On Tue 30-08-11 11:13:50, Curt Wohlgemuth wrote:
> Hi Jan:
>
> >> +static const char *wb_stats_labels[WB_REASON_MAX] = {
> >> + [WB_REASON_BALANCE_DIRTY] = "page: balance_dirty_pages",
> >> + [WB_REASON_BACKGROUND] = "page: background_writeout",
> >> + [WB_REASON_TRY_TO_FREE_PAGES] = "page: try_to_free_pages",
> >> + [WB_REASON_SYNC] = "page: sync",
> >> + [WB_REASON_PERIODIC] = "page: periodic",
> >> + [WB_REASON_FDATAWRITE] = "page: fdatawrite",
> >> + [WB_REASON_LAPTOP_TIMER] = "page: laptop_periodic",
> >> + [WB_REASON_FREE_MORE_MEM] = "page: free_more_memory",
> >> + [WB_REASON_FS_FREE_SPACE] = "page: fs_free_space",
> >> +};
> > I don't think it's good to have two enum->string translation tables for
> > reasons. That's prone to errors which is in fact proven by the fact that
> > you ommitted FORKER_THREAD reason here.
>
> Ah, thanks for catching the omitted reason. I assume you mean the
> table above, and
>
> +#define show_work_reason(reason)
>
> from the patch 2/3 (in the trace events file). Hmm, that could be a
> challenge, given the limitations on what you can do in trace macros.
> I'll think on this though.
Ah right, with trace points it's not so simple. I guess we could have
an array with labels as a global symbol and dereference it in tracepoints.
Honza
--
Jan Kara <jack@xxxxxxx>
SUSE Labs, CR
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
