919263 - <input type=range> — "Assertion failure: typeIsRange || aAttribute == nsGkAtoms::value (why?)"

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Jesse Ruderman]

Attachment: testcase

Assertion failure: typeIsRange || aAttribute == nsGkAtoms::value (why?), at layout/forms/nsRangeFrame.cpp:700

Comment 1

[Jesse Ruderman]

Attachment: stack

Comment 2

[Jonathan Watt [:jwatt]]

Attachment: patch

This assertion is just bogus. We can easily and legitimately get into this state since script can change these attributes immediately after changing the 'type' attribute, in which case the change hints will not have been processed yet, so any frame reconstruction for the type change will not yet have occurred.

Comment 3

[Daniel Holbert [:dholbert]]

Comment on attachment 809337 [details] [diff] [review]
patch

># HG changeset patch
># Parent 7d0e956bd79a734ed8360eefc58b1338f6eab56b
># User Jonathan Watt <[e-mail]>
>Bug 919263 - <input type=range> â "Assertion failure: typeIsRange || aAttribute == nsGkAtoms::value (why?)". r=dholbert

Odd "a^" character there --------^ in commit message - get rid of that.

More importantly, fix the commit message -- "describe the change, not the problem".  Maybe change it to "remove bogus assertion from nsRangeFrame.cpp"?

>diff --git a/layout/forms/nsRangeFrame.cpp b/layout/forms/nsRangeFrame.cpp
>-      MOZ_ASSERT(typeIsRange || aAttribute == nsGkAtoms::value, "why?");
>+      // If script changed the 'input's type before setting these attributes
>+      // then we don't need to do anything since we are going to be reframed.

s/'input's/<input>'s/

Comment 4

[Daniel Holbert [:dholbert]]

(In reply to Daniel Holbert [:dholbert] from comment #3)
> >Bug 919263 - <input type=range> â "Assertion failure: typeIsRange || aAttribute == nsGkAtoms::value (why?)". r=dholbert
> 
> Odd "a^" character there --------^ in commit message - get rid of that.

(Hmm, interesting; looks like that only shows up in the "Edit Attachment As Comment" view; it renders as a long dash if I view the patch file directly. Maybe that's a bugzilla bug, then. But anyway, as long as you're fixing the commit message text, might as well swap that out for a friendlier character.)

Comment 5

[Daniel Holbert [:dholbert]]

(For reference, Jesse says the a^ thing is bug 477442.)

Comment 6

[Jonathan Watt [:jwatt]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/d9c1802e45d9

Comment 7

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/d9c1802e45d9