ARM: Add shift operations to the type recording binary operation stub

ARM: Add shift operations to the type recording binary operation stub

Committed: http://code.google.com/p/v8/source/detail?r=6737

[Mads Ager (chromium), 2011-02-10 16:27:45]

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc
File src/arm/code-stubs-arm.cc (right):

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:2778: // Use only the 5 least significant bits of the
shift count.
You have a macro assembler thing for this that you use above?

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:2784: __ and_(r2, r2, Operand(0x1f));
Ditto?

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:2798: __ and_(r2, r2, Operand(0x1f));
And here.

http://codereview.chromium.org/6471023/diff/5/src/arm/full-codegen-arm.cc
File src/arm/full-codegen-arm.cc (right):

http://codereview.chromium.org/6471023/diff/5/src/arm/full-codegen-arm.cc#new...
src/arm/full-codegen-arm.cc:1616: __ CallStub(&stub);
Do you have to be careful that you do not accidentally get something after this
call that marks it as having inlined smi code?

http://codereview.chromium.org/6471023/diff/5/src/arm/lithium-arm.cc
File src/arm/lithium-arm.cc (right):

http://codereview.chromium.org/6471023/diff/5/src/arm/lithium-arm.cc#newcode189
src/arm/lithium-arm.cc:189: case Token::SHL: return "sal-t";
sal -> shl

[Søren Thygesen Gjesse, 2011-02-10 20:22:17]

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc
File src/arm/code-stubs-arm.cc (right):

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:2778: // Use only the 5 least significant bits of the
shift count.
On 2011/02/10 16:27:45, Mads Ager wrote:
> You have a macro assembler thing for this that you use above?

The one above takes a smi, but here we have an int32. Added
GetLeastBitsFromInt32 to the macro assembler for clarity.

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:2784: __ and_(r2, r2, Operand(0x1f));
On 2011/02/10 16:27:45, Mads Ager wrote:
> Ditto?

Ditto.

http://codereview.chromium.org/6471023/diff/5/src/arm/code-stubs-arm.cc#newco...
src/arm/code-stubs-arm.cc:2798: __ and_(r2, r2, Operand(0x1f));
On 2011/02/10 16:27:45, Mads Ager wrote:
> And here.

Ditto.

http://codereview.chromium.org/6471023/diff/5/src/arm/full-codegen-arm.cc
File src/arm/full-codegen-arm.cc (right):

http://codereview.chromium.org/6471023/diff/5/src/arm/full-codegen-arm.cc#new...
src/arm/full-codegen-arm.cc:1616: __ CallStub(&stub);
On 2011/02/10 16:27:45, Mads Ager wrote:
> Do you have to be careful that you do not accidentally get something after
this
> call that marks it as having inlined smi code?

Thats true, used EmitCallIC(stub.GetCode(), NULL) as there currently is no
inlining of smi binary operaions in the ARM full codegen.

http://codereview.chromium.org/6471023/diff/5/src/arm/lithium-arm.cc
File src/arm/lithium-arm.cc (right):

http://codereview.chromium.org/6471023/diff/5/src/arm/lithium-arm.cc#newcode189
src/arm/lithium-arm.cc:189: case Token::SHL: return "sal-t";
On 2011/02/10 16:27:45, Mads Ager wrote:
> sal -> shl

Done.

[Mads Ager (chromium), 2011-02-10 20:23:28]

LGTM