[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the entry trampoline.

[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}

Committed: https://crrev.com/39738bc90537a015bdaa4898478edaf88b021d11
Cr-Commit-Position: refs/heads/master@{#36310}

[rmcilroy, 2016-05-12 10:51:00]

rmcilroy@chromium.org changed reviewers:
+ mstarzinger@chromium.org, oth@chromium.org

[rmcilroy, 2016-05-12 10:51:01]

Turned out pretty clean I think. Let me know what you think and if happy with it
I'll port to the other architectures.

[rmcilroy, 2016-05-12 10:51:34]

Patchset #1 (id:1) has been deleted

[oth, 2016-05-12 12:34:30]

Cool, lgtm.

[Michael Starzinger, 2016-05-13 09:39:09]

LGTM. Awesome, I like it!

[rmcilroy, 2016-05-13 15:26:13]

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

[commit-bot: I haz the power, 2016-05-13 15:26:20]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/40001

[commit-bot: I haz the power, 2016-05-13 15:44:13]

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

[commit-bot: I haz the power, 2016-05-13 15:44:14]

Dry run: Try jobs failed on following builders:
  v8_linux64_asan_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_asan_rel_ng/buil...)
  v8_linux64_asan_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_asan_rel_ng_trig...)
  v8_linux_nodcheck_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng/bu...)
  v8_linux_nodcheck_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng_tr...)

[rmcilroy, 2016-05-13 16:27:46]

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

[commit-bot: I haz the power, 2016-05-13 16:27:48]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/60001

[commit-bot: I haz the power, 2016-05-13 16:48:28]

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

[commit-bot: I haz the power, 2016-05-13 16:48:29]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/1647)
  v8_mac_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng_triggered/bui...)
  v8_win_nosnap_shared_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)
  v8_win_nosnap_shared_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)

[rmcilroy, 2016-05-13 20:55:24]

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

[commit-bot: I haz the power, 2016-05-13 20:55:31]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/80001

[commit-bot: I haz the power, 2016-05-13 21:14:01]

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

[commit-bot: I haz the power, 2016-05-13 21:14:02]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/1651)
  v8_mac_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng_triggered/bui...)

[rmcilroy, 2016-05-16 11:51:09]

Patchset #3 (id:60001) has been deleted

[rmcilroy, 2016-05-16 11:51:26]

Patchset #3 (id:80001) has been deleted

[rmcilroy, 2016-05-16 11:51:33]

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

[commit-bot: I haz the power, 2016-05-16 11:51:36]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/100001

[rmcilroy, 2016-05-16 11:53:26]

Description was changed from

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

BUG=v8:4280
LOG=N
==========

to

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and 
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this 
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N
==========

[commit-bot: I haz the power, 2016-05-16 11:54:25]

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

[commit-bot: I haz the power, 2016-05-16 11:54:26]

Dry run: Try jobs failed on following builders:
  v8_linux_mipsel_compile_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_mipsel_compile_rel...)
  v8_mac_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/1671)
  v8_win_compile_dbg on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/1...)

[rmcilroy, 2016-05-16 12:37:53]

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

[commit-bot: I haz the power, 2016-05-16 12:37:58]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/110028
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/110028

[commit-bot: I haz the power, 2016-05-16 12:42:47]

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

[commit-bot: I haz the power, 2016-05-16 12:42:48]

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

[rmcilroy, 2016-05-16 12:49:52]

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

[commit-bot: I haz the power, 2016-05-16 12:49:59]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/130001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/130001

[commit-bot: I haz the power, 2016-05-16 12:54:07]

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

[commit-bot: I haz the power, 2016-05-16 12:54:08]

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

[rmcilroy, 2016-05-16 13:18:55]

Patchset #5 (id:130001) has been deleted

[rmcilroy, 2016-05-16 13:19:03]

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

[commit-bot: I haz the power, 2016-05-16 13:19:14]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/150001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/150001

[rmcilroy, 2016-05-16 13:21:02]

There were some issues with deopt to Ignition, so I had to make a few more
changes. The plus side is that now we no longer need seperate
InterpreterNotifyDeoptXXX functions, we can reuse the same builtins used by FCG
:). In doing this, I've renamed FullCodegenerator::State to
Deoptimizer::BailoutState (which is where most of the mechanical changes in
full-codegen[-arch] come from.

PTAL, thanks.

[commit-bot: I haz the power, 2016-05-16 13:42:01]

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

[commit-bot: I haz the power, 2016-05-16 13:42:02]

Dry run: This issue passed the CQ dry run.

[Michael Starzinger, 2016-05-17 14:41:53]

https://codereview.chromium.org/1969423002/diff/150001/src/deoptimizer.h
File src/deoptimizer.h (right):

https://codereview.chromium.org/1969423002/diff/150001/src/deoptimizer.h#newc...
src/deoptimizer.h:418: return NULL;
nit: s/NULL/nullptr/

https://codereview.chromium.org/1969423002/diff/150001/src/full-codegen/full-...
File src/full-codegen/full-codegen.h (left):

https://codereview.chromium.org/1969423002/diff/150001/src/full-codegen/full-...
src/full-codegen/full-codegen.h:31: enum State {
If we were to add "using Deoptimizer::BailoutState::TOS_REGISTER" and "using
Deoptimizer::BailoutState::NO_REGISTERS" here, would that allow us to just
reference it without specifying the full namespace everywhere? Just makes all
call-sites more concise. WDYT?

