Make browser process a singleton service

Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is used to package any existing global services previously
associated with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407, 594852
TEST=chrome --mash functions normally, standalone mash runner functions normally with chrome, chrome non-mash functions normally, browser_tests work with and without --run-in-mash
TBR=tsepez@chromium.org

Review-Url: https://codereview.chromium.org/2695803004
Cr-Commit-Position: refs/heads/master@{#450750}
Committed: https://chromium.googlesource.com/chromium/src/+/73b9848f38e00bf1b9f48cfa2da7f8cda7c41599

[Ken Rockot(use gerrit already), 2017-02-14 04:03:24]

Description was changed from

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to service
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser. As such code which
previously connected to content_browser has been fixed to do something
less weird.

An "ash" service is now embedded in the browser process when not
running in mash mode - this allows simplification of ash client code
as there is no longer a need to switch the target service name at
runtime when connecting to ash interfaces.

[TODO: Hook up Keyboard and Launchable interfaces in some sane way
 - probably via a separate packaged singleton private to chrome -
 and delete ChromeInterfaceFactory]

BUG=649407
==========

to

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser. As such code which
previously connected to content_browser has been fixed to do something
less weird.

An "ash" service is now embedded in the browser process when not
running in mash mode - this allows simplification of ash client code
as there is no longer a need to switch the target service name at
runtime when connecting to ash interfaces.

[TODO: Hook up Keyboard and Launchable interfaces in some sane way
 - probably via a separate packaged singleton private to chrome -
 and delete ChromeInterfaceFactory]

BUG=649407
==========

[Ken Rockot(use gerrit already), 2017-02-14 04:05:15]

Description was changed from

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser. As such code which
previously connected to content_browser has been fixed to do something
less weird.

An "ash" service is now embedded in the browser process when not
running in mash mode - this allows simplification of ash client code
as there is no longer a need to switch the target service name at
runtime when connecting to ash interfaces.

[TODO: Hook up Keyboard and Launchable interfaces in some sane way
 - probably via a separate packaged singleton private to chrome -
 and delete ChromeInterfaceFactory]

BUG=649407
==========

to

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser in order to avoid
races with the content_browser instance created (now asynchronously)
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces.

 - TODO: Hook up Keyboard and Launchable interfaces in some sane way.
   probably via a separate packaged singleton private to chrome.
   delete ChromeInterfaceFactory

BUG=649407
==========

[Ken Rockot(use gerrit already), 2017-02-14 07:01:36]

Description was changed from

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser in order to avoid
races with the content_browser instance created (now asynchronously)
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces.

 - TODO: Hook up Keyboard and Launchable interfaces in some sane way.
   probably via a separate packaged singleton private to chrome.
   delete ChromeInterfaceFactory

BUG=649407
==========

to

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces.

 - TODO: Hook up Keyboard and Launchable interfaces in some sane way.
   probably via a separate packaged singleton private to chrome.
   delete ChromeInterfaceFactory

BUG=649407
==========

[Ken Rockot(use gerrit already), 2017-02-14 22:31:36]

Description was changed from

==========
Make browser process a singleton service

[WIP]

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces.

 - TODO: Hook up Keyboard and Launchable interfaces in some sane way.
   probably via a separate packaged singleton private to chrome.
   delete ChromeInterfaceFactory

BUG=649407
==========

to

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
==========

[Ken Rockot(use gerrit already), 2017-02-14 22:33:40]

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

[commit-bot: I haz the power, 2017-02-14 22:34:26]

Dry run: 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, 2017-02-14 22:39:08]

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

[commit-bot: I haz the power, 2017-02-14 22:39:08]

Dry run: Try jobs failed on following builders:
  ios-device on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device/builds...)
  ios-device-xcode-clang on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device-xcode-...)
  ios-simulator on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator/bui...)
  ios-simulator-xcode-clang on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator-xco...)

[Ken Rockot(use gerrit already), 2017-02-14 22:46:12]

Description was changed from

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is then used to package any existing global services previously
assocaited with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
==========

to

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is used to package any existing global services previously
associated with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
TEST=chrome --mash functions normally, standalone mash runner functions normally
with chrome, chrome non-mash functions normally, browser_tests work with and
without --run-in-mash
==========

[Ken Rockot(use gerrit already), 2017-02-14 22:53:09]

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

[commit-bot: I haz the power, 2017-02-14 22:54:03]

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

[Ken Rockot(use gerrit already), 2017-02-14 22:54:21]

rockot@chromium.org changed reviewers:
+ sky@chromium.org

[Ken Rockot(use gerrit already), 2017-02-14 22:54:22]

Scott, since Ben is OOO you win the prestigious honor of reviewing this CL!
Please take a look at everything

[commit-bot: I haz the power, 2017-02-14 22:59:18]

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

[commit-bot: I haz the power, 2017-02-14 22:59:19]

Dry run: Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)
  mac_chromium_compile_dbg_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_comp...)
  mac_chromium_rel_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_rel_...)

