Revert of Ensure consistent return ordering in base::WaitableEvent::WaitMany

Revert of Ensure consistent return ordering in base::WaitableEvent::WaitMany (patchset #5 id:80001 of https://codereview.chromium.org/2758853002/ )

Reason for revert:
Causing extra time for chrome_public_test_apk, telemetry_perf_unittests etc.

Original issue's description:
> Ensure consistent return ordering in base::WaitableEvent::WaitMany
> 
> The Windows implementation (via WaitForMultipleObjects) guarantees that
> if multiple objects are signaled, the one with the smallest index is
> returned on wakeup. This is a useful guarantee to make, as callers are
> granted the ability to rotate objects as needed to avoid potential
> starvation of lower-frequency events.
> 
> The POSIX implementation of WaitMany does not support this behavior, as
> precedence is determined exclusively by WaitableEvent address order
> rather than input array order. The behavior today is a result of the
> implementation detail that we must lock the WaitableEvent kernels in a
> globally consistent order, and this makes it effectively impossible for
> a caller to control signaling priority.
> 
> This CL changes the POSIX implementation to guarantee the same return
> behavior as the Windows implementation, allowing the WaitMany API to in
> turn guarantee left-to-right precedence of the WaitableEvents with which
> it's called.
> 
> This has some minor additional cost as we no longer short-circuit on the
> first signaled event we find, and we always have to iterate over the array
> twice now. This cost is negligible in practice as WaitMany is only ever
> used in production with a small (n < 5) number of events.
> 
> BUG=698460
> 
> Review-Url: https://codereview.chromium.org/2758853002
> Cr-Commit-Position: refs/heads/master@{#457967}
> Committed: https://chromium.googlesource.com/chromium/src/+/926b72bf8aba665d906fe411a11a9f587335c1bb

TBR=dcheng@chromium.org,rockot@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=698460
Review-Url: https://codereview.chromium.org/2761993004
Cr-Commit-Position: refs/heads/master@{#458602}
Committed: https://chromium.googlesource.com/chromium/src/+/a0a9840a39b61caca179486dcf20927fe99ad4a6

[BigBossZhiling, 2017-03-21 22:06:22]

The CQ bit was checked by hzl@google.com

[BigBossZhiling, 2017-03-21 22:06:23]

Created Revert of Ensure consistent return ordering in
base::WaitableEvent::WaitMany

[commit-bot: I haz the power, 2017-03-21 22:06:47]

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, 2017-03-21 22:06:48]

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

[commit-bot: I haz the power, 2017-03-21 22:06:49]

No L-G-T-M from a valid reviewer yet.
CQ run can only be started once the patch has received an L-G-T-M from a full
committer.
Even if an L-G-T-M may have been provided, it was from a non-committer,_not_ a
full super star committer.
Committers are members of the group "project-chromium-committers".
Note that this has nothing to do with OWNERS files.

[Ken Rockot(use gerrit already), 2017-03-21 22:07:23]

LGTM to experimentally determine whether this is actually having a performence
impact on telemetry

[BigBossZhiling, 2017-03-21 22:08:51]

The CQ bit was checked by hzl@google.com

[commit-bot: I haz the power, 2017-03-21 22:09:14]

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

[dcheng, 2017-03-21 23:02:26]

lgtm

[commit-bot: I haz the power, 2017-03-21 23:24:38]

CQ is committing da patch.

Bot data: {"patchset_id": 1, "attempt_start_ts": 1490134131378420, "parent_rev":
"b7f329f1bcb344b270da73f0ef23650198886a0f", "commit_rev":
"a0a9840a39b61caca179486dcf20927fe99ad4a6"}

[commit-bot: I haz the power, 2017-03-21 23:26:44]

Description was changed from

==========
Revert of Ensure consistent return ordering in base::WaitableEvent::WaitMany
(patchset #5 id:80001 of https://codereview.chromium.org/2758853002/ )

Reason for revert:
Causing extra time for chrome_public_test_apk, telemetry_perf_unittests etc.

