Reland: chromeos: Convert ash VPNDelegate interface to mojo

Reland: chromeos: Convert ash VPNDelegate interface to mojo

Relanding because previous version crrev.com/434004 was missing an entry in
chrome_content_browser_manifest_overlay.json

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
TBR=tsepez@chromium.org for copying manifest line from ash/mus/manifest.json
to chrome_content_browser_manifest_overlay.json

Committed: https://crrev.com/c469dd5d3ceede70f70a1e86b97c711696b1fa49
Cr-Commit-Position: refs/heads/master@{#434079}

[James Cook, 2016-11-18 23:14:14]

Description was changed from

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider). We can
collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
==========

to

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
==========

[James Cook, 2016-11-18 23:14:44]

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

[commit-bot: I haz the power, 2016-11-18 23:15:43]

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

[James Cook, 2016-11-18 23:15:46]

jamescook@chromium.org changed reviewers:
+ rockot@chromium.org, stevenjb@chromium.org

[James Cook, 2016-11-18 23:15:47]

stevenjb, please take a look.

rockot, can you take a quick look at vpn_list.h/cc and vpn_list_client.h/cc to
see if I'm doing the right thing with mojo interface bindings?

[commit-bot: I haz the power, 2016-11-18 23:19:55]

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

[commit-bot: I haz the power, 2016-11-18 23:19:55]

Dry run: Try jobs failed on following builders:
  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_...)

[James Cook, 2016-11-18 23:23:24]

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

[commit-bot: I haz the power, 2016-11-18 23:24:32]

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, 2016-11-18 23:28:48]

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

[commit-bot: I haz the power, 2016-11-18 23:28:50]

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

[stevenjb, 2016-11-19 00:04:58]

https://codereview.chromium.org/2513673004/diff/40001/ash/common/system/chrom...
File ash/common/system/chromeos/network/vpn_list.cc (right):

https://codereview.chromium.org/2513673004/diff/40001/ash/common/system/chrom...
ash/common/system/chromeos/network/vpn_list.cc:61: void
VpnList::ShowAddPage(const VPNProvider::Key& key) {
This is kind of an odd method to have in a class named simply "VpnList".

I'd actually rather see the mojo call on the same interface that we use to show
Settings. (Ultimately all networking UI is going to be attached to Settings
anyway). Does that seem reasonable?

https://codereview.chromium.org/2513673004/diff/40001/ash/public/interfaces/v...
File ash/public/interfaces/vpn_list.mojom (right):

https://codereview.chromium.org/2513673004/diff/40001/ash/public/interfaces/v...
ash/public/interfaces/vpn_list.mojom:17: SetClient(VpnListClient client);
This just seems confusing to me. It seems like the Show methods should just be
part of whatever client handles "show UI" calls (e.g. "show settings"). There's
no real need to associate that with the model.

https://codereview.chromium.org/2513673004/diff/40001/ash/public/interfaces/v...
ash/public/interfaces/vpn_list.mojom:32: ShowThirdPartyAddNetworkPage(string
extension_id);
In theory we could make this a single method, with an empty extension_id
indicating "built in"...

[James Cook, 2016-11-19 00:18:30]

rockot, please hold off. Per stevenjb's suggestions I'm going to move the "Show
Add Network" stuff over to the existing ash::mojom::SystemTrayClient interface,
which will eliminate a bunch of the mojo stuff from this CL.

[stevenjb, 2016-11-19 00:18:58]

https://codereview.chromium.org/2513673004/diff/2/chrome/browser/ui/ash/vpn_l...
File chrome/browser/ui/ash/vpn_list_client.cc (right):

https://codereview.chromium.org/2513673004/diff/2/chrome/browser/ui/ash/vpn_l...
chrome/browser/ui/ash/vpn_list_client.cc:168:
vpn_list->SetThirdPartyVpnProviders(std::move(third_party_providers));
Also, if we eliminate the UI calls, we wouldn't need to invert the relationship,
i.e. this class could be the server and the UI can request the list and listen
to updates. If we do it this way then aren't we assuming that the client already
exists?

[James Cook, 2016-11-19 00:49:18]

https://codereview.chromium.org/2513673004/diff/2/chrome/browser/ui/ash/vpn_l...
File chrome/browser/ui/ash/vpn_list_client.cc (right):

https://codereview.chromium.org/2513673004/diff/2/chrome/browser/ui/ash/vpn_l...
chrome/browser/ui/ash/vpn_list_client.cc:168:
vpn_list->SetThirdPartyVpnProviders(std::move(third_party_providers));
On 2016/11/19 00:18:58, stevenjb wrote:
> Also, if we eliminate the UI calls, we wouldn't need to invert the
relationship,
> i.e. this class could be the server and the UI can request the list and listen
> to updates. If we do it this way then aren't we assuming that the client
already
> exists?

