[hatari-devel] Freeze / abort in Talk2Talk 2 demo?
Per Almered
per at brainfish.net
Mon Jul 12 14:56:14 CEST 2010
I can try to re-save that particular song in a newer version of
maxYMiser, and if you want, I could deliver the source code for
the "readme" program too, just let me know if that would help!
Per.
> -----Ursprungligt meddelande-----
> Från: hatari-devel-bounces at lists.berlios.de
> [mailto:hatari-devel-bounces at lists.berlios.de] För
> npomarede at corp.free.fr
> Skickat: den 12 juli 2010 14:41
> Till: Mailing list for Hatari developers
> Ämne: Re: [hatari-devel] Freeze / abort in Talk2Talk 2 demo?
>
>
> On Mon, 12 Jul 2010, Anders Eriksson wrote:
>
> > On Mon, 12 Jul 2010, Eero Tamminen wrote:
> >
> >> I tried to debug the issue, but CMake by default makes
> non-debuggable
> >> binaries. Main thing Gdb told is that Hatari goes often to
> >> Video_InterruptHandler_EndLine().
> >>
> >> After re-compiling Hatari with asserts and -g (+ still using -O3),
> >> reading the demo text + scrolling up & down many times, I got an
> >> assert:
> >> hatari: /home/eero/work/hatari/src/cycInt.c:378:
> >> CycInt_AddRelativeInterruptWithOffset: Assertion
> `CycleTime >= 0' failed.
> >>
> >> Can somebody reproduce this (freeze, or abort when asserts are
> >> enabled)?
> >
> > Hi,
> >
> > I have a feeling it's not the scrolling, it will crash even
> if you let
> > it be
> > for itself. I have some suspicion it might be the maxYMiser
> soundrout,
> > because I've seen this in other (older though) maxYMiser
> songs. But just a
> > hunch.
>
> Hi
>
>
> hmmm, crashes here too, even without scrolling that much. As
> there's no
> border removal / resolution changes, I suspect this may be a
> problem in
> the STE DMA sound emulation, which uses internal hatari's
> timers too (as
> it happened once in the EPSS demo).
> I will try to capture some traces, but this will be difficult
> as it's not
> easily reproduceable and the demo can run quite a lot before freezing.
>
> Nicolas
>
> _______________________________________________
> hatari-devel mailing list
> hatari-devel at lists.berlios.de
> https://lists.berlios.de/mailman/listinfo/hata> ri-devel
>
>
> __________ Information from ESET NOD32 Antivirus,
> version of virus signature database 5271 (20100712) __________
>
> The message was checked by ESET NOD32 Antivirus.
>
http://www.eset.com
More information about the hatari-devel
mailing list