[counters] Making counter properly reentrant.

[counters] Making runtime counters reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate stack-allocated linked
list to track the timer stack solves this issue. This is a temporary workaround
with the limitations of the counter system in mind. Eventually we will move to
the trace-based system for these kind of statistics.

BUG=v8:4770
LOG=n

Committed: https://crrev.com/5e468666b7ff0893b6c9614b4c70a6f7b1dcb027
Cr-Commit-Position: refs/heads/master@{#34208}

[Camillo Bruni, 2016-02-12 13:40:06]

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

[Camillo Bruni, 2016-02-12 13:40:07]

Hi Jaro, PTAL
I had to adapt the counters again... they didn't handle reentrant
builtins/runtime functions very well ;). Hope this is more or less readable.

[Jarin, 2016-02-13 19:28:20]

nice, lgtm. Have you tried to measure the overhead of the book-keeping?

Funnily, I was thinking about the reentrancy problem when reviewing the CL that
introduced the proper accounting, and I managed to convince myself that it's
fine. Ah, well...

https://codereview.chromium.org/1695733002/diff/1/src/counters.cc
File src/counters.cc (right):

https://codereview.chromium.org/1695733002/diff/1/src/counters.cc#newcode276
src/counters.cc:276: current_timer_link_ = new TimerLink(counter,
current_timer_link_);
Nit: Allocating on the C++ heap for every runtime call makes me a bit uneasy. 
Maybe we could link on-stack records?

https://codereview.chromium.org/1695733002/diff/1/src/vm-state-inl.h
File src/vm-state-inl.h (right):

https://codereview.chromium.org/1695733002/diff/1/src/vm-state-inl.h#newcode58
src/vm-state-inl.h:58: 
Nit: Extra newline?

[Camillo Bruni, 2016-02-15 17:26:57]

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

[commit-bot: I haz the power, 2016-02-15 17:27:07]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1695733002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1695733002/40001

[commit-bot: I haz the power, 2016-02-15 17:31:18]

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

[commit-bot: I haz the power, 2016-02-15 17:31:19]

Dry run: Try jobs failed on following builders:
  v8_linux_nodcheck_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel/build...)
  v8_win_compile_dbg on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/1...)

[Camillo Bruni, 2016-02-15 19:32:14]

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

[commit-bot: I haz the power, 2016-02-15 19:32:51]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1695733002/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1695733002/100001

[commit-bot: I haz the power, 2016-02-15 19:36:41]

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

[commit-bot: I haz the power, 2016-02-15 19:36:42]

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...)

[Camillo Bruni, 2016-02-22 20:06:21]

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

[commit-bot: I haz the power, 2016-02-22 20:06:31]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1695733002/140001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1695733002/140001

[commit-bot: I haz the power, 2016-02-22 20:10:57]

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

[commit-bot: I haz the power, 2016-02-22 20:10:58]

Dry run: Try jobs failed on following builders:
  v8_linux_arm64_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm64_rel/builds/1...)
  v8_win_nosnap_shared_compile_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_compil...)

[Camillo Bruni, 2016-02-22 21:34:58]

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

[commit-bot: I haz the power, 2016-02-22 21:35:03]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1695733002/160001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1695733002/160001

[commit-bot: I haz the power, 2016-02-22 22:01:45]

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

[commit-bot: I haz the power, 2016-02-22 22:01:46]

Dry run: This issue passed the CQ dry run.

[Camillo Bruni, 2016-02-23 08:14:40]

Description was changed from

==========
[counters] Making counter properly reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate linked list to track
the timers solves this issue.

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

to

==========
[counters] Making counter properly reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate linked list to track
the timers solves this issue.

BUG=v8:4770
LOG=n
==========

[Camillo Bruni, 2016-02-23 08:20:05]

Description was changed from

==========
[counters] Making counter properly reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate linked list to track
the timers solves this issue.

BUG=v8:4770
LOG=n
==========

to

==========
[counters] Making runtime counters reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate linked list to track
the timers solves this issue. This is a temporary workaround with the
limitations of the counter system in mind. Eventually we will move to the
trace-based system for these kind of statistics.

BUG=v8:4770
LOG=n
==========

[Camillo Bruni, 2016-02-23 08:21:49]

Description was changed from

==========
[counters] Making runtime counters reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate linked list to track
the timers solves this issue. This is a temporary workaround with the
limitations of the counter system in mind. Eventually we will move to the
trace-based system for these kind of statistics.

BUG=v8:4770
LOG=n
==========

to

==========
[counters] Making runtime counters reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate stack-allocated linked
list to track the timer stack solves this issue. This is a temporary workaround
with the limitations of the counter system in mind. Eventually we will move to
the trace-based system for these kind of statistics.

BUG=v8:4770
LOG=n
==========

[Camillo Bruni, 2016-02-23 08:21:55]

The CQ bit was checked by cbruni@chromium.org

[Camillo Bruni, 2016-02-23 08:21:55]

The patchset sent to the CQ was uploaded after l-g-t-m from jarin@chromium.org
Link to the patchset: https://codereview.chromium.org/1695733002/#ps180001
(title: "fixing comment")

[commit-bot: I haz the power, 2016-02-23 08:22:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1695733002/180001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1695733002/180001

[commit-bot: I haz the power, 2016-02-23 08:45:56]

Committed patchset #10 (id:180001)

[commit-bot: I haz the power, 2016-02-23 08:46:27]

Description was changed from

==========
[counters] Making runtime counters reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate stack-allocated linked
list to track the timer stack solves this issue. This is a temporary workaround
with the limitations of the counter system in mind. Eventually we will move to
the trace-based system for these kind of statistics.

BUG=v8:4770
LOG=n
==========

to

==========
[counters] Making runtime counters reentrant.

So far counters did not work when they were reentrant and thus would lead to
wrong bookkeeping of the counter stack. Using a separate stack-allocated linked
list to track the timer stack solves this issue. This is a temporary workaround
with the limitations of the counter system in mind. Eventually we will move to
the trace-based system for these kind of statistics.

BUG=v8:4770
LOG=n

Committed: https://crrev.com/5e468666b7ff0893b6c9614b4c70a6f7b1dcb027
Cr-Commit-Position: refs/heads/master@{#34208}
==========

[commit-bot: I haz the power, 2016-02-23 08:46:28]

Patchset 10 (id:??) landed as
https://crrev.com/5e468666b7ff0893b6c9614b4c70a6f7b1dcb027
Cr-Commit-Position: refs/heads/master@{#34208}