Handlify the runtime lookup of CallIC and KeyedCallIC.

Handlify the runtime lookup of CallIC and KeyedCallIC.


R=ulan@chromium.org
BUG=
TEST=

Committed: http://code.google.com/p/v8/source/detail?r=9701

[Kevin Millikin (Chromium), 2011-10-19 07:20:30]

http://codereview.chromium.org/8345038/diff/1/src/ic.cc
File src/ic.cc (right):

http://codereview.chromium.org/8345038/diff/1/src/ic.cc#newcode530
src/ic.cc:530: UpdateCaches(&lookup, state, extra_ic_state, object, name);
I stopped short of handlifying the UpdateCaches and the stub cache for call's,
because it's a bit more complicated, so I'll do it as a separate change.

With this change, UpdateCaches will still fail without performing GC if there is
an allocation failure.

http://codereview.chromium.org/8345038/diff/1/src/objects.cc
File src/objects.cc (right):

http://codereview.chromium.org/8345038/diff/1/src/objects.cc#newcode539
src/objects.cc:539: Handle<Object> Object::GetProperty(Handle<Object> object,
I can't decide if I like the isolate argument or not.  It seems a bit nicer that
the caller doesn't have to worry about it.  What do you think?

[ulan, 2011-10-19 08:48:12]

LGTM with comments.

http://codereview.chromium.org/8345038/diff/1/src/api.cc
File src/api.cc (right):

http://codereview.chromium.org/8345038/diff/1/src/api.cc#newcode3099
src/api.cc:3099: &ignored);
I thought we had a convention that we put each argument in a separate line if
they all do not fit in single line.

http://codereview.chromium.org/8345038/diff/1/src/objects.cc
File src/objects.cc (right):

http://codereview.chromium.org/8345038/diff/1/src/objects.cc#newcode539
src/objects.cc:539: Handle<Object> Object::GetProperty(Handle<Object> object,
On 2011/10/19 07:20:30, Kevin Millikin wrote:
> I can't decide if I like the isolate argument or not.  It seems a bit nicer
that
> the caller doesn't have to worry about it.  What do you think?
We are using an implicit invariant that the 'isolate_' member of all IC and stub
compilers is equal to the expression below. If the invariant will stay valid in
future (i.e. other people cannot accidentally break it), then I would also
prefer to drop the isolate argument.

http://codereview.chromium.org/8345038/diff/1/src/runtime.cc
File src/runtime.cc (right):

http://codereview.chromium.org/8345038/diff/1/src/runtime.cc#newcode8133
src/runtime.cc:8133: JSFunction::CompileLazy(function, CLEAR_EXCEPTION);
Do we assume after this line that the function is compiled regardless of
CompileLazy result?