http://www.iro.umontreal.ca/~gambit/doc/gambit-c.html#mapping-of-types says:
In the case of the struct, union and type types, the internal pointer points to a copy of the C data type being represented. When an instance of one of these types is converted from C to Scheme, a block of memory is allocated from the C heap and initialized with the instance and then a foreign object is allocated from the Scheme heap and initialized with the pointer to this copy. [...]
Does this mean that existing foreign objects are immutable?
Afficher les réponses par date
On Thu, Oct 8, 2009 at 9:41 PM, Julian Scheid julians37@googlemail.com wrote:
http://www.iro.umontreal.ca/~gambit/doc/gambit-c.html#mapping-of-types says:
In the case of the struct, union and type types, the internal pointer points to a copy of the C data type being represented. When an instance of one of these types is converted from C to Scheme, a block of memory is allocated from the C heap and initialized with the instance and then a foreign object is allocated from the Scheme heap and initialized with the pointer to this copy. [...]
I've made a test case that shows they are mutable when accessed by a pointer. Does this mean the paragraph above only applies when accessing a struct by value?
This also brings me to another question, I'm somewhat confused about ___alloc_rc.
James Long had an issue [1] that I could reproduce and was able to fix by using __alloc_rc instead of malloc. Also, 4.5.2's gambit.h has ___USE_RC_ALLOCATION_FOR_FOREIGN defined.
At the same time, I have examples that seem to work fine with regular new/malloc. Is that just a lucky accident?
If ___alloc_rc was indeed required, how could the FFI use foreign objects that have been created outside of Gambit's control?
It would be great if someone could shed some light on this, thanks in advance.
[1] http://github.com/jlongster/gambit-iphone-example/blob/master/lib/ffi/osx.sc...
-
Julian Scheid