ARM: replace RegExpCEntryStub with DirectCEntryStub.

ARM: replace RegExpCEntryStub with DirectCEntryStub.

RegExpCEntryStub is therefore removed.

BUG=none
TEST=none
R=bmeurer@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=16618

[Rodolph Perfetta, 2013-09-10 10:53:03]

RegExpCEntryStub and DirectCEntryStub provide the same functionality.
DirectCEntryStub has also been updated to be more friendly with branch
predictors.

[Benedikt Meurer, 2013-09-10 11:07:06]

LGTM with nits.

https://codereview.chromium.org/23468015/diff/1/src/arm/regexp-macro-assemble...
File src/arm/regexp-macro-assembler-arm.cc (right):

https://codereview.chromium.org/23468015/diff/1/src/arm/regexp-macro-assemble...
src/arm/regexp-macro-assembler-arm.cc:1067: if (stack_alignment < kPointerSize)
stack_alignment = kPointerSize;
Nit: I think we should simply ASSERT(IsAligned(stack_alignment, kPointerSize))
here. Otherwise to be 100% correct, we would have to make sure that kPointerSize
is a multiple of stack_alignment.

https://codereview.chromium.org/23468015/diff/1/src/arm/regexp-macro-assemble...
src/arm/regexp-macro-assembler-arm.cc:1082: if (OS::ActivationFrameAlignment()
!= 0) {
Nit: We can also safely assert that the stack_alignment is never 0.

[Rodolph Perfetta, 2013-09-10 13:50:00]

https://codereview.chromium.org/23468015/diff/1/src/arm/regexp-macro-assemble...
File src/arm/regexp-macro-assembler-arm.cc (right):

https://codereview.chromium.org/23468015/diff/1/src/arm/regexp-macro-assemble...
src/arm/regexp-macro-assembler-arm.cc:1067: if (stack_alignment < kPointerSize)
stack_alignment = kPointerSize;
On 2013/09/10 11:07:06, Benedikt Meurer wrote:
> Nit: I think we should simply ASSERT(IsAligned(stack_alignment, kPointerSize))
> here. Otherwise to be 100% correct, we would have to make sure that
kPointerSize
> is a multiple of stack_alignment.

Done.

https://codereview.chromium.org/23468015/diff/1/src/arm/regexp-macro-assemble...
src/arm/regexp-macro-assembler-arm.cc:1082: if (OS::ActivationFrameAlignment()
!= 0) {
On 2013/09/10 11:07:06, Benedikt Meurer wrote:
> Nit: We can also safely assert that the stack_alignment is never 0.

Done.

[Rodolph Perfetta, 2013-09-10 13:50:39]

Committed patchset #1 manually as r16618 (presubmit successful).