[turbofan] Handle impossible types (Type::None()) in the backend.

[turbofan] Handle impossible types (Type::None()) in the backend.

BUG=chromium:630611
Committed: https://crrev.com/a81d19d583baea17ea0aa70609c17ad18ce3f269
Cr-Commit-Position: refs/heads/master@{#37994}

[Jarin, 2016-07-22 13:58:10]

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

[commit-bot: I haz the power, 2016-07-22 13:58:20]

Dry run: 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, 2016-07-22 14:03:16]

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

[commit-bot: I haz the power, 2016-07-22 14:03:17]

Dry run: Try jobs failed on following builders:
  v8_linux_arm64_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm64_rel_ng/build...)
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/5516)

[Jarin, 2016-07-22 14:05:37]

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

[commit-bot: I haz the power, 2016-07-22 14:05:46]

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

[Jarin, 2016-07-22 14:06:28]

jarin@chromium.org changed reviewers:
+ bmeurer@chromium.org

[Jarin, 2016-07-22 14:06:29]

Could you take a look, please?

(I feel terrible about this.)

[commit-bot: I haz the power, 2016-07-22 14:10:15]

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

[commit-bot: I haz the power, 2016-07-22 14:10:17]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/5520)

[Jarin, 2016-07-22 20:21:02]

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

[commit-bot: I haz the power, 2016-07-22 20:21:07]

Dry run: 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, 2016-07-22 20:25:57]

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

[commit-bot: I haz the power, 2016-07-22 20:25:58]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/5531)

[Jarin, 2016-07-22 20:31:55]

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

[commit-bot: I haz the power, 2016-07-22 20:32:01]

Dry run: 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, 2016-07-22 20:37:11]

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

[commit-bot: I haz the power, 2016-07-22 20:37:12]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/5534)

[Jarin, 2016-07-22 20:55:17]

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

[commit-bot: I haz the power, 2016-07-22 20:55:21]

Dry run: 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, 2016-07-22 20:57:31]

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

[commit-bot: I haz the power, 2016-07-22 20:57:32]

Dry run: Try jobs failed on following builders:
  v8_android_arm_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_android_arm_compile_rel/...)
  v8_linux_arm_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel_ng/builds/...)
  v8_linux_dbg_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_dbg_ng/builds/9552)
  v8_linux_mipsel_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_mipsel_compile_rel...)
  v8_linux_nodcheck_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng/bu...)
  v8_linux_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_rel_ng/builds/9517)

[Jarin, 2016-07-22 21:20:53]

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

[commit-bot: I haz the power, 2016-07-22 21:21:01]

Dry run: 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, 2016-07-22 21:46:03]

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

[commit-bot: I haz the power, 2016-07-22 21:46:03]

Dry run: This issue passed the CQ dry run.

[Benedikt Meurer, 2016-07-24 17:23:41]

I feel your pain. But I have a suggestion how to make it less severe.

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/instructi...
File src/compiler/instruction-selector.cc (right):

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/instructi...
src/compiler/instruction-selector.cc:1326: #if V8_TARGET_ARCH_64_BIT
How about a simpler approach?

Emit(kArchImpossible, g.DefineAsConstant(node,
Constant(static_cast<intptr_t>(0))));

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/opcodes.h
File src/compiler/opcodes.h (right):

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/opcodes.h...
src/compiler/opcodes.h:379: V(ImpossibleToBit)            \
Hm, why not have a single Impossible machine/common operator, with a
MachineRepresentation parameter? That would also simplify the instruction
selection code.

[Jarin, 2016-07-24 18:15:40]

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/instructi...
File src/compiler/instruction-selector.cc (right):

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/instructi...
src/compiler/instruction-selector.cc:1326: #if V8_TARGET_ARCH_64_BIT
On 2016/07/24 17:23:41, Benedikt Meurer wrote:
> How about a simpler approach?
> 
> Emit(kArchImpossible, g.DefineAsConstant(node,
> Constant(static_cast<intptr_t>(0))));

I tried that, did not compile on Mac. See the diff from patchset 4.

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/opcodes.h
File src/compiler/opcodes.h (right):

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/opcodes.h...
src/compiler/opcodes.h:379: V(ImpossibleToBit)            \
On 2016/07/24 17:23:41, Benedikt Meurer wrote:
> Hm, why not have a single Impossible machine/common operator, with a
> MachineRepresentation parameter? That would also simplify the instruction
> selection code.

It just seemed that parametrized machine ops are discouraged. I think I prefer
this explicit way - it is nice when the output representation is obvious from
the opcode.

[Benedikt Meurer, 2016-07-25 03:36:50]

The CQ bit was checked by bmeurer@chromium.org

[Benedikt Meurer, 2016-07-25 03:36:51]

lgtm

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/instructi...
File src/compiler/instruction-selector.cc (right):

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/instructi...
src/compiler/instruction-selector.cc:1326: #if V8_TARGET_ARCH_64_BIT
Mhm, that's weird.

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/opcodes.h
File src/compiler/opcodes.h (right):

https://codereview.chromium.org/2177483002/diff/100001/src/compiler/opcodes.h...
src/compiler/opcodes.h:379: V(ImpossibleToBit)            \
I'd have made this a common operator in the same spirit as Phi/Select that know
their representation. But acknowledge your decision.

[commit-bot: I haz the power, 2016-07-25 03:36:58]

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, 2016-07-25 04:00:24]

Committed patchset #6 (id:100001)

[commit-bot: I haz the power, 2016-07-25 04:03:08]

Description was changed from

==========
[turbofan] Handle impossible types (Type::None()) in the backend.

BUG=chromium:630611
==========

to

==========
[turbofan] Handle impossible types (Type::None()) in the backend.

BUG=chromium:630611

Committed: https://crrev.com/a81d19d583baea17ea0aa70609c17ad18ce3f269
Cr-Commit-Position: refs/heads/master@{#37994}
==========

[commit-bot: I haz the power, 2016-07-25 04:03:09]

Patchset 6 (id:??) landed as
https://crrev.com/a81d19d583baea17ea0aa70609c17ad18ce3f269
Cr-Commit-Position: refs/heads/master@{#37994}