Dunno if this was clear in my other CL, but there's one other reason to invert
the relationship: Ash needs to be able to decide, synchronously when the system
tray menu opens, whether or not to show the row for VPNs row. So it needs to
have the list of VPNs before that, otherwise we have to wait for an IPC
round-trip to chrome (which might have a busy UI thread) before we can spawn the
menu. Regardless of how we think about client and server, ash still needs a
master list, and still needs updates from chrome.

I think it's reasonable to assume that the client (ash) always exists when
chrome starts up. That's consistent with how Chrome runs on other platforms --
the "OS" provides system UI, and chrome pokes into it to do stuff like show dock
icons.

[James Cook, 2016-11-22 02:02:04]

Patchset #6 (id:90001) has been deleted

[James Cook, 2016-11-22 02:02:29]

jamescook@chromium.org changed reviewers:
- rockot@chromium.org

[James Cook, 2016-11-22 02:03:25]

Steven, please take another look. I moved the "show UI" calls to the existing
system tray mojo interface as we discussed earlier.

[James Cook, 2016-11-22 02:03:33]

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

[commit-bot: I haz the power, 2016-11-22 02:04:24]

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, 2016-11-22 02:08:11]

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

[commit-bot: I haz the power, 2016-11-22 02:08:12]

Dry run: Try jobs failed on following builders:
  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_...)

[James Cook, 2016-11-22 02:29:39]

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

[commit-bot: I haz the power, 2016-11-22 02:30:15]

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, 2016-11-22 02:35:49]

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

[commit-bot: I haz the power, 2016-11-22 02:35:50]

Dry run: Try jobs failed on following builders:
  chromeos_amd64-generic_chromium_compile_only_ng on
master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromeos_amd64-...)
  chromeos_daisy_chromium_compile_only_ng on master.tryserver.chromium.linux
(JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromeos_daisy_...)
  chromeos_x86-generic_chromium_compile_only_ng on
master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromeos_x86-ge...)

[James Cook, 2016-11-22 17:18:40]

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

[commit-bot: I haz the power, 2016-11-22 17:19:08]

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

[stevenjb, 2016-11-22 17:19:48]

LGTM!

https://codereview.chromium.org/2513673004/diff/130001/ash/common/system/chro...
File ash/common/system/chromeos/network/vpn_list.h (right):

https://codereview.chromium.org/2513673004/diff/130001/ash/common/system/chro...
ash/common/system/chromeos/network/vpn_list.h:73: void
BindRequest(mojom::VpnListRequest request);
Just for my edification: Why isn't this (and bindings_) part of mojom::VpnList?

[James Cook, 2016-11-22 17:37:30]

https://codereview.chromium.org/2513673004/diff/130001/ash/common/system/chro...
File ash/common/system/chromeos/network/vpn_list.h (right):

https://codereview.chromium.org/2513673004/diff/130001/ash/common/system/chro...
ash/common/system/chromeos/network/vpn_list.h:73: void
BindRequest(mojom::VpnListRequest request);
On 2016/11/22 17:19:48, stevenjb wrote:
> Just for my edification: Why isn't this (and bindings_) part of
mojom::VpnList?

This method (and bindings_) can actually exist anywhere. This works as follows:
1. Some process (say, chrome) makes a mojo connection to ash.
2. Ash sees an inbound mojo connection request. The app runs an OnConnect
method. See
https://cs.chromium.org/chromium/src/ash/mus/window_manager_application.cc?dr...
3. OnConnect registers all the mojo interfaces that are allowed to be used by
that connection.
https://cs.chromium.org/chromium/src/ash/common/mojo_interface_factory.cc?dr&...
4. Later, when chrome requests to use the VpnList interface, ash needs to "bind"
that request to the code/class that implements the interface. The request comes
in on the IO thread and there's a base::Bind callback that binds the request to
an object instance.
https://cs.chromium.org/chromium/src/ash/common/mojo_interface_factory.cc?dr&...
5. The |binding_| stores the mapping of interface request to object instance. It
can live anywhere. Sometimes it is a BindingSet if there is the possibility that
more than one external client would connect to the same object instance.

In ash we store the bindings in the implementation classes themselves.

In chrome we sometimes do that, and sometimes store them elsewhere. See
https://cs.chromium.org/chromium/src/chrome/browser/chromeos/chrome_interface...

I'm not sure if one is necessarily better than the other.

[James Cook, 2016-11-22 17:39:25]

jamescook@chromium.org changed reviewers:
+ sky@chromium.org, tsepez@chromium.org

[James Cook, 2016-11-22 17:39:26]

sky, can I get OWNERS for
chrome/browser/ui/views/ash/chrome_browser_main_extra_parts_ash.cc/h ?

tsepez, can I get OWNERS for ash/public/interfaces/system_tray.mojom and
vpn_list.mojom?

