Modify the process-switching load code to support POST payloads on loads
Categories
(Core :: DOM: Core & HTML, enhancement, P3)
Tracking
()
Tracking | Status | |
---|---|---|
firefox66 | --- | fixed |
People
(Reporter: relaas, Assigned: nika)
References
(Blocks 2 open bugs)
Details
Attachments
(7 files)
(deleted),
text/x-phabricator-request
|
Details | |
(deleted),
text/x-phabricator-request
|
Details | |
(deleted),
text/x-phabricator-request
|
Details | |
(deleted),
text/x-phabricator-request
|
Details | |
(deleted),
text/x-phabricator-request
|
Details | |
(deleted),
text/x-phabricator-request
|
Details | |
(deleted),
text/x-phabricator-request
|
Details |
Updated•6 years ago
|
Reporter | ||
Updated•6 years ago
|
Assignee | ||
Updated•6 years ago
|
Updated•6 years ago
|
Assignee | ||
Comment 2•6 years ago
|
||
Assignee | ||
Comment 3•6 years ago
|
||
Assignee | ||
Comment 4•6 years ago
|
||
Assignee | ||
Comment 5•6 years ago
|
||
Assignee | ||
Comment 6•6 years ago
|
||
Assignee | ||
Comment 7•6 years ago
|
||
Assignee | ||
Comment 8•6 years ago
|
||
Comment 9•6 years ago
|
||
(In reply to :Nika Layzell from comment #8)
As I mention in some of the above comments, I am running into issues due to
the channel being closed by the original TabChild as it is shut down during
the process swap. It seems as though the redirect process doesn't proceed
far enough before the new TabParent is provided for the redirect to proceed
smoothly.Valentin, do you know what changes would need to be made to allow this to
work? I have tried to figure out how the HTTP loading code works, but I am
not yet confident in my understanding & don't know what I would need to set
up.
From what I recall, the old channel is deleted by HttpChannelParentListener::OnRedirectResult() [1]
This is the big-picture of how redirects work. [2]
I'm thinking we should delay destroying the old docshell/tabchild until the redirect is completed, no? Or maybe add a way of destroying the tabChild that doesn't close the parent-side of the channel, so it gets a chance to complete the redirect?
[1] https://searchfox.org/mozilla-central/rev/76fe4bb385348d3f45bbebcf69ba8c7283dfcec7/netwerk/protocol/http/HttpChannelParentListener.cpp#322
[2] https://gist.github.com/valenting/ce444c31bb8ef2f53b9acfc3f59feedb
Comment 11•6 years ago
|
||
What's the review priority for me here?
Updated•6 years ago
|
Assignee | ||
Comment 13•6 years ago
|
||
Assignee | ||
Comment 14•6 years ago
|
||
(In reply to Honza Bambas (:mayhemer) from comment #12)
see c11
This is fairly high priority with the new revision just posted. These patches now past at least some tests & I would like to start being able to use the new process changing capabilities. It'd also be really nice to have your input as to how exactly the interception should be performed, especially considering the limitations which I'm thinking currently exist. If you have a good solution for working around those issues please let me know :-)
Comment 15•6 years ago
|
||
OK, I will do this as the first thing the next week (Monday).
Comment 16•6 years ago
|
||
Comment 17•6 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/b0b61b6a4014
https://hg.mozilla.org/mozilla-central/rev/881438d34f09
https://hg.mozilla.org/mozilla-central/rev/c320f8c70785
https://hg.mozilla.org/mozilla-central/rev/47630659e925
https://hg.mozilla.org/mozilla-central/rev/ae82e6f22e29
https://hg.mozilla.org/mozilla-central/rev/909797e624a8
https://hg.mozilla.org/mozilla-central/rev/4e5e0c80bae2
Updated•6 years ago
|
Updated•6 years ago
|
Description
•