Re: [PATCH] Documentation/trace: Correcting and extendingtracepoint documentation

[Rob Landley]

On 08/22/2013 04:49:31 PM, Zoltan Kiss wrote:
> The sample missed the moving of the header files into the events  
> subdirectory.
> I've also extended it based on the existing headers, and mentioned  
> the tiny
> but important role of CREATE_TRACE_POINTS.
>
> Signed-off-by: Zoltan Kiss <zoltan.kiss@xxxxxxxxxx>
> ---
>  Documentation/trace/tracepoints.txt |   19 +++++++++++++++++--
>  1 file changed, 17 insertions(+), 2 deletions(-)
>
> diff --git a/Documentation/trace/tracepoints.txt  
> b/Documentation/trace/tracepoints.txt
> index da49437..e8e3c4b 100644
> --- a/Documentation/trace/tracepoints.txt
> +++ b/Documentation/trace/tracepoints.txt
> @@ -40,7 +40,13 @@ Two elements are required for tracepoints :
>
>  In order to use tracepoints, you should include linux/tracepoint.h.
>
> -In include/trace/subsys.h :
> +In include/trace/events/subsys.h :
> +
> +#undef TRACE_SYSTEM
> +#define TRACE_SYSTEM subsys

That addition I can sort of see, I guess?

> +#if !defined(_TRACE_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
> +#define _TRACE_SUBSYS_H

But this makes no sense to me: why is it needed? (I.E. why must it be  
block copied into each _user_ of tracepoints?)

>  #include <linux/tracepoint.h>
>
> @@ -48,10 +54,16 @@ DECLARE_TRACE(subsys_eventname,
>  	TP_PROTO(int firstarg, struct task_struct *p),
>  	TP_ARGS(firstarg, p));
>
> +#endif /* _TRACE_SUBSYS_H */
> +
> +/* This part must be outside protection */
> +#include <trace/define_trace.h>
> +

Why? (Both why do you need to #include a header outside a multiple  
inclusion guard, and why is the additional header needed at all in  
_every_ subsystem trace header?)

>  In subsys/file.c (where the tracing statement must be added) :
>
> -#include <trace/subsys.h>
> +#include <trace/events/subsys.h>
>
> +#define CREATE_TRACE_POINTS
>  DEFINE_TRACE(subsys_eventname);
>
>  void somefct(void)
> @@ -72,6 +84,9 @@ Where :
>  - TP_ARGS(firstarg, p) are the parameters names, same as found in the
>    prototype.
>
> +- if you use the header in multiple source files, #define  
> CREATE_TRACE_POINTS
> +  should appear only in one source file
> +
>  Connecting a function (probe) to a tracepoint is done by providing a
>  probe (function to call) for the specific tracepoint through
>  register_trace_subsys_eventname().  Removing a probe is done through

I guess the documentation isn't at fault if the tracepoint subsystem is  
suddenly becoming a lot more repetitive and complicated for no readily  
apparent reason, but did anybody ask _why_ this thing changed? Can we  
document why the extra redundancy is needed?

Rob--
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/