Fix heap get for split heap

[JojoS62]

Summary of changes

This is a fix for the field heap_stats.reserved_size in the mbed heap statistics. With split heap, the heap can contain two heap areas. In this case, mbed_heap_size changes to be the 2nd heap, and mbed_heap_size_0 will become the first heap part.
The heap statistics did took that change into account, and always used mbed_heap_size for the total heap size.

This PR adds a variable mbed_heap_size_total to reflect the total size, regardless if one or two heaps are used.

Impact of changes

Heap functions are declared as weak, so a custom implementation could be used for even more heap areas or heap in external SDRAM. In this case, the variable mbed_heap_size_total must be calculated in software_init_hook() in mbed_boot_gcc_arm.c for a correct output in the heap statistics.

Migration actions required

Documentation

---

Pull request type

[x] Patch update (Bug fix / Target update / Docs update / Test update / Refactor)
[] Feature update (New feature / Functionality change / New API)
[] Major update (Breaking change E.g. Return code change / API behaviour change)

---

Test results

[] No Tests required for this change (E.g docs only update)
[] Covered by existing mbed-os tests (Greentea or Unittest)
[] Tests / results supplied as part of this PR

---

Reviewers

---

[multiplemonomials]

@JojoS62 Have you run the heap size Greentea test to make sure there are no breakages? Also, maybe we should update the heap stats test to test for this fix.

Other than that, looks good!

[JojoS62]

I haven’t used the tests for long time, I remember it is difficult with a custom target.
My patch doesn’t change the alloc code, sbrk uses the start/end addresses and not the heap size.

[JojoS62]

The following tests passed:
        test-mbed-platform-stats-heap
        test-mbed-rtos-heap-and-stack
        test-mbed-storage-blockdevice-heap_block_device

theses tests have passed on a NUCLEO_H743ZI2 with split_heap.
But the damned git rebase command...