kernel: Add os:cmd/2 with max_size option

git cherry-pick 55e929c4ed5cd854038c18697123ea94948ebf35

3 files changed, 80 insertions, 18 deletions

diff --git a/lib/kernel/doc/src/os.xml b/lib/kernel/doc/src/os.xml
index 739ac35d2a..a6a63448b7 100644
--- a/lib/kernel/doc/src/os.xml
+++ b/lib/kernel/doc/src/os.xml
@@ -38,17 +38,35 @@
       most platforms.</p>
   </description>
 
+  <datatypes>
+    <datatype>
+      <name name="os_command"/>
+    </datatype>
+    <datatype>
+      <name name="os_command_opts"/>
+      <desc>
+        <p>Options for <seealso marker="#cmd/2"><c>os:cmd/2</c></seealso></p>
+        <taglist>
+          <tag><c>max_size</c></tag>
+          <item>
+            <p>The maximum size of the data returned by the <c>os:cmd</c> call.
+            See the <seealso marker="#cmd/2"><c>os:cmd/2</c></seealso>
+            documentation for more details.</p>
+          </item>
+        </taglist>
+      </desc>
+    </datatype>
+  </datatypes>
+
   <funcs>
     <func>
       <name name="cmd" arity="1"/>
+      <name name="cmd" arity="2"/>
       <fsummary>Execute a command in a shell of the target OS.</fsummary>
       <desc>
         <p>Executes <c><anno>Command</anno></c> in a command shell of the
-          target OS,
-          captures the standard output of the command, and returns this
-          result as a string. This function is a replacement of
-          the previous function <c>unix:cmd/1</c>; they are equivalent on a
-          Unix platform.</p>
+          target OS, captures the standard output of the command,
+          and returns this result as a string.</p>
         <p><em>Examples:</em></p>
         <code type="none">
 LsOut = os:cmd("ls"), % on unix platform
@@ -57,6 +75,21 @@ DirOut = os:cmd("dir"), % on Win32 platform</code>
           called from another program (for example, <c>os:cmd/1</c>)
           can differ, compared with the standard output of the command
           when called directly from an OS command shell.</p>
+        <p><c>os:cmd/2</c> was added in kernel-5.5 (OTP-20.2.1). It makes it
+          possible to pass an options map as the second argument in order to
+          control the behaviour of <c>os:cmd</c>. The possible options are:
+        </p>
+        <taglist>
+          <tag><c>max_size</c></tag>
+          <item>
+            <p>The maximum size of the data returned by the <c>os:cmd</c> call.
+              This option is a safety feature that should be used when the command
+              executed can return a very large, possibly infinite, result.</p>
+            <code type="none">
+> os:cmd("cat /dev/zero", #{ max_size => 20 }).
+[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0]</code>
+          </item>
+        </taglist>
       </desc>
     </func>
 
diff --git a/lib/kernel/src/os.erl b/lib/kernel/src/os.erl
index a90a6dcca7..dbb28c1210 100644
--- a/lib/kernel/src/os.erl
+++ b/lib/kernel/src/os.erl
@@ -21,7 +21,7 @@
 
 %% Provides a common operating system interface.
 
--export([type/0, version/0, cmd/1, find_executable/1, find_executable/2]).
+-export([type/0, version/0, cmd/1, cmd/2, find_executable/1, find_executable/2]).
 
 -include("file.hrl").
 
@@ -32,6 +32,11 @@
          putenv/2, system_time/0, system_time/1,
 	 timestamp/0, unsetenv/1]).
 
+-type os_command() :: atom() | io_lib:chars().
+-type os_command_opts() :: #{ max_size => non_neg_integer() | infinity }.
+
+-export_type([os_command/0, os_command_opts/0]).
+
 -spec getenv() -> [string()].
 
