Make ToBooleanStub more consistent across platforms.

Make ToBooleanStub more consistent across platforms.

The declaration of the ToBoolean class moved to the platform-independent part
and its implementations are now structurally very similar. This is just an
intermediate cleanup step to add type recording at the call site.

Note that the MIPS implementation has not really been touched, so it should
continue to work, too.
Committed: http://code.google.com/p/v8/source/detail?r=8359

[Mads Ager (chromium), 2011-06-22 07:43:24]

LGTM with the ARM (and probably MIPS) conditions changed.

http://codereview.chromium.org/7218012/diff/1/src/arm/code-stubs-arm.cc
File src/arm/code-stubs-arm.cc (right):

http://codereview.chromium.org/7218012/diff/1/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:1661: __ Ret(gt);
Not your change, but this should be ge, right? Otherwise we will not handle
JS_VALUE_TYPE == FIRST_SPEC_OBJECT_TYPE here. I think we will handle it
correctly below because it is not a string and not a heap number, but let's do
it here instead.

http://codereview.chromium.org/7218012/diff/1/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:1665: __ b(&not_string, gt);
Again, this should be ge to accound for FIRST_NONSTRING_TYPE not being a string
(it is a map, so it shouldn't happen, but still).

[Sven Panne, 2011-06-22 08:02:59]

http://codereview.chromium.org/7218012/diff/1/src/arm/code-stubs-arm.cc
File src/arm/code-stubs-arm.cc (right):

http://codereview.chromium.org/7218012/diff/1/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:1661: __ Ret(gt);
On 2011/06/22 07:43:24, Mads Ager wrote:
> Not your change, but this should be ge, right? Otherwise we will not handle
> JS_VALUE_TYPE == FIRST_SPEC_OBJECT_TYPE here. I think we will handle it
> correctly below because it is not a string and not a heap number, but let's do
> it here instead.

Done, and for the corresponding test in code-stubs-mips.cc, too.

http://codereview.chromium.org/7218012/diff/1/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:1665: __ b(&not_string, gt);
On 2011/06/22 07:43:24, Mads Ager wrote:
> Again, this should be ge to accound for FIRST_NONSTRING_TYPE not being a
string
> (it is a map, so it shouldn't happen, but still).

Done, and for the corresponding test in code-stubs-mips.cc, too.