[Ken Rockot(use gerrit already), 2017-02-14 23:01:39]

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

[commit-bot: I haz the power, 2017-02-14 23:03:06]

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

[sky, 2017-02-15 00:35:26]

https://codereview.chromium.org/2695803004/diff/60001/chrome/app/BUILD.gn
File chrome/app/BUILD.gn (right):

https://codereview.chromium.org/2695803004/diff/60001/chrome/app/BUILD.gn#new...
chrome/app/BUILD.gn:407:
":chrome_content_packaged_services_manifest_overlay_for_mash",
Should all the mash related stuff be guarded by enable_package_mash_services?

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/browser_...
File chrome/browser/browser_resources.grd (right):

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/browser_...
chrome/browser/browser_resources.grd:628: <include
name="IDR_CHROME_CONTENT_PACKAGED_SERVICES_MANIFEST_OVERLAY"
file="${root_gen_dir}\chrome\app\chrome_content_packaged_services_manifest_overlay.json"
type="BINDATA" use_base_dir="false"/>
Only include if enable_package_mash_services?

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/chrome_c...
File chrome/browser/chrome_content_browser_client.cc (right):

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/chrome_c...
chrome/browser/chrome_content_browser_client.cc:473: const char
kChromeServiceName[] = "chrome";
Add description.

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/chrome_c...
chrome/browser/chrome_content_browser_client.cc:475: class ChromeService
Please add description, also, I would be inclined to use ChromeOS in the name to
reinforce this is only for chromeos.

[Ken Rockot(use gerrit already), 2017-02-15 01:34:04]

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

[commit-bot: I haz the power, 2017-02-15 01:34:30]

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

[Ken Rockot(use gerrit already), 2017-02-15 01:38:05]

https://codereview.chromium.org/2695803004/diff/60001/chrome/app/BUILD.gn
File chrome/app/BUILD.gn (right):

https://codereview.chromium.org/2695803004/diff/60001/chrome/app/BUILD.gn#new...
chrome/app/BUILD.gn:407:
":chrome_content_packaged_services_manifest_overlay_for_mash",
On 2017/02/15 at 00:35:26, sky wrote:
> Should all the mash related stuff be guarded by enable_package_mash_services?

Sure, done

https://codereview.chromium.org/2695803004/diff/60001/chrome/app/BUILD.gn#new...
chrome/app/BUILD.gn:407:
":chrome_content_packaged_services_manifest_overlay_for_mash",
On 2017/02/15 at 00:35:26, sky wrote:
> Should all the mash related stuff be guarded by enable_package_mash_services?

Done where possible. For now the catalog in //mash/BUILD.gn still needs to
reference the for_mash manifest target, but //mash/BUILD.gn otherwise doesn't
know about enable_package_mash_services, so I've left the manifest targets
defined unconditionally. I could teach it about that flag but that kinda seems
like it's just making the existing layering violation worse. WDYT?

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/browser_...
File chrome/browser/browser_resources.grd (right):

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/browser_...
chrome/browser/browser_resources.grd:628: <include
name="IDR_CHROME_CONTENT_PACKAGED_SERVICES_MANIFEST_OVERLAY"
file="${root_gen_dir}\chrome\app\chrome_content_packaged_services_manifest_overlay.json"
type="BINDATA" use_base_dir="false"/>
On 2017/02/15 at 00:35:26, sky wrote:
> Only include if enable_package_mash_services?

Nope, this is used in and out of the mash environment. For example chrome
packages image_decoder and preferences services here in all environments.

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/chrome_c...
File chrome/browser/chrome_content_browser_client.cc (right):

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/chrome_c...
chrome/browser/chrome_content_browser_client.cc:473: const char
kChromeServiceName[] = "chrome";
On 2017/02/15 at 00:35:26, sky wrote:
> Add description.

Done

https://codereview.chromium.org/2695803004/diff/60001/chrome/browser/chrome_c...
chrome/browser/chrome_content_browser_client.cc:475: class ChromeService
On 2017/02/15 at 00:35:26, sky wrote:
> Please add description, also, I would be inclined to use ChromeOS in the name
to reinforce this is only for chromeos.

