Make room in Code flags to support compiled stubs.

Make room in Code flags to support compiled stubs.

R=mstarzinger@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=12008

[Michael Starzinger, 2012-07-06 16:22:19]

First round of comments.

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects-inl.h
File src/objects-inl.h (right):

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects-inl.h#n...
src/objects-inl.h:3069: StubMajorKeyField::update(previous, major));
Indentation seems off. Can we also pull this out into a local "updated" variable
to be consistent with the other setters?

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects-inl.h#n...
src/objects-inl.h:3294: READ_UINT32_FIELD(this, kKindSpecificFlags1Offset)) +
Token::EQ;
Can we move the addition of Token::EQ into CompareIC::ComputeOperation(), so
that all the magic is in there?

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects-inl.h#n...
src/objects-inl.h:3300: Token::IsCompareOp(static_cast<Token::Value>(value));
There is probably an ASSERT missing around that call.

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects-inl.h#n...
src/objects-inl.h:3301: value -= Token::EQ;
Can we move the substraction of Token::EQ into ICCompareStub::FinishCode to have
all the magic in there.

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects.h
File src/objects.h (right):

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects.h#newco...
src/objects.h:4531: static const int kHeaderPaddingStart =
kKindSpecificFlags2Offset;
Hmm, I am not sure, is that really the right offset?

https://chromiumcodereview.appspot.com/10701106/diff/5002/src/objects.h#newco...
src/objects.h:4538: // Byte offsets within kKindSpecificFlagsOffset.
Adapt the comment to include the number.

[danno, 2012-07-06 16:36:00]

Feedback addressed

http://codereview.chromium.org/10701106/diff/5002/src/objects-inl.h
File src/objects-inl.h (right):

http://codereview.chromium.org/10701106/diff/5002/src/objects-inl.h#newcode3069
src/objects-inl.h:3069: StubMajorKeyField::update(previous, major));
On 2012/07/06 16:22:19, Michael Starzinger wrote:
> Indentation seems off. Can we also pull this out into a local "updated"
variable
> to be consistent with the other setters?

Done.

http://codereview.chromium.org/10701106/diff/5002/src/objects-inl.h#newcode3294
src/objects-inl.h:3294: READ_UINT32_FIELD(this, kKindSpecificFlags1Offset)) +
Token::EQ;
On 2012/07/06 16:22:19, Michael Starzinger wrote:
> Can we move the addition of Token::EQ into CompareIC::ComputeOperation(), so
> that all the magic is in there?

Done.

http://codereview.chromium.org/10701106/diff/5002/src/objects-inl.h#newcode3300
src/objects-inl.h:3300: Token::IsCompareOp(static_cast<Token::Value>(value));
On 2012/07/06 16:22:19, Michael Starzinger wrote:
> There is probably an ASSERT missing around that call.

Done.

http://codereview.chromium.org/10701106/diff/5002/src/objects-inl.h#newcode3301
src/objects-inl.h:3301: value -= Token::EQ;
On 2012/07/06 16:22:19, Michael Starzinger wrote:
> Can we move the substraction of Token::EQ into ICCompareStub::FinishCode to
have
> all the magic in there.

Done.

http://codereview.chromium.org/10701106/diff/5002/src/objects.h
File src/objects.h (right):

http://codereview.chromium.org/10701106/diff/5002/src/objects.h#newcode4531
src/objects.h:4531: static const int kHeaderPaddingStart =
kKindSpecificFlags2Offset;
On 2012/07/06 16:22:19, Michael Starzinger wrote:
> Hmm, I am not sure, is that really the right offset?

Done.

http://codereview.chromium.org/10701106/diff/5002/src/objects.h#newcode4538
src/objects.h:4538: // Byte offsets within kKindSpecificFlagsOffset.
On 2012/07/06 16:22:19, Michael Starzinger wrote:
> Adapt the comment to include the number.

Done.

[Michael Starzinger, 2012-07-06 16:52:12]

LGTM (with a few final nits).

https://chromiumcodereview.appspot.com/10701106/diff/9002/src/objects.h
File src/objects.h (right):

https://chromiumcodereview.appspot.com/10701106/diff/9002/src/objects.h#newco...
src/objects.h:4557: // KindSpecificFlags1 layout (STUB and OPTIMIZED)
Can we change "OPTIMIZED" to "OPTIMIZED_FUNCTION" so that it shows up when
grepping for it?

https://chromiumcodereview.appspot.com/10701106/diff/9002/src/objects.h#newco...
src/objects.h:4609: // KindSpecificFlags2 layout (STUB and OPTIMIZED)
Likewise.

https://chromiumcodereview.appspot.com/10701106/diff/9002/src/objects.h#newco...
src/objects.h:4625: // KindSpecificFlags[1] layout (FUNCTION)
Can we drop the square brackets?

[danno, 2012-07-06 22:05:30]

http://codereview.chromium.org/10701106/diff/9002/src/objects.h
File src/objects.h (right):

http://codereview.chromium.org/10701106/diff/9002/src/objects.h#newcode4557
src/objects.h:4557: // KindSpecificFlags1 layout (STUB and OPTIMIZED)
On 2012/07/06 16:52:12, Michael Starzinger wrote:
> Can we change "OPTIMIZED" to "OPTIMIZED_FUNCTION" so that it shows up when
> grepping for it?

Done.

http://codereview.chromium.org/10701106/diff/9002/src/objects.h#newcode4609
src/objects.h:4609: // KindSpecificFlags2 layout (STUB and OPTIMIZED)
On 2012/07/06 16:52:12, Michael Starzinger wrote:
> Likewise.

Done.

http://codereview.chromium.org/10701106/diff/9002/src/objects.h#newcode4625
src/objects.h:4625: // KindSpecificFlags[1] layout (FUNCTION)
On 2012/07/06 16:52:12, Michael Starzinger wrote:
> Can we drop the square brackets?

Done.