POV-Ray : Newsgroups : povray.unofficial.patches : Megapov crashes when post-processing large images Server Time
2 Sep 2024 00:20:01 EDT (-0400)
  Megapov crashes when post-processing large images (Message 1 to 6 of 6)  
From: Gilles Tran
Subject: Megapov crashes when post-processing large images
Date: 14 Sep 2000 13:55:54
Message: <39C110A4.D911B571@inapg.inra.fr>
Just had the problem a few times now... When rendering a large image
(3200 x 2400) with PP enabled (tested with soft_glow and focal_blur),
Megapov (Windows) crashes at the beginning of the post-process phase.
There is no problem on small images. Before it crashes, the following
message is repeated on screen :

unknow error 17
attempt to free null pointer (file: c:\pov\mypov\source\povray.c  line
1941)

The following scene causes it for example :
#version unofficial Megapov 0.5;
global_settings{post_process{focal_blur{2,2,5,1}}}
camera{  location -x*10   direction z   right     4/3*x   look_at 0}
light_source{  100*x   color 1  }
sphere { 0.0, 1 texture {pigment{color 0.5} }}

command line +w3200 +h2400 +f -d

Configuration : Megapov 0.5 on WinNT (126 Mb RAM, PIII 733 MHz)

G.


Post a reply to this message

From: Gilles Tran
Subject: Re: Megapov crashes when post-processing large images
Date: 14 Sep 2000 14:33:22
Message: <39C11976.31EB5753@inapg.inra.fr>
Well, further investigation reveals that setting the swap file size to 450
Mb (instead of 140...) fixes the problem. Just an insufficient memory
problem after all.
G.


Post a reply to this message

From: Warp
Subject: Re: Megapov crashes when post-processing large images
Date: 15 Sep 2000 02:49:35
Message: <39c1c67f@news.povray.org>
Gilles Tran <tra### [at] inapginrafr> wrote:
: Well, further investigation reveals that setting the swap file size to 450
: Mb (instead of 140...) fixes the problem. Just an insufficient memory
: problem after all.

  It shouldn't crash the program though (ie. trying to free a null pointer
is always a bug).

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):_;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/


Post a reply to this message

From: Gilles Tran
Subject: Re: Megapov crashes (again)
Date: 16 Sep 2000 04:53:15
Message: <39C33484.148B3FD8@inapg.inra.fr>
Gilles Tran wrote:

> Well, further investigation reveals that setting the swap file size to 450
> Mb (instead of 140...) fixes the problem. Just an insufficient memory
> problem after all.
> G.

There's still a problem. Even after the swap file increase, the post process
starts OK, processes some of the image and then... crashes NT, a good old blue
screen of death. Wild guess : this must be one of these things called "memory
leak".

G.


Post a reply to this message

From: Christoph Hormann
Subject: Re: Megapov crashes (again)
Date: 16 Sep 2000 06:12:13
Message: <39C3477D.A02FFCB@schunter.etc.tu-bs.de>
Gilles Tran wrote:
> 
> There's still a problem. Even after the swap file increase, the post process
> starts OK, processes some of the image and then... crashes NT, a good old blue
> screen of death. Wild guess : this must be one of these things called "memory
> leak".
> 
> G.

I just tested a large scene with post process myself because i'm also using NT,
did not finish it, because after 20 minutes of PP, it had finished mere 7%.  It
did not crash until then, but the swapfile got 540 Mb large.  

In fact NT should not crash because of an application memory leak (it maybe does
though :-) but in any case, megapov's memory use during post processing seems
not very efficient.  

Christoph

--
Christoph Hormann <chr### [at] gmxde>
Homepage: http://www.schunter.etc.tu-bs.de/~chris/


Post a reply to this message

From: Gilles Tran
Subject: Re: Megapov crashes (again)
Date: 16 Sep 2000 06:57:46
Message: <39C351B3.75005024@inapg.inra.fr>
Christoph Hormann wrote:

> I just tested a large scene with post process myself because i'm also using NT,
> did not finish it, because after 20 minutes of PP, it had finished mere 7%.  It
> did not crash until then, but the swapfile got 540 Mb large.
>
> In fact NT should not crash because of an application memory leak (it maybe does
> though :-) but in any case, megapov's memory use during post processing seems not
> very efficient.

Possibly I should raise the swap even higher. The strange thing is that I rendered
at least one scene with PP on Megapov 0.3 with a large image size before and didn't
have this problem (but this was a different computer and I don't remember the swap
file setup).
G.


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.