Description
3fcddd6#diff-414c8f59bff0d1b63680b64763d8c529R370 started invoking the isBorderBox test 100% of the time instead of only when necessary. In my case this causes a series of .css({width}) calls to cause a "forced reflow" (that's the chrome terminology in the performance tool).
Updating the width/height set to avoid computing isBorderBox unless necessary fixes it for me locally, but makes the code a bit uglier:
6582c6582
< isBorderBox = jQuery.css( elem, "boxSizing", false, styles ) === "border-box",
---
> isBorderBox,
6587c6587
< isBorderBox,
---
> isBorderBox = jQuery.css( elem, "boxSizing", false, styles ) === "border-box",
6593c6593
< if ( isBorderBox && support.scrollboxSize() === styles.position ) {
---
> if ( support.scrollboxSize() === styles.position && (undefined === isBorderBox ? (isBorderBox = jQuery.css( elem, "boxSizing", false, styles ) === "border-box") : isBorderBox) ) {
However so far I have been unable to reproduce this in jsbin, so there must be another condition that I'm not understanding such as a specific position type on the element or parent element?
Here's an example where a series of ng-style="{width: X}" in AngularJS causes this (and jQuery 3.1 does not cause it):

Link to test case
TODO
Description
3fcddd6#diff-414c8f59bff0d1b63680b64763d8c529R370 started invoking the
isBorderBoxtest 100% of the time instead of only when necessary. In my case this causes a series of.css({width})calls to cause a "forced reflow" (that's the chrome terminology in the performance tool).Updating the width/height
setto avoid computingisBorderBoxunless necessary fixes it for me locally, but makes the code a bit uglier:However so far I have been unable to reproduce this in jsbin, so there must be another condition that I'm not understanding such as a specific position type on the element or parent element?
Here's an example where a series of

ng-style="{width: X}"in AngularJS causes this (and jQuery 3.1 does not cause it):Link to test case
TODO