Re: [mi] organize possible exec async mi oc command reasons

[Daniel Jacobowitz <drow@false.org>]

On Thu, Mar 24, 2005 at 10:46:02AM -0500, Bob Rossi wrote:
> Hi,
> 
> I am going to attempt to organize the possible MI exec async output
> commands. The current problem is that it is not easy to understand the
> possible number of 'exec async output commands' that GDB/MI can return
> to the FE. I would like to simply this, and eventually add a testcase
> for each case. This is a start at enumerating all of the case's and
> giving the user a minimal amount of information.
> 
> The design is to have mi-common.h be the interface to the rest of 
> GDB. Also, MI developers can have a place to look at to determine all
> possible supported command types.
> 
> Eli, if you can think of a better way for me to add to the doco, let me
> know.
> 
> BTW, someone please tell me if the testsuite diff is OK, for some reason
> there are minor differences between a fresh checkout and my patched
> tree, although, I think it has nothing to do with my changes.
> 
> I've attached the added files and the testsuite diff. The patch is
> below.

I don't have much comment about the patch; it seems like a plausible
idea.  However, please pay attention to the coding and formatting
standards:

- ChangeLog entries are capitalized and end with periods.
- Function definitions have the function name in the first column.
- Function braces are in the first column.
- There's a space before function argument lists.
- Function argument lists have to be indented normally; an argument
  on a new line goes at the same depth as the first argument on the
  previous line.
- Comments don't use multiple leading *s.

and so forth.

I sometimes consider myself the pedantic guardian of GDB's source code
formatting :-)

-- 
Daniel Jacobowitz
CodeSourcery, LLC