Adds the ability for the renderer to create the mojo channel

Adds the ability for the renderer to create the mojo channel

This is initiated and owned by RenderProcessHost. I've added a
method to RenderProcessHostImpl to create the channel. No one is
calling it yet, that will come after this.

RenderProcessHostImpl::CreateMojoChannel initiates the connection and
sends an IPC message to the renderer. The renderer than creates its
end of the connection.

End to end test will come once I've added all the pieces.

BUG=none
TEST=none
R=darin@chromium.org, tsepez@chromium.org, viettrungluu@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257342

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257505

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257691

[sky, 2014-03-13 22:24:30]

darin: all
tsepez: ipc
vtl: content/common/mojo/mojo_channel_init (and anything else you want to look
at)

[viettrungluu, 2014-03-13 23:02:35]

LGTM w/nits on mojo_channel_init.*. If I get around to it, I'll look at the
other files too (but if I don't get back to you, please go ahead with Darin and
Tom's reviews).

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
File content/common/mojo/mojo_channel_init.cc (right):

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.cc:50: DCHECK(io_thread_task_runner_.get()
== NULL);  // Should only init once.
Note: I imagine that even if we remove the scoped_refptr's implicit conversion
to a dumb pointer, we'll keep it testable (like scoped_ptr is), so you should be
able to do just |DCHECK(!io_thread_task_runner_);|.

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
File content/common/mojo/mojo_channel_init.h (right):

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.h:20: namespace embedder{
nit: missingspace

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.h:63: mojo::ScopedHandle handle_;
"handle" -> "bootstrap_message_pipe"?

(I guess I should make the embedder API returned a C++ ScopedMessagePipeHandle
too.)

[Tom Sepez, 2014-03-13 23:05:35]

Messages LGTM.

https://codereview.chromium.org/195993010/diff/20001/content/renderer/render_...
File content/renderer/render_thread_impl.cc (right):

https://codereview.chromium.org/195993010/diff/20001/content/renderer/render_...
content/renderer/render_thread_impl.cc:396: // MojoMessageFilter deletes itself
as necessary.
nit: can we avoid this pattern and use AddFilter instead?  It looks wrong
despite the comment.

[sky, 2014-03-13 23:10:49]

https://codereview.chromium.org/195993010/diff/20001/content/renderer/render_...
File content/renderer/render_thread_impl.cc (right):

https://codereview.chromium.org/195993010/diff/20001/content/renderer/render_...
content/renderer/render_thread_impl.cc:396: // MojoMessageFilter deletes itself
as necessary.
On 2014/03/13 23:05:36, Tom Sepez wrote:
> nit: can we avoid this pattern and use AddFilter instead?  It looks wrong
> despite the comment. 

I actually did this first, then pulled out my hair trying to figure out why
MojoMessageFilter wasn't getting the message. Turns out this happens because the
message is a control message (it isn't associated with a particular view) and
control messages don't go to the filters.

MojoMessageFilter is a bit confusingly named since it's a
RenderProcessHostObserver and not a filter, but I couldn't think of a better
name.

[sky, 2014-03-13 23:14:47]

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
File content/common/mojo/mojo_channel_init.cc (right):

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.cc:50: DCHECK(io_thread_task_runner_.get()
== NULL);  // Should only init once.
On 2014/03/13 23:02:36, viettrungluu wrote:
> Note: I imagine that even if we remove the scoped_refptr's implicit conversion
> to a dumb pointer, we'll keep it testable (like scoped_ptr is), so you should
be
> able to do just |DCHECK(!io_thread_task_runner_);|.

Done.

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
File content/common/mojo/mojo_channel_init.h (right):

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.h:20: namespace embedder{
On 2014/03/13 23:02:36, viettrungluu wrote:
> nit: missingspace

Done.

https://codereview.chromium.org/195993010/diff/20001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.h:63: mojo::ScopedHandle handle_;
On 2014/03/13 23:02:36, viettrungluu wrote:
> "handle" -> "bootstrap_message_pipe"?
> 
> (I guess I should make the embedder API returned a C++ ScopedMessagePipeHandle
> too.)

Done.