[commit-bot: I haz the power, 2016-11-22 17:45:41]

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

[commit-bot: I haz the power, 2016-11-22 17:45:41]

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

[James Cook, 2016-11-22 17:48:20]

Patchset #8 (id:140001) has been deleted

[James Cook, 2016-11-22 17:48:28]

Patchset #2 (id:20001) has been deleted

[stevenjb, 2016-11-22 17:50:12]

Thanks for the detailed explanation!

It does however seem like there is a lot of boilerplate code to do this:

https://cs.chromium.org/search/?q=bindings_.AddBinding%5C(this,+std::move%5C(...

It might be nice if there was a helper template class, e.g. public
MojoBindings<mojom::Foo>.



On Tue, Nov 22, 2016 at 10:37 AM, <jamescook@chromium.org> wrote:

>
> https://codereview.chromium.org/2513673004/diff/130001/
> ash/common/system/chromeos/network/vpn_list.h
> File ash/common/system/chromeos/network/vpn_list.h (right):
>
> https://codereview.chromium.org/2513673004/diff/130001/
> ash/common/system/chromeos/network/vpn_list.h#newcode73
> ash/common/system/chromeos/network/vpn_list.h:73: void
> BindRequest(mojom::VpnListRequest request);
> On 2016/11/22 17:19:48, stevenjb wrote:
> > Just for my edification: Why isn't this (and bindings_) part of
> mojom::VpnList?
>
> This method (and bindings_) can actually exist anywhere. This works as
> follows:
> 1. Some process (say, chrome) makes a mojo connection to ash.
> 2. Ash sees an inbound mojo connection request. The app runs an
> OnConnect method. See
> https://cs.chromium.org/chromium/src/ash/mus/window_
> manager_application.cc?dr&sq=package:chromium&rcl=1479817139&l=147
> 3. OnConnect registers all the mojo interfaces that are allowed to be
> used by that connection.
> https://cs.chromium.org/chromium/src/ash/common/mojo_
> interface_factory.cc?dr&sq=package:chromium&rcl=1479817139&l=50
> 4. Later, when chrome requests to use the VpnList interface, ash needs
> to "bind" that request to the code/class that implements the interface.
> The request comes in on the IO thread and there's a base::Bind callback
> that binds the request to an object instance.
> https://cs.chromium.org/chromium/src/ash/common/mojo_
> interface_factory.cc?dr&sq=package:chromium&rcl=1479817139&l=38
> 5. The |binding_| stores the mapping of interface request to object
> instance. It can live anywhere. Sometimes it is a BindingSet if there is
> the possibility that more than one external client would connect to the
> same object instance.
>
> In ash we store the bindings in the implementation classes themselves.
>
> In chrome we sometimes do that, and sometimes store them elsewhere. See
> https://cs.chromium.org/chromium/src/chrome/browser/
> chromeos/chrome_interface_factory.cc?dr&q=chrome_interface_&sq=package:
> chromium&l=157
>
> I'm not sure if one is necessarily better than the other.
>
> https://codereview.chromium.org/2513673004/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[Tom Sepez, 2016-11-22 18:10:03]

.mojom OWNERS LGTM

[sky, 2016-11-22 20:50:06]

LGTM

[James Cook, 2016-11-22 21:13:35]

The CQ bit was checked by jamescook@chromium.org

[James Cook, 2016-11-22 21:13:36]

The patchset sent to the CQ was uploaded after l-g-t-m from
stevenjb@chromium.org
Link to the patchset: https://codereview.chromium.org/2513673004/#ps160001
(title: "fix linux compile")

[commit-bot: I haz the power, 2016-11-22 21:14:03]

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, 2016-11-22 22:18:05]

CQ is committing da patch.

Bot data: {"patchset_id": 160001, "attempt_start_ts": 1479849215505740,
"parent_rev": "7e66dc9f0237d9bc618ec26c3ec1cc118da34127", "commit_rev":
"8ae68851e371415f39f58a6af385496ae8fd577e"}

[commit-bot: I haz the power, 2016-11-22 22:18:30]

Description was changed from

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
==========

to

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
==========

[commit-bot: I haz the power, 2016-11-22 22:18:31]

Committed patchset #7 (id:160001)

[commit-bot: I haz the power, 2016-11-22 22:20:46]

Description was changed from

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
==========

to

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest

Committed: https://crrev.com/56e0ddccb50e0b062ea6fea8640c6dcea9ea59a7
Cr-Commit-Position: refs/heads/master@{#434004}
==========

[commit-bot: I haz the power, 2016-11-22 22:20:47]

Patchset 7 (id:??) landed as
https://crrev.com/56e0ddccb50e0b062ea6fea8640c6dcea9ea59a7
Cr-Commit-Position: refs/heads/master@{#434004}

[James Cook, 2016-11-23 00:10:39]

A revert of this CL (patchset #7 id:160001) has been created in
https://codereview.chromium.org/2522613004/ by jamescook@chromium.org.

