Summary: | WebPageProxy::updatePlayingMediaDidChange should protect from a null m_userMediaPermissionRequestManager | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | youenn fablet <youennf> | ||||
Component: | WebRTC | Assignee: | youenn fablet <youennf> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | commit-queue, eric.carlson, webkit-bug-importer, youennf | ||||
Priority: | P2 | Keywords: | InRadar | ||||
Version: | WebKit Nightly Build | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
youenn fablet
2019-10-07 03:19:26 PDT
Created attachment 380312 [details]
Patch
Comment on attachment 380312 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=380312&action=review > Source/WebKit/UIProcess/WebPageProxy.cpp:8309 > + ASSERT(m_userMediaPermissionRequestManager); If we expect m_userMediaPermissionRequestManager to be NULL some of the time, do we really want an ASSERT here? (In reply to Eric Carlson from comment #3) > Comment on attachment 380312 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=380312&action=review > > > Source/WebKit/UIProcess/WebPageProxy.cpp:8309 > > + ASSERT(m_userMediaPermissionRequestManager); > > If we expect m_userMediaPermissionRequestManager to be NULL some of the > time, do we really want an ASSERT here? By design, when nulling m_userMediaPermissionRequestManager, we set back the capture state to none. For the capture state to be different from None, m_userMediaPermissionRequestManager should in theory be recreated. So we do not expect it to be null but we cannot ignore the fact that a wrong IPC message may happen. Comment on attachment 380312 [details] Patch Clearing flags on attachment: 380312 Committed r250773: <https://trac.webkit.org/changeset/250773> All reviewed patches have been landed. Closing bug. |