This is a compile failure in the test because feature::kSitePerProcess moved from content/public to chrome/ in r545649.  I'm trying to put together a fix on the M66 branch for it.

CL started: https://chromium-review.googlesource.com/c/chromium/src/+/1012572
Compiling locally to verify.

Am I looking at the wrong thing? The problem I see is about mini_installer_test.

Comment 5: Good question.  Abdul showed me a different error via chat, but I'm not seeing it in the URL above.  I'm trying to address this one:

"../../chrome/browser/policy/site_isolation_policy_browsertest.cc:206:57: error: no member named 'kSitePerProcess' in namespace 'features'
   scoped_feature_list_.InitAndEnableFeature(features::kSitePerProcess);"

Abdul, which URL did that error get reported on?  (It does make sense that it would fail to compile in that way.) 

I'm pretty sure this is related to rolling depot_tools recipes on the branch, to fix other issues. Unfortunately, that roll also brought in some other recipe changes, specifically:

https://chromium-review.googlesource.com/c/chromium/tools/build/+/986882/5/scripts/slave/recipe_modules/chromium_tests/steps.py

which added the mini_installer_test target. But that target isn't in the BUILD.gn files on the 3359 branch. Since we need the depot_tool roll, I think we need to also cherry-pick the relevant BUILD.gn changes to add that target. +mmeade, how much trouble would that be?

Oh, I'm fixing a compile failure on go/betabuilders:
https://uberchromegw.corp.google.com/i/official.desktop.continuous/builders/win64%20beta/builds/3333

Abdul, maybe that's a different bug report?

That beta 3333 failure is bogus. It was killed because we restarted the master.

After chatting with Abdul, it turns out there are no builders failing with this error, but the compile error was spotted by Sahel in a local build.  The fix is needed on the M66 branch to compile, so I'll land it.

Sounds like most of this bug is about a different issue, though, so I'll reassign to mmoss@ for that one.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/06de8e2145bd47ff76ae5b047348101a58c62d96

commit 06de8e2145bd47ff76ae5b047348101a58c62d96
Author: Charlie Reis <creis@chromium.org>
Date: Fri Apr 13 21:31:36 2018

Fix M66 compile error in site_isolation_policy_browsertest.cc.

The merge in https://chromium-review.googlesource.com/c/chromium/src/+/1012137
caused a compile error on the M66 branch due to kSitePerProcess moving
from content/public to chrome/ in r545649 (67.0.3381.0).

BUG= 832898 ,  827821 
TBR=lukasza, pastarmovj

Change-Id: I1e339cac100d9fcec24015e0ba8364da9e1b7140
Reviewed-on: https://chromium-review.googlesource.com/1012572
Reviewed-by: Charlie Reis <creis@chromium.org>
Cr-Commit-Position: refs/branch-heads/3359@{#707}
Cr-Branched-From: 66afc5e5d10127546cc4b98b9117aff588b5e66b-refs/heads/master@{#540276}
[modify] https://crrev.com/06de8e2145bd47ff76ae5b047348101a58c62d96/chrome/browser/policy/site_isolation_policy_browsertest.cc

Now Linux is failing as well, I am assuming this should fail on Mac as well.

Nope, mini_installer_tests is Windows-only, so wouldn't affect other platforms. Please create separate bugs for any other failures.

Since the failure was due to "error: no member named 'kSitePerProcess' in namespace 'features'" , I have assumed it's the same will create the new bug. Thank you Michael.

For reference, my CL from comment 11 should fix  issue 832965  (which pbommana filed in comment 15).

I think mmoss@ is still looking at the mini_installer_tests Windows issue.

I'm working on getting the BUILD files updated.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c1f8c1ca34053f6cef1f85afcc03e6ab36c3eda5

commit c1f8c1ca34053f6cef1f85afcc03e6ab36c3eda5
Author: Michael Moss <mmoss@google.com>
Date: Sat Apr 14 00:32:04 2018

Add dummy mini_installer_tests target to 3359 branch.

This fixes the issue where recipes replaced:
  def compile_targets(self, _):
      return ['mini_installer', 'next_version_mini_installer']

with:
  def compile_targets(self, _):
      return ['mini_installer_tests']

but there was no corresponding 'mini_installer_tests' target on the 3359
branch. This adds a dummy target which just recreates the old behavior of
building the other two targets.

BUG= 832898 
R=abdulsyed@google.com, mmeade@google.com

Change-Id: Iab9ded7eaec4821d1f44014b2e472a4c420768fc
Reviewed-on: https://chromium-review.googlesource.com/1013281
Reviewed-by: Michael Moss <mmoss@chromium.org>
Cr-Commit-Position: refs/branch-heads/3359@{#712}
Cr-Branched-From: 66afc5e5d10127546cc4b98b9117aff588b5e66b-refs/heads/master@{#540276}
[modify] https://crrev.com/c1f8c1ca34053f6cef1f85afcc03e6ab36c3eda5/chrome/installer/mini_installer/BUILD.gn

It's still building, but https://uberchromegw.corp.google.com/i/official.desktop.continuous/builders/win64%20stable/builds/828 is past the "missing target" compile error from before, so hopefully this is good now.