GitLab

[Context]
This is a part of an effort to achieve MBI. We want to avoid per-thread task runner as much as possible.

[Previous code]
RenderFrameImpl::GetInterfaceProvider was using per-thread default task runner because task_runner was not specified.

[What this CL does]
Use per-frame task runner for RenderFrameImpl::GetInterfaceProvider instead of per-thread task runner.

[Rlated stack trace]
base::SequencedTaskRunnerHandle::Get()
mojo::internal::GetTaskRunnerToUseFromUserProvidedTaskRunner()
mojo::internal::BindingStateBase::BindInternal()
mojo::internal::BindingState<>::Bind()
mojo::Receiver<>::Bind()
mojo::Receiver<>::Receiver()
mojo::ReceiverSetBase<>::Entry::Entry()
std::__Cr::make_unique<>()
mojo::ReceiverSetBase<>::AddImpl()
mojo::ReceiverSetBase<>::Add()
content::RenderFrameImpl::GetInterfaceProvider()
content::mojom::FrameStubDispatch::Accept()

Bug: 1105403
Change-Id: I9a9c8110ed0746b80cab30604f7ac6f6707211e4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2465607Reviewed-by: Kentaro Hara <haraken@chromium.org>
Reviewed-by: Kouhei Ueno <kouhei@chromium.org>
Commit-Queue: Minoru Chikamune <chikamune@chromium.org>
Cr-Commit-Position: refs/heads/master@{#816393}