Re: fio: when not using a job file, timeouts are not applied

[Chris Worley <worleys@xxxxxxxxx>]

Do you mean to be using --timeout or --runtime?

On Mon, Jul 18, 2011 at 6:02 PM, Daniel Farina <daniel@xxxxxxxxxx> wrote:
> On Mon, Jul 18, 2011 at 4:49 PM, Chris Worley <worleys@xxxxxxxxx> wrote:
>> Could you add-in "--loops=10000"?
>
> No timeout occurs.
>
> FYI, I have poked around at init.c, and feel that somehow
> fill_def_thread isn't doing its job when job files are not being used:
>
> int parse_options(int argc, char *argv[])
> {
>        int job_files, i;
>
>        f_out = stdout;
>        f_err = stderr;
>
>        fio_options_fill_optstring();
>        fio_options_dup_and_init(l_opts);
>
>        if (setup_thread_area())
>                return 1;
>        if (fill_def_thread())
>                return 1;
>
>        job_files = parse_cmd_line(argc, argv);
>
>        for (i = 0; i < job_files; i++) {
>                if (fill_def_thread())
>                        return 1;
>                if (parse_jobs_ini(ini_file[i], i))
>                        return 1;
>                free(ini_file[i]);
>        }
>
> I think something is fishy here with the order of events (see
> parse_cmd_line, which interprets the timeout, but fill_def_thread is
> not called after that point unless there are job files). gdb seemed to
> confirm that the value wasn't getting copied to the thread-local
> memory.  (I also did try the non-forked based backend to make sure it
> wasn't some mechanism problem there, if memory serves).
>
> --
> fdr
>
--
To unsubscribe from this list: send the line "unsubscribe fio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html