> ?MODULE:F(Opts) end, OptGenList); @@ -5356,6 +5395,7 @@ repeat_for_opts(F, [Atom | Tail], AccList) when is_atom(Atom) -> repeat_for_opts(F, [repeat_for_opts_atom2list(Atom) | Tail ], AccList). repeat_for_opts_atom2list(all_types) -> [set,ordered_set,bag,duplicate_bag]; -repeat_for_opts_atom2list(write_concurrency) -> [{write_concurrency,false},{write_concurrency,true}]. +repeat_for_opts_atom2list(write_concurrency) -> [{write_concurrency,false},{write_concurrency,true}]; +repeat_for_opts_atom2list(read_concurrency) -> [{read_concurrency,false},{read_concurrency,true}]. -- cgit v1.2.3