Three possible minor possible optimizations on ARM.

Alexandre


http://codereview.chromium.org/7350021/diff/16001/src/arm/lithium-codegen-arm.cc
File src/arm/lithium-codegen-arm.cc (right):

http://codereview.chromium.org/7350021/diff/16001/src/arm/lithium-codegen-arm.cc#newcode2458
src/arm/lithium-codegen-arm.cc:2458: __ add(elements, elements,
operand);
If key is constant this add instruction can be merged with the one
below.

http://codereview.chromium.org/7350021/diff/16001/src/arm/lithium-codegen-arm.cc#newcode3306
src/arm/lithium-codegen-arm.cc:3306: __ add(elements, elements,
operand);
If the key is constant, these two add instructions can be merged into
one.

http://codereview.chromium.org/7350021/diff/16001/src/arm/lithium-codegen-arm.cc#newcode3314
src/arm/lithium-codegen-arm.cc:3314: __ Vmov(value,
FixedDoubleArray::canonical_not_the_hole_nan_as_double());
You can spare the previous branch by making this vmov conditional.

http://codereview.chromium.org/7350021/

--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev

Reply via email to