Windows Qt5 binary

[Tomaz Canabrava]

Salva,

Can you run subsurface in vangrind --tool=callgrind , hit the issue,
then kill subsurface after a minute and send me the log that it will
generate?

Tomaz

On Mon, Jun 9, 2014 at 3:43 PM, Salvador Cuñat <salvador.cunat at gmail.com> wrote:
> On Mon, Jun 09, 2014 at 11:09:31AM -0700, Dirk Hohndel wrote:
>
>> Yeah, I mentioned that last night. I broke 'something' but haven't quite
>> figured out what. Only appears to happen on Windows, though.
>>
> No. In my linux machine it also fails but don't show so loooong number,
> maybe because it hangs.
>
>
>> > In Linux (debian wheezy on 32b machine with Gnome3), when cursor enters
>> > the profile tab, subsurface *almost* hangs out (not a total crash). And
>> > it seems related the issue 2.- in WIN7 (win doesn't hangs as it's on a
>> > newer 64 b machine, with better processor) so it seems a calculus
>> > problem.
>>
>> Do you have TTS/NDL calculation enabled? But it should self-disable if it
>> takes too long. Can you figure out where it spends all its time?
>>
>> > P.D. A final run on test dives shows this problem only happens when
>> > there is gas consumption during the dive, so ...
>>
>> Hmm - can you say more? Which test dive exhibits the problem?
>> What exactly does it mean to "*almost* hang"? How long a delay? What is
>> happening during that delay?
>
> I would put my money it's trying to calculate the SAC. The issue arises
> in *all* my real dives, but *only* in test dives involving gas
> consumption (e.g. dives/test30.xml). We are talking about a delay in
> minutes, where nothing happens (subsurface is completely dead). I think
> the difference relays in my machines, not in the O.S., as there is a big
> processor difference between the linux 32b and the win7 64b one; the
> newer one ends the calculus while the slower never reachs the end (well,
> never is not true, it's only that my patience ends first ;-) ).
>
> I think the fault is somewhere between 125ddd955c04, 11380a5deb071 and
> bbe62f756a0c.
>
> Regards.
>
> Salva.
> _______________________________________________
> subsurface mailing list
> subsurface at hohndel.org
> http://lists.hohndel.org/cgi-bin/mailman/listinfo/subsurface