On Sun, 14 Jan 2024 23:17:11 +0900
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx> wrote:
> >
> > Looks good to me and tested.
> >
> > Reviewed-by: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> > Tested-by: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
>
> Sorry, I should cancel these. I found this test did not pass if I set
> the function tracer on.
>
> / # cd /sys/kernel/tracing/
> /sys/kernel/tracing # echo function > current_tracer
> /sys/kernel/tracing # /mnt/map_test
> TAP version 13
> 1..2
> # Starting 2 tests from 2 test cases.
> # RUN map.subbuf_size_4k.meta_page_check ...
> # map_test.c:174:meta_page_check:Expected self->meta->entries (15293) == 16 (16)
> # meta_page_check: Test terminated by assertion
> # FAIL map.subbuf_size_4k.meta_page_check
> not ok 1 map.subbuf_size_4k.meta_page_check
> # RUN map.subbuf_size_8k.meta_page_check ...
> # map_test.c:174:meta_page_check:Expected self->meta->entries (15270) == 16 (16)
> # meta_page_check: Test terminated by assertion
> # FAIL map.subbuf_size_8k.meta_page_check
> not ok 2 map.subbuf_size_8k.meta_page_check
> # FAILED: 0 / 2 tests passed.
> # Totals: pass:0 fail:2 xfail:0 xpass:0 skip:0 error:0
>
> Does this depend on "nop" tracer?
Yes.
> > > +
> > > +static int tracefs_reset(void)
> > > +{
> > > + if (__tracefs_write_int(TRACEFS_ROOT"/tracing_on", 0))
> > > + return -1;
> > > + if (__tracefs_write_int(TRACEFS_ROOT"/trace", 0))
> > > + return -1;
> > > + if (__tracefs_write(TRACEFS_ROOT"/set_event", ""))
> > > + return -1;
Looks like the below is needed:
if (__traces_write(TRACEFS_ROOT"/current_tracer", "nop"))
return -1;
> > > +
> > > + return 0;
> > > +}
> > > +
-- Steve