[darin (slow to review), 2014-03-14 22:03:36]

https://codereview.chromium.org/195993010/diff/60001/content/common/mojo/mojo...
File content/common/mojo/mojo_channel_init.cc (right):

https://codereview.chromium.org/195993010/diff/60001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.cc:21: static
base::LazyInstance<base::Lock> lock = LAZY_INSTANCE_INITIALIZER;
hmm, to be we don't have an equivalent of pthread_once(). perhaps an alternative
is:

struct Initializer {
  Initializer() {
    mojo::embedder::Init();
  }
};
static base::lazyInstance<Initializer>::Leaky initializer =
    LAZY_INSTANCE_INITIALIZER;

void InitMojo() {
  initializer.Get();
}

I'm not sure if it matters, but it saves you from needing to allocate a lock.
This just relies on the LazyInstance to run the constructor exactly once. (This
does not result in malloc being called.)

I agree it would probably be good to move this function to some earlier part of
startup when we know we only have one thread active.

https://codereview.chromium.org/195993010/diff/60001/content/content_renderer...
File content/content_renderer.gypi (right):

https://codereview.chromium.org/195993010/diff/60001/content/content_renderer...
content/content_renderer.gypi:307: 'renderer/mojo_message_filter.cc',
nit: we might want to pro-actively introduce a renderer/mojo/ subdirectory. i
predict it will fill up!

https://codereview.chromium.org/195993010/diff/60001/content/renderer/mojo_me...
File content/renderer/mojo_message_filter.cc (right):

https://codereview.chromium.org/195993010/diff/60001/content/renderer/mojo_me...
content/renderer/mojo_message_filter.cc:47:
ChildProcess::current()->current()->io_message_loop_proxy());
nit: did you mean to call current()->current()? isn't the second call
unnecessary?

https://codereview.chromium.org/195993010/diff/60001/content/renderer/render_...
File content/renderer/render_thread_impl.cc (right):

https://codereview.chromium.org/195993010/diff/60001/content/renderer/render_...
content/renderer/render_thread_impl.cc:404: new MojoMessageFilter(this);
Don't you need to call AddFilter? Oh, wait... I see. This class is not an
IPC::ChannelProxy::MessageFilter subclass. You should probably give this class a
different name, as *MessageFilter is usually reserved for things that are
subclasses of that interface.

Maybe call it MojoRenderProcessObserver?

[darin (slow to review), 2014-03-14 22:04:10]

LGTM with above issues addressed.

[sky, 2014-03-14 22:45:11]

https://codereview.chromium.org/195993010/diff/60001/content/common/mojo/mojo...
File content/common/mojo/mojo_channel_init.cc (right):

https://codereview.chromium.org/195993010/diff/60001/content/common/mojo/mojo...
content/common/mojo/mojo_channel_init.cc:21: static
base::LazyInstance<base::Lock> lock = LAZY_INSTANCE_INITIALIZER;
On 2014/03/14 22:03:36, darin wrote:
> hmm, to be we don't have an equivalent of pthread_once(). perhaps an
alternative
> is:
> 
> struct Initializer {
>   Initializer() {
>     mojo::embedder::Init();
>   }
> };
> static base::lazyInstance<Initializer>::Leaky initializer =
>     LAZY_INSTANCE_INITIALIZER;
> 
> void InitMojo() {
>   initializer.Get();
> }
> 
> I'm not sure if it matters, but it saves you from needing to allocate a lock.
> This just relies on the LazyInstance to run the constructor exactly once.
(This
> does not result in malloc being called.)
> 
> I agree it would probably be good to move this function to some earlier part
of
> startup when we know we only have one thread active.

Done.

https://codereview.chromium.org/195993010/diff/60001/content/content_renderer...
File content/content_renderer.gypi (right):

https://codereview.chromium.org/195993010/diff/60001/content/content_renderer...
content/content_renderer.gypi:307: 'renderer/mojo_message_filter.cc',
On 2014/03/14 22:03:36, darin wrote:
> nit: we might want to pro-actively introduce a renderer/mojo/ subdirectory. i
> predict it will fill up!

