diff options
Diffstat (limited to 'lib/tools')
| -rw-r--r-- | lib/tools/doc/src/cover.xml | 29 | ||||
| -rw-r--r-- | lib/tools/doc/src/cover_chapter.xml | 7 | ||||
| -rw-r--r-- | lib/tools/src/cover.erl | 34 |
3 files changed, 60 insertions, 10 deletions
diff --git a/lib/tools/doc/src/cover.xml b/lib/tools/doc/src/cover.xml index 323bd0dda8..0a3302bda5 100644 --- a/lib/tools/doc/src/cover.xml +++ b/lib/tools/doc/src/cover.xml @@ -270,6 +270,8 @@ defaults to <c>function</c>.</p> <p>If <c>Module</c> is not Cover compiled, the function returns <c>{error,{not_cover_compiled,Module}}</c>.</p> + <p>HINT: It is possible to issue multiple analyse_to_file commands at + the same time. </p> </desc> </func> <func> @@ -307,6 +309,33 @@ <c>.beam</c> file, or in <c>../src</c> relative to that directory. If no source code is found, <c>,{error,no_source_code_found}</c> is returned.</p> + <p>HINT: It is possible to issue multiple analyse_to_file commands at + the same time. </p> + </desc> + </func> + <func> + <name>async_analyse_to_file(Module) -> </name> + <name>async_analyse_to_file(Module,Options) -> </name> + <name>async_analyse_to_file(Module, OutFile) -> </name> + <name>async_analyse_to_file(Module, OutFile, Options) -> pid()</name> + <fsummary>Asynchronous call to analyse_to_file.</fsummary> + <type> + <v>Module = atom()</v> + <v>OutFile = string()</v> + <v>Options = [Option]</v> + <v>Option = html</v> + <v>Error = {not_cover_compiled,Module} | {file,File,Reason} | no_source_code_found | not_main_node</v> + <v> File = string()</v> + <v> Reason = term()</v> + </type> + <desc> + <p>This function works exactly the same way as + <seealso marker="#analyse_to_file-1">analyse_to_file</seealso> except + that it is asynchronous instead of synchronous. The spawned process + will link with the caller when created. If an <c>Error</c> occurs + while doing the cover analysis the process will crash with the same + error reason as <seealso marker="#analyse_to_file-1">analyse_to_file</seealso> + would return.</p> </desc> </func> <func> diff --git a/lib/tools/doc/src/cover_chapter.xml b/lib/tools/doc/src/cover_chapter.xml index b4f7919183..92a790c34e 100644 --- a/lib/tools/doc/src/cover_chapter.xml +++ b/lib/tools/doc/src/cover_chapter.xml @@ -403,6 +403,13 @@ ok database contains information about each executable line in each Cover compiled module, performance decreases proportionally to the size and number of the Cover compiled modules.</p> + <p>To improve performance when analysing cover results it is possible + to do multiple calls to <seealso marker="cover#analyse-1">analyse</seealso> + and <seealso marker="cover#analyse_to_file-1">analyse_to_file</seealso> + at once. You can also use the + <seealso marker="cover#async_analyse_to_file-1">async_analyse_to_file</seealso> + convenience function. + </p> </section> <section> diff --git a/lib/tools/src/cover.erl b/lib/tools/src/cover.erl index 50a812aa09..cc4f75f2e8 100644 --- a/lib/tools/src/cover.erl +++ b/lib/tools/src/cover.erl @@ -35,23 +35,37 @@ %% remote_process_loop/1. %% %% TABLES -%% Each nodes has an ets table named 'cover_internal_data_table' -%% (?COVER_TABLE). This table contains the coverage data and is -%% continously updated when cover compiled code is executed. +%% Each nodes has two tables: cover_internal_data_table (?COVER_TABLE) and. +%% cover_internal_clause_table (?COVER_CLAUSE_TABLE). +%% ?COVER_TABLE contains the bump data i.e. the data about which lines +%% have been executed how many times. +%% ?COVER_CLAUSE_TABLE contains information about which clauses in which modules +%% cover is currently collecting statistics. %% -%% The main node owns a table named -%% 'cover_collected_remote_data_table' (?COLLECTION_TABLE). This table -%% contains data which is collected from remote nodes (either when a -%% remote node is stopped with cover:stop/1 or when analysing. When -%% analysing, data is even moved from the ?COVER_TABLE on the main -%% node to the ?COLLECTION_TABLE. +%% The main node owns tables named +%% 'cover_collected_remote_data_table' (?COLLECTION_TABLE) and +%% 'cover_collected_remote_clause_table' (?COLLECTION_CLAUSE_TABLE). +%% These tables contain data which is collected from remote nodes (either when a +%% remote node is stopped with cover:stop/1 or when analysing). When +%% analysing, data is even moved from the COVER tables on the main +%% node to the COLLECTION tables. %% %% The main node also has a table named 'cover_binary_code_table' %% (?BINARY_TABLE). This table contains the binary code for each cover %% compiled module. This is necessary so that the code can be loaded %% on remote nodes that are started after the compilation. %% - +%% PARELLALISM +%% To take advantage of SMP when doing the cover analysis both the data +%% collection and analysis has been parallelized. One process is spawned for +%% each node when collecting data, and on the remote node when collecting data +%% one process is spawned per module. +%% +%% When analyzing data it is possible to issue multiple analyse(_to_file)/X +%% calls at once. They are however all calls (for backwardscompatability +%% reasons) so the user of cover will have to spawn several processes to to the +%% calls ( or use async_analyse_to_file ). +%% %% External exports -export([start/0, start/1, |