Re: RFA: Make cli-out follow gdb_stdout

[Andrew Cagney <ac131313 at ges dot redhat dot com>]

> This mail is followed by several others in which Andrew and Daniel argue about
> the proposed patch...
>
> Andrew says
>  >>So the assertion:
>  >>
>  >>      global uiout->stream->ui_file == global gdb_stdout
>  >>
>  >>doesn't hold :-(
>
> But I think there is a confusion at that point
> (at least after reading  of the thread, I was not really sure that 
> it was clear).
> My impression is that the current way the cli_out_new works is not correct 
> because if you change gdb_stdout, this change is not reflected in 
> the uiout global var. Thus the above assertion 
> does not hold in the current old code, but if I understood Daneil's patch correctly
> that patch should have solved that issue.
This is sitting in my in-tray.  Yes, the assertion doesn't hold.

>    Thus I am very disappointed by the final outcome of this thread...
> I really tought that Daniel's idea was the way to go,
> as uiout is not known in all source files, but gdb_stdout is.
>
> Also, I didn't understand the intent of Andrew concerning this issue.
> Andrew, could you try to reformulate your point of view?
There is always a tradeoff between allowing an immediate ``quick fix'' 
and seeking a better medium to long term solution.  cf:

/* Keep structure return address in this variable.  FIXME:  This is a 
horrid kludge which should not be allowed to continue living.  This only 
allows a single nested call to a structure-returning function.  Come on, 
guys!  -- gnu@cygnus.com, Aug 92  */

static CORE_ADDR rs6000_struct_return_address;

One immediate thing to fix would be catch_exceptions() which is breaking 
its contract -- it at least should keep the uiout and uifile in sync.  I 
think, if that is fixed, then I think it will be possible to fix the 
problems you've encountered.

enjoy,
Andrew