Which CL regressed this again?  It wasn't in M57, right?

The CL causing this regression in M58 is crrev.com/2669663002.

We merged the CL to M57 beta then switched back. So M57 is unaffected.

Great, glad to hear M57 isn't affected - thanks!  Marking this Pri-1 for M58 to indicate we really don't want to ship this regression (it's subtle but pretty bad for drawing apps).

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac

commit be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac
Author: lanwei <lanwei@chromium.org>
Date: Tue Feb 21 18:08:10 2017

Touchmoves should not be suppressed if the touchstart is consumed

In a touch sequence starting with a touchstart, if the touchstart is consumed
(prevent-defaulted), we will not suppress the following touchmoves even if
they are in the slop regions.

BUG= 692702 

Review-Url: https://codereview.chromium.org/2701553002
Cr-Commit-Position: refs/heads/master@{#451796}

[modify] https://crrev.com/be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac/third_party/WebKit/LayoutTests/TestExpectations
[modify] https://crrev.com/be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac/third_party/WebKit/LayoutTests/fast/events/pointerevents/multi-pointer-event-in-slop-region.html
[add] https://crrev.com/be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac/third_party/WebKit/LayoutTests/fast/events/pointerevents/pointer-event-consumed-touchstart-in-slop-region.html
[modify] https://crrev.com/be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac/third_party/WebKit/LayoutTests/fast/events/pointerevents/pointer-event-in-slop-region.html
[modify] https://crrev.com/be77e3f3b19b056f8ef9c5464ebadcc2d5bf25ac/third_party/WebKit/Source/core/input/TouchEventManager.cpp

How does it affect mouse events PointerEvents?
Do we need to call e.preventDefault() on touchstart event to prevent slop suppression even when using a digitizer pen or a mouse?

In Chrome 56 slop suppression is affecting pen and mouse in paint apps.

There is no slop suppression for non-touch events. Perhaps you should open a new bug and provide details there.

Any PointerEvent based app that used to be affected by slop suppression will be fixed on M58, see  Issue 593061 . Please comment on that bug if your problems are not fixed in M58.