Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-10-20 08:45:46 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
712,989 Commits 615 Branches 98 Tags 5.9 GiB
3b72f1fb2f
Commit Graph

20328 Commits

Author SHA1 Message Date
Csoregi Natalia
39be65564e Backed out changeset 05e436fbc095 (bug 1644945) for bustage on Linux on DocumentLoadListener.cpp. CLOSED TREE 2020-06-12 15:47:40 +03:00
Matt Woodrow
39ef406847 Bug 1644945 - Implement a wrapper nsIParentChannel for ParentProcessDocumentChannel so that we can use the same code in DocumentLoadListener. r=jya 
Differential Revision: https://phabricator.services.mozilla.com/D79241
 2020-06-11 16:21:45 +00:00
Jean-Yves Avenard
7a9999e7b7 Bug 1641737 - P7. Make the DLL and PPDC promises use direct task dispatch. r=mattwoodrow 
We attempt to reduce the number of event loop iterations, which would bring us back to a similar behaviour to the pre-DocumentChannel days.

This prevent some tests to have increase intermittent failures.

While ideally, we would prefer to fix or re-write those tests properly, the extra work would block us for too long.

All those increased intermittents already have bug logged about them; so we aren't hiding dust under the rust.

Differential Revision: https://phabricator.services.mozilla.com/D78489
 2020-06-12 05:57:02 +00:00
Jean-Yves Avenard
a0ac7bafae Bug 1641737 - P6. Remove the need for an extra promise for when a DocumentChannel would connect. r=mattwoodrow 
The earlier changes make this unnecessary. The load will start and will automatically resume once the PDocumentChannelParent::RedirectToRealChannel promise gets resolved when a DocumentChannel claim the DocumentLoadListener.

Differential Revision: https://phabricator.services.mozilla.com/D78488
 2020-06-12 05:56:54 +00:00
Jean-Yves Avenard
dfdacc5993 Bug 1641737 - P5. Remove unnecessary method. r=mattwoodrow 
Differential Revision: https://phabricator.services.mozilla.com/D78210
 2020-06-12 05:56:46 +00:00
Jean-Yves Avenard
d08ba26968 Bug 1641737 - P3. Streamline cleanup so it's all done in the same place. r=mattwoodrow 
Differential Revision: https://phabricator.services.mozilla.com/D78180
 2020-06-12 05:56:36 +00:00
Jean-Yves Avenard
2137f67a30 Bug 1641737 - P2. Use MozPromise with DocumentLoadListener and remove cycle. r=mattwoodrow. 
Differential Revision: https://phabricator.services.mozilla.com/D78179
 2020-06-12 05:56:28 +00:00
Jean-Yves Avenard
a943b8c10a Bug 1644009 - P12. Make STS thread supports nsIDirectTaskDispatcher interface. r=valentin,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D79105
 2020-06-12 05:10:47 +00:00
Jean-Yves Avenard
d3f369570f Bug 1644009 - P3. Revert "Bug 1592488 - P8. Ensure AbstractThread exists with STS thread." r=valentin,necko-reviewers 
DirectTask dispatching will now be directly handled by nsThread.

