186754 - "overflow" overrides z-index.

Status: RESOLVED FIXED

(Core :: Web Painting, defect, P1)

Description

[Laurent Martelli]

User-Agent:       Mozilla/5.0 Galeon/1.2.7 (X11; Linux i686; U;) Gecko/20021204 Debian/1.2.7-4
Build Identifier: Mozilla/5.0 Galeon/1.2.7 (X11; Linux i686; U;) Gecko/20021204 Debian/1.2.7-4

If an element A with a fixed position and z-index=1 is inside another element B
with z-index=0, A does not receive mouse events such as "onclick".

Reproducible: Always

Steps to Reproduce:
1.
2.
3.

Comment 1

[Ruslan Ismailov]

Of course it is, it is effect of "overflow: auto;" properity. If you change it
to hidden or remove it, all be as you wish. But auto there is mean visible, so
bigger area with z-index: 0 overflow lesser area with z-index: 1.

Comment 2

[Laurent Martelli]

Yet, I find it strange that an element with a bugger z-index is "behind" one
with a lower z-index. CSS2 specs says "Boxes with greater stack levels are
always formatted in front of boxes with lower stack levels". So there's a
contradiction somewhere.

Comment 3

[Ruslan Ismailov]

I think this bug should be named: "z-index vs overflow: visible priority"

Comment 4

[Olivier Cahagne]

please reopen if you disagree.

*** This bug has been marked as a duplicate of 78087 ***

Comment 5

[Ruslan Ismailov]

Attachment: Overflow:Auto vs z-index

Sorry, I little misguided. Where is ther root of all overflow:auto bugs? I just
like to show, that overflow:auto is not working properly. Look at this testase
to see difference between Auto, Hidden and Visible. All 3 states should be
visible, because of z-index:1

Comment 6

[Vladimir Ermakov]

Sounds like css to me. I dont know if this is a dupe or not, ian, can you take a
look?

Comment 7

[Ruslan Ismailov]

It is not CSS, I sure. It is bug inside overflow:auto. Please, reopen this bug
-- it have nothing with negative z-index bug.

Comment 8

[Vladimir Ermakov]

reopening per coments.

Comment 9

[Vladimir Ermakov]

 Moving to layout, which really does sound more appropriate. I can confirm the
bug  on Win2k build 2002-12-20-08-trunk, OS-All. This is definately a bug
because it used to work on 7.01 build, at least on mac.

Comment 10

[Ruslan Ismailov]

Attachment: Overflow: Auto and z-index case

This testcase show, that Overflow: Auto boxes are on top of all other boxes,
even if z-index made another order.

Comment 11

[Christopher Hoess (gone)]

*** Bug 188502 has been marked as a duplicate of this bug. ***

Comment 12

[Boris Zbarsky [:bzbarsky]]

views.

Comment 13

[Boris Zbarsky [:bzbarsky]]

resummarizing to make the problem clear

Comment 14

[Ruslan Ismailov]

Attachment: Overflow:Auto and Scroll overrided z-index

Small change of previous testcase, included overflow:scroll box.

Comment 15

[Ruslan Ismailov]

I think I found -- it is backfire from bug 39621. 

Comment 16

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Attachment: fix

The problem is that scrolling frames don't get their content parent hooked up
correctly when their view is created. This patch makes nsCSSFrameConstructor
pass the content-parent frame down into BuildScrollFrame so that the scrolling
view gets its zparent set right.

Comment 17

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 112571 [details] [diff] [review]
fix

we need to fix this file to not suck....

Comment 18

[Hixie (not reading bugmail)]

Does this patch also fix
   http://www.hixie.ch/tests/adhoc/css/box/absolute/overflow/010.html
...?

Comment 19

[Robert O'Callahan (:roc) (email my personal email if necessary)]

No, that testcase is not fixed.

That is a weird one...

Comment 20

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Attachment: Updating patch to trunk

Comment 21

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Hixie: that testcase is fixed by my patch in bug 182107.

Comment 22

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Fix checked in.