Re: [Chicken-users] Segfault with large data-structures (bug)

[Jim Ursetto]

On Feb 4, 2013, at 2:28 PM, Felix wrote:

> 
> Perhaps, but I really don't see a problem of allowing a limit
> on heap allocation in the runtime system.

I think a segfault is an appropriate response to OOM, but I wonder
if it's possible to panic() instead if the heap size can't be increased
as needed.

I also wonder why csi hangs at 100% CPU instead of segfaulting.
Seems like a further indication that panic() should be used.

Currently a panic() is done if heap size is fixed, but not
if it hits the max -- fixed is probably a special case of max,
but I haven't looked very closely.

Jim