[x64] Use vcvtlsi2sd when AVX is enabled

[x64] Use vcvtlsi2sd when AVX is enabled

BUG=v8:4406
LOG=N

Committed: https://crrev.com/adcbe619a959fe1d8f21d06fbf5984868c4f6b9a
Cr-Commit-Position: refs/heads/master@{#31276}

Committed: https://crrev.com/7cdcf0b1bc45777b48923e9758a3a30e8620c1a1
Cr-Commit-Position: refs/heads/master@{#31315}

[alph, 2015-10-15 05:12:12]

alph@chromium.org changed reviewers:
+ bmeurer@chromium.org, danno@chromium.org, yurys@chromium.org

[Benedikt Meurer, 2015-10-15 05:13:47]

LGTM with nits/questions.

https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64.cc
File src/x64/macro-assembler-x64.cc (right):

https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64...
src/x64/macro-assembler-x64.cc:796: xorps(dst, dst);
Nit: Use xorpd for consistency here?

https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64...
src/x64/macro-assembler-x64.cc:808: xorps(dst, dst);
Nit: Use xorpd for consistency here?

[alph, 2015-10-15 05:20:07]

https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64.cc
File src/x64/macro-assembler-x64.cc (right):

https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64...
src/x64/macro-assembler-x64.cc:796: xorps(dst, dst);
On 2015/10/15 05:13:47, Benedikt Meurer wrote:
> Nit: Use xorpd for consistency here?

Ok.
I didn't want to touch legacy code in this patch.
In general I plan to replace all these xorps'es across entire code. I remember
from my previous compiler experience that mixing float types has some
performance penalties on certain AMD CPUs.

[Benedikt Meurer, 2015-10-15 05:20:37]

On 2015/10/15 05:20:07, alph wrote:
>
https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64.cc
> File src/x64/macro-assembler-x64.cc (right):
> 
>
https://codereview.chromium.org/1404903004/diff/1/src/x64/macro-assembler-x64...
> src/x64/macro-assembler-x64.cc:796: xorps(dst, dst);
> On 2015/10/15 05:13:47, Benedikt Meurer wrote:
> > Nit: Use xorpd for consistency here?
> 
> Ok.
> I didn't want to touch legacy code in this patch.
> In general I plan to replace all these xorps'es across entire code. I remember
> from my previous compiler experience that mixing float types has some
> performance penalties on certain AMD CPUs.

Ok, sounds good.

[alph, 2015-10-15 05:30:57]

The CQ bit was checked by alph@chromium.org

[commit-bot: I haz the power, 2015-10-15 05:31:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1404903004/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1404903004/1

[commit-bot: I haz the power, 2015-10-15 05:53:21]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-10-15 05:53:32]

Patchset 1 (id:??) landed as
https://crrev.com/adcbe619a959fe1d8f21d06fbf5984868c4f6b9a
Cr-Commit-Position: refs/heads/master@{#31276}

[alph, 2015-10-15 06:08:08]

A revert of this CL (patchset #1 id:1) has been created in
https://codereview.chromium.org/1396283004/ by alph@chromium.org.

The reason for reverting is: Caused a crash on Windows.

[alph, 2015-10-15 16:48:08]

The CQ bit was checked by alph@chromium.org

[alph, 2015-10-15 16:48:08]

The patchset sent to the CQ was uploaded after l-g-t-m from bmeurer@chromium.org
Link to the patchset: https://codereview.chromium.org/1404903004/#ps20001
(title: "Fix the test on Win64")

[commit-bot: I haz the power, 2015-10-15 16:48:17]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1404903004/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1404903004/20001

[commit-bot: I haz the power, 2015-10-15 17:34:41]

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

[commit-bot: I haz the power, 2015-10-15 17:34:41]

Try jobs failed on following builders:
  v8_win_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel/builds/10674)

[alph, 2015-10-15 17:41:17]

The CQ bit was checked by alph@chromium.org

[commit-bot: I haz the power, 2015-10-15 17:41:30]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1404903004/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1404903004/20001

[commit-bot: I haz the power, 2015-10-15 17:50:07]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-10-15 17:50:48]

Patchset 2 (id:??) landed as
https://crrev.com/7cdcf0b1bc45777b48923e9758a3a30e8620c1a1
Cr-Commit-Position: refs/heads/master@{#31315}