Uncommit and shrink semi-spaces only on low allocation rate.

Uncommit and shrink semi-spaces only on low allocation rate.

BUG=chromium:481811
LOG=n

Committed: https://crrev.com/f034d6d835edff6d57f2bc8ea691feb21866e798
Cr-Commit-Position: refs/heads/master@{#28585}

[Hannes Payer (out of office), 2015-05-21 15:09:12]

hpayer@chromium.org changed reviewers:
+ ulan@chromium.org

[Hannes Payer (out of office), 2015-05-21 15:14:25]

We could also move this code into the GC epilogue to also uncommit when there is
no idle time (e.g.  on mobile). WDYT?

[Hannes Payer (out of office), 2015-05-21 15:14:26]

We could also move this code into the GC epilogue to also uncommit when there is
no idle time (e.g.  on mobile). WDYT?

[ulan, 2015-05-21 15:21:25]

lgtm

> We could also move this code into the GC epilogue to also uncommit when there
is
> no idle time (e.g.  on mobile). WDYT?

Yes and the same for scaling allocation factor based on throughput.

https://codereview.chromium.org/1148633005/diff/1/src/heap/gc-tracer.cc
File src/heap/gc-tracer.cc (right):

https://codereview.chromium.org/1148633005/diff/1/src/heap/gc-tracer.cc#newco...
src/heap/gc-tracer.cc:626: return static_cast<size_t>(allocated_bytes /
kThroughputTimeFrame);
(+ 1) to avoid returning 0 if allocated_bytes < kThroughputTimeFrame?

https://codereview.chromium.org/1148633005/diff/1/src/heap/heap.cc
File src/heap/heap.cc (right):

https://codereview.chromium.org/1148633005/diff/1/src/heap/heap.cc#newcode4570
src/heap/heap.cc:4570: return
heap_state.current_new_space_allocation_throughput_in_bytes_per_ms <
check for heap_state.current_new_space_allocation_throughput_in_bytes_per_ms ==
0 and be conservative in that case?

[Hannes Payer (out of office), 2015-05-21 17:17:03]

https://codereview.chromium.org/1148633005/diff/1/src/heap/gc-tracer.cc
File src/heap/gc-tracer.cc (right):

https://codereview.chromium.org/1148633005/diff/1/src/heap/gc-tracer.cc#newco...
src/heap/gc-tracer.cc:626: return static_cast<size_t>(allocated_bytes /
kThroughputTimeFrame);
On 2015/05/21 15:21:25, ulan wrote:
> (+ 1) to avoid returning 0 if allocated_bytes < kThroughputTimeFrame?

Done.

https://codereview.chromium.org/1148633005/diff/1/src/heap/heap.cc
File src/heap/heap.cc (right):

https://codereview.chromium.org/1148633005/diff/1/src/heap/heap.cc#newcode4570
src/heap/heap.cc:4570: return
heap_state.current_new_space_allocation_throughput_in_bytes_per_ms <
On 2015/05/21 15:21:25, ulan wrote:
> check for heap_state.current_new_space_allocation_throughput_in_bytes_per_ms
==
> 0 and be conservative in that case?

Good point! Done.

[Hannes Payer (out of office), 2015-05-21 17:17:48]

The CQ bit was checked by hpayer@chromium.org

[Hannes Payer (out of office), 2015-05-21 17:17:49]

The patchset sent to the CQ was uploaded after l-g-t-m from ulan@chromium.org
Link to the patchset: https://codereview.chromium.org/1148633005/#ps20001
(title: " ")

[commit-bot: I haz the power, 2015-05-21 17:18:02]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1148633005/20001

[commit-bot: I haz the power, 2015-05-21 18:22:36]

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

[commit-bot: I haz the power, 2015-05-21 18:22:37]

Try jobs failed on following builders:
  v8_linux_arm_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/4438)

[Hannes Payer (out of office), 2015-05-22 07:17:31]

The CQ bit was checked by hpayer@chromium.org

[Hannes Payer (out of office), 2015-05-22 07:17:31]

The patchset sent to the CQ was uploaded after l-g-t-m from ulan@chromium.org
Link to the patchset: https://codereview.chromium.org/1148633005/#ps40001
(title: " ")

[commit-bot: I haz the power, 2015-05-22 07:17:41]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1148633005/40001

[commit-bot: I haz the power, 2015-05-22 08:05:03]

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

[commit-bot: I haz the power, 2015-05-22 08:05:04]

Try jobs failed on following builders:
  v8_linux_arm_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/4447)

[Hannes Payer (out of office), 2015-05-22 11:12:39]

The CQ bit was checked by hpayer@chromium.org

[Hannes Payer (out of office), 2015-05-22 11:12:40]

The patchset sent to the CQ was uploaded after l-g-t-m from ulan@chromium.org
Link to the patchset: https://codereview.chromium.org/1148633005/#ps60001
(title: " ")

[commit-bot: I haz the power, 2015-05-22 11:12:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1148633005/60001

[commit-bot: I haz the power, 2015-05-22 11:43:16]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2015-05-22 11:43:32]

Patchset 4 (id:??) landed as
https://crrev.com/f034d6d835edff6d57f2bc8ea691feb21866e798
Cr-Commit-Position: refs/heads/master@{#28585}

[Michael Achenbach, 2015-05-22 13:45:40]

machenbach@chromium.org changed reviewers:
+ ishell@chromium.org, machenbach@chromium.org

[Michael Achenbach, 2015-05-22 13:45:41]

Possible that this broke predictability?
http://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20predictable/b...