Re: [PATCH v3 2/2] kdb: Simplify kdb_defcmd macro logic

[Doug Anderson]

Hi,

On Thu, May 13, 2021 at 4:29 AM Sumit Garg <sumit.garg@xxxxxxxxxx> wrote:
>
> Switch to use a linked list instead of dynamic array which makes
> allocation of kdb macro and traversing the kdb macro commands list
> simpler.
>
> Suggested-by: Daniel Thompson <daniel.thompson@xxxxxxxxxx>
> Signed-off-by: Sumit Garg <sumit.garg@xxxxxxxxxx>
> ---
>  kernel/debug/kdb/kdb_main.c | 107 +++++++++++++++++++-----------------
>  1 file changed, 58 insertions(+), 49 deletions(-)
>
> diff --git a/kernel/debug/kdb/kdb_main.c b/kernel/debug/kdb/kdb_main.c
> index de685b2a8ce7..ce7f4c71992d 100644
> --- a/kernel/debug/kdb/kdb_main.c
> +++ b/kernel/debug/kdb/kdb_main.c
> @@ -654,13 +654,16 @@ static void kdb_cmderror(int diag)
>   *     zero for success, a kdb diagnostic if error
>   */
>  struct kdb_macro_t {
> -       int count;
> -       bool usable;
> -       kdbtab_t cmd;
> -       char **command;
> +       kdbtab_t cmd;                   /* Macro command name */

It's more than just the name, right?


> +       struct list_head commands;      /* Associated command list */

I get confused between the two different usages of "command". Can we
call the individual entries of a macro "statements". So this would be:

struct list_head statements; /* Associated statement list */

...and the structure below would get renamed as well.

>  };
> +
> +struct kdb_macro_cmd_t {
> +       char *cmd;                      /* Command name */

This isn't the "name" of the command, is it? It's the actual statement
that the user entered?


Other than that, this looks like a nice patch to me.


-Doug