allocator cleanup: remove dependencies on allocator from all targets

allocator cleanup: remove dependencies on allocator from all targets

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator)
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.

BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/115ceedf72a6b2be9a7b3d60f7cd36afa0192dde
Cr-Commit-Position: refs/heads/master@{#372167}

[Primiano Tucci (use gerrit), 2016-01-27 16:25:40]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator for all targets

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

to

==========
allocator cleanup: remove dependencies on allocator for all targets

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[Primiano Tucci (use gerrit), 2016-01-27 19:24:59]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator for all targets

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator for all targets


Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build:
  https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy changes. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[Primiano Tucci (use gerrit), 2016-01-27 20:29:43]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator for all targets


Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build:
  https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy changes. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator from all targets

This CL concludes the allocator cleanup removing the extra dependencies
from various targets (% base) to allocator.

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build:
  https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy changes. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[Primiano Tucci (use gerrit), 2016-01-27 20:30:22]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator from all targets

This CL concludes the allocator cleanup removing the extra dependencies
from various targets (% base) to allocator.

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build:
  https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy changes. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator from all targets

This CL concludes the allocator cleanup removing the extra dependencies
from various targets (% base) to allocator.

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy changes. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[Primiano Tucci (use gerrit), 2016-01-27 20:32:40]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator from all targets

This CL concludes the allocator cleanup removing the extra dependencies
from various targets (% base) to allocator.

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy changes. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator from all targets

This CL concludes the allocator cleanup removing the extra dependencies
from various targets (% base) to allocator.

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[Primiano Tucci (use gerrit), 2016-01-27 20:38:28]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator from all targets

This CL concludes the allocator cleanup removing the extra dependencies
from various targets (% base) to allocator.

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator from all targets

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[Primiano Tucci (use gerrit), 2016-01-27 20:38:28]

primiano@chromium.org changed reviewers:
+ wfh@chromium.org

[Primiano Tucci (use gerrit), 2016-01-27 20:41:02]

Will, can I ask you to look just at the ninja file diff in the commit message
and see if that looks sane to you?
I will find reviewers for the CL itself.

[Primiano Tucci (use gerrit), 2016-01-28 14:41:26]

primiano@chromium.org changed reviewers:
+ petrcermak@chromium.org

[Primiano Tucci (use gerrit), 2016-01-28 14:41:27]

petrcermak@ can I ask you a first round of eyeballing on this? I know that you
are not an owner of these folders, but I'd appreciate a committer to do a first
sanity check to prevent I did something too stupid.
As visible this CL is extremely mechanical. What I am doing here is removing all
dependencies from allocator target (Both in gyp and GN) from everything % base.

Can I haz a 2nd pair of eyes on this? I think what I am asking is to check that
I didn't mistakenly dropped or altered other conditions other than allocator.

Thanks very much!

[petrcermak, 2016-01-28 14:58:51]

I've throroughly checked through the whole patch and the only changes done by
this patch are:

  * Remove dependency on allocator.gyp (many places)
  * Remove dependency on base.gyp (3 places)
  * Remove --rdynamic flag

As long as that is intended, you have a non-owner (i.e. worthless) LGTM from me
:-)

Petr

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/platform/blink_platform_tests.gyp (left):

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/platform/blink_platform_tests.gyp:58:
'<(DEPTH)/base/base.gyp:base',
Is the dependency on base supposed to be removed as well?

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/platform/blink_platform_tests.gyp:109:
'<(DEPTH)/base/base.gyp:base',
ditto

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/wtf/wtf_tests.gyp (left):

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/wtf/wtf_tests.gyp:52: '<(DEPTH)/base/base.gyp:base',
ditto

[Primiano Tucci (use gerrit), 2016-01-28 15:20:03]

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/platform/blink_platform_tests.gyp (left):

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/platform/blink_platform_tests.gyp:58:
'<(DEPTH)/base/base.gyp:base',
On 2016/01/28 14:58:50, petrcermak wrote:
> Is the dependency on base supposed to be removed as well?

Yup it is redundant as they get it via test_support_base. There is really no
reason why they should additionally depend on base, but only on some platforms.
But thanks for double-checking. This is precisely the reason why I asked for a
review.

