Some cleanup fixes

Some cleanup fixes

* Consolidated CopyJSObject and CopyJSObjectWithAllocationSite
* Factory.h helper for struct maps.
* BuildFastLiteral shouldn't create allocation sites in pretenuring mode.

BUG=
R=mstarzinger@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=16855

[mvstanton, 2013-09-19 15:33:43]

Hi guys, here most of the changes from a review of the allocsite code Toon and I
did Monday, and comments from Michael on Wednesday. Thx for having a look,
--Michael

[Michael Starzinger, 2013-09-19 16:45:25]

LGTM with nit.

https://codereview.chromium.org/24255005/diff/6001/src/heap.h
File src/heap.h (left):

https://codereview.chromium.org/24255005/diff/6001/src/heap.h#oldcode1010
src/heap.h:1010: // Allocates a new utility object in the old generation.
I liked that comment, can we get it back?

[mvstanton, 2013-09-20 09:19:15]

thx, I added the comment back!
--Michael

https://codereview.chromium.org/24255005/diff/6001/src/heap.h
File src/heap.h (left):

https://codereview.chromium.org/24255005/diff/6001/src/heap.h#oldcode1010
src/heap.h:1010: // Allocates a new utility object in the old generation.
On 2013/09/19 16:45:25, Michael Starzinger wrote:
> I liked that comment, can we get it back?

absolutely, it was removed because for a while I was allowing creation of the
structs in the new generation, but as we discussed with always_allocate(), the
scenario is not worth it.

[mvstanton, 2013-09-20 09:27:55]

Committed patchset #4 manually as r16855 (presubmit successful).