[Michael Starzinger, 2016-05-17 14:48:19]

Still LGTM modulo comments of course.

[rmcilroy, 2016-05-17 15:20:05]

https://codereview.chromium.org/1969423002/diff/150001/src/deoptimizer.h
File src/deoptimizer.h (right):

https://codereview.chromium.org/1969423002/diff/150001/src/deoptimizer.h#newc...
src/deoptimizer.h:418: return NULL;
On 2016/05/17 14:41:53, Michael Starzinger wrote:
> nit: s/NULL/nullptr/

Done.

https://codereview.chromium.org/1969423002/diff/150001/src/full-codegen/full-...
File src/full-codegen/full-codegen.h (left):

https://codereview.chromium.org/1969423002/diff/150001/src/full-codegen/full-...
src/full-codegen/full-codegen.h:31: enum State {
On 2016/05/17 14:41:53, Michael Starzinger wrote:
> If we were to add "using Deoptimizer::BailoutState::TOS_REGISTER" and "using
> Deoptimizer::BailoutState::NO_REGISTERS" here, would that allow us to just
> reference it without specifying the full namespace everywhere? Just makes all
> call-sites more concise. WDYT?

As discussed offline - needed to use typedef and kept BailoutState:: in call
sites.

[rmcilroy, 2016-05-17 15:20:13]

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

[commit-bot: I haz the power, 2016-05-17 15:20:19]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/170001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/170001

[commit-bot: I haz the power, 2016-05-17 15:51:28]

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

[commit-bot: I haz the power, 2016-05-17 15:51:29]

Dry run: This issue passed the CQ dry run.

[rmcilroy, 2016-05-17 16:41:53]

The CQ bit was checked by rmcilroy@chromium.org

[rmcilroy, 2016-05-17 16:41:54]

The patchset sent to the CQ was uploaded after l-g-t-m from oth@chromium.org,
mstarzinger@chromium.org
Link to the patchset: https://codereview.chromium.org/1969423002/#ps170001
(title: "Review comments")

[commit-bot: I haz the power, 2016-05-17 16:41:59]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/170001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/170001

[commit-bot: I haz the power, 2016-05-17 16:44:42]

Description was changed from

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and 
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this 
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N
==========

to

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and 
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this 
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N
==========

[commit-bot: I haz the power, 2016-05-17 16:44:44]

Committed patchset #6 (id:170001)

[commit-bot: I haz the power, 2016-05-17 16:46:41]

Description was changed from

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and 
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this 
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N
==========

to

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
==========

[commit-bot: I haz the power, 2016-05-17 16:46:42]

Patchset 6 (id:??) landed as
https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}

[Michael Achenbach, 2016-05-17 19:43:46]

A revert of this CL (patchset #6 id:170001) has been created in
https://codereview.chromium.org/1986353002/ by machenbach@chromium.org.

The reason for reverting is: Breaks 
https://build.chromium.org/p/client.v8.ports/builders/V8%20Linux%20-%20arm%20...

Might only affect pure release builds?.

[rmcilroy, 2016-05-18 07:44:42]

Description was changed from

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
==========

to

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
==========

[rmcilroy, 2016-05-18 07:46:22]

On 2016/05/17 19:43:46, Michael Achenbach (slow) wrote:
> A revert of this CL (patchset #6 id:170001) has been created in
> https://codereview.chromium.org/1986353002/ by mailto:machenbach@chromium.org.
> 
> The reason for reverting is: Breaks 
>
https://build.chromium.org/p/client.v8.ports/builders/V8%20Linux%20-%20arm%20...
> 
> Might only affect pure release builds?.

Thanks for the revert Michael. You were right, this only happened on release
builds, and only on Arm.

I landed a fix to the bug in Turbofan which caused the bot to fail in
https://codereview.chromium.org/1988623003/

I will reland this now and keep an eye on the Arm bot.

[rmcilroy, 2016-05-18 07:46:57]

The CQ bit was checked by rmcilroy@chromium.org

[commit-bot: I haz the power, 2016-05-18 07:47:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1969423002/170001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1969423002/170001

[commit-bot: I haz the power, 2016-05-18 07:50:08]

Description was changed from

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
==========

to

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
==========

[commit-bot: I haz the power, 2016-05-18 07:50:10]

Committed patchset #6 (id:170001)

[commit-bot: I haz the power, 2016-05-18 07:52:18]

Description was changed from

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
==========

to

==========
[Interpreter] Remove InterpreterExitTrampoline and replace with returning to the
entry trampoline.

In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.

This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.

BUG=v8:4280
LOG=N

Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}

Committed: https://crrev.com/39738bc90537a015bdaa4898478edaf88b021d11
Cr-Commit-Position: refs/heads/master@{#36310}
==========

[commit-bot: I haz the power, 2016-05-18 07:52:20]

Patchset 6 (id:??) landed as
https://crrev.com/39738bc90537a015bdaa4898478edaf88b021d11
Cr-Commit-Position: refs/heads/master@{#36310}