Differential Revision: https://phabricator.services.mozilla.com/D79087
 2020-06-12 05:09:40 +00:00
Razvan Maries
6425307954 Backed out 2 changesets (bug 1642667) for perma failures on test_peerConnection_basicAudioNATRelay.html. CLOSED TREE 
Backed out changeset 94d3e7706178 (bug 1642667)
Backed out changeset f17d5aa45ecc (bug 1642667)
 2020-06-12 00:46:08 +03:00
Andrea Marchesini
c1c4f9f3ca Bug 1642667 - Isolate alt-srv and connection pool per first-party when privacy.partition.network_state is set to true - part 2 - tests, r=dragana,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78083
 2020-06-10 13:38:24 +00:00
Andrea Marchesini
6a5e156133 Bug 1642667 - Isolate alt-srv and connection pool per first-party when privacy.partition.network_state is set to true - part 1 - implementation, r=dragana,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78082
 2020-06-10 13:38:26 +00:00
Henri Sivonen
bb89387206 Bug 1644694 - Do not call mozilla::dom::FallbackEncoding::FromLocale() in nsIndexedToHTML.cpp. r=valentin 
Differential Revision: https://phabricator.services.mozilla.com/D79100
 2020-06-11 12:48:30 +00:00
Michal Novotny
743c33fd15 Bug 1569715 - Don't lookup CORS preflight cache when request bypasses HTTP cache, r=mayhemer,necko-reviewers,valentin 
Differential Revision: https://phabricator.services.mozilla.com/D78075
 2020-06-11 10:44:11 +00:00
Matt Woodrow
28d755ee31 Bug 1625866 - Implement nsIProgressEventSink on DocumentLoadListener. r=farre 
Differential Revision: https://phabricator.services.mozilla.com/D77982
 2020-06-08 15:15:17 +00:00
Matt Woodrow
28bef71f71 Bug 1643196 - Add 'scriptable' to nsIMultiPartChannelListener. r=mayhemer,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78175
 2020-06-10 14:03:50 +00:00
Dragana Damjanovic
f5681bc3e2 Bug 1644377 - Put Draft-28 behind a pref. r=michal,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D79031
 2020-06-10 15:48:43 +00:00
Dragana Damjanovic
a886b3e138 Bug 1644377 - Use only the first Http3 versionfrom AltSvc and add X_Firefox_Http3 header. r=michal,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D79070
 2020-06-10 10:24:49 +00:00
Dragana Damjanovic
fb111b765c Bug 1644377 - Add support for multiple http3 versions. r=michal,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78995
 2020-06-10 16:17:13 +00:00
Jason Kratzer
481d77f84c Bug 1558539 - Remove unnecessary call to mutator.Finalize(). r=decoder,necko-reviewers,valentin 
Differential Revision: https://phabricator.services.mozilla.com/D79018
 2020-06-10 16:19:09 +00:00
Valentin Gosu
b8e5f41b3f Bug 1644478 - Don't increase priority for TRR service channel r=mayhemer,necko-reviewers 
This was done before using a different thread for the TRR channel
and it didn't actually make a difference in performance, so
maybe we should remove this code.

Differential Revision: https://phabricator.services.mozilla.com/D78939
 2020-06-10 15:53:15 +00:00
Alexis Beingessner
b9bb5e1106 Bug 1642724 - convert network.http.sendOriginHeader to a StaticPref. r=KrisWright,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78911
 2020-06-10 01:18:52 +00:00
Andrea Marchesini
995aafe3a7 Bug 1643288 - Isolate HTTP channel authentication per first-party when privacy.partition.network_state is set to true - part 2 - nsIHttpProtocolHandler.authCacheKeys, r=mayhemer,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78304
 2020-06-10 14:44:08 +00:00
Andrea Marchesini
6bf8d9a1b5 Bug 1643288 - Isolate HTTP channel authentication per first-party when privacy.partition.network_state is set to true, r=mayhemer,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78303
 2020-06-10 14:43:46 +00:00
Narcis Beleuzu
86aa1a058e Backed out 2 changesets (bug 1642667) for mda failures on test_peerConnection_basicAudioNATRelay.html . CLOSED TREE 
Backed out changeset c1672891baf5 (bug 1642667)
Backed out changeset 4745600e205a (bug 1642667)
 2020-06-10 16:34:45 +03:00
Andrea Marchesini
68f456af30 Bug 1642971 - Support FTP scheme for cookies, r=mayhemer 
Differential Revision: https://phabricator.services.mozilla.com/D78042
 2020-06-10 12:28:54 +00:00
Andrea Marchesini
3ba071fd19 Bug 1642667 - Isolate alt-srv and connection pool per first-party when privacy.partition.network_state is set to true - part 2 - tests, r=dragana,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78083
 2020-06-09 11:55:21 +00:00
Andrea Marchesini
9cc4f89021 Bug 1642667 - Isolate alt-srv and connection pool per first-party when privacy.partition.network_state is set to true - part 1 - implementation, r=dragana,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78082
 2020-06-10 11:32:05 +00:00
Simon Giesecke
82dc9b2271 Bug 1642949 - Replace uses of RemoveElementAt by RemoveLastElement/PopLastElement where possible. r=necko-reviewers,froydnj 
Differential Revision: https://phabricator.services.mozilla.com/D78027
 2020-06-10 10:46:14 +00:00
Butkovits Atila
e3dce68834 Backed out 3 changesets (bug 1643289, bug 1642949) for causing failure at test_headless_screenshot.html. CLOSED TREE 
Backed out changeset 98c420f73380 (bug 1643289)
Backed out changeset 9447ea8910aa (bug 1643289)
Backed out changeset 0c827da9d847 (bug 1642949)
 2020-06-10 10:07:23 +03:00
ssengupta
1529705e20 Bug 1641825 - P2 - GTest cases added for nsBufferedInputStream to test AsyncWait/AsyncLengthWait behaviour when stream closed r=baku,necko-reviewers,mayhemer 
Depends on D77433

