 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Mark Wagner <mar### [at] gte net> wrote:
: As big as the compiler will allow.
Take into account that the program might _reserve_ the space for the stack
at run time. That is, if you set a 100 Megs stack, the program may need
100 Megs of memory for the stack + the memory for the rest of the things.
--
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 13 Jan 2000 04:51:09 -0500, Nieminen Juha wrote:
>Mark Wagner <mar### [at] gtenet> wrote:
>: As big as the compiler will allow.
>
> Take into account that the program might _reserve_ the space for the stack
>at run time. That is, if you set a 100 Megs stack, the program may need
>100 Megs of memory for the stack + the memory for the rest of the things.
Not under win32, as I understand it. Win32 has two numbers governing stack
size, a reserved amount set aside in the page table at runtime, and a committed
amount that's actually mapped. As the stack grows, Windows allocates more pages
to it. Also, for those skilled at hex editing, the reserved and committed stack
space are just numbers in the executable file, so they can be changed rather
easily without recompiling.
The official POV has a reserve/commit of 80000/10000 (hex). Megapov is set to
100000/1000 (hex). So if this is a stack problem, it would hit the official
version too, though a little higher commit wouldn't hurt megapov in the
performance department.
--
These are my opinions. I do NOT speak for the POV-Team.
The superpatch: http://www2.fwi.com/~parkerr/superpatch/
My other stuff: http://www2.fwi.com/~parkerr/traces.html
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I also have a crash problem with photons, the problem
isn't so much the photons or the scene, but when it runs
out of memory with MegaPov, the only thing I can do is
physically turn the machine off and on again.
When I run out of mem with official POV, it just gives
me a nice little message saying that it's run out of
memory, so I do something else instead, or make some
changes to the scene. But MegaPov completely freezes
the box, I've tried this about three times, but
Linux doesn't like being turned off like that, so I
havn't tried it recently. Looking at my sig, it was
in fact five days ago since I tried it.
Info
RH6.0, MegaPov, over 64M Ram 1G of swap available.
Scene, David Fontain's torus with a disco ball
inside and lots of photons.
Can provide the Pov file if required.
--
Cheers
Steve email mailto:sjl### [at] ndirectcouk
%HAV-A-NICEDAY Error not enough coffee 0 pps.
web http://www.ndirect.co.uk/~sjlen/
or http://start.at/zero-pps
9:17pm up 5 days, 5:28, 6 users, load average: 5.39, 5.24, 5.11
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I didn't change anything about memory handling, so I don't know why this
would act differently than the official version. Did you compile your own
official version, or download a binary?
-Nathan
Steve <sjl### [at] zero-ppslocaldomain> wrote...
> When I run out of mem with official POV, it just gives
> me a nice little message saying that it's run out of
> memory, so I do something else instead, or make some
> changes to the scene. But MegaPov completely freezes
> the box, I've tried this about three times, but
> Linux doesn't like being turned off like that, so I
> havn't tried it recently. Looking at my sig, it was
> in fact five days ago since I tried it.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sat, 29 Jan 2000 00:50:22 -0500, Nathan Kopp <Nat### [at] Koppcom> wrote:
>I didn't change anything about memory handling, so I don't know why this
>would act differently than the official version. Did you compile your own
>official version, or download a binary?
No, didn't compile, d/l'd the binaries.
my mem looks like this, in units of KBs.
[sjlen@zero-pps pov]$ free -k
total used free shared buffers cached
Mem: 63136 60884 2252 16692 1312 32616
-/+ buffers/cache: 26956 36180
Swap: 989704 32332 957372
When my current render has finished I'll try rendering something in Megapov
that I know runs out of mem in official POV, and then run it again in
Pov and see what happens. (as this scene doesn't use any new Megapov features).
Scene below:
#declare phd=1;
#include "colors.inc"
global_settings{
adc_bailout 0.003922
ambient_light <1.0,1.0,1.0>
assumed_gamma 2.2
max_trace_level 50
photons{
gather 20,100
radius 0.1*phd,2,0.1*phd
autostop 0
//jitter .4
expand_thresholds 0.2, 40
}
}
#declare Strobe =
#declare TheGap=5;
#declare MaxAng=359;
union{
#local outerctr=0;
#while(outerctr<MaxAng)
#local ctr=0;
#while(ctr<MaxAng)
box{
<-0.1,-0.1,-0.1>
<0.1,0.1,0.1>
translate z*-2.085
rotate <ctr,outerctr,0>
}
#local ctr=ctr+TheGap;
#end
#local outerctr=outerctr+TheGap;
#end
pigment{Silver}
finish{
phong 0.8
phong_size 50
reflection 1.0
specular 1.0
roughness 0.0001
ambient 0.0
diffuse 0.0
}
}
object{
Strobe
translate<0,5,10>
scale 0.5
photons{
separation 0.01*phd
refraction off
reflection on
}
translate z*-10
rotate y*-90
}
torus{
10,2
hollow
photons{
separation 0.01*phd
refraction off
reflection on
}
pigment{ color White }
finish { ambient 0
diffuse .1
reflection .9
specular 0.01
}
}
camera{
location <0,0,-10>
look_at <1,0,-10> }
light_source { <0,0,-10> Red
photons{
reflection on
refraction off
}
}
light_source { <0,0,-10> Green rotate 120*y
photons{
reflection on
refraction off
}
}
light_source { <0,0,-10> Blue rotate 240*y
photons{
reflection on
refraction off
}
}
///////////////////////////////////////////////
Hope this helps.
