On Wed, 29 Mar 2023 at 01:15, Tony Pavlov via Sdcc-user < sdcc-user@lists.sourceforge.net> wrote:
> Hello Adrian, > > Wednesday, March 29, 2023, 5:52:01 AM, you wrote: > > just implement your own __sdcc_heap_init which set __sdcc_heap_free->next > to something like SP - <max_stack_size> instead of HEAP_END. > Yup, but then: * without recompiling/linking in an alternative definition for heap.s, i'd have a 1k wasted default heap, and * my CP/M hacks need to know about the internals of the heap allocator, like the struct definitions and such which aren't public. I did the above hack, but I instead wrote an sdcc_heap_add() that took another block of RAM to add to the free list. I had to duplicate the heap header_t definition though. If it were something in the library then it'd be super easy to just call to add regions that aren't fixed. (This seems to be a semi-common question on larger RAM 8 bit targets, so it'd likely solve this cleanly for a bunch of folks.) -adrian
_______________________________________________ Sdcc-user mailing list Sdcc-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sdcc-user
