Do we need to invalidate during DocumentViewerImpl::Destroy() ?

RESOLVED FIXED in mozilla8

Status

()

Core
Layout
RESOLVED FIXED
10 years ago
6 years ago

People

(Reporter: mats, Assigned: mats)

Tracking

Trunk
mozilla8
Points:
---

Firefox Tracking Flags

(Not tracked)

Details

(Whiteboard: [inbound])

Attachments

(2 attachments, 1 obsolete attachment)

(Assignee)

Description

10 years ago
Created attachment 311013 [details]
stack

See attached stack, it seems a bit unnecessary (and fragile) to do that.

One way of fixing it would be to call SetEnableRendering() in
DocumentViewerImpl::Show()/Hide().
(Assignee)

Comment 1

10 years ago
Created attachment 311014 [details] [diff] [review]
wip
Comment on attachment 311014 [details] [diff] [review]
wip

I like this
Attachment #311014 - Flags: superreview+
Attachment #311014 - Flags: review+
Assignee: nobody → mats.palmgren
Doesn't that disable rendering for the whole viewmanager tree, though?  That doesn't seem like the right thing to be doing, does it?
Comment on attachment 311014 [details] [diff] [review]
wip

of course, you're right :-(.

The safest thing to do here would be to just have nsViewManager::UpdateView drop invalidation on the floor if the view observer has been disconnected.
Attachment #311014 - Flags: superreview-
Attachment #311014 - Flags: superreview+
Attachment #311014 - Flags: review-
Attachment #311014 - Flags: review+
(Assignee)

Comment 5

6 years ago
Created attachment 546883 [details] [diff] [review]
wip2

Like so?  It passed on Try and I tested it manually on Linux/WinXP/OSX and
I didn't see any problems.
Attachment #311014 - Attachment is obsolete: true
Attachment #546883 - Flags: review?(roc)
Attachment #546883 - Flags: review?(roc) → review+
(Assignee)

Comment 6

6 years ago
http://hg.mozilla.org/integration/mozilla-inbound/rev/1e772f5fbc24
Whiteboard: [inbound]
Target Milestone: --- → mozilla8
http://hg.mozilla.org/mozilla-central/rev/1e772f5fbc24
Status: NEW → RESOLVED
Last Resolved: 6 years ago
Resolution: --- → FIXED
Why do we want to do this? If we have a change to the view tree that affects the rendering of the window why shouldn't we invalidate?
If the view manager has no observer then it's not connected to a visible presentation, I guess.
That document doesn't care about the invalidate, but a parent document might need to repaint if a child document goes away.
Wouldn't the child document going away already have triggered the invalidation?
You need to log in before you can comment on or make changes to this bug.