 getenv() -> erlang:nif_error(undef).
@@ -223,15 +228,21 @@ extensions() ->
 
 %% Executes the given command in the default shell for the operating system.
 -spec cmd(Command) -> string() when
-      Command :: atom() | io_lib:chars().
+      Command :: os_command().
 cmd(Cmd) ->
+    cmd(Cmd, #{ }).
+
+-spec cmd(Command, Options) -> string() when
+      Command :: os_command(),
+      Options :: os_command_opts().
+cmd(Cmd, Opts) ->
     validate(Cmd),
     {SpawnCmd, SpawnOpts, SpawnInput, Eot} = mk_cmd(os:type(), Cmd),
     Port = open_port({spawn, SpawnCmd}, [binary, stderr_to_stdout,
                                          stream, in, hide | SpawnOpts]),
     MonRef = erlang:monitor(port, Port),
     true = port_command(Port, SpawnInput),
-    Bytes = get_data(Port, MonRef, Eot, []),
+    Bytes = get_data(Port, MonRef, Eot, [], 0, maps:get(max_size, Opts, infinity)),
     demonitor(MonRef, [flush]),
     String = unicode:characters_to_list(Bytes),
     if  %% Convert to unicode list if possible otherwise return bytes
@@ -282,12 +293,13 @@ validate1([List|Rest]) when is_list(List) ->
 validate1([]) ->
     ok.
 
-get_data(Port, MonRef, Eot, Sofar) ->
+get_data(Port, MonRef, Eot, Sofar, Size, Max) ->
     receive
 	{Port, {data, Bytes}} ->
-            case eot(Bytes, Eot) of
+            case eot(Bytes, Eot, Size, Max) of
                 more ->
-                    get_data(Port, MonRef, Eot, [Sofar,Bytes]);
+                    get_data(Port, MonRef, Eot, [Sofar, Bytes],
+                             Size + byte_size(Bytes), Max);
                 Last ->
                     catch port_close(Port),
                     flush_until_down(Port, MonRef),
@@ -298,13 +310,16 @@ get_data(Port, MonRef, Eot, Sofar) ->
 	    iolist_to_binary(Sofar)
     end.
 
-eot(_Bs, <<>>) ->
+eot(_Bs, <<>>, _Size, _Max) ->
     more;
-eot(Bs, Eot) ->
+eot(Bs, Eot, Size, Max) ->
     case binary:match(Bs, Eot) of
-        nomatch -> more;
-        {Pos, _} ->
-            binary:part(Bs,{0, Pos})
+        nomatch when Size + byte_size(Bs) < Max ->
+            more;
+        {Pos, _} when Size + Pos < Max ->
+            binary:part(Bs,{0, Pos});
+        _ ->
+            binary:part(Bs,{0, Max - Size})
     end.
 
 %% When port_close returns we know that all the
diff --git a/lib/kernel/test/os_SUITE.erl b/lib/kernel/test/os_SUITE.erl
index e76d6ec482..70752e6ee6 100644
--- a/lib/kernel/test/os_SUITE.erl
+++ b/lib/kernel/test/os_SUITE.erl
@@ -25,7 +25,8 @@
 -export([space_in_cwd/1, quoting/1, cmd_unicode/1, space_in_name/1, bad_command/1,
 	 find_executable/1, unix_comment_in_command/1, deep_list_command/1,
          large_output_command/1, background_command/0, background_command/1,
-         message_leak/1, close_stdin/0, close_stdin/1, perf_counter_api/1]).
+         message_leak/1, close_stdin/0, close_stdin/1, max_size_command/1,
+         perf_counter_api/1]).
 
 -include_lib("common_test/include/ct.hrl").
 
@@ -37,7 +38,7 @@ all() ->
     [space_in_cwd, quoting, cmd_unicode, space_in_name, bad_command,
      find_executable, unix_comment_in_command, deep_list_command,
      large_output_command, background_command, message_leak,
-     close_stdin, perf_counter_api].
+     close_stdin, max_size_command, perf_counter_api].
 
 groups() ->
     [].
@@ -312,6 +313,19 @@ close_stdin(Config) ->
 
     "-1" = os:cmd(Fds).
 
+max_size_command(_Config) ->
+
+    Res20 = os:cmd("cat /dev/zero", #{ max_size => 20 }),
+    20 = length(Res20),
+
+    Res0 = os:cmd("cat /dev/zero", #{ max_size => 0 }),
+    0 = length(Res0),
+
+    Res32768 = os:cmd("cat /dev/zero", #{ max_size => 32768 }),
+    32768 = length(Res32768),
+
+    ResHello = os:cmd("echo hello", #{ max_size => 20 }),
+    6 = length(ResHello).
 
 %% Test that the os:perf_counter api works as expected
 perf_counter_api(_Config) ->