Import WLOG no longer working

Linus Torvalds torvalds at linux-foundation.org
Fri Aug 14 14:37:10 PDT 2020


On Fri, Aug 14, 2020 at 1:55 PM Salvador Cuñat <salvador.cunat at gmail.com> wrote:
>
> Dive num = 394 -- ev type = 11
> ev->gas,index = 1
> Dive num = 398 -- ev type = 11
> ev->gas,index = 1
> Dive num = 401 -- ev type = 25
> ev->gas,index = 0
> Dive num = 401 -- ev type = 6
> Dive num = 401 -- ev type = 3
> Dive num = 401 -- ev type = 3
> Dive num = 401 -- ev type = 3
> Dive num = 401 -- ev type = 6
> Dive num = 401 -- ev type = 25
> ev->gas,index = 270305728
> Violación de segmento
>
> You were right, off course Dirk, look at the last line before the segfault.
> I've pasted the previous lines because they are wrong too ... Dive 401 has just a single event type 25, following events are wrong and, most probably come from following dives.

Ok, so the first ev type = 25 looks sane, with a gas index of zero.

And the other events are goarbage and don't show up in the wlog data?

ev type = 3 is ascent warning, and type 6 is some transmitter warning
(SAMPLE_EVENT_TRANSMITTER?)

> Yes, seems completely random.

That SAMPLE_EVENT_TRANSMITTER looks odd too. I've never seen it, but
who knows. Might be some odd translation artifact, but it does exist
in the uwatec_memomouse_parser, so maybe it's real.

Is there any way we could reproduce this locally?

But another thing you could try is to just run your load under
valgrind. That might pinpoint the problem..

               Linus


More information about the subsurface mailing list