The reason for reverting is: Breaks the VPN list in classic ash due to a missing
entry in chrome's content browser manifest overlay. (I probably didn't catch
this in manual testing due to stale generated manifest files from an earlier
version of this patch.)

I will re-land with a fixed manifest file.
.

[James Cook, 2016-11-23 00:44:50]

Description was changed from

==========
chromeos: Convert ash VPNDelegate interface to mojo

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest

Committed: https://crrev.com/56e0ddccb50e0b062ea6fea8640c6dcea9ea59a7
Cr-Commit-Position: refs/heads/master@{#434004}
==========

to

==========
Reland: chromeos: Convert ash VPNDelegate interface to mojo

Relanding because previous version crrev.com/434004 was missing an entry in
chrome_content_browser_manifest_overlay.json

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
TBR=tsepez@chromium.org for copying manifest line from ash/mus/manifest.json
to chrome_content_browser_manifest_overlay.json
==========

[James Cook, 2016-11-23 00:45:28]

The CQ bit was checked by jamescook@chromium.org

[James Cook, 2016-11-23 00:45:29]

The patchset sent to the CQ was uploaded after l-g-t-m from tsepez@chromium.org,
stevenjb@chromium.org, sky@chromium.org
Link to the patchset: https://codereview.chromium.org/2513673004/#ps180001
(title: "fix manifest for reland")

[commit-bot: I haz the power, 2016-11-23 00:46:40]

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, 2016-11-23 01:58:02]

CQ is committing da patch.

Bot data: {"patchset_id": 180001, "attempt_start_ts": 1479861928472110,
"parent_rev": "a6726a839b13c17111a4659aaf2d87de39057fc1", "commit_rev":
"8c6fa55a4d23cb38121b4c1fede2b546698cccf7"}

[commit-bot: I haz the power, 2016-11-23 01:58:29]

Description was changed from

==========
Reland: chromeos: Convert ash VPNDelegate interface to mojo

Relanding because previous version crrev.com/434004 was missing an entry in
chrome_content_browser_manifest_overlay.json

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
TBR=tsepez@chromium.org for copying manifest line from ash/mus/manifest.json
to chrome_content_browser_manifest_overlay.json
==========

to

==========
Reland: chromeos: Convert ash VPNDelegate interface to mojo

Relanding because previous version crrev.com/434004 was missing an entry in
chrome_content_browser_manifest_overlay.json

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
TBR=tsepez@chromium.org for copying manifest line from ash/mus/manifest.json
to chrome_content_browser_manifest_overlay.json
==========

[commit-bot: I haz the power, 2016-11-23 01:58:31]

Committed patchset #8 (id:180001)

[commit-bot: I haz the power, 2016-11-23 02:00:07]

Description was changed from

==========
Reland: chromeos: Convert ash VPNDelegate interface to mojo

Relanding because previous version crrev.com/434004 was missing an entry in
chrome_content_browser_manifest_overlay.json

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
TBR=tsepez@chromium.org for copying manifest line from ash/mus/manifest.json
to chrome_content_browser_manifest_overlay.json
==========

to

==========
Reland: chromeos: Convert ash VPNDelegate interface to mojo

Relanding because previous version crrev.com/434004 was missing an entry in
chrome_content_browser_manifest_overlay.json

Under mustash the ash shell runs in a different process than chrome browser.
Create mojo interfaces to allow chrome to tell ash about extension-based
third-party VPN providers and to allow ash to ask chrome to open VPN config
UI.

This CL maintains the existing ash::VPNProvider struct rather than replacing
it with ash::mojom::ThirdPartyVpnProvider because the two express different
concepts (the former includes the built-in OpenVPN/L2TP provider and is used
in several places in UI code). We can collapse them later.

BUG=651148
TEST=added to ash_unittests VpnListTest
TBR=tsepez@chromium.org for copying manifest line from ash/mus/manifest.json
to chrome_content_browser_manifest_overlay.json

Committed: https://crrev.com/c469dd5d3ceede70f70a1e86b97c711696b1fa49
Cr-Commit-Position: refs/heads/master@{#434079}
==========

[commit-bot: I haz the power, 2016-11-23 02:00:09]

Patchset 8 (id:??) landed as
https://crrev.com/c469dd5d3ceede70f70a1e86b97c711696b1fa49
Cr-Commit-Position: refs/heads/master@{#434079}