Child process shouldn't send events to parent if a11y isn't enabled in parent

NEW
Assigned to

Status

()

Core
Disability Access APIs
P3
normal
2 years ago
2 years ago

People

(Reporter: lsocks, Assigned: lsocks)

Tracking

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(1 attachment)

(Assignee)

Description

2 years ago
Child process will try to send IPC messages to parent when parent process doesn't have a11y enabled but child does (e.g. marionette-e10s tests)
(Assignee)

Comment 1

2 years ago
Created attachment 8762717 [details] [diff] [review]
check if a11y is enabled in parent first before sending events
Assignee: nobody → lorien
Attachment #8762717 - Flags: review?(tbsaunde+mozbugs)
(Assignee)

Updated

2 years ago
Summary: Child process shouldn't send IPC messages if a11y isn't enabled in parent → Child process shouldn't send events to parent if a11y isn't enabled in parent
Comment on attachment 8762717 [details] [diff] [review]
check if a11y is enabled in parent first before sending events

r=me with the first comment fixed.

> Accessible::HandleAccEvent(AccEvent* aEvent)
> {
>   NS_ENSURE_ARG_POINTER(aEvent);
> 
>-  if (IPCAccessibilityActive() && Document()) {
>+  if (IPCAccessibilityActive() && Document() && GetAccService()->IsEnabledInChrome()) {

you missed the IPCAccessibilityActive() check in DocManager.cpp which would have the effect of not creating any DocAccessibleParents (the NotificationController change prevents sub document ones).

>+
>+  virtual void SetEnabledInChrome() = 0;

could use a comment.  Also yuk this interface should be merged with nsAccessibilityService itself, but that requires either a static cast or fixing how the service is started which I don't think needs to be done now.
Attachment #8762717 - Flags: review?(tbsaunde+mozbugs) → review+

Updated

2 years ago
Priority: -- → P2

Updated

2 years ago
Priority: P2 → P3
You need to log in before you can comment on or make changes to this bug.