--
Cheers
Steve email mailto:sjl### [at] ndirectcouk
%HAV-A-NICEDAY Error not enough coffee 0 pps.
web http://www.ndirect.co.uk/~sjlen/
or http://start.at/zero-pps
3:43am up 1 day, 2:12, 4 users, load average: 2.23, 2.22, 2.13
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Dear Steve san:
I make test your scene file and get report from memory short error.
My application is MPW MrC version(based official mac 3.1g r2).
I set 128MB to application and system vertial memory is on,
(Power mac G3 DT233 memory 160MB VM set to 320MB).
Report is like this:
Creating bounding slabs.
Scene contains 5185 frame level objects; 0 infinite.
Creating vista buffer.
Creating light buffers...
Creating Photon Maps.
Building Photon Maps.........................................................
Slab building error.
Ran out of application memory; Increase POV-Ray application memory size.
Reclaiming memory
-- [Finish] File=photonsbug8.pov,Time=00.1.30 4:27:46 PM
----------------------------------------------------------------------------
POV-Ray Mac Memory Cache Statistics
Cache hits: 25467755 Cache misses: 816
Total calls to malloc: 25468571 Cache hit rate: 99%
----------------------------------------------------------------------------
Y.Tanabe
Kobe,Japan
Steve wrote:
> On Sat, 29 Jan 2000 00:50:22 -0500, Nathan Kopp <Nat### [at] Koppcom> wrote:
> >I didn't change anything about memory handling, so I don't know why this
> >would act differently than the official version. Did you compile your own
> >official version, or download a binary?
>
> No, didn't compile, d/l'd the binaries.
>
> my mem looks like this, in units of KBs.
>
> [sjlen@zero-pps pov]$ free -k
> total used free shared buffers cached
> Mem: 63136 60884 2252 16692 1312 32616
> -/+ buffers/cache: 26956 36180
> Swap: 989704 32332 957372
>
> When my current render has finished I'll try rendering something in Megapov
> that I know runs out of mem in official POV, and then run it again in
> Pov and see what happens. (as this scene doesn't use any new Megapov features).
>
> Scene below:
> Hope this helps.
>
> --
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sun, 30 Jan 2000 16:49:46 +0900, Y.Tanabe <tec### [at] khrimorjp> wrote:
>Dear Steve san:
>
>I make test your scene file and get report from memory short error.
>My application is MPW MrC version(based official mac 3.1g r2).
>I set 128MB to application and system vertial memory is on,
>(Power mac G3 DT233 memory 160MB VM set to 320MB).
>
>Report is like this:
>
>Creating bounding slabs.
>Scene contains 5185 frame level objects; 0 infinite.
>
>Creating vista buffer.
>Creating light buffers...
>Creating Photon Maps.
>Building Photon Maps.........................................................
>Slab building error.
>Ran out of application memory; Increase POV-Ray application memory size.
>Reclaiming memory
>-- [Finish] File=photonsbug8.pov,Time=00.1.30 4:27:46 PM
>----------------------------------------------------------------------------
>POV-Ray Mac Memory Cache Statistics
>Cache hits: 25467755 Cache misses: 816
>Total calls to malloc: 25468571 Cache hit rate: 99%
>----------------------------------------------------------------------------
Thanks, I'll have to wait untill my machine has finished rendering it's current
scene before I try this again. The current render has been going for more than
30 hours, so I don't want the machine to crash just yet.
--
Cheers
Steve email mailto:sjl### [at] ndirectcouk
%HAV-A-NICEDAY Error not enough coffee 0 pps.
web http://www.ndirect.co.uk/~sjlen/
or http://start.at/zero-pps
2:31pm up 1 day, 13:01, 4 users, load average: 2.29, 2.25, 2.10
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sat, 29 Jan 2000 00:50:22 -0500, Nathan Kopp <Nat### [at] Koppcom> wrote:
>I didn't change anything about memory handling, so I don't know why this
>would act differently than the official version. Did you compile your own
>official version, or download a binary?
>
Right I've now done some extensive experimentation.
I took a scene that will render in Official pov but requires
more memory than I have, so it's just a standard scene with
millions of objects. In both official and Mega it gives
me a message telling me that it's run out of memory, no problem.
But when I render a scene with photons and it runs out of memory
while creating the photon data the box freezes completely.
I had to go out (yes actually outside the house) today so I set
the render running with xmegapov -Imyscene.pov and left it for
a few hours, it froze, so I thought give it a few hours and it
might sort itself out, but after the screen had cutout I couldn't
even get my display back by pressing the shift key the way I
normally do. Before losing the screen and after the freeze
the mouse pointer wouldn't move, I couldn't Alt-Ctrl-Fn to another
session, I couldn't Alt-Ctrl-Backspace out of X.
One more thing that I will try is running a scene that has photons,
but runs out of memory before it gets to the point where it starts
to generate the photon data. I will try this soon, and post the
results in this thread.
--
Cheers
Steve email mailto:sjl### [at] ndirectcouk
%HAV-A-NICEDAY Error not enough coffee 0 pps.
web http://www.ndirect.co.uk/~sjlen/
or http://start.at/zero-pps
6:37pm up 4:09, 2 users, load average: 1.44, 1.22, 1.14
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On a scene that runns out of memory before starting to
generate the photon data the machine doesn't freeze.
--
Cheers
Steve email mailto:sjl### [at] ndirectcouk
%HAV-A-NICEDAY Error not enough coffee 0 pps.
web http://www.ndirect.co.uk/~sjlen/
or http://start.at/zero-pps
8:53pm up 6:26, 2 users, load average: 1.60, 1.52, 2.03
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Steve wrote in message ...
>On a scene that runns out of memory before starting to
>generate the photon data the machine doesn't freeze.
Sounds like a recursive call to Error(). Someone who has a debugger might
want to check this out.
Mark
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