Differential Revision: https://phabricator.services.mozilla.com/D77799
 2020-06-10 06:39:07 +00:00
ssengupta
76c2c39353 Bug 1641825 - P1 - nsBufferedInputStream does not return error and executes callback on AsyncWait/AsyncLengthWait if stream is closed r=baku,necko-reviewers,mayhemer 
Differential Revision: https://phabricator.services.mozilla.com/D77433
 2020-06-10 06:38:44 +00:00
Chris Fronk
2c6fc2f557 Bug 1644237 - Fix crash in BackgroundFileSaver r=valentin,necko-reviewers 
Previous implementation on Finish thread was shutdown but not set to
null, as background task queue cannot be shutdown need to add nullptr
check around access in notify worker.

Differential Revision: https://phabricator.services.mozilla.com/D78845
 2020-06-10 06:23:58 +00:00
Simon Giesecke
d419f0ff08 Bug 1642949 - Replace uses of RemoveElementAt by RemoveLastElement/PopLastElement where possible. r=necko-reviewers,froydnj 
Differential Revision: https://phabricator.services.mozilla.com/D78027
 2020-06-10 05:49:28 +00:00
Andy Grover
2475b7008d Bug 1644532 - Update to Neqo 0.4.1 r=dragana,necko-reviewers 
Default to using draft 27.

Differential Revision: https://phabricator.services.mozilla.com/D78960
 2020-06-09 19:39:47 +00:00
Junior Hsu
58f2c8bc94 Bug 1638513 - P3 strip request body header for preflight, r=valentin,necko-reviewers 
(a) Steal nsCORSListenerProxy::UpdateType since the whole usage is removed
    by bug 1312864. Now I'm using it to indicate if the headers should be
	stripped.
(b) Intentionally to let HttpBaseChannel strip the header to save one
    headers copy. It's set via idl, hence I don't think it's moveable.

Differential Revision: https://phabricator.services.mozilla.com/D78831
 2020-06-09 18:38:50 +00:00
Junior Hsu
c967db3c67 Bug 1638513 - P2 let nsIHttpChannel implement shouldStripRequestBodyHeader, r=valentin,necko-reviewers 
We need to extract the common method into the idl since the connection between necko
and XHR is idl. Need to import the whole necko if we do something like
`#include "HttpBaseChannel.h"

Differential Revision: https://phabricator.services.mozilla.com/D78830
 2020-06-09 18:20:15 +00:00
Junior Hsu
b7d7ea4350 Bug 1638513 - P1 extract parsing method, r=valentin,necko-reviewers 
Differential Revision: https://phabricator.services.mozilla.com/D78829
 2020-06-09 18:19:42 +00:00
Haik Aftandilian
b1f63ced7e Bug 1632845 - Firefox gets stuck when quit from the dock if the user cancels quitting from a beforeunload prompt r=Gijs,spohl,necko-reviewers,geckoview-reviewers,dragana,agi 
Return NSTerminateCancel from applicationShouldTerminate when the user chooses to "Stay on Page".

