Re: [PATCH 3/3] tools lib traceevent: Introduce pevent_strerror

[Steven Rostedt]

On Sat, 2012-06-16 at 07:25 +0900, Namhyung Kim wrote:
> Steven Rostedt <rostedt@xxxxxxxxxxx> writes:
> 
> > On Tue, 2012-06-12 at 16:42 +0900, Namhyung Kim wrote:
> >
> >> +/*
> >> + * This must have a same ordering as the enum pevent_errno.
> >> + */
> >> +static const char * const pevent_error_str[] = {
> >> +	"failed to allocate memory",
> >> +	"failed to parse event",
> >> +	"failed to read event id",
> >> +	"failed to read event format",
> >> +	"failed to read event print fmt",
> >> +	"failed to allocate field name for ftrace",
> >> +};
> >> +
> >
> > Here's a little macro trick to keep the strings and enums always in
> > sync:
> >
> > #define __PEVENT_ERRNO_CODES \
> >  _C(PEVENT_ERRNO__MEM_ALLOC_FAILED,	"failed to allocate memory"),	\
> >  _C(PEVENT_ERRNO__PARSE_EVENT_FAILED,	"failed to parse event"),	\
> >  _C(PEVENT_ERRNO__READ_ID_FAILED,	"failed to read event id"),	\
> >  _C(PEVENT_ERRNO__READ_FORMAT_FAILED,	"failed to read event format"),	\
> >  _C(PEVENT_ERRNO__READ_PRINT_FAILED,	"failed to read event print fmt"),\
> >  _C(PEVENT_ERRNO__OLD_FTRACE_ARG_FAILED,"failed to allocate field name for ftrace"),
> >
> > #undef _C
> > #define _C(a,b) b
> > static const char * const pevent_error_str[] = { __PEVENT_ERROR_CODES };
> >
> 
> Cool. Maybe even shorter:
> 
> #define __PEVENT_ERRNO_CODES \
>  _C(MEM_ALLOC_FAILED,	"failed to allocate memory"),	\
>  _C(PARSE_EVENT_FAILED,	"failed to parse event"),	\
>  _C(READ_ID_FAILED,	"failed to read event id"),	\
>  _C(READ_FORMAT_FAILED,	"failed to read event format"),	\
>  _C(READ_PRINT_FAILED,	"failed to read event print fmt"), \
>  _C(OLD_FTRACE_ARG_FAILED, "failed to allocate field name for ftrace"),
> 
> #undef _C
> #define _C(a,b) PEVENT_ERRNO__ ## a

I think you wanted b on this one. ;-)

> static const char * const pevent_error_str[] = { __PEVENT_ERROR_CODES };
> 
> But it make them less grep-able?

Yeah, that is a problem, although we could perhaps teach ctags and etags
how to find them.

-- Steve


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/