[turbofan] Fix various issues with --turbo-inlining enabled.

[turbofan] Fix various issues with --turbo-inlining enabled.

This is in preparation to enabling --turbo-inlining by default, fixing
various issues when general purpose inlining is running against our
entire test suite.

R=bmeurer@chromium.org
BUG=v8:4493
LOG=n

Committed: https://crrev.com/e6a923ab5680bf48b34083992408be85ab437dc4
Cr-Commit-Position: refs/heads/master@{#31294}

[Michael Starzinger, 2015-10-15 11:05:23]

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

[Benedikt Meurer, 2015-10-15 11:18:23]

LGTM with nit.

https://codereview.chromium.org/1407533004/diff/20001/src/compiler/js-inlinin...
File src/compiler/js-inlining-heuristic.cc (right):

https://codereview.chromium.org/1407533004/diff/20001/src/compiler/js-inlinin...
src/compiler/js-inlining-heuristic.cc:59: if (!p.feedback().IsValid()) return
NoChange();  // No feedback.
Nit: Can we set call count to zero in this case? I'd prefer to also be able to
inline call sites that were never executed so far based on other criteria.

[Michael Starzinger, 2015-10-15 11:29:30]

https://codereview.chromium.org/1407533004/diff/20001/src/compiler/js-inlinin...
File src/compiler/js-inlining-heuristic.cc (right):

https://codereview.chromium.org/1407533004/diff/20001/src/compiler/js-inlinin...
src/compiler/js-inlining-heuristic.cc:59: if (!p.feedback().IsValid()) return
NoChange();  // No feedback.
On 2015/10/15 11:18:23, Benedikt Meurer wrote:
> Nit: Can we set call count to zero in this case? I'd prefer to also be able to
> inline call sites that were never executed so far based on other criteria.

Done. As discussed offline, I used {-1} instead of {0} to be in sync with the
default value that CallICNexus::ExtractCallCount returns. Left a comment in that
regard as well.

[Michael Starzinger, 2015-10-15 11:47:48]

The CQ bit was checked by mstarzinger@chromium.org

[Michael Starzinger, 2015-10-15 11:47:49]

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/1407533004/#ps40001
(title: "Addressed comment.")

[commit-bot: I haz the power, 2015-10-15 11:47:59]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1407533004/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1407533004/40001

[commit-bot: I haz the power, 2015-10-15 12:01:58]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-10-15 12:02:12]

Patchset 3 (id:??) landed as
https://crrev.com/e6a923ab5680bf48b34083992408be85ab437dc4
Cr-Commit-Position: refs/heads/master@{#31294}