Fix materialization of arguments objects with unknown values.

Fix materialization of arguments objects with unknown values.

This fixes the deoptimizer to materialize arguments objects of correct
length even in cases where the actual argument values are unknown and
were optimized away by Crankshaft. This can happen if only the length
property or the identity of an arguments object is used.

R=svenpanne@chromium.org
BUG=chromium:163530
TEST=mjsunit/regress/regress-crbug-163530

Committed: http://code.google.com/p/v8/source/detail?r=13763

[Sven Panne, 2013-02-27 14:27:23]

LGTM with a nit. General note: AddToTranslation's signature with 3 bools and 2
ints is becoming a bit fragile and easy to use incorrectly. OTOH, insisting on
separate enums for bools while using plain old ints for more than 100 different
things all over v8 is ridiculous, too, so I think the CL is OK. :-)

https://codereview.chromium.org/12335132/diff/1/test/mjsunit/regress/regress-...
File test/mjsunit/regress/regress-crbug-163530.js (right):

https://codereview.chromium.org/12335132/diff/1/test/mjsunit/regress/regress-...
test/mjsunit/regress/regress-crbug-163530.js:63: object.a = function A(x,y,z) {
nit: spaces after commas here and below

[Michael Starzinger, 2013-02-27 14:34:01]

Addressed comments. Landing.

I agree that the signature of AddToTranslation() is getting complex and hard to
use. But in absence of a better alternative and since this method is also
tightly coupled with WriteTranslation() which is the only call-site, I also
agree that it's probably OK for this particular CL.

https://codereview.chromium.org/12335132/diff/1/test/mjsunit/regress/regress-...
File test/mjsunit/regress/regress-crbug-163530.js (right):

https://codereview.chromium.org/12335132/diff/1/test/mjsunit/regress/regress-...
test/mjsunit/regress/regress-crbug-163530.js:63: object.a = function A(x,y,z) {
On 2013/02/27 14:27:23, Sven Panne wrote:
> nit: spaces after commas here and below

Done. Fixed throughout this file.

[Michael Starzinger, 2013-02-27 14:38:01]

Committed patchset #3 manually as r13763 (presubmit successful).