209087 - Content interfaces' out parameters should use ptr-to-ptrs instead of ref-to-ptrs

Status: RESOLVED FIXED

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

Description

[Christopher Aillon (sabbatical, not receiving bugmail)]

Touching nsIAttribute, nsI*Content, nsIDocument, nsIDocumentViewer,
nsINameSpace*, nsINodeInfo*

Johnny said he's okay with this.  Patch to follow.

Comment 1

[Christopher Aillon (sabbatical, not receiving bugmail)]

Attachment: Patch (tar.gz)

Comment 2

[Christopher Aillon (sabbatical, not receiving bugmail)]

Attachment: Patch v2 (tar.gz)

Also deCOMtaminate three methods on nsINodeInfo.

Comment 3

[Christopher Aillon (sabbatical, not receiving bugmail)]

Attachment: Patch v3 (tar.gz)

Incorporates some comments from Johnny:

- use HasAttr instead of GetAttr in one place where we don't care about the
value.
- use swap() in tight loops
- foo->bar(getter_AddRefs(foo)); may break on some platform/compilers

Comment 4

[Christopher Aillon (sabbatical, not receiving bugmail)]

Comment on attachment 125545 [details]
Patch v3 (tar.gz)

Rheeeet!

Comment 5

[Johnny Stenback (:jst)]

Comment on attachment 125545 [details]
Patch v3 (tar.gz)

r+sr=jst

Comment 6

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: patch to fix crash regression in nsGenericHTMLElement::GetBaseURL

The changes to GetBaseURL caused a crash in the (eHTMLUnit_String ==
aBaseHref.GetUnit()) case because |url| went out of scope before |*aBaseURL|
was |AddRef|ed.  It fixes a crash I'm seeing on a mail message (and probably on
many other testcases).	I'm going to check it in without review in the hopes
that it makes the 9pm builds.

Comment 7

[Johnny Stenback (:jst)]

Comment on attachment 125616 [details] [diff] [review]
patch to fix crash regression in nsGenericHTMLElement::GetBaseURL

sr=jst

Comment 8

[Christopher Aillon (sabbatical, not receiving bugmail)]

Ah, thanks for fixing that silly mistake, David.

Marking this as fixed.