On thinking about this some more, populating GlobalRequestID is the best way to do this.

This entails propagating the GlobalRequestID all the way to WillProcessResponse(NavigationURLLoaderImplCore::NotifyResponseStarted, NavigationURLLoaderImpl::NotifyResponseStarted, etc).

The duped bug is for a feature request in core PlzNavigate code. But this feature is still not implemented, judging by the TODOs and early exits: https://cs.chromium.org/chromium/src/chrome/browser/data_use_measurement/chrome_data_use_ascriber.cc?type=cs&q=%22//+TODO(ryansturm):+Handle+PlzNavigate+(http://crbug/664233).%22&l=57

I believe that https://codereview.chromium.org/2537593002 plumbed the actual GlobalRequestID for plzNavigate, so I will investigate and fix the TODOs in ChromeDataUseAscriber code.

What's the status of this?

Raj is looking at this.

Adding Proj-PlzNavigate, which we are aiming to launch in M60.

Ascribing URLRequest to RenderFrameHost should happen based on GlobalRequestID. ResourceRequestInfo::GetRenderFrameForRequest() should not be used in GetOrCreateDataUseRecorderEntry().

To be clear, mainframe and subframe navigations will only have GlobalRequestID populated in the request when it starts (RenderFrameHost ID will not be populated). This is true for PlzNavigate and non PlzNavigate cases.

For subresource requests RenderFrameHost ID should be populated, and ascription should use that.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/975ba1a456087d8781f61754b747c66dab30c97f

commit 975ba1a456087d8781f61754b747c66dab30c97f
Author: tbansal <tbansal@chromium.org>
Date: Wed May 31 20:40:30 2017

PlzNavigate support in data use ascriber

With PlzNavigate, RenderFrameHost will not be populated in mainframe
requests. So global request ID alone should be used for mapping those
requests to main renderframe, when navigation commits later. This CL is
based off of Chromium CL 2874353002.

BUG= 664233 
TBR=rajendrant@chromium.org

Review-Url: https://codereview.chromium.org/2913063002
Cr-Commit-Position: refs/heads/master@{#476006}

[modify] https://crrev.com/975ba1a456087d8781f61754b747c66dab30c97f/chrome/browser/data_use_measurement/chrome_data_use_ascriber.cc
[modify] https://crrev.com/975ba1a456087d8781f61754b747c66dab30c97f/chrome/browser/data_use_measurement/chrome_data_use_ascriber_unittest.cc
[modify] https://crrev.com/975ba1a456087d8781f61754b747c66dab30c97f/components/data_reduction_proxy/core/browser/data_reduction_proxy_compression_stats.cc
[modify] https://crrev.com/975ba1a456087d8781f61754b747c66dab30c97f/components/data_reduction_proxy/core/browser/data_reduction_proxy_data_use_observer.cc

This is fixed. https://codereview.chromium.org/2914923002/ should also have pointed to this bug.