1425455 - Black screen with "Checker Plus for Gmail" & "Amazon Assistant for Firefox" extension after landing patches from bug #1035091

Status: VERIFIED WORKSFORME

(Core :: CSS Parsing and Computation, defect, P3)

Description

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

STR:
1. Install "Checker Plus for Gmail™" extension ( https://addons.mozilla.org/en-US/firefox/addon/checker-plus-gmail/ )
2. Open some GMail message
3. Scroll it
and see black screen

Other reports are in:
bug #1422274 comment #2
bug #1422274 comment #12
bug #1422274 comment #14


Workaround:
Set "layout.css.moz-document.content.enabled" preference to "true" in about:config


Regression caused by:
bug #1035091

Regression range pushlog:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=4ad2150e73c0b19f1f6264a4b9ad6927986d43ce&tochange=37e0bd919af057d44c5c1410458c0f00a3653c11

Comment 1

[Emilio Cobos Álvarez (:emilio)]

Hi Jason, I see you're the developer of this extension: How are you using @-moz-document? Is there any kind of add-on API that would suit your add-on better?

Comment 2

[Jason Savard]

As a workaround, I'm using this css directive to explicitly set the background to white in my extension. Note, I'm targeting Firefox only because this issue does not exist on Chrome.

@-moz-document url-prefix() {
	#inbox, #openEmailScrollArea {background-color:white}
	body.background-skin #inbox, #openEmailScrollArea {background-color:rgba(255,255,255,0.1)}
}

I also mentioned it in this ticket: https://bugzilla.mozilla.org/show_bug.cgi?id=1417286#c9

Comment 3

[fireattack [:fireattack]]

per discussion, this seems to be the exactly same bug as bug 1417286, Feel free to combine them.

Also it affects another addon "Amazon assistant" (check bug 1417286's comment), if we want to stick on this one, please CC their dev: [e-mail]

Comment 5

[Andrew Overholt [:overholt]]

Anything we should do here, other than waiting on extension changes?

Comment 6

[Emilio Cobos Álvarez (:emilio)]

It's not 100% clear to me whether this is due to removing @-moz-document breaking a workaround for the black screen bug (potentially a graphics bug?), in which case we should figure out that one, or if it's a real CSS engine bug or something like that, in which case I could look into it.

From the earlier comments I suspect it's the first, but confirmation would be nice.

Comment 7

[Andrew Overholt [:overholt]]

(In reply to Emilio Cobos Álvarez [:emilio] from comment #6)
> It's not 100% clear to me whether this is due to removing @-moz-document
> breaking a workaround for the black screen bug (potentially a graphics
> bug?), in which case we should figure out that one, or if it's a real CSS
> engine bug or something like that, in which case I could look into it.

Who can do that?

Comment 8

[Andrew Overholt [:overholt]]

(In reply to Emilio Cobos Álvarez [:emilio] from comment #6)
> It's not 100% clear to me whether this is due to removing @-moz-document
> breaking a workaround for the black screen bug

I had somehow missed comment 0's thoroughness:

(In reply to Virtual_ManPL [:Virtual] - (please needinfo? me - so I will see your comment/reply/question/etc.) from comment #0)
> Set "layout.css.moz-document.content.enabled" preference to "true" in
> about:config

Does that answer your question, Emilio?

Comment 9

[Emilio Cobos Álvarez (:emilio)]

(In reply to Andrew Overholt [:overholt] from comment #8)
> > Set "layout.css.moz-document.content.enabled" preference to "true" in
> > about:config
> 
> Does that answer your question, Emilio?

No, I know that enabling moz-document "fixes" the issue, but my reading of the other comments is that that @-moz-document rule is just working around the real issue, which looks like a gfx issue. Please correct me if I'm wrong.

Comment 10

[Andrew Overholt [:overholt]]

Ah, I see now that the root cause graphics bug is bug 1414033.

Comment 11

[Markus Stange [:mstange]]

(In reply to Andrew Overholt [:overholt] from comment #10)
> Ah, I see now that the root cause graphics bug is bug 1414033.

I don't think that's true. What made you think so?

Comment 12

[Andrew Overholt [:overholt]]

(In reply to Markus Stange [:mstange] from comment #11)
> (In reply to Andrew Overholt [:overholt] from comment #10)
> > Ah, I see now that the root cause graphics bug is bug 1414033.
> 
> I don't think that's true. What made you think so?

Hmm, I can't remember. But maybe because bug 1400259 comment 46 indicates the symptom here (the black background thing) is caused by "A container layer is holding on to its intermediate surface even though its child layers have changed." And the commit message there says "The bug has something to do with ContainerLayer nesting changes being mishandled: a new ContainerLayer for the transform is being inserted around the container layer for the opacity, which has an intermediate surface. This change makes the outer ContainerLayer permanent so that the dynamic insertion case is not hit." which I perhaps mis-intepreted as being bug 1414033.

At any rate, is there anything we can do to avoid people working around the black background symptom with an @-moz-document?

Comment 13

[Jet Villegas (inactive)]

[ Triage 2017/02/20: P3 ]

Comment 14

[Julien Cristau [:jcristau]]

fix-optional for 59 per P3 priority

Comment 15

[Liz Henry (:lizzard) (relman/hg->git project)]

This got triaged by the team back in January. Are there workarounds in these extensions at this point or is the problem still active?

Comment 16

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

Redirecting this question to extension developer.

Comment 17

[Liz Henry (:lizzard) (relman/hg->git project)]

Too late to fix in 63. We could still take a patch for 65 and potentially for 64.

Comment 18

[Neha Kochar [:neha]]

Bulk change for all regression bugs with status-firefox67 as 'fix-optional' to be marked 'affected' for status-firefox68.

Comment 19

[Patricia Lawless]

Bulk change to wontfix for 68 (P3+ carryover with needinfo).

Comment 20

[Liz Henry (:lizzard) (relman/hg->git project)]

Happy to take a patch for 70 but since this is triaged and set to P3 priority I'm setting it as fix-optional.
That will remove the bug from weekly regression triage.

Comment 21

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

I'm marking this bug as WORKSFORME, as I can't reproduce this issue anymore.

Comment 22

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

• clearing needinfo