Cleanup some of the range uses in ModI/DivI.

Cleanup some of the range uses in ModI/DivI.

BUG=v8:3204
LOG=y
R=svenpanne@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=19796

[Benedikt Meurer, 2014-03-10 12:56:42]

PTAL

[Sven Panne, 2014-03-11 11:32:42]

LGTM with nits

https://codereview.chromium.org/191293013/diff/50001/src/a64/lithium-codegen-...
File src/a64/lithium-codegen-a64.cc (right):

https://codereview.chromium.org/191293013/diff/50001/src/a64/lithium-codegen-...
src/a64/lithium-codegen-a64.cc:2621: Abs(divisor) != 1) {
Hmmm, I consciously did not use Abs here (and similar places) because I wanted
to avoid Abs(kMinInt). This "works" currently, but not if we add an ASSERT in
Abs later.

I would prefer the previous version, but it's up to you...

https://codereview.chromium.org/191293013/diff/50001/src/hydrogen-instructions.h
File src/hydrogen-instructions.h (right):

https://codereview.chromium.org/191293013/diff/50001/src/hydrogen-instruction...
src/hydrogen-instructions.h:858: bool RangeCanInclude(int value) const {
Can't we kill this, too?

[Benedikt Meurer, 2014-03-11 11:56:35]

https://codereview.chromium.org/191293013/diff/50001/src/a64/lithium-codegen-...
File src/a64/lithium-codegen-a64.cc (right):

https://codereview.chromium.org/191293013/diff/50001/src/a64/lithium-codegen-...
src/a64/lithium-codegen-a64.cc:2621: Abs(divisor) != 1) {
On 2014/03/11 11:32:42, Sven Panne wrote:
> Hmmm, I consciously did not use Abs here (and similar places) because I wanted
> to avoid Abs(kMinInt). This "works" currently, but not if we add an ASSERT in
> Abs later.
> 
> I would prefer the previous version, but it's up to you...

Ok, makes sense.

https://codereview.chromium.org/191293013/diff/50001/src/hydrogen-instructions.h
File src/hydrogen-instructions.h (right):

https://codereview.chromium.org/191293013/diff/50001/src/hydrogen-instruction...
src/hydrogen-instructions.h:858: bool RangeCanInclude(int value) const {
I'm on it...

[Benedikt Meurer, 2014-03-11 11:57:39]

Committed patchset #6 manually as r19796 (presubmit successful).