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/platform/blink_platform_tests.gyp:109:
'<(DEPTH)/base/base.gyp:base',
On 2016/01/28 14:58:50, petrcermak wrote:
> ditto
Ack.

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/wtf/wtf_tests.gyp (left):

https://codereview.chromium.org/1647453002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/wtf/wtf_tests.gyp:52: '<(DEPTH)/base/base.gyp:base',
On 2016/01/28 14:58:50, petrcermak wrote:
> ditto

Acknowledged.

[brettw, 2016-01-28 19:11:14]

brettw@chromium.org changed reviewers:
+ brettw@chromium.org

[brettw, 2016-01-28 19:11:14]

rs lgtm

[Will Harris, 2016-01-28 19:22:59]

On 2016/01/28 19:11:14, brettw wrote:
> rs lgtm

yup lgtm. the only large change is the GN, Linux, component build, as you said,
but as long as base_unittests still has the right dependency (which it does)
then should be fine.

[Primiano Tucci (use gerrit), 2016-01-28 19:29:45]

The CQ bit was checked by primiano@chromium.org

[Primiano Tucci (use gerrit), 2016-01-28 19:29:47]

The patchset sent to the CQ was uploaded after l-g-t-m from
petrcermak@chromium.org
Link to the patchset: https://codereview.chromium.org/1647453002/#ps40001
(title: "pure rebase")

[Primiano Tucci (use gerrit), 2016-01-28 19:31:37]

+thakis FYI as he followed all this cleanup

[commit-bot: I haz the power, 2016-01-28 19:33:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1647453002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1647453002/40001

[commit-bot: I haz the power, 2016-01-28 21:11:03]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator from all targets

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator from all targets

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[commit-bot: I haz the power, 2016-01-28 21:11:04]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-01-28 21:12:05]

Description was changed from

==========
allocator cleanup: remove dependencies on allocator from all targets

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our 
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator) 
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.


BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
allocator cleanup: remove dependencies on allocator from all targets

Overview of the cleanup:
------------------------
- In the context of the discussion in https://goo.gl/K2m649 our
  illustrious base/ owners suggested that having base being the only
  target depending on allocator (and having the other targets inherit
  that recursively) would be a more scalable solution, as opposite to
  having to remember to add an exec -> allocator to each target.
- This base -> allocator dep landed in crrev.com/1616793003.
- After that CL, until this point, many targets got two paths that lead
  to allocator:
    1. The indirect one via base (content_shell -> base -> allocator)
       which is what we want to preserve.
    2. The direct one (content_shell -> allocator) which is the
       inconsistent one we want to drop.

This CL gets rid of all the instances of 2.
The nice property of this sequencing of CLs is that the effect of this
large change on the final .ninja files is minimal. See details below.

Effect on the produced ninja files:
-----------------------------------

GYP, Linux, static build: https://paste.ee/p/RaJLj
  Just some small reordering of include paths.
  Dropping --rdynamic (as expected) in keyboard_unittests

GN, Linux, static build: https://paste.ee/p/zYtrQ
  Various targets lose the dependency on allocator.stamp (expected, the
  order is enforced by depending on base which depends on allocator.stamp)

GN, Android (both component and static are similar): https://paste.ee/p/Iq6DD
  As above.

GN, Linux, component build: https://paste.ee/p/jHUmv
  This is the most juicy change. As expected all the ODR goes away and
  random targets stop re-linking tcmalloc .o files.

GYP, Win, static: https://paste.ee/p/j0IjL
  Extremely minimal changes: All.ninja stops depending on libcmt.lib.
  Sounds fine as base (which depends on that) is definitely depending on libcmt.

GN, Win, static: https://paste.ee/p/Q6zTo
  Minimal change, like Linux GN, dropping dependencies on the .stamp files.

BUG=564618
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/115ceedf72a6b2be9a7b3d60f7cd36afa0192dde
Cr-Commit-Position: refs/heads/master@{#372167}
==========

[commit-bot: I haz the power, 2016-01-28 21:12:07]

Patchset 3 (id:??) landed as
https://crrev.com/115ceedf72a6b2be9a7b3d60f7cd36afa0192dde
Cr-Commit-Position: refs/heads/master@{#372167}