Re: [patchv2 8/11] TUI: source "file" -> "fullname"

[Doug Evans <dje@google.com>]

On Sun, Jan 27, 2013 at 2:36 PM, Jan Kratochvil
<jan.kratochvil@redhat.com> wrote:
> Hi,
>
> this is a new part in this v2 series.
>
> It was tested current "gdb -tui" is really incorrect and it gets fixed by this
> patch:
>         cd gdb.linespec/
>         cd base/one/
>         g++ -o ../../1.o -c thefile.cc -Wall -g
>         cd ../two
>         g++ -o ../../2.o -c thefile.cc -Wall -g
>         cd  ../..
>         g++ -o 0 lspec.cc -Wall -g 1.o 2.o
>         ../../gdb-tuiN -tui ./0
>         ../../gdb-tuiY -tui ./0
>         b m
>         b n
>         r
>         c
> With this patch GDB displays the first thefile.cc for m() and the second
> thefile.cc for n(); without this patch it forgot to display n().
>
> It expects that if !ui_out_is_mi_like_p and !ui_source_list it is really TUI
> and no other uiout interface; I hope it is safe assumption and there will be
> no new interfaces besides current CLI + TUI + MI.

Re: "no new interfaces": Perhaps not immediately relevant here, but I
think it's something we need to keep in mind.
I gather people are thinking of putting GUI-like things on top of gdb
via the Python API.
So it's really CLI + TUI + MI + potentially-Python.