Revert of Oilpan: Decommit backing storage of CallbackStacks

Revert of Oilpan: Decommit backing storage of CallbackStacks (patchset #7 id:120001 of https://codereview.chromium.org/1686943002/ )

Reason for revert:
This CL will cause page faults at every marking phase. Marking must be super fast, so we need to reconsider the approach.

Original issue's description:
> Oilpan: Decommit backing storage of CallbackStacks
>
> CallbackStacks are used only while Oilpan's GC is doing marking & weak processing.
> However, currently each CallbackStack continues retaining one Block forever.
>
> This wastes memory a lot. Oilpan has 4 global CallbackStacks and 1 CallbackStack per thread.
> Each Block consumes 8192 * sizeof(Item) = 128 KB. This means that Oilpan wastes 128 KB * (4 + # of threads).
> When I start Chrome's new tab page, it creates 18 CallbackStacks,
> meaning that it wastes 2.3 MB of memory.
>
> This CL removes the waste by discarding system pages of the Block after finishing every GC phase.
>
> BUG=
>
> Committed: https://crrev.com/34393519671e62e5a66602736899b94ad33dd45a
> Cr-Commit-Position: refs/heads/master@{#374674}

TBR=keishi@chromium.org,sigbjornf@opera.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=

Committed: https://crrev.com/d06acbf2093b7ffb8fb03b721debdd06ffcabafc
Cr-Commit-Position: refs/heads/master@{#374868}

[haraken, 2016-02-11 06:31:12]

Created Revert of Oilpan: Decommit backing storage of CallbackStacks

[commit-bot: I haz the power, 2016-02-11 06:32:25]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689823003/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689823003/1

[commit-bot: I haz the power, 2016-02-11 06:33:11]

Description was changed from

==========
Revert of Oilpan: Decommit backing storage of CallbackStacks (patchset #7
id:120001 of https://codereview.chromium.org/1686943002/ )

Reason for revert:
This CL will cause page faults at every marking phase. Marking must be super
fast, so we need to reconsider the approach.


Original issue's description:
> Oilpan: Decommit backing storage of CallbackStacks
> 
> CallbackStacks are used only while Oilpan's GC is doing marking & weak
processing.
> However, currently each CallbackStack continues retaining one Block forever.
> 
> This wastes memory a lot. Oilpan has 4 global CallbackStacks and 1
CallbackStack per thread.
> Each Block consumes 8192 * sizeof(Item) = 128 KB. This means that Oilpan
wastes 128 KB * (4 + # of threads).
> When I start Chrome's new tab page, it creates 18 CallbackStacks,
> meaning that it wastes 2.3 MB of memory.
> 
> This CL removes the waste by discarding system pages of the Block after
finishing every GC phase.
> 
> BUG=
> 
> Committed: https://crrev.com/34393519671e62e5a66602736899b94ad33dd45a
> Cr-Commit-Position: refs/heads/master@{#374674}

TBR=keishi@chromium.org,sigbjornf@opera.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=
==========

to

==========
Revert of Oilpan: Decommit backing storage of CallbackStacks (patchset #7
id:120001 of https://codereview.chromium.org/1686943002/ )

Reason for revert:
This CL will cause page faults at every marking phase. Marking must be super
fast, so we need to reconsider the approach.


Original issue's description:
> Oilpan: Decommit backing storage of CallbackStacks
> 
> CallbackStacks are used only while Oilpan's GC is doing marking & weak
processing.
> However, currently each CallbackStack continues retaining one Block forever.
> 
> This wastes memory a lot. Oilpan has 4 global CallbackStacks and 1
CallbackStack per thread.
> Each Block consumes 8192 * sizeof(Item) = 128 KB. This means that Oilpan
wastes 128 KB * (4 + # of threads).
> When I start Chrome's new tab page, it creates 18 CallbackStacks,
> meaning that it wastes 2.3 MB of memory.
> 
> This CL removes the waste by discarding system pages of the Block after
finishing every GC phase.
> 
> BUG=
> 
> Committed: https://crrev.com/34393519671e62e5a66602736899b94ad33dd45a
> Cr-Commit-Position: refs/heads/master@{#374674}

TBR=keishi@chromium.org,sigbjornf@opera.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=
==========

[commit-bot: I haz the power, 2016-02-11 06:33:12]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-02-16 22:34:53]

Description was changed from

==========
Revert of Oilpan: Decommit backing storage of CallbackStacks (patchset #7
id:120001 of https://codereview.chromium.org/1686943002/ )

Reason for revert:
This CL will cause page faults at every marking phase. Marking must be super
fast, so we need to reconsider the approach.


Original issue's description:
> Oilpan: Decommit backing storage of CallbackStacks
> 
> CallbackStacks are used only while Oilpan's GC is doing marking & weak
processing.
> However, currently each CallbackStack continues retaining one Block forever.
> 
> This wastes memory a lot. Oilpan has 4 global CallbackStacks and 1
CallbackStack per thread.
> Each Block consumes 8192 * sizeof(Item) = 128 KB. This means that Oilpan
wastes 128 KB * (4 + # of threads).
> When I start Chrome's new tab page, it creates 18 CallbackStacks,
> meaning that it wastes 2.3 MB of memory.
> 
> This CL removes the waste by discarding system pages of the Block after
finishing every GC phase.
> 
> BUG=
> 
> Committed: https://crrev.com/34393519671e62e5a66602736899b94ad33dd45a
> Cr-Commit-Position: refs/heads/master@{#374674}

TBR=keishi@chromium.org,sigbjornf@opera.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=
==========

to

==========
Revert of Oilpan: Decommit backing storage of CallbackStacks (patchset #7
id:120001 of https://codereview.chromium.org/1686943002/ )

Reason for revert:
This CL will cause page faults at every marking phase. Marking must be super
fast, so we need to reconsider the approach.

Original issue's description:
> Oilpan: Decommit backing storage of CallbackStacks
>
> CallbackStacks are used only while Oilpan's GC is doing marking & weak
processing.
> However, currently each CallbackStack continues retaining one Block forever.
>
> This wastes memory a lot. Oilpan has 4 global CallbackStacks and 1
CallbackStack per thread.
> Each Block consumes 8192 * sizeof(Item) = 128 KB. This means that Oilpan
wastes 128 KB * (4 + # of threads).
> When I start Chrome's new tab page, it creates 18 CallbackStacks,
> meaning that it wastes 2.3 MB of memory.
>
> This CL removes the waste by discarding system pages of the Block after
finishing every GC phase.
>
> BUG=
>
> Committed: https://crrev.com/34393519671e62e5a66602736899b94ad33dd45a
> Cr-Commit-Position: refs/heads/master@{#374674}

TBR=keishi@chromium.org,sigbjornf@opera.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=

Committed: https://crrev.com/d06acbf2093b7ffb8fb03b721debdd06ffcabafc
Cr-Commit-Position: refs/heads/master@{#374868}
==========

[commit-bot: I haz the power, 2016-02-16 22:34:54]

Patchset 1 (id:??) landed as
https://crrev.com/d06acbf2093b7ffb8fb03b721debdd06ffcabafc
Cr-Commit-Position: refs/heads/master@{#374868}