Narrow Bisect info : 
https://chromium.googlesource.com/chromium/src/+log/12f4051195bd555d7449b331db556bbff0aae1cd..5e7860a0564bb5d86d6b9d98022cdadd3f190d54?pretty=fuller&n=100

Suspecting: r472440 from Narrow bisect

@krb : Could you please help to reassign if your change is not the cause for this change.

This particular change was an intermittent change. The final version landed a week or 2 later. With it, I'm seeing the cursor homed - never in the middle of the text. This still doesn't look right, though, so I'll take a closer look.

krb@, is the current state a P-1 targetting M-60?  If so, the merge window for M-60 is rapidly closing...

I'm (hopefully) about to check this into head, but I'm concerned about merging it. This bug is a regression, but it's not a P1. (Tabbing into the Omnibox may put the cursor in the wrong place.)

My concern is that, if I don't also merge dependent changes, or don't test it well enough, it may just generate another bug, equal or worse.

I will take a look, merge the change privately, and test it, but I'm curious what your opinion is (about the priority vs risk.)

> I'm curious what your opinion is (about the priority vs risk.)
This is best a question for the person in charge of releases.  From reading other bugs, I think they're drawing a strong line this late in the process, e.g., from yesterday:
"We're cracking down on merges for this milestone and this bug doesn't have enough impact/severity to justify merging at this stage of the release.  If you feel strongly this should be merged, you can add the Merge-Request-60 label back and we'll re-review."

It sounds like you don't feel strongly and are concerned about the risks, so no?

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/3dc51db95f327e4f86962c94d6d5455ee75cf433

commit 3dc51db95f327e4f86962c94d6d5455ee75cf433
Author: Kevin Bailey <krb@chromium.org>
Date: Thu Jun 29 14:43:43 2017

[omnibox] Select all when we revert all

If you look at all the times we call RevertAll(), they either also call
SelectAll(true), or wouldn't mind if we did. To simplify both
RevertAll() and the calling code, we now make this call unconditionally.
It happens to fix an issue also, where a stale cursor is produced when
tabbing into the Omnibox.

Bug:  735802 
Change-Id: Ib70f3ffd88555ef6aa600e17f111f22442e94eb0
Reviewed-on: https://chromium-review.googlesource.com/548985
Commit-Queue: Kevin Bailey <krb@chromium.org>
Reviewed-by: Trent Apted <tapted@chromium.org>
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/master@{#483362}
[modify] https://crrev.com/3dc51db95f327e4f86962c94d6d5455ee75cf433/chrome/browser/autocomplete/autocomplete_browsertest.cc
[modify] https://crrev.com/3dc51db95f327e4f86962c94d6d5455ee75cf433/chrome/browser/ui/cocoa/omnibox/omnibox_view_mac.mm
[modify] https://crrev.com/3dc51db95f327e4f86962c94d6d5455ee75cf433/chrome/browser/ui/omnibox/omnibox_view_browsertest.cc
[modify] https://crrev.com/3dc51db95f327e4f86962c94d6d5455ee75cf433/chrome/browser/ui/views/omnibox/omnibox_view_views.cc
[modify] https://crrev.com/3dc51db95f327e4f86962c94d6d5455ee75cf433/chrome/browser/ui/views/omnibox/omnibox_view_views_browsertest.cc
[modify] https://crrev.com/3dc51db95f327e4f86962c94d6d5455ee75cf433/components/omnibox/browser/omnibox_edit_model.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5604cb88fc88d2c3b7886702e8253b21ceea600e

commit 5604cb88fc88d2c3b7886702e8253b21ceea600e
Author: Kevin Bailey <krb@chromium.org>
Date: Fri Jul 07 18:28:44 2017

Revert "[omnibox] Select all when we revert all"

This reverts commit 3dc51db95f327e4f86962c94d6d5455ee75cf433.

