rpict: out of memory in ecalloc

Hi all,

I grabbed and recompiled with the latest HEAD as per the previous thread. I started a scene and came back to find that it had died with the following error reported:

rpict: out of memory in ecalloc

I am using rpiece to run the scene on a machine equipped with 2GB of memory. I do not know if this is related to the latest changes or some other problem.

-Jack

···

--
# John E. de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

That's really strange. It seems that ecalloc() is only called for one purpose, which is to allocate nodes in the .cal expression parse tree. This is a very small amount of memory, so either you just happened to run out of memory when reading in a new .cal file or else it was caught in a loop reading the .cal file in over and over again.

I don't see how the changes I made to the source routines could be affecting this, unless you have so many light sources that it gobbled up all your VM before loading in the.cal file. That seems incredibly unlikely, since it would have failed on a larger memory request, and you would need over 250,000 light sources to use up a GByte of RAM.

Maybe Schorch has some ideas -- he also checked in a bunch of changes, though they were all ANSIfication fixes that shouldn't have caused problems.

Can you rerun the process, just to see if it dies at the same point?

-Greg

···

From: Jack de Valpine <[email protected]>
Date: January 2, 2004 10:54:10 AM PST

Hi all,

I grabbed and recompiled with the latest HEAD as per the previous thread. I started a scene and came back to find that it had died with the following error reported:

rpict: out of memory in ecalloc

I am using rpiece to run the scene on a machine equipped with 2GB of memory. I do not know if this is related to the latest changes or some other problem.

-Jack

Hi Greg,

Thanks, I am rerunning the process to see what happens. I will note that, although I have successfully rendered this scene before, it is with an older version of Radiance (probably somewhere between 3.4 and 3.5 in vintage), so it sound's like this could likely be something completely unrelated to the source routines. There are a limited number of lights in the scene, so I do not believe that this would be a problem.

As a side note. Would it be possible to have the routine(s) ecalloc and others report the name of the .cal that a process dies on, if indeed this is the problem?

-Jack

Greg Ward wrote:

···

That's really strange. It seems that ecalloc() is only called for one purpose, which is to allocate nodes in the .cal expression parse tree. This is a very small amount of memory, so either you just happened to run out of memory when reading in a new .cal file or else it was caught in a loop reading the .cal file in over and over again.

I don't see how the changes I made to the source routines could be affecting this, unless you have so many light sources that it gobbled up all your VM before loading in the.cal file. That seems incredibly unlikely, since it would have failed on a larger memory request, and you would need over 250,000 light sources to use up a GByte of RAM.

Maybe Schorch has some ideas -- he also checked in a bunch of changes, though they were all ANSIfication fixes that shouldn't have caused problems.

Can you rerun the process, just to see if it dies at the same point?

-Greg

From: Jack de Valpine <[email protected]>
Date: January 2, 2004 10:54:10 AM PST

Hi all,

I grabbed and recompiled with the latest HEAD as per the previous thread. I started a scene and came back to find that it had died with the following error reported:

rpict: out of memory in ecalloc

I am using rpiece to run the scene on a machine equipped with 2GB of memory. I do not know if this is related to the latest changes or some other problem.

-Jack

_______________________________________________
Radiance-dev mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-dev

--
# John E. de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

Ok, new error now at same point in process (same percentage complete):

rpict: system - out of memory in getfunc: Cannot allocate memory

-Jack

Jack de Valpine wrote:

···

Hi Greg,

Thanks, I am rerunning the process to see what happens. I will note that, although I have successfully rendered this scene before, it is with an older version of Radiance (probably somewhere between 3.4 and 3.5 in vintage), so it sound's like this could likely be something completely unrelated to the source routines. There are a limited number of lights in the scene, so I do not believe that this would be a problem.

As a side note. Would it be possible to have the routine(s) ecalloc and others report the name of the .cal that a process dies on, if indeed this is the problem?

-Jack

Greg Ward wrote:

That's really strange. It seems that ecalloc() is only called for one purpose, which is to allocate nodes in the .cal expression parse tree. This is a very small amount of memory, so either you just happened to run out of memory when reading in a new .cal file or else it was caught in a loop reading the .cal file in over and over again.

I don't see how the changes I made to the source routines could be affecting this, unless you have so many light sources that it gobbled up all your VM before loading in the.cal file. That seems incredibly unlikely, since it would have failed on a larger memory request, and you would need over 250,000 light sources to use up a GByte of RAM.

Maybe Schorch has some ideas -- he also checked in a bunch of changes, though they were all ANSIfication fixes that shouldn't have caused problems.

Can you rerun the process, just to see if it dies at the same point?

-Greg

From: Jack de Valpine <[email protected]>
Date: January 2, 2004 10:54:10 AM PST

Hi all,

I grabbed and recompiled with the latest HEAD as per the previous thread. I started a scene and came back to find that it had died with the following error reported:

rpict: out of memory in ecalloc

I am using rpiece to run the scene on a machine equipped with 2GB of memory. I do not know if this is related to the latest changes or some other problem.

-Jack

_______________________________________________
Radiance-dev mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-dev

--
# John E. de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction