[turbofan] Improve typing rule for modulus.

[turbofan] Improve typing rule for modulus.

Infer a more precise type even in case where NaN and/or -0 is a possible
outcome of the operation, and use this more precise type to improve code
generation for the modulus itself by trying harder to stick to Word32
operations instead of going to Float64, and also optimize the pattern
where we compare the output of x % y to some non-zero integer constant
K, in which case we can truncate the output of x % y to Word32 if the
type of x % y is Signed32/Unsigned32 \/ NaN \/ MinusZero, as NaN and
MinusZero will both be truncated to zero, which cannot match the non
zero constant K.

R=jarin@chromium.org

Committed: https://crrev.com/17a7f5839684872b72d4fe9e6d2cfb5876e85af7
Cr-Commit-Position: refs/heads/master@{#38267}

[Benedikt Meurer, 2016-08-02 18:52:15]

The CQ bit was checked by bmeurer@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-02 18:52:20]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Benedikt Meurer, 2016-08-02 19:06:00]

Hey Jaro,

Here's some nice optimization for the various modulus operators in the Kraken
crypto benchmarks. Gives a nice speedup with the loop variable analysis
(avoiding the soft deopts or the expensive Float64Mods altogether).
Please take a look.

Thanks,
Benedikt

[commit-bot: I haz the power, 2016-08-02 19:14:06]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-02 19:14:07]

Dry run: This issue passed the CQ dry run.

[Jarin, 2016-08-02 20:02:09]

lgtm

https://codereview.chromium.org/2202413002/diff/1/src/compiler/simplified-low...
File src/compiler/simplified-lowering.cc (right):

https://codereview.chromium.org/2202413002/diff/1/src/compiler/simplified-low...
src/compiler/simplified-lowering.cc:680: bool OneInputCannotBe(Node* node, Type*
type) {
I have to say this: Somehow, I see you adding new helpers, and at the same time
complaining that there are too many helpers...

[Benedikt Meurer, 2016-08-03 03:55:09]

The CQ bit was checked by bmeurer@chromium.org

[Benedikt Meurer, 2016-08-03 03:55:10]

The patchset sent to the CQ was uploaded after l-g-t-m from jarin@chromium.org
Link to the patchset: https://codereview.chromium.org/2202413002/#ps20001
(title: "Some cleanup plus use type feedback properly.")

[commit-bot: I haz the power, 2016-08-03 03:55:21]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-03 04:39:02]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-08-03 04:41:44]

Description was changed from

==========
[turbofan] Improve typing rule for modulus.

Infer a more precise type even in case where NaN and/or -0 is a possible
outcome of the operation, and use this more precise type to improve code
generation for the modulus itself by trying harder to stick to Word32
operations instead of going to Float64, and also optimize the pattern
where we compare the output of x % y to some non-zero integer constant
K, in which case we can truncate the output of x % y to Word32 if the
type of x % y is Signed32/Unsigned32 \/ NaN \/ MinusZero, as NaN and
MinusZero will both be truncated to zero, which cannot match the non
zero constant K.

R=jarin@chromium.org
==========

to

==========
[turbofan] Improve typing rule for modulus.

Infer a more precise type even in case where NaN and/or -0 is a possible
outcome of the operation, and use this more precise type to improve code
generation for the modulus itself by trying harder to stick to Word32
operations instead of going to Float64, and also optimize the pattern
where we compare the output of x % y to some non-zero integer constant
K, in which case we can truncate the output of x % y to Word32 if the
type of x % y is Signed32/Unsigned32 \/ NaN \/ MinusZero, as NaN and
MinusZero will both be truncated to zero, which cannot match the non
zero constant K.

R=jarin@chromium.org

Committed: https://crrev.com/17a7f5839684872b72d4fe9e6d2cfb5876e85af7
Cr-Commit-Position: refs/heads/master@{#38267}
==========

[commit-bot: I haz the power, 2016-08-03 04:41:45]

Patchset 2 (id:??) landed as
https://crrev.com/17a7f5839684872b72d4fe9e6d2cfb5876e85af7
Cr-Commit-Position: refs/heads/master@{#38267}