Re: Strange change between bash-3.2 and bash-4.0

[Roman Rakus]

Roman Rakus wrote:
> Hi
> there is simple script:
> #!/bin/sh
> echo foo|cat /nosuchfile
> echo "here: $?"
>
> executed by sh -e
> In bash-3.2 we got:
> cat: /nosuchfile: No such file or directory
>
> but in bash-4.0:
> cat: /nosuchfile: No such file or directory
> here: 1
>
>
> From execute_cmd.c:
>
> /* 10/6/2008 -- added test for pipe_in and pipe_out because they indicate
> the presence of a pipeline, and (until Posix changes things), a
> pipeline failure should not cause the parent shell to exit on an
> unsuccessful return status, even in the presence of errexit.. */
>
> I'm not sure what Posix says
>
> but it breaks the:
> -e  Exit immediately if a command exits with a non-zero status.
>
> For reference see https://bugzilla.redhat.com/show_bug.cgi?id=483385
>
> The fix:
> diff -up bash-4.0-rc1/execute_cmd.c.err bash-4.0-rc1/execute_cmd.c
> --- bash-4.0-rc1/execute_cmd.c.err 2009-01-31 20:05:38.000000000 +0200
> +++ bash-4.0-rc1/execute_cmd.c 2009-01-31 20:05:46.000000000 +0200
> @@ -764,7 +764,7 @@ execute_command_internal (command, async
>
> if (ignore_return == 0 && invert == 0 &&
> ((posixly_correct && interactive == 0 && special_builtin_failed) ||
> - (exit_immediately_on_error && pipe_in == NO_PIPE && pipe_out == 
> NO_PIPE && exec_result != EXECUTION_SUCCESS)))
> + (exit_immediately_on_error && exec_result != EXECUTION_SUCCESS)))
> {
> last_command_exit_value = exec_result;
> run_pending_traps ();
>
>
> RR
If this stays as it is, bash will never exit immediately if a command 
(in this case pipeline) exits with a non-zero status.
RR