Re: [PATCH 3/3] gdb/record: minor clean, remove some unneeded arguments

[Keith Seitz <keiths@redhat.com>]

On 4/15/24 07:19, Andrew Burgess wrote:
> I spotted that the two functions:
> 
>    record_full_open_1
>    record_full_core_open_1
> 
> both took two arguments, neither of which are used.

Nice catch. I've looked over this series, and I just have one really
trivial -- if not silly -- comment.

> diff --git a/gdb/record-full.c b/gdb/record-full.c
> index 4c3667f48ba..4ef85b0cdb6 100644
> --- a/gdb/record-full.c
> +++ b/gdb/record-full.c
> @@ -957,11 +957,14 @@ static void record_full_init_record_breakpoints (void);
>   /* Open the process record target.  */
>   
>   static void
> -record_full_open (const char *name, int from_tty)
> +record_full_open (const char *args, int from_tty)
>   {
>     if (record_debug)
>       gdb_printf (gdb_stdlog, "Process record: record_full_open\n");
>   
> +  if (args != nullptr)
> +    error (_("trailing junk: '%s'"), args);
> +
>     record_preopen ();
>   
>     /* Reset */

While I am no fan of this type of error, I understand the motive
in terms of consistency. Grep'ping for this string in gdb gives
numerous wordings of this type of error:

ada-lang.c:    error (_("Junk at end of expression"));
ada-lang.c:    error (_("Junk at end of arguments."));
break-catch-exec.c:    error (_("Junk at end of arguments."));
break-catch-fork.c:    error (_("Junk at end of arguments."));
break-catch-throw.c:    error (_("Junk at end of arguments."));
breakpoint.c:	    error (_("Junk at end of expression"));
breakpoint.c:		    error (_("Junk at end of expression"));
breakpoint.c:	    error (_("Junk after thread keyword."));
breakpoint.c:	    error (_("Junk after inferior keyword."));
breakpoint.c:	    error (_("Junk after task keyword."));
breakpoint.c:	error (_("Junk at end of arguments."));
breakpoint.c:		error (_("Junk after thread keyword."));
breakpoint.c:		error (_("Junk after task keyword."));
breakpoint.c:    error (_("Junk at end of command."));
breakpoint.c:    error (_("Junk at end of arguments."));
breakpoint.c:	error (_("Junk at end of arguments."));
btrace.c:    error (_("Junk after argument: %s."), arg);
linespec.c:    error (_("Junk at end of line specification: %s"), string);
linespec.c:    error (_("Junk at end of line specification: %s"), string);
parse.c:    error (_("Junk after end of expression."));
psymtab.c:	error (_("Junk at end of command"));
record-btrace.c:    error (_("Trailing junk: '%s'."), args);
record-btrace.c:    error (_("Trailing junk: '%s'."), args);
record-btrace.c:	error (_("Trailing junk: '%s'."), args + l2);
record-btrace.c:	error (_("Trailing junk: '%s'."), args + l1);
record-btrace.c:    error (_("Trailing junk: '%s'."), args);
record.c:    error (_("Junk after argument: %s."), arg);
record.c:    error (_("Junk after argument: %s."), arg);
record.c:    error (_("Junk after argument: %s."), arg);
symfile.c:	error (_("Junk after %s"), argv[1]);
symfile.c:	error (_("Junk after %s"), argv[0]);
symmisc.c:	error (_("Junk at end of command"));
symmisc.c:	error (_("Junk at end of command"));

So i guess as far as the record feature goes, the error message should
at least be "Trailing junk: %s". [Yeah, that is one heck of a silly nit,
isn't it?]

Otherwise, LGTM.

Keith