Original issue's description:
> Ensure consistent return ordering in base::WaitableEvent::WaitMany
> 
> The Windows implementation (via WaitForMultipleObjects) guarantees that
> if multiple objects are signaled, the one with the smallest index is
> returned on wakeup. This is a useful guarantee to make, as callers are
> granted the ability to rotate objects as needed to avoid potential
> starvation of lower-frequency events.
> 
> The POSIX implementation of WaitMany does not support this behavior, as
> precedence is determined exclusively by WaitableEvent address order
> rather than input array order. The behavior today is a result of the
> implementation detail that we must lock the WaitableEvent kernels in a
> globally consistent order, and this makes it effectively impossible for
> a caller to control signaling priority.
> 
> This CL changes the POSIX implementation to guarantee the same return
> behavior as the Windows implementation, allowing the WaitMany API to in
> turn guarantee left-to-right precedence of the WaitableEvents with which
> it's called.
> 
> This has some minor additional cost as we no longer short-circuit on the
> first signaled event we find, and we always have to iterate over the array
> twice now. This cost is negligible in practice as WaitMany is only ever
> used in production with a small (n < 5) number of events.
> 
> BUG=698460
> 
> Review-Url: https://codereview.chromium.org/2758853002
> Cr-Commit-Position: refs/heads/master@{#457967}
> Committed:
https://chromium.googlesource.com/chromium/src/+/926b72bf8aba665d906fe411a11a...

TBR=dcheng@chromium.org,rockot@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=698460
==========

to

==========
Revert of Ensure consistent return ordering in base::WaitableEvent::WaitMany
(patchset #5 id:80001 of https://codereview.chromium.org/2758853002/ )

Reason for revert:
Causing extra time for chrome_public_test_apk, telemetry_perf_unittests etc.

Original issue's description:
> Ensure consistent return ordering in base::WaitableEvent::WaitMany
> 
> The Windows implementation (via WaitForMultipleObjects) guarantees that
> if multiple objects are signaled, the one with the smallest index is
> returned on wakeup. This is a useful guarantee to make, as callers are
> granted the ability to rotate objects as needed to avoid potential
> starvation of lower-frequency events.
> 
> The POSIX implementation of WaitMany does not support this behavior, as
> precedence is determined exclusively by WaitableEvent address order
> rather than input array order. The behavior today is a result of the
> implementation detail that we must lock the WaitableEvent kernels in a
> globally consistent order, and this makes it effectively impossible for
> a caller to control signaling priority.
> 
> This CL changes the POSIX implementation to guarantee the same return
> behavior as the Windows implementation, allowing the WaitMany API to in
> turn guarantee left-to-right precedence of the WaitableEvents with which
> it's called.
> 
> This has some minor additional cost as we no longer short-circuit on the
> first signaled event we find, and we always have to iterate over the array
> twice now. This cost is negligible in practice as WaitMany is only ever
> used in production with a small (n < 5) number of events.
> 
> BUG=698460
> 
> Review-Url: https://codereview.chromium.org/2758853002
> Cr-Commit-Position: refs/heads/master@{#457967}
> Committed:
https://chromium.googlesource.com/chromium/src/+/926b72bf8aba665d906fe411a11a...

TBR=dcheng@chromium.org,rockot@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=698460

Review-Url: https://codereview.chromium.org/2761993004
Cr-Commit-Position: refs/heads/master@{#458602}
Committed:
https://chromium.googlesource.com/chromium/src/+/a0a9840a39b61caca179486dcf20...
==========

[commit-bot: I haz the power, 2017-03-21 23:26:46]

Committed patchset #1 (id:1) as
https://chromium.googlesource.com/chromium/src/+/a0a9840a39b61caca179486dcf20...

[Ken Rockot(use gerrit already), 2017-03-22 19:33:42]

A revert of this CL (patchset #1 id:1) has been created in
https://codereview.chromium.org/2773463002/ by rockot@chromium.org.

The reason for reverting is: Even if this did negatively impact telemetry
performance, we'd want the WaitMany change for the sake of API correctness.

However, after leaving the revert in place for a day, it doesn't look like this
CL was the culprit anyway..