Re: [MI] Duplicate --thread-group flag not detected

[Joel Brobecker <brobecker@adacore.com>]

> You make it hard to say no :-)

:-D

> So here's my attempt at new tests for these cases.  I wasn't sure
> where to put it so I created a new mi-general.exp test which should
> eventually test the parsing of general MI syntax, including the
> --thread-group, --thread, --frame, --all, --reverse, flags.

OK. I propose we treat this patch as separate. Please just go ahead
with the code part, while we review the testing part.

> The problem is that there is another bug that makes the new tests
> fail, so I had to temporarily comment them out:
> http://sourceware.org/ml/gdb-patches/2010-11/msg00436.html

Let's not comment them out, but mark them KFAIL instead, if we can.


> 2010-12-05  Marc Khouzam  <marc.khouzam@ericsson.com>
> 
> 	* gdb.mi/mi-general.exp: New file.

Overal, the test seems fine, but I have limited experimence with
MI testcases.

> +set testfile "basics"
> +set srcfile ${testfile}.c
> +set binfile ${objdir}/${subdir}/mi-basics

I believe we decided in the past that it was a bad idea to share the same
example program between testcases. One reason is that, if we split the
gdb.mi testcase in multiple groups to allow more parallel testing, we
might run into troubles.  However, I am just saying that for your info.
It's not your problem, and I know that you have little time for this
testcase, and I already appreciate the effort you put into it. So it's
fine to leave it as is.

> +    mi_gdb_test "18-break-insert --thread-group i1 bogus" \
> +             "18\\^error,msg=\"Function \\\\\"bogus\\\\\" not defined.\"" \
> +             "Valid --thread-group flag"

Do we want to test the MI command with a sequence number. I know that
they are allowed, but aren't they obsolete?

Other than that, I don't have any other recommendation.

-- 
Joel