Re: [PATCH 01/14] tools lib traceevent: Get rid of malloc_or_die() in show_error()

From: Namhyung Kim
Date: Tue Dec 10 2013 - 19:40:43 EST


On Tue, 10 Dec 2013 14:30:18 +0900, Namhyung Kim wrote:
> On Tue, 10 Dec 2013 14:01:44 +0900, Namhyung Kim wrote:
>> On Mon, 9 Dec 2013 21:14:10 -0500, Steven Rostedt wrote:
>>> On Tue, 10 Dec 2013 11:03:44 +0900
>>> Namhyung Kim <namhyung@xxxxxxxxxx> wrote:
>>>
>>>> What about returning error code rather than string? This way we won't
>>>> worry about the allocation of the error string itself.
>>>>
>>>> But the downside of it is loosing a positional info of the error.
>>>> Hmm.. what about using a static buffer in pevent for it then?
>>>
>>> A static buffer may be the solution. Never need to worry about
>>> allocating it on error, as it will already be allocated. And we can add
>>> APIs to print it nicely.
>>>
>>> Perhaps call it
>>>
>>> pevent->filter_error_buffer
>>>
>>> ?
>
> Hmm.. thinking about it twice, if it's only for filter functions
> wouldn't it be better moving it to event_filter rather than pevent?
>
> filter->error_buffer
>

One more thinking, if we agree on converting to return error code, does
pevent_filter_add_filter_str() also need to be changed not to receive
the third "error_str" argument?

And should we extend the error code to include the return value of
pevent_filter_match() too? If not, it seems we need to pass another
argument to receive the actual error code in case of FILTER_ERROR.

I'm saying these here since they might require interface/signature
change so will affect existing users like trace-cmd.

Thanks,
Namhyung
--
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/