diff options
| author | Sverker Eriksson <[email protected]> | 2013-01-23 16:46:08 +0100 |
|---|---|---|
| committer | Sverker Eriksson <[email protected]> | 2013-01-23 16:46:08 +0100 |
| commit | b55ba329ba25ba4322efdf16053280a5f7dd1a5d (patch) | |
| tree | 0029b4e0d85ce3ae4990a5cdb42d5353d9f3e756 | |
| parent | 0ab216f3dc96890b50f7430895ad5d7f6251f129 (diff) | |
| download | otp-b55ba329ba25ba4322efdf16053280a5f7dd1a5d.tar.gz otp-b55ba329ba25ba4322efdf16053280a5f7dd1a5d.tar.bz2 otp-b55ba329ba25ba4322efdf16053280a5f7dd1a5d.zip | |
erts,stdlib: Increase number of locks for write_concurrency
| -rw-r--r-- | erts/emulator/beam/erl_db_hash.h | 2 | ||||
| -rw-r--r-- | lib/stdlib/doc/src/ets.xml | 10 |
2 files changed, 8 insertions, 4 deletions
diff --git a/erts/emulator/beam/erl_db_hash.h b/erts/emulator/beam/erl_db_hash.h index cddd8dfadd..73c4078fc5 100644 --- a/erts/emulator/beam/erl_db_hash.h +++ b/erts/emulator/beam/erl_db_hash.h @@ -33,7 +33,7 @@ typedef struct hash_db_term { DbTerm dbterm; /* The actual term */ } HashDbTerm; -#define DB_HASH_LOCK_CNT 16 +#define DB_HASH_LOCK_CNT 64 typedef struct db_table_hash_fine_locks { union { erts_smp_rwmtx_t lck; diff --git a/lib/stdlib/doc/src/ets.xml b/lib/stdlib/doc/src/ets.xml index abaf64fb91..44c050a0d3 100644 --- a/lib/stdlib/doc/src/ets.xml +++ b/lib/stdlib/doc/src/ets.xml @@ -932,7 +932,8 @@ ets:select(Table,MatchSpec),</code> If set to <c>true</c>, 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 sequential access and concurrent reader performance. + at the expense of memory consumption and the performance of + sequential access and concurrent reading. The <c>write_concurrency</c> option can be combined with the <seealso marker="#new_2_read_concurrency">read_concurrency</seealso> option. You typically want to combine these when large concurrent @@ -944,8 +945,11 @@ ets:select(Table,MatchSpec),</code> <seealso marker="#concurrency">atomicy and isolation</seealso>. Functions that makes such promises over several objects (like <c>insert/2</c>) will gain less (or nothing) from this option.</p> - <p>Table type <c>ordered_set</c> is not affected by this option in current - implementation.</p> + <p>In current implementation, table type <c>ordered_set</c> is not + affected by this option. Also, the memory consumption inflicted by + both <c>write_concurrency</c> and <c>read_concurrency</c> is a + constant overhead per table. This overhead can be especially large + when both options are combined.</p> </item> <item> <marker id="new_2_read_concurrency"></marker> |