Done

[commit-bot: I haz the power, 2017-02-15 04:14:45]

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

[commit-bot: I haz the power, 2017-02-15 04:14:46]

Dry run: Try jobs failed on following builders:
  linux_chromium_chromeos_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Ken Rockot(use gerrit already), 2017-02-15 08:31:52]

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

[commit-bot: I haz the power, 2017-02-15 08:31:59]

Dry run: 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, 2017-02-15 09:43:43]

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

[commit-bot: I haz the power, 2017-02-15 09:43:44]

Dry run: This issue passed the CQ dry run.

[sky, 2017-02-15 16:22:56]

LGTM

[Ken Rockot(use gerrit already), 2017-02-15 16:31:54]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2017-02-15 16:32:59]

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, 2017-02-15 16:43:12]

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

[commit-bot: I haz the power, 2017-02-15 16:43:14]

Try jobs failed on following builders:
  chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[Ken Rockot(use gerrit already), 2017-02-15 17:11:38]

rockot@chromium.org changed reviewers:
+ jam@chromium.org

[Ken Rockot(use gerrit already), 2017-02-15 17:11:39]

+jam for content

[jam, 2017-02-15 18:24:25]

lgtm

[Ken Rockot(use gerrit already), 2017-02-15 18:32:49]

Description was changed from

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is used to package any existing global services previously
associated with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
TEST=chrome --mash functions normally, standalone mash runner functions normally
with chrome, chrome non-mash functions normally, browser_tests work with and
without --run-in-mash
==========

to

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is used to package any existing global services previously
associated with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
TEST=chrome --mash functions normally, standalone mash runner functions normally
with chrome, chrome non-mash functions normally, browser_tests work with and
without --run-in-mash
TBR=tsepez@chromium.org
==========

[Ken Rockot(use gerrit already), 2017-02-15 18:33:01]

TBR tsepez for constant addition in mojom

[Ken Rockot(use gerrit already), 2017-02-15 18:33:03]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2017-02-15 18:34:08]

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, 2017-02-15 18:58:07]

CQ is committing da patch.

Bot data: {"patchset_id": 100001, "attempt_start_ts": 1487183583276370,
"parent_rev": "5e8ee75a86c1c36b53c1a6b5ec28d89b7bb6d4ad", "commit_rev":
"73b9848f38e00bf1b9f48cfa2da7f8cda7c41599"}

[commit-bot: I haz the power, 2017-02-15 18:58:43]

Description was changed from

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is used to package any existing global services previously
associated with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
TEST=chrome --mash functions normally, standalone mash runner functions normally
with chrome, chrome non-mash functions normally, browser_tests work with and
without --run-in-mash
TBR=tsepez@chromium.org
==========

to

==========
Make browser process a singleton service

Changes the primordial browser process service manager connection
to identify as a singleton service named content_packaged_services.
This is used to package any existing global services previously
associated with the global root content_browser connection.

The root content_browser connection still exists in order to serve
as a client process host for non-renderer child processes.

One result of this change is that no services other than content_*
are allowed to connect directly to content_browser, in order to avoid
racing with the (now asynchronous) content_browser instance creation
by ServiceManagerContext. As such, code which previously connected to
content_browser has also been fixed to do something less weird. To wit:

 - An "ash" service is now embedded in the browser process when not
   running in mash mode - this allows simplification of ash client code
   as there is no longer a need to switch the target service name at
   runtime when connecting to ash interfaces
 - A new "chrome" service is packaged in content_packaged_services by
   within Chrome. This hosts a mash::mojom::Launchable interface, allowing
   Chrome to be launched within mash by connecting to "chrome"
 - ChromeInterfaceFactory is deleted

Somewhat unrelated but because this CL exposed a timing issue between packaged
service connection and main thread browser initialization, this also removes the
unnecessary blocking of startup on service manager connection (see change in
chrome_browser_main_extra_parts_views.cc.)

BUG=649407,594852
TEST=chrome --mash functions normally, standalone mash runner functions normally
with chrome, chrome non-mash functions normally, browser_tests work with and
without --run-in-mash
TBR=tsepez@chromium.org

Review-Url: https://codereview.chromium.org/2695803004
Cr-Commit-Position: refs/heads/master@{#450750}
Committed:
https://chromium.googlesource.com/chromium/src/+/73b9848f38e00bf1b9f48cfa2da7...
==========

[commit-bot: I haz the power, 2017-02-15 18:58:44]

Committed patchset #6 (id:100001) as
https://chromium.googlesource.com/chromium/src/+/73b9848f38e00bf1b9f48cfa2da7...