[compiler] Remove context value input from JSLoadContext and JSStoreContext.

[compiler] Remove context value input from JSLoadContext and JSStoreContext.

JS operators always have an implicit context input, so just use that instead.

BUG=

Committed: https://crrev.com/9ef7ab1e2e2f4dfcfa16172b3588b760bc3c7461
Cr-Commit-Position: refs/heads/master@{#41392}

[neis, 2016-11-30 09:51:02]

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

[commit-bot: I haz the power, 2016-11-30 09:51:09]

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

[neis, 2016-11-30 10:02:48]

Description was changed from

==========
WIP

COMMIT=false
BUG=
==========

to

==========
[compiler] Remove context value input from JSLoadContext and JSStoreContext.

JS operators always have an implicit context input, so just use that instead.

BUG=
==========

[neis, 2016-11-30 10:02:49]

neis@chromium.org changed reviewers:
+ mstarzinger@chromium.org

[neis, 2016-11-30 10:02:59]

ptal

[neis, 2016-11-30 10:04:02]

Ignore the changes in src/compiler/js-generic-lowering.cc, they are already in a
separate CL.

[commit-bot: I haz the power, 2016-11-30 10:16:31]

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

[commit-bot: I haz the power, 2016-11-30 10:16:31]

Dry run: This issue passed the CQ dry run.

[Michael Starzinger, 2016-11-30 12:43:28]

LGTM with one comment to address. Nice! I like it!

https://codereview.chromium.org/2541813002/diff/1/src/compiler/js-typed-lower...
File src/compiler/js-typed-lowering.cc (right):

https://codereview.chromium.org/2541813002/diff/1/src/compiler/js-typed-lower...
src/compiler/js-typed-lowering.cc:1539: node->InsertInput(jsgraph()->zone(), 0,
context);
nit: First inserting and then removing an input is most likely less efficient
then just replacing all inputs. Also I think readability is improved by listing
all new inputs. I was thinking along the lines of ...

node->ReplaceInput(node, 0, context);
node->ReplaceInput(node, 1, value);
node->ReplaceInput(node, 2, effect)

The "effect" is already at the right place, but needs to be updated with the new
effect produced above. The "control" should already be in the right place and
can remain unchanged. The "value" should be loaded at the top of the function
before any mutation.

https://codereview.chromium.org/2541813002/diff/1/src/compiler/js-typed-lower...
src/compiler/js-typed-lowering.cc:1545: return Changed(node);  // control??
explanation: The {JSStoreContext} operators are marked as taking one control
input whereas {JSLoadContext} take zero control inputs. Hence the asymmetry
here.

[neis, 2016-11-30 15:15:12]

On 2016/11/30 12:43:28, Michael Starzinger wrote:
> LGTM with one comment to address. Nice! I like it!
> 
>
https://codereview.chromium.org/2541813002/diff/1/src/compiler/js-typed-lower...
> File src/compiler/js-typed-lowering.cc (right):
> 
>
https://codereview.chromium.org/2541813002/diff/1/src/compiler/js-typed-lower...
> src/compiler/js-typed-lowering.cc:1539: node->InsertInput(jsgraph()->zone(),
0,
> context);
> nit: First inserting and then removing an input is most likely less efficient
> then just replacing all inputs. Also I think readability is improved by
listing
> all new inputs. I was thinking along the lines of ...
> 
> node->ReplaceInput(node, 0, context);
> node->ReplaceInput(node, 1, value);
> node->ReplaceInput(node, 2, effect)
> 
> The "effect" is already at the right place, but needs to be updated with the
new
> effect produced above. The "control" should already be in the right place and
> can remain unchanged. The "value" should be loaded at the top of the function
> before any mutation.

Done.
 
>
https://codereview.chromium.org/2541813002/diff/1/src/compiler/js-typed-lower...
> src/compiler/js-typed-lowering.cc:1545: return Changed(node);  // control??
> explanation: The {JSStoreContext} operators are marked as taking one control
> input whereas {JSLoadContext} take zero control inputs. Hence the asymmetry
> here.

Thanks!

[neis, 2016-11-30 15:17:10]

The CQ bit was checked by neis@chromium.org

[neis, 2016-11-30 15:17:10]

The patchset sent to the CQ was uploaded after l-g-t-m from
mstarzinger@chromium.org
Link to the patchset: https://codereview.chromium.org/2541813002/#ps40001
(title: "Rebase.")

[commit-bot: I haz the power, 2016-11-30 15:17:22]

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, 2016-11-30 15:45:38]

CQ is committing da patch.

Bot data: {"patchset_id": 40001, "attempt_start_ts": 1480519030435110,
"parent_rev": "fc3ac4a212b87806588e630a83bd61bec2d98a14", "commit_rev":
"4f5e6714ba9044e72574b71f3d1f3e12c11a7f83"}

[commit-bot: I haz the power, 2016-11-30 15:45:44]

Description was changed from

==========
[compiler] Remove context value input from JSLoadContext and JSStoreContext.

JS operators always have an implicit context input, so just use that instead.

BUG=
==========

to

==========
[compiler] Remove context value input from JSLoadContext and JSStoreContext.

JS operators always have an implicit context input, so just use that instead.

BUG=
==========

[commit-bot: I haz the power, 2016-11-30 15:45:44]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-11-30 15:46:12]

Description was changed from

==========
[compiler] Remove context value input from JSLoadContext and JSStoreContext.

JS operators always have an implicit context input, so just use that instead.

BUG=
==========

to

==========
[compiler] Remove context value input from JSLoadContext and JSStoreContext.

JS operators always have an implicit context input, so just use that instead.

BUG=

Committed: https://crrev.com/9ef7ab1e2e2f4dfcfa16172b3588b760bc3c7461
Cr-Commit-Position: refs/heads/master@{#41392}
==========

[commit-bot: I haz the power, 2016-11-30 15:46:13]

Patchset 3 (id:??) landed as
https://crrev.com/9ef7ab1e2e2f4dfcfa16172b3588b760bc3c7461
Cr-Commit-Position: refs/heads/master@{#41392}