Tested the issue on windows 10 & 7 using chrome M60 #60.0.3112.78 and canary M62 #62.0.3172.0 and observed as attached in screencast.

Observed that in firefox there is no white gap shown in different zooming and chrome has white gap in different zooming.

@info-- Could you please check attached screencast and confirm us the expected behavior  is same as the firefox to triage the issue better.

Thanks!

Deleted: 750490.mp4 967 KB

	750490.mp4 967 KB View Download

@hdodda: Yes I can confirm this, in Mozilla Firefox is no white gap, I also tested it :)

Thank you for providing more feedback. Adding requester "hdodda@chromium.org" to the cc list and removing "Needs-Feedback" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

I have created a fiddle with the test case here https://jsfiddle.net/2x4txvb5/

Able to repro this on chrome 59.0.3071.115 on both Windows 10 and Mac 10.12.6, but only at zoom levels of 125% and 175%. Unable to repro this on canary 62.0.3173.0 

Reassigning to Blink>Paint as this seems like it could be a painting issue. The CSS seems to be correct here.

Able to reproduce the issue on mac os 10.12.6 , ubuntu 14.04 and windows 7 using M60 #60.0.3112.78 and canary M62 #62.0.3173.2 .

This is a regression issue broken in M58.

Using the per-revision bisect providing the bisect results,
Good build: 58.0.3011.0(Revision: 449903).
Bad build: 58.0.3013.0 (Revision: 450530).

You are probably looking for a change made after 449942 (known good), but no later than 449943 (first known bad).

CHANGELOG URL:

The script might not always return single CL as suspectas some perf builds might get missing due to failure.

 https://chromium.googlesource.com/chromium/src/+log/d6df828820be4b0f22b9bed959fb2bd93f84b331..934becac5daa91ea979fb66e4ae21761ca11ebc9

From the CL above, assigning the issue to the concern owner 

@karlo- Could you please check whether this is caused with respect to your change, if not please help us in assigning it to the right owner.

Review-Url: https://codereview.chromium.org/2640143005

Thanks!

Not a P1. Yes, it's annoying, but it does not prevent WebSite functioning or lead to missing content.

It's almost impossible to fix this too.

Why isn't it possible to fix it?

It's hard to fix because we would need to modify the sub-pixel snapping behavior such that this case worked while not breaking others. In particular, background images might develop gaps if we rounded out to fill the gap in this case. The context of one border abutting another is very hard to detect, so we can't special case situations like this.

Arguably the outer border should snap the same way as the inner one if it is in the same place, but I'm guessing they are not actually in exactly the same place (because one does not draw on top of the other).