Refactor ExternalReference::isolate_address() to not rely on Isolate::Current().

Refactor ExternalReference::isolate_address() to not rely on Isolate::Current().

R=svenpanne@chromium.org
BUG=

Committed: https://code.google.com/p/v8/source/detail?r=14400

[Sven Panne, 2013-04-23 13:18:08]

https://codereview.chromium.org/14429003/diff/1/src/arm/regexp-macro-assemble...
File src/arm/regexp-macro-assembler-arm.cc (right):

https://codereview.chromium.org/14429003/diff/1/src/arm/regexp-macro-assemble...
src/arm/regexp-macro-assembler-arm.cc:383: __ mov(r3,
Operand(ExternalReference::isolate_address(masm_->isolate())));
Add a

   Isolate* RegExpMacroAssemblerARM::isolate() const { return masm_->isolate();
}

here and at the other RegExpMacroAssemblers, we did similar things for other
classes. It's a good thing that the caller doesn't need to know the details
where to fetch the Isolate.

https://codereview.chromium.org/14429003/diff/1/src/arm/stub-cache-arm.cc
File src/arm/stub-cache-arm.cc (right):

https://codereview.chromium.org/14429003/diff/1/src/arm/stub-cache-arm.cc#new...
src/arm/stub-cache-arm.cc:1298: __ Move(scratch3(),
Handle<Object>(callback->data(), masm()->isolate()));
StubCompiler has an isolate() method, so use this here and elsewhere.

https://codereview.chromium.org/14429003/diff/1/src/x64/deoptimizer-x64.cc
File src/x64/deoptimizer-x64.cc (right):

https://codereview.chromium.org/14429003/diff/1/src/x64/deoptimizer-x64.cc#ne...
src/x64/deoptimizer-x64.cc:656: Isolate* isolate = masm()->isolate();
Again: Add an isolate() method to EntryGenerator (here and on other platforms).

[Yang, 2013-04-23 14:45:31]

On 2013/04/23 13:18:08, Sven Panne wrote:
>
https://codereview.chromium.org/14429003/diff/1/src/arm/regexp-macro-assemble...
> File src/arm/regexp-macro-assembler-arm.cc (right):
> 
>
https://codereview.chromium.org/14429003/diff/1/src/arm/regexp-macro-assemble...
> src/arm/regexp-macro-assembler-arm.cc:383: __ mov(r3,
> Operand(ExternalReference::isolate_address(masm_->isolate())));
> Add a
> 
>    Isolate* RegExpMacroAssemblerARM::isolate() const { return
masm_->isolate();
> }
> 
> here and at the other RegExpMacroAssemblers, we did similar things for other
> classes. It's a good thing that the caller doesn't need to know the details
> where to fetch the Isolate.
> 
> https://codereview.chromium.org/14429003/diff/1/src/arm/stub-cache-arm.cc
> File src/arm/stub-cache-arm.cc (right):
> 
>
https://codereview.chromium.org/14429003/diff/1/src/arm/stub-cache-arm.cc#new...
> src/arm/stub-cache-arm.cc:1298: __ Move(scratch3(),
> Handle<Object>(callback->data(), masm()->isolate()));
> StubCompiler has an isolate() method, so use this here and elsewhere.
> 
> https://codereview.chromium.org/14429003/diff/1/src/x64/deoptimizer-x64.cc
> File src/x64/deoptimizer-x64.cc (right):
> 
>
https://codereview.chromium.org/14429003/diff/1/src/x64/deoptimizer-x64.cc#ne...
> src/x64/deoptimizer-x64.cc:656: Isolate* isolate = masm()->isolate();
> Again: Add an isolate() method to EntryGenerator (here and on other
platforms).

Addressed comments. Please take another look.

[Sven Panne, 2013-04-24 06:29:46]

LGTM with a few nits.

https://codereview.chromium.org/14429003/diff/6001/src/arm/regexp-macro-assem...
File src/arm/regexp-macro-assembler-arm.h (right):

https://codereview.chromium.org/14429003/diff/6001/src/arm/regexp-macro-assem...
src/arm/regexp-macro-assembler-arm.h:227: inline Isolate* isolate() { return
masm_->isolate(); }
Remove "inline", add "const".

https://codereview.chromium.org/14429003/diff/6001/src/ia32/regexp-macro-asse...
File src/ia32/regexp-macro-assembler-ia32.h (right):

https://codereview.chromium.org/14429003/diff/6001/src/ia32/regexp-macro-asse...
src/ia32/regexp-macro-assembler-ia32.h:200: inline Isolate* isolate() const {
return masm_->isolate(); }
I don't think "inline" is really needed here, we add this far too often, anyway.

https://codereview.chromium.org/14429003/diff/6001/src/mips/regexp-macro-asse...
File src/mips/regexp-macro-assembler-mips.h (right):

https://codereview.chromium.org/14429003/diff/6001/src/mips/regexp-macro-asse...
src/mips/regexp-macro-assembler-mips.h:232: inline Isolate* isolate() { return
masm_->isolate(); }
Remove "inline", add "const.

https://codereview.chromium.org/14429003/diff/6001/src/x64/regexp-macro-assem...
File src/x64/regexp-macro-assembler-x64.h (right):

https://codereview.chromium.org/14429003/diff/6001/src/x64/regexp-macro-assem...
src/x64/regexp-macro-assembler-x64.h:275: inline Isolate* isolate() const {
return masm_.isolate(); }
Remove "inline".

[Yang, 2013-04-24 07:38:53]

On 2013/04/24 06:29:46, Sven Panne wrote:
> LGTM with a few nits.
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/arm/regexp-macro-assem...
> File src/arm/regexp-macro-assembler-arm.h (right):
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/arm/regexp-macro-assem...
> src/arm/regexp-macro-assembler-arm.h:227: inline Isolate* isolate() { return
> masm_->isolate(); }
> Remove "inline", add "const".
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/ia32/regexp-macro-asse...
> File src/ia32/regexp-macro-assembler-ia32.h (right):
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/ia32/regexp-macro-asse...
> src/ia32/regexp-macro-assembler-ia32.h:200: inline Isolate* isolate() const {
> return masm_->isolate(); }
> I don't think "inline" is really needed here, we add this far too often,
anyway.
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/mips/regexp-macro-asse...
> File src/mips/regexp-macro-assembler-mips.h (right):
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/mips/regexp-macro-asse...
> src/mips/regexp-macro-assembler-mips.h:232: inline Isolate* isolate() { return
> masm_->isolate(); }
> Remove "inline", add "const.
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/x64/regexp-macro-assem...
> File src/x64/regexp-macro-assembler-x64.h (right):
> 
>
https://codereview.chromium.org/14429003/diff/6001/src/x64/regexp-macro-assem...
> src/x64/regexp-macro-assembler-x64.h:275: inline Isolate* isolate() const {
> return masm_.isolate(); }
> Remove "inline".

Done.

[Yang, 2013-04-24 07:39:54]

Committed patchset #3 manually as r14400 (presubmit successful).