Add missing StorePointer/StoreSmi in ContextScope.

Add missing StorePointer/StoreSmi in ContextScope.

R=iposva@google.com

Committed: https://code.google.com/p/dart/source/detail?r=41592

[Ivan Posva, 2014-10-29 05:17:28]

https://codereview.chromium.org/673253002/diff/40001/runtime/vm/raw_object.h
File runtime/vm/raw_object.h (right):

https://codereview.chromium.org/673253002/diff/40001/runtime/vm/raw_object.h#...
runtime/vm/raw_object.h:1217: RawObject** data() { OPEN_ARRAY_START(RawObject*,
RawObject*); }
Should we have the non-const data() accessor in this class at all. When is it
ever needed apart in the from()/to() calculation?

https://codereview.chromium.org/673253002/diff/40001/runtime/vm/raw_object.h#...
runtime/vm/raw_object.h:1219: const VariableDesc* VariableDescAddr(intptr_t
index) const {
Why did you have to move this functionality to RawContextScope from
ContextScope?

[koda, 2014-10-29 20:10:27]

https://codereview.chromium.org/673253002/diff/40001/runtime/vm/raw_object.h
File runtime/vm/raw_object.h (right):

https://codereview.chromium.org/673253002/diff/40001/runtime/vm/raw_object.h#...
runtime/vm/raw_object.h:1217: RawObject** data() { OPEN_ARRAY_START(RawObject*,
RawObject*); }
On 2014/10/29 05:17:27, Ivan Posva wrote:
> Should we have the non-const data() accessor in this class at all. When is it
> ever needed apart in the from()/to() calculation?

Changed from/to to use VariableDescAddr.

PTAL

https://codereview.chromium.org/673253002/diff/40001/runtime/vm/raw_object.h#...
runtime/vm/raw_object.h:1219: const VariableDesc* VariableDescAddr(intptr_t
index) const {
On 2014/10/29 05:17:27, Ivan Posva wrote:
> Why did you have to move this functionality to RawContextScope from
> ContextScope?

I wanted to keep the tricky representation details of the
struct/pointers/padding issues contained in this file. For instance, the person
addressing the above TODO will likely want to change VariableDescAddr too.

(And in the updated CL it's also needed for from/to.)

[koda, 2014-11-05 23:01:01]

Ping.

[Ivan Posva, 2014-11-07 00:18:12]

LGTM with comment.

-Ivan

https://codereview.chromium.org/673253002/diff/80001/runtime/vm/raw_object.h
File runtime/vm/raw_object.h (right):

https://codereview.chromium.org/673253002/diff/80001/runtime/vm/raw_object.h#...
runtime/vm/raw_object.h:1228: return reinterpret_cast<RawObject**>(end) - 1;
Pointer arithmetic?

[koda, 2014-11-07 00:51:08]

On 2014/11/07 00:18:12, Ivan Posva wrote:
> LGTM with comment.
> 
> -Ivan
> 
> https://codereview.chromium.org/673253002/diff/80001/runtime/vm/raw_object.h
> File runtime/vm/raw_object.h (right):
> 
>
https://codereview.chromium.org/673253002/diff/80001/runtime/vm/raw_object.h#...
> runtime/vm/raw_object.h:1228: return reinterpret_cast<RawObject**>(end) - 1;
> Pointer arithmetic?

Then, would you also prefer removing pointer arithmetic from all the visitors
that do:

void VisitPointers(RawObject** first, RawObject** last) {
  for (RawObject** current = first; current <= last; current++) {
    ...   

?

[koda, 2014-11-07 01:51:23]

Committed patchset #5 (id:80001) manually as r41592 (presubmit successful).