[mlir][EmitC]Expand the MemRefToEmitC pass - Lowering `AllocOp` (#148257)

This aims to lower `memref.alloc` to `emitc.call_opaque “malloc” ` or
`emitc.call_opaque “aligned_alloc” `
From:
```
module{
  func.func @allocating() {
  %alloc_5 = memref.alloc() : memref<999xi32>
  return
  }
}
```

To:
```
module {
  emitc.include <"stdlib.h">
  func.func @allocating() {
    %0 = emitc.call_opaque "sizeof"() {args = [i32]} : () -> !emitc.size_t
    %1 = "emitc.constant"() <{value = 999 : index}> : () -> index
    %2 = emitc.mul %0, %1 : (!emitc.size_t, index) -> !emitc.size_t
    %3 = emitc.call_opaque "malloc"(%2) : (!emitc.size_t) -> !emitc.ptr<!emitc.opaque<"void">>
    %4 = emitc.cast %3 : !emitc.ptr<!emitc.opaque<"void">> to !emitc.ptr<i32>
    return
  }
}
```
Which is then translated as:
```
#include <stdlib.h>
void allocating() {
  size_t v1 = sizeof(int32_t);
  size_t v2 = 999;
  size_t v3 = v1 * v2;
  void* v4 = malloc(v3);
  int32_t* v5 = (int32_t*) v4;
  return;
}
```

0 files changed, 0 insertions, 0 deletions