Done.

https://codereview.chromium.org/195993010/diff/60001/content/renderer/mojo_me...
File content/renderer/mojo_message_filter.cc (right):

https://codereview.chromium.org/195993010/diff/60001/content/renderer/mojo_me...
content/renderer/mojo_message_filter.cc:47:
ChildProcess::current()->current()->io_message_loop_proxy());
On 2014/03/14 22:03:36, darin wrote:
> nit: did you mean to call current()->current()? isn't the second call
> unnecessary?

My mistake. Done.

https://codereview.chromium.org/195993010/diff/60001/content/renderer/render_...
File content/renderer/render_thread_impl.cc (right):

https://codereview.chromium.org/195993010/diff/60001/content/renderer/render_...
content/renderer/render_thread_impl.cc:404: new MojoMessageFilter(this);
On 2014/03/14 22:03:36, darin wrote:
> Don't you need to call AddFilter? Oh, wait... I see. This class is not an
> IPC::ChannelProxy::MessageFilter subclass. You should probably give this class
a
> different name, as *MessageFilter is usually reserved for things that are
> subclasses of that interface.
> 
> Maybe call it MojoRenderProcessObserver?

Done.

[sky, 2014-03-14 22:45:17]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-14 22:45:34]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/80001

[commit-bot: I haz the power, 2014-03-14 22:47:23]

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

[commit-bot: I haz the power, 2014-03-14 22:47:24]

Try jobs failed on following builders:
  tryserver.chromium on linux_chromium_clang_dbg

[sky, 2014-03-14 23:33:31]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-14 23:33:36]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/80001

[commit-bot: I haz the power, 2014-03-14 23:49:50]

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

[commit-bot: I haz the power, 2014-03-14 23:49:51]

Retried try job too often on ios_dbg_simulator for step(s) base_unittests,
compile, components_unittests, content_unittests, crypto_unittests,
net_unittests, sql_unittests, ui_unittests, url_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=ios_dbg_si...

[sky, 2014-03-14 23:54:14]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-14 23:54:17]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/100001

[commit-bot: I haz the power, 2014-03-14 23:55:48]

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

[commit-bot: I haz the power, 2014-03-14 23:55:49]

Try jobs failed on following builders:
  tryserver.chromium on linux_chromium_clang_dbg

[sky, 2014-03-15 01:37:48]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-15 01:38:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/100001

[commit-bot: I haz the power, 2014-03-15 03:05:18]

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

[commit-bot: I haz the power, 2014-03-15 03:05:18]

Retried try job too often on win_rel for step(s) browser_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[sky, 2014-03-15 04:47:08]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-15 04:47:15]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/100001

[commit-bot: I haz the power, 2014-03-15 05:49:55]

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

[commit-bot: I haz the power, 2014-03-15 05:49:56]

Retried try job too often on win_rel for step(s) browser_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[sky, 2014-03-15 14:09:57]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-15 14:10:00]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/100001

[commit-bot: I haz the power, 2014-03-15 19:47:59]

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

[commit-bot: I haz the power, 2014-03-15 19:48:00]

Retried try job too often on win_rel for step(s) unit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[sky, 2014-03-15 21:07:42]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2014-03-15 21:07:49]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sky@chromium.org/195993010/100001

[sky, 2014-03-15 22:00:31]

Committed patchset #6 manually as r257342 (presubmit successful).

[jamesr, 2014-03-15 22:48:15]

A revert of this CL has been created in
https://codereview.chromium.org/201283002/ by jamesr@chromium.org.

The reason for reverting is: Broke linux bot:

