Fix issue with -0 in Maps

Fix issue with -0 in Maps

Because we generated a different hash code for 0 and -0 we ended up
not even getting to the SameValueZero check.

BUG=v8:3906
LOG=N
R=adamk

Committed: https://crrev.com/925364f5b447b42c5af748896e552d6e918adca3
Cr-Commit-Position: refs/heads/master@{#26787}

[arv (Not doing code reviews), 2015-02-20 19:19:08]

PTAL

https://codereview.chromium.org/947443005/diff/1/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/947443005/diff/1/src/objects.cc#newcode801
src/objects.cc:801: if (IsSmi()) {
IsNumber() checked IsSmi() || IsHeapNumber()

If it is a Smi it cannot be a NaN or -0.

https://codereview.chromium.org/947443005/diff/1/src/objects.cc#newcode803
src/objects.cc:803: uint32_t hash =
ComputeLongHash(double_to_uint64(static_cast<double>(num)));
I tried a few other approaches here but it is important that we compute the same
hash code here as in the heap number case.

[adamk, 2015-02-20 19:50:04]

https://codereview.chromium.org/947443005/diff/1/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/947443005/diff/1/src/objects.cc#newcode803
src/objects.cc:803: uint32_t hash =
ComputeLongHash(double_to_uint64(static_cast<double>(num)));
On 2015/02/20 19:19:08, arv wrote:
> I tried a few other approaches here but it is important that we compute the
same
> hash code here as in the heap number case.

This is fine for now, and the right approach for this patch (changing only one
thing at a time) but soon I'll be creating a patch to use a 32-bit hash function
for Smis, and have code in the IsHeapNumber branch to compute the same 32-bit
hash for Smi-representable doubles.

https://codereview.chromium.org/947443005/diff/1/src/objects.cc#newcode809
src/objects.cc:809: if (i::IsMinusZero(num)) num = 0;
Nit: shouldn't need the "i::" prefix here.

https://codereview.chromium.org/947443005/diff/1/test/mjsunit/es6/map-set-min...
File test/mjsunit/es6/map-set-minus-zero.js (right):

https://codereview.chromium.org/947443005/diff/1/test/mjsunit/es6/map-set-min...
test/mjsunit/es6/map-set-minus-zero.js:1: // Copyright 2014 the V8 project
authors. All rights reserved.
Nit: 2015

https://codereview.chromium.org/947443005/diff/1/test/mjsunit/es6/map-set-min...
test/mjsunit/es6/map-set-minus-zero.js:7: var t = new Map();
Please add a Set codepath as well.

https://codereview.chromium.org/947443005/diff/1/test/mjsunit/es6/map-set-min...
test/mjsunit/es6/map-set-minus-zero.js:20: assertEquals(t.size, 0);
Please invert all the assertEquals() calls in this file: format is
assertEquals(expected, actual).

[arv (Not doing code reviews), 2015-02-20 20:01:00]

Add set test and cleanup test

[arv (Not doing code reviews), 2015-02-20 20:01:33]

fix year

[arv (Not doing code reviews), 2015-02-20 20:03:18]

rename file

[arv (Not doing code reviews), 2015-02-20 20:04:12]

remove extra file

[arv (Not doing code reviews), 2015-02-20 20:04:30]

Patchset #2 (id:20001) has been deleted

[arv (Not doing code reviews), 2015-02-20 20:04:37]

Patchset #2 (id:40001) has been deleted

[arv (Not doing code reviews), 2015-02-20 20:04:44]

Patchset #2 (id:60001) has been deleted

[arv (Not doing code reviews), 2015-02-20 20:05:34]

PTAL

https://codereview.chromium.org/947443005/diff/1/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/947443005/diff/1/src/objects.cc#newcode809
src/objects.cc:809: if (i::IsMinusZero(num)) num = 0;
On 2015/02/20 19:50:03, adamk wrote:
> Nit: shouldn't need the "i::" prefix here.

It is needed. There is a class function with the same name.

[adamk, 2015-02-20 20:11:24]

lgtm

[arv (Not doing code reviews), 2015-02-20 20:26:36]

The CQ bit was checked by arv@chromium.org

[commit-bot: I haz the power, 2015-02-20 20:27:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/947443005/80001

[commit-bot: I haz the power, 2015-02-20 20:27:37]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/947443005/80001

[commit-bot: I haz the power, 2015-02-20 21:02:41]

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

[commit-bot: I haz the power, 2015-02-20 21:02:42]

Failed to apply the patch.

[commit-bot: I haz the power, 2015-02-20 21:02:46]

Committed patchset #2 (id:80001)

[commit-bot: I haz the power, 2015-02-20 21:03:09]

Patchset 2 (id:??) landed as
https://crrev.com/925364f5b447b42c5af748896e552d6e918adca3
Cr-Commit-Position: refs/heads/master@{#26787}