Re: [PATCH 03/25] Class-fy tdesc_reg tdesc_type and tdesc_feature

[Simon Marchi <simon.marchi@polymtl.ca>]

On 2017-06-12 10:41, Yao Qi wrote:
> This patch class-fies them, adding ctor, dtor, and deleting
> copy ctor and assignment operator.
> 
> gdb:
> 
> 2017-05-20  Yao Qi  <yao.qi@linaro.org>
> 
> 	* target-descriptions.c (tdesc_reg): Add ctor, dtor.
> 	Delete copy ctor and assignment operator.
> 	(tdesc_type): Likewise.
> 	(tdesc_feature): Likewise.
> 	(tdesc_free_reg): Remove.
> 	(tdesc_create_reg): Use new.
> 	(tdesc_free_type): Remove.
> 	(tdesc_create_vector): Use new.
> 	(tdesc_create_union): Likewise.
> 	(tdesc_create_flags): Likewise.
> 	(tdesc_create_enum): Likewise.
> 	(tdesc_free_feature): Delete.
> 	(free_target_description): Use delete.
> ---
>  gdb/target-descriptions.c | 200 
> +++++++++++++++++++++++-----------------------
>  1 file changed, 100 insertions(+), 100 deletions(-)
> 
> diff --git a/gdb/target-descriptions.c b/gdb/target-descriptions.c
> index 9a7e2dd..e2dcd1d 100644
> --- a/gdb/target-descriptions.c
> +++ b/gdb/target-descriptions.c
> @@ -48,6 +48,32 @@ DEF_VEC_O(property_s);
> 
>  typedef struct tdesc_reg
>  {
> +public:

Since you are using the "struct" keyword, members are public by default, 
so you don't have to put "public:".  Same things for other types.

Otherwise it LGTM.  Of course, more c++ification is always possible, but 
I think it's fine to go in small increments.

Just FYI, I tested this on top of my "poison-xnew" branch, which points 
out the invalid use of XNEW/xfree on non-POD types, and it built fine.

Simon