FAILED: c++ -Wl,-z,now -Wl,-z,relro -pthread -Wl,-z,noexecstack -fPIC -L.
-Wl,-uIsHeapProfilerRunning,-uProfilerStart
-Wl,-u_Z21InitialMallocHook_NewPKvj,-u_Z22InitialMallocHook_MMapPKvS0_jiiix,-u_Z22InitialMallocHook_SbrkPKvi
-Wl,-u_Z21InitialMallocHook_NewPKvm,-u_Z22InitialMallocHook_MMapPKvS0_miiil,-u_Z22InitialMallocHook_SbrkPKvl
-Wl,-u_ZN15HeapLeakChecker12IgnoreObjectEPKv,-u_ZN15HeapLeakChecker14UnIgnoreObjectEPKv
-m32 -Wl,-O1 -Wl,--as-needed -Wl,--gc-sections -Wl,-rpath=\$ORIGIN/lib/
-Wl,-rpath-link=lib/ -o content_perftests -Wl,--start-group
obj/content/browser/renderer_host/input/content_perftests.input_router_impl_perftest.o
obj/content/common/content_perftests.cc_messages_perftest.o
obj/content/test/content_perftests.run_all_perftests.o
obj/content/libcontent_browser.a obj/content/libcontent_common.a
obj/content/libtest_support_content.a obj/base/libtest_support_base.a
obj/cc/libcc.a obj/testing/libgtest.a obj/testing/perf/libperf_test.a
obj/ui/gfx/libgfx.a obj/ui/gfx/libgfx_geometry.a obj/base/libbase.a
obj/base/libbase_static.a obj/base/allocator/liballocator_extension_thunks.a
obj/third_party/modp_b64/libmodp_b64.a
obj/base/third_party/dynamic_annotations/libdynamic_annotations.a
obj/base/libsymbolize.a obj/base/libxdg_mime.a
obj/third_party/libevent/libevent.a obj/components/libtracing.a obj/ipc/libipc.a
obj/net/libnet.a obj/base/libbase_i18n.a obj/third_party/icu/libicui18n.a
obj/third_party/icu/libicuuc.a obj/third_party/icu/libicudata.a
obj/crypto/libcrcrypto.a obj/net/third_party/nss/libcrssl.a obj/sdch/libsdch.a
obj/third_party/zlib/libchrome_zlib.a obj/url/liburl_lib.a
obj/build/linux/libgio.a obj/skia/libskia_library.a obj/skia/libskia_opts.a
obj/skia/libskia_opts_ssse3.a obj/third_party/sfntly/libsfntly.a
obj/skia/libskia_chrome.a obj/skia/libskia_chrome_opts.a
obj/third_party/libjingle/libjingle.a
obj/third_party/libjingle/libjingle_p2p_constants.a
obj/ui/accessibility/libaccessibility.a obj/third_party/libpng/libpng.a
obj/third_party/libjpeg_turbo/libjpeg_turbo.a obj/ui/accessibility/libax_gen.a
obj/tools/json_schema_compiler/libapi_gen_util.a obj/ui/base/libui_base.a
obj/ui/events/libevents_base.a obj/ui/events/libdom4_keycode_converter.a
obj/ui/events/libevents.a obj/ui/shell_dialogs/libshell_dialogs.a
obj/ui/aura/libaura.a obj/gpu/libcommand_buffer_client.a
obj/gpu/libcommand_buffer_common.a obj/gpu/command_buffer/libgles2_utils.a
obj/gpu/libcommand_buffer_service.a obj/gpu/libdisk_cache_proto.a
obj/third_party/protobuf/libprotobuf_lite.a obj/third_party/re2/libre2.a
obj/third_party/smhasher/libcityhash.a obj/ui/gl/libgl_wrapper.a
obj/third_party/angle/src/libtranslator.a
obj/third_party/angle/src/libpreprocessor.a obj/gpu/libgles2_cmd_helper.a
obj/gpu/libgpu_config.a obj/build/linux/libpci.a
obj/third_party/libXNVCtrl/libXNVCtrl.a obj/gpu/libgpu_ipc.a
obj/ui/compositor/libcompositor.a obj/media/libmedia.a
obj/third_party/opus/libopus.a obj/media/libshared_memory_support.a
obj/media/libshared_memory_support_sse.a obj/third_party/ffmpeg/libffmpeg.a
obj/third_party/libvpx/libvpx.a obj/third_party/libvpx/libvpx_asm_offsets_vp8.a
obj/third_party/libvpx/libvpx_intrinsics_mmx.a
obj/third_party/libvpx/libvpx_intrinsics_sse2.a
obj/third_party/libvpx/libvpx_intrinsics_ssse3.a obj/media/libmedia_asm.a
obj/media/libmedia_mmx.a obj/media/libmedia_sse.a obj/media/libmedia_sse2.a
obj/gpu/libgles2_c_lib.a obj/gpu/libgles2_implementation.a
obj/gpu/skia_bindings/libgpu_skia_bindings.a
obj/third_party/WebKit/Source/platform/libblink_platform.a
obj/third_party/WebKit/Source/wtf/libwtf.a
obj/third_party/WebKit/Source/platform/libblink_common.a
obj/third_party/libwebp/libwebp_dec.a obj/third_party/libwebp/libwebp_dsp.a
