add a little more checking to proxy subtree removal

RESOLVED FIXED in Firefox 43

Status

()

defect
RESOLVED FIXED
4 years ago
4 years ago

People

(Reporter: tbsaunde, Assigned: tbsaunde)

Tracking

unspecified
mozilla43
Points:
---
Dependency tree / graph

Firefox Tracking Flags

(firefox43 fixed)

Details

Attachments

(1 attachment)

Assignee

Description

4 years ago
No description provided.
Comment on attachment 8660049 [details] [diff] [review]
add a little more checking to proxy subtree removal

Review of attachment 8660049 [details] [diff] [review]:
-----------------------------------------------------------------

r=me with one change requested inline below:

::: accessible/ipc/DocAccessibleParent.cpp
@@ +109,5 @@
> +  // We shouldn't actually need this because mAccessibles shouldn't have an
> +  // entry for the document itself, but it doesn't hurt to be explicit.
> +  if (!aRootID) {
> +    NS_ERROR("trying to hide entire document?");
> +    return false;

Please return true here. Returning false destroys the world I think.
Attachment #8660049 - Flags: review?(dbolter) → review+
Assignee: nobody → tbsaunde+mozbugs
Assignee

Comment 3

4 years ago
(In reply to David Bolter [:davidb] from comment #2)
> Comment on attachment 8660049 [details] [diff] [review]
> add a little more checking to proxy subtree removal
> 
> Review of attachment 8660049 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> r=me with one change requested inline below:
> 
> ::: accessible/ipc/DocAccessibleParent.cpp
> @@ +109,5 @@
> > +  // We shouldn't actually need this because mAccessibles shouldn't have an
> > +  // entry for the document itself, but it doesn't hurt to be explicit.
> > +  if (!aRootID) {
> > +    NS_ERROR("trying to hide entire document?");
> > +    return false;
> 
> Please return true here. Returning false destroys the world I think.

which we want to do here (at least until we find it exposes a bunch of content side bugs we need to hack around)
Maybe Maybe MOZ_DIAGNOSTIC_ASSERT(aRoodID) is good?
https://hg.mozilla.org/mozilla-central/rev/6d8da59723c3
Status: NEW → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla43
You need to log in before you can comment on or make changes to this bug.