Re: [PATCH v6] [gdb/mi] Don't treat references to compound values as "simple".

[Gareth Rees <grees@undo.io>]

Tom Tromey wrote:
>
> >>>>> "Gareth" == Gareth Rees via Gdb-patches <gdb-patches@sourceware.org> writes:
>
> Gareth> The '--simple-values' argument to '-stack-list-arguments' and similar
> Gareth> GDB/MI commands does not take reference types into account, so that
> Gareth> references to arbitrarily large structures are considered "simple" and
> Gareth> printed. This means that the '--simple-values' argument cannot be used
> Gareth> by IDEs when tracing the stack due to the time taken to print large
> Gareth> structures passed by reference.
>
> This looks good to me.  I haven't been following the progress of this
> patch -- have the documentation changes been reviewed?  If so, please
> check it in.

Following up again. I don't have commit rights, so I can't check it in
myself. Could you check it in for me? But before you do, I just want
to check that you're happy with the solution I've picked here.

There are two ways to fix the problem of '--simple-values' printing
references to compound values:

First, if it's acceptable to modify the behaviour of the
'--simple-values' option, then we can suppress the printing of
references to compound types, as in this patch. This solution was
preferred by Andrew Burgess [1] [2] [3] out of concern about the
maintenance cost of accumulating backwards-compatible but arguably
broken features. I added an argument that the original behaviour was a
bug, and that it remains hard to use correctly [4].

Second, if we have to preserve the behaviour of the '--simple-values'
option, then we need to add a new option. This solution was preferred
by Eli Zaretskii [5] [6] out of concern about backwards compatibility.

I've proposed a patch that applies the first solution, but I could
work with either. So I need someone to break the tie between Andrew
and Eli. Take a look at the cited messages and let me know if you're
happy, and if you are, then commit the patch.

[1] https://sourceware.org/pipermail/gdb-patches/2022-September/191747.html
[2] https://sourceware.org/pipermail/gdb-patches/2023-March/197736.html
[3] https://sourceware.org/pipermail/gdb-patches/2023-March/197894.html
[4] https://sourceware.org/pipermail/gdb-patches/2023-March/197799.html
[5] https://sourceware.org/pipermail/gdb-patches/2022-September/191754.html
[6] https://sourceware.org/pipermail/gdb-patches/2023-March/197798.html