obj/third_party/libwebp/libwebp_utils.a obj/third_party/libwebp/libwebp_demux.a
obj/third_party/libwebp/libwebp_enc.a obj/third_party/ots/libots.a
obj/third_party/brotli/libbrotli.a obj/third_party/qcms/libqcms.a
obj/v8/tools/gyp/libv8_base.ia32.a obj/v8/tools/gyp/libv8_snapshot.a
obj/third_party/iccjpeg/libiccjpeg.a
obj/third_party/harfbuzz-ng/libharfbuzz-ng.a
obj/third_party/WebKit/Source/web/libblink_web.a
obj/third_party/WebKit/Source/core/libwebcore_dom.a
obj/third_party/WebKit/Source/heap/libblink_heap.a
obj/third_party/WebKit/Source/heap/libblink_heap_asm_stubs.a
obj/third_party/libxml/libxml2.a obj/third_party/libxslt/libxslt.a
obj/third_party/sqlite/libsqlite3.a
obj/third_party/WebKit/Source/core/libwebcore_html.a
obj/third_party/WebKit/Source/core/libwebcore_remaining.a
obj/third_party/WebKit/Source/core/libwebcore_rendering.a
obj/third_party/WebKit/Source/core/libwebcore_svg.a
obj/third_party/WebKit/Source/core/libwebcore_generated.a obj/gin/libgin.a
obj/third_party/WebKit/Source/modules/libmodules.a
obj/webkit/common/gpu/libwebkit_gpu.a obj/webkit/common/libwebkit_common.a
obj/webkit/libwebkit_storage_browser.a obj/sql/libsql.a
obj/third_party/leveldatabase/libleveldatabase.a
obj/third_party/snappy/libsnappy.a obj/webkit/libwebkit_storage_common.a
obj/mojo/libmojo_environment_chromium.a
obj/mojo/libmojo_environment_chromium_impl.a obj/mojo/libmojo_common_lib.a
obj/mojo/libmojo_system_impl.a obj/ppapi/libppapi_shared.a
obj/ui/surface/libsurface.a obj/google_apis/libgoogle_apis.a
obj/third_party/zlib/google/libzip.a obj/third_party/zlib/libminizip.a
obj/ui/events/libgesture_detection.a obj/ui/snapshot/libsnapshot.a
obj/content/browser/speech/proto/libspeech_proto.a obj/net/libhttp_server.a
obj/printing/libprinting.a obj/sandbox/libsandbox_services.a
obj/sandbox/libsuid_sandbox_client.a obj/sandbox/libseccomp_bpf.a
obj/sandbox/libseccomp_bpf_helpers.a libyuv.a obj/jingle/libjingle_glue.a
obj/third_party/webrtc/modules/libdesktop_capture.a
obj/third_party/webrtc/system_wrappers/source/libsystem_wrappers.a
obj/third_party/webrtc/modules/libdesktop_capture_differ_sse2.a
obj/sandbox/libc_urandom_override.a obj/ppapi/libppapi_ipc.a
obj/third_party/flac/libflac.a obj/third_party/speex/libspeex.a
obj/dbus/libdbus.a obj/net/libnet_test_support.a obj/testing/libgmock.a
obj/net/tools/tld_cleanup/libtld_cleanup_util.a
obj/base/allocator/liballocator.a obj/net/libnet_with_v8.a
obj/ui/events/libevents_test_support.a obj/ui/gfx/libgfx_test_support.a
obj/content/libcontent_app_both.a obj/content/libcontent_child.a
obj/webkit/child/libwebkit_child.a obj/ui/native_theme/libnative_theme.a
obj/content/libcontent_gpu.a obj/content/libcontent_ppapi_plugin.a
obj/content/libcontent_renderer.a
obj/webkit/renderer/compositor_bindings/libwebkit_compositor_bindings.a
obj/webkit/renderer/compositor_bindings/libwebkit_compositor_support.a
obj/mojo/libmojo_js_bindings_lib.a obj/third_party/libjingle/libjingle_webrtc.a
obj/third_party/libjingle/libjingle_webrtc_common.a
obj/third_party/libsrtp/libsrtp.a obj/third_party/webrtc/modules/libmedia_file.a
obj/third_party/webrtc/modules/libvideo_capture_module.a
obj/third_party/webrtc/modules/libwebrtc_utility.a
obj/third_party/webrtc/modules/libaudio_coding_module.a
obj/third_party/webrtc/modules/libCNG.a
obj/third_party/webrtc/common_audio/libcommon_audio.a
obj/third_party/webrtc/common_audio/libcommon_audio_sse2.a
obj/third_party/webrtc/modules/libG711.a
obj/third_party/webrtc/modules/libG722.a
obj/third_party/webrtc/modules/libiLBC.a
obj/third_party/webrtc/modules/libiSAC.a
obj/third_party/webrtc/modules/libiSACFix.a
obj/third_party/webrtc/modules/libPCM16B.a
obj/third_party/webrtc/modules/libNetEq.a
obj/third_party/webrtc/modules/libwebrtc_opus.a
