Re: [PATCH] fault-inject: Improve debug flexibility

[Akinobu Mita]

Hi Dmitry,

2014-08-19 17:51 GMT+09:00 Dmitry Monakhov <dmonakhov@xxxxxxxxxx>:
> Current debug levels are not optimal. Especially if one want to
> provoke big numbers of faults(broken device simulator) then any verbose
> level will produce giant numbers of identical logging messages. For that
> reason it is reasonable to have DUMP_ONCE flag, which will be cleared
> after first failure. Also the only way to understand which fail_attr
> was triggered is by means of investigating stack trace, which is also
> sub-optimal. Let's just dump name of this fail_attr.

Something like this feature looks very useful.  But instead of giving a
magic flag (1024) to existing 'verbose' parameter, I think creating new
parameter like 'verbose_ratelimit' is more useful.

        # dump only once in five seconds
        echo 5 > /sys/kernel/debug/fail_make_request/verbose_ratelimit

        # dump only once in a day
        echo $((24 * 60 * 60)) > \
                /sys/kernel/debug/fail_make_request/verbose_ratelimit
--
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/