Reason for revert: For some reason, this change causes a 25% increase in GL rendering for Windows. See  crbug.com/739303 

Original change's description:
> [omnibox] Select all when we revert all
> 
> If you look at all the times we call RevertAll(), they either also call
> SelectAll(true), or wouldn't mind if we did. To simplify both
> RevertAll() and the calling code, we now make this call unconditionally.
> It happens to fix an issue also, where a stale cursor is produced when
> tabbing into the Omnibox.
> 
> Bug:  735802 
> Change-Id: Ib70f3ffd88555ef6aa600e17f111f22442e94eb0
> Reviewed-on: https://chromium-review.googlesource.com/548985
> Commit-Queue: Kevin Bailey <krb@chromium.org>
> Reviewed-by: Trent Apted <tapted@chromium.org>
> Reviewed-by: Peter Kasting <pkasting@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#483362}

TBR=pkasting@chromium.org,tapted@chromium.org,krb@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug:  735802 
Change-Id: I4d075e5c5a20707f74bd2d330d466f67a3241ce8
Reviewed-on: https://chromium-review.googlesource.com/563482
Reviewed-by: Kevin Bailey <krb@chromium.org>
Commit-Queue: Kevin Bailey <krb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#484992}
[modify] https://crrev.com/5604cb88fc88d2c3b7886702e8253b21ceea600e/chrome/browser/autocomplete/autocomplete_browsertest.cc
[modify] https://crrev.com/5604cb88fc88d2c3b7886702e8253b21ceea600e/chrome/browser/ui/cocoa/omnibox/omnibox_view_mac.mm
[modify] https://crrev.com/5604cb88fc88d2c3b7886702e8253b21ceea600e/chrome/browser/ui/omnibox/omnibox_view_browsertest.cc
[modify] https://crrev.com/5604cb88fc88d2c3b7886702e8253b21ceea600e/chrome/browser/ui/views/omnibox/omnibox_view_views.cc
[modify] https://crrev.com/5604cb88fc88d2c3b7886702e8253b21ceea600e/chrome/browser/ui/views/omnibox/omnibox_view_views_browsertest.cc
[modify] https://crrev.com/5604cb88fc88d2c3b7886702e8253b21ceea600e/components/omnibox/browser/omnibox_edit_model.cc

We've analyzed the performance regression seen in  Issue 739303  and concluded that it's not real, and caused by a flaw in the frame_times and mean_frame_times metrics on Windows. krb@, let's re-land your CL and close  Issue 739303  as WontFix.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/3cda6c9c7ee5274ea800bce409691a15a949fc2c

commit 3cda6c9c7ee5274ea800bce409691a15a949fc2c
Author: Kevin Bailey <krb@chromium.org>
Date: Thu Aug 03 23:20:18 2017

[omnibox] Select-all after revert-all, in more cases

If you look at all the times we call RevertAll(), they either also call
SelectAll(true), or wouldn't mind if we did. We wish to standardize on
this policy. It happens to fix an issue also, where a stale cursor is
produced when tabbing into the Omnibox.

Bug:  735802 ,  750716 
Change-Id: I5aace2b656f4241f2760acadc2a156dbd5fc31f4
Reviewed-on: https://chromium-review.googlesource.com/572206
Reviewed-by: Avi Drissman <avi@chromium.org>
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Commit-Queue: Kevin Bailey <krb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#491874}
[modify] https://crrev.com/3cda6c9c7ee5274ea800bce409691a15a949fc2c/chrome/browser/ui/cocoa/omnibox/omnibox_view_mac.mm
[modify] https://crrev.com/3cda6c9c7ee5274ea800bce409691a15a949fc2c/chrome/browser/ui/omnibox/omnibox_view_browsertest.cc
[modify] https://crrev.com/3cda6c9c7ee5274ea800bce409691a15a949fc2c/chrome/browser/ui/views/omnibox/omnibox_view_views.cc
[modify] https://crrev.com/3cda6c9c7ee5274ea800bce409691a15a949fc2c/components/omnibox/browser/omnibox_edit_model.cc