obj/third_party/webrtc/modules/libacm2.a
obj/third_party/webrtc/modules/libNetEq4.a
obj/third_party/webrtc/modules/libwebrtc_video_coding.a
obj/third_party/webrtc/modules/libwebrtc_i420.a
obj/third_party/webrtc/common_video/libcommon_video.a
obj/third_party/webrtc/modules/video_coding/utility/libvideo_coding_utility.a
obj/third_party/webrtc/modules/video_coding/codecs/vp8/libwebrtc_vp8.a
obj/third_party/webrtc/modules/libvideo_render_module.a
obj/third_party/usrsctp/libusrsctplib.a
obj/third_party/libjingle/libpeerconnection.a
obj/third_party/webrtc/video_engine/libvideo_engine_core.a
obj/third_party/webrtc/modules/librtp_rtcp.a
obj/third_party/webrtc/modules/libpaced_sender.a
obj/third_party/webrtc/modules/libremote_bitrate_estimator.a
obj/third_party/webrtc/modules/remote_bitrate_estimator/librbe_components.a
obj/third_party/webrtc/modules/libbitrate_controller.a
obj/third_party/webrtc/modules/libvideo_processing.a
obj/third_party/webrtc/modules/libvideo_processing_sse2.a
obj/third_party/webrtc/voice_engine/libvoice_engine.a
obj/third_party/webrtc/modules/libaudio_conference_mixer.a
obj/third_party/webrtc/modules/libaudio_processing.a
obj/third_party/webrtc/modules/libaudioproc_debug_proto.a
obj/third_party/webrtc/modules/libaudio_processing_sse2.a
obj/third_party/webrtc/modules/libaudio_device.a obj/ppapi/libppapi_host.a
obj/ppapi/libppapi_proxy.a obj/content/libcontent_utility.a
obj/content/libcontent_worker.a obj/cc/libcc_test_support.a
obj/gpu/libgpu_unittest_utils.a obj/ui/gl/libgl_unittest_utils.a
obj/ppapi/libppapi_unittest_shared.a obj/ipc/libtest_support_ipc.a
obj/ui/aura/libaura_test_support.a
obj/ui/compositor/libcompositor_test_support.a obj/ui/libui_test_support.a
lib/libmojo_system.so -Wl,--end-group -lrt -ldl -lgmodule-2.0 -lgobject-2.0
-lgthread-2.0 -lglib-2.0 -lX11 -lXi -lXrandr -lnss3 -lnssutil3 -lsmime3 -lplds4
-lplc4 -lnspr4 -lgconf-2 -lgio-2.0 -lresolv -lfontconfig -lfreetype
-lpangocairo-1.0 -lcairo -lpangoft2-1.0 -lpango-1.0 -lexpat -lXcursor -lXext
-lXfixes -lXrender -lXcomposite -lasound -lXdamage -lXtst -lcups -lgssapi_krb5
-lkrb5 -lk5crypto -lcom_err -lgnutls -lgcrypt -lz -lpthread -lcrypt -lm
-L/lib/i386-linux-gnu -lcap -ludev -ldbus-1
obj/content/common/mojo/content_common.mojo_channel_init.o: In function
`content::MojoChannelInit::Init(int, scoped_refptr<base::TaskRunner>)':
mojo_channel_init.cc:(.text._ZN7content15MojoChannelInit4InitEi13scoped_refptrIN4base10TaskRunnerEE+0x230):
undefined reference to `MojoClose'
obj/content/common/mojo/content_common.mojo_channel_init.o: In function
`content::MojoChannelInit::~MojoChannelInit()':
mojo_channel_init.cc:(.text._ZN7content15MojoChannelInitD2Ev+0x124): undefined
reference to `MojoClose'
mojo_channel_init.cc:(.text._ZN7content15MojoChannelInitD2Ev+0x134): undefined
reference to `MojoClose'
obj/mojo/system/embedder/mojo_system_impl.embedder.o: In function
`mojo::embedder::Init()':
embedder.cc:(.text._ZN4mojo8embedder4InitEv+0x30): undefined reference to
`mojo::Core::Init(mojo::Core*)'
obj/mojo/system/embedder/mojo_system_impl.embedder.o: In function
`mojo::embedder::CreateChannel(mojo::embedder::ScopedPlatformHandle,
scoped_refptr<base::TaskRunner>, base::Callback<void
(mojo::embedder::ChannelInfo*)>)':
embedder.cc:(.text._ZN4mojo8embedder13CreateChannelENS0_20ScopedPlatformHandleE13scoped_refptrIN4base10TaskRunnerEENS3_8CallbackIFvPNS0_11ChannelInfoEEEE+0x29):
undefined reference to `mojo::Core::Get()'
obj/mojo/system/mojo_system_impl.core_impl.o: In function
`mojo::system::CoreImpl::~CoreImpl()':
core_impl.cc:(.text._ZN4mojo6system8CoreImplD2Ev+0x4b): undefined reference to
`mojo::Core::~Core()'
collect2: ld returned 1 exit status
ninja: build stopped: subcommand failed.

