Re: [PATCH v4 1/4] gdb: Introduce setting construct within cmd_list_element

[Lancelot SIX <lsix@lancelotsix.com>]

> 
> I think the change below should fix it.

Hi,

Sorry for this. I did rebase just few days ago and let this one slip
through.  I’ll fix this and address your other comments shortly.

> 
> Since it's one of those
> settings that reverts the value if the new value is invalid, it could
> eventually be changed to use a getter/setter, so that the setter simply
> rejects the new value if it's not valid.

I’ll do that in separate commit once this series is pushed if this ok
with you. For the moment I’ll just fix the current compile error
with something similar to what you propose bellow.

Thanks a lot for the review.

Best,
Lancelot.

> 
> 
> From 96df63a390451e0bbdfc74784ea09a3aa360656d Mon Sep 17 00:00:00 2001
> From: Simon Marchi <simon.marchi@polymtl.ca>
> Date: Thu, 30 Sep 2021 08:26:41 -0400
> Subject: [PATCH] fix
> 
> Change-Id: Id78157cacf2de9cfacba1f19dea17b0f7b65aa0f
> ---
>  gdb/bt-utils.c | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/gdb/bt-utils.c b/gdb/bt-utils.c
> index 79e6e090d428..2054e68c6e5a 100644
> --- a/gdb/bt-utils.c
> +++ b/gdb/bt-utils.c
> @@ -29,15 +29,15 @@ gdb_internal_backtrace_set_cmd (const char *args, int from_tty,
>  				cmd_list_element *c)
>  {
>    gdb_assert (c->type == set_cmd);
> -  gdb_assert (c->var_type == var_boolean);
> -  gdb_assert (c->var != nullptr);
> +  gdb_assert (c->var->type () == var_boolean);
> +  gdb_assert (c->var.has_value ());
>  
>  #ifndef GDB_PRINT_INTERNAL_BACKTRACE
> -  bool *var_ptr = (bool *) c->var;
> +  bool val = c->var->get<bool> ();
>  
> -  if (*var_ptr)
> +  if (val)
>      {
> -      *var_ptr = false;
> +      c->var->set<bool> (false);
>        error (_("support for this feature is not compiled into GDB"));
>      }
>  #endif
> -- 
> 2.33.0
>