Closed
Bug 949666
Opened 11 years ago
Closed 10 years ago
[e10s] Refactor file picker code to work better across processes
Categories
(Core :: DOM: Core & HTML, defect)
Tracking
()
RESOLVED
FIXED
mozilla36
Tracking | Status | |
---|---|---|
e10s | m8+ | --- |
People
(Reporter: billm, Assigned: evilpie)
References
Details
Attachments
(1 file, 1 obsolete file)
6.16 KB,
patch
|
billm
:
review+
|
Details | Diff | Splinter Review |
I'm filing this bug to cover a few things: 1. The file picker dialog should return file handles, not just paths. That's the most reasonable way to get it working with sandboxing. 2. The code to determine the last directory where you used the file picker runs in the content process, but it makes much more sense for it to run in the parent. 3. The current proxied implementation of the file picker uses ::Show, which is model. We should switch to an API that uses ::Open.
Reporter | ||
Comment 2•10 years ago
|
||
Well, we still have problem #2. We could deal with that in some other way though. Maybe we just need to make the content pref service work in the child. Or we could move that code to the parent.
Flags: needinfo?(wmccloskey)
Updated•10 years ago
|
tracking-e10s:
--- → +
Updated•10 years ago
|
Assignee | ||
Comment 3•10 years ago
|
||
I think most stuff here is fixed, but problem #2 is still quite annoying, especially when using bugzilla. Renominating.
Reporter | ||
Comment 4•10 years ago
|
||
Blake implemented the content pref service a while ago, so we just need to remove the conditionals around that code. Want to take this Tom?
Assignee | ||
Comment 5•10 years ago
|
||
Reporter | ||
Comment 6•10 years ago
|
||
Comment on attachment 8519443 [details] [diff] [review] Pass displayDirectory to parent Review of attachment 8519443 [details] [diff] [review]: ----------------------------------------------------------------- Looks great. Thanks! ::: dom/html/HTMLInputElement.cpp @@ +708,2 @@ > mInput->OwnerDoc(), lastUsedDir); > + printf("result: %u\n", uint32_t(rv)); Looks like you left this in by accident.
Attachment #8519443 -
Flags: review?(wmccloskey) → review+
Assignee | ||
Comment 7•10 years ago
|
||
https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=e8c639228134 https://hg.mozilla.org/integration/mozilla-inbound/rev/1a2c37d55e5e
Comment 8•10 years ago
|
||
sorry had to back this out for test failures like https://treeherder.mozilla.org/ui/logviewer.html#?job_id=3721157&repo=mozilla-inbound (seems this was also in the try run but not sure how this was caused by this push)
Flags: needinfo?(evilpies)
Comment 9•10 years ago
|
||
also seems https://treeherder.mozilla.org/ui/logviewer.html#?job_id=3721153&repo=mozilla-inbound is related
Assignee | ||
Comment 10•10 years ago
|
||
I guess we should put back the content process check, but just for b2g.
Flags: needinfo?(evilpies)
Updated•10 years ago
|
Assignee | ||
Comment 11•10 years ago
|
||
I put back the ContentProcess check for b2g. Looks like this fixes the issue on try: https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=ad1d0c7b82bb
Attachment #8519443 -
Attachment is obsolete: true
Attachment #8521705 -
Flags: review?(wmccloskey)
Reporter | ||
Updated•10 years ago
|
Attachment #8521705 -
Flags: review?(wmccloskey) → review+
Assignee | ||
Comment 12•10 years ago
|
||
I made an other try push. It looks really clean, and not a single b2g mochitest failure https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=b4cbcf12b411 https://hg.mozilla.org/integration/mozilla-inbound/rev/c9e1814d648d
Comment 13•10 years ago
|
||
https://hg.mozilla.org/mozilla-central/rev/c9e1814d648d
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla36
Updated•5 years ago
|
Component: HTML: Form Submission → DOM: Core & HTML
You need to log in
before you can comment on or make changes to this bug.
Description
•