http://build.chromium.org/p/chromium/builders/Linux/builds/48261/steps/compil....

[sky, 2014-03-17 21:52:40]

Ok, I figured out a workaround for the linker bug. Diff between patchsets 7 and
11 to see changes. Will reland shortly.

[sky, 2014-03-17 22:55:33]

Committed patchset #11 manually as r257505 (presubmit successful).

[viettrungluu, 2014-03-18 01:09:02]

Sorry, had to revert (at the request of kareng), r257565.

It broke the Linux Official builds (32- and 64-bit) -- some packaging stuff. I
believe that libmojo_system.so needs to be added to some list somewhere.

[Vadim Sh., 2014-03-18 01:10:38]

On 2014/03/17 22:55:33, sky wrote:
> Committed patchset #11 manually as r257505 (presubmit successful).

FYI this CL broke isolated tests on Linux:
http://build.chromium.org/p/chromium.swarm/builders/Linux%20Swarm%20Tests/bui...
It introduces libmojo_system.so runtime dependency that is not isolated with the
test.

chromium.linux used to catch this kind of regressions but it doesn't anymore,
here's a bug to fix this particular regression: https://crbug.com/353403

Note that currently CI waterfall and linux try server doesn't rely on isolates,
so it's ok NOT to revert this change. I'll try to prepare a separate CL to fix
*.isolate's to include libmojo_system.so.

[Vadim Sh., 2014-03-18 01:24:32]

Oh... The CL got reverted.

Can you please verify browser_tests, interactive_ui_tests and unit_tests are
able to start in isolated mode before committing this patch again?
Please refer to this page for how to run isolated tests locally:
http://www.chromium.org/developers/testing/isolated-testing/for-swes

You may need to modify corresponding *.isolate files to make it work.

[sky, 2014-03-18 17:06:25]

Until build bots are straightened out I'm going to add a use_mojo gyp_define.
I'll make it conditionally true as the bots get straightened out. Diff between
12 and 13 to see the changes.

[viettrungluu, 2014-03-18 17:11:07]

Still lgtm

[sky, 2014-03-18 18:03:12]

Committed patchset #13 manually as r257691 (presubmit successful).