Differential Revision: https://phabricator.services.mozilla.com/D78839
 2020-06-09 18:20:36 +00:00
Razvan Maries
caec984d54 Backed out 7 changesets (bug 1641737) assertion failures. CLOSED TREE 
Backed out changeset 7f7b98339065 (bug 1641737)
Backed out changeset 32fba417ebd0 (bug 1641737)
Backed out changeset abd9cd77f3cb (bug 1641737)
Backed out changeset e3bf9a45db6a (bug 1641737)
Backed out changeset 94d47578009c (bug 1641737)
Backed out changeset 28bca5c5b8b4 (bug 1641737)
Backed out changeset 0296d0f6c1d1 (bug 1641737)
 2020-06-09 18:55:08 +03:00
Razvan Maries
2ff52580c4 Backed out 2 changesets (bug 1640105) for perma failures on remoteautomation.py. CLOSED TREE 
Backed out changeset e12bb339406c (bug 1640105)
Backed out changeset f66cacb5c833 (bug 1640105)
 2020-06-09 17:38:23 +03:00
Valentin Gosu
841562eab6 Bug 1640724 - head_trr.js: Emit DoH response when we got all the bytes instead of "end" event r=dragana,necko-reviewers 
This patch is a workaround for an issue that causes an intermittent failure
in test_trr_case_sensitivity.js.
I was able to reproduce the bug locally with some logging, but not with
nsHttp:5 or under rr, which made it difficult to diagnose the root cause.
What I was able to determine using logging and timestamps in the nodejs code
is that we would not get the "end" event for a request, to which we were
reacting to send back the DoH response. The request has a timer, and
eventually that timer would fire and cancel the request. At that point
we would see the "end" event in nodejs but it's too late to actually process
the response.

It's not clear if this was a bug in Firefox's HTTP2 implementation
(maybe the end isn't signaled properly) or in nodejs.

This fix makes it so we send back the response when the number of bytes
specified in contentLength reaches the server.
We should investigate in a follow-up bug if there's another Necko bug
involved here.

Differential Revision: https://phabricator.services.mozilla.com/D78676
 2020-06-09 13:17:28 +00:00
Kershaw Chang
1137b41ea1 Bug 1640105 - Don't launch socket process during shutdown r=dragana 
Depends on D77303

Differential Revision: https://phabricator.services.mozilla.com/D77304
 2020-06-09 12:20:35 +00:00
Kershaw Chang
dad33c111e Bug 1640105 - Initialize necessary services eariler r=dragana 
Initialize http handler, DNS service, and IOService earlier to avoid crash.

Differential Revision: https://phabricator.services.mozilla.com/D77303
 2020-06-09 12:19:30 +00:00
Jean-Yves Avenard
067a5a74d1 Bug 1641737 - P7. Make the DLL and PPDC promises use direct task dispatch. r=mattwoodrow 
We attempt to reduce the number of event loop iterations, which would bring us back to a similar behaviour to the pre-DocumentChannel days.

This prevent some tests to have increase intermittent failures.

While ideally, we would prefer to fix or re-write those tests properly, the extra work would block us for too long.

All those increased intermittents already have bug logged about them; so we aren't hiding dust under the rust.

Differential Revision: https://phabricator.services.mozilla.com/D78489
 2020-06-09 07:24:23 +00:00
Jean-Yves Avenard
55dc18c387 Bug 1641737 - P6. Remove the need for an extra promise for when a DocumentChannel would connect. r=mattwoodrow 
The earlier changes make this unnecessary. The load will start and will automatically resume once the PDocumentChannelParent::RedirectToRealChannel promise gets resolved when a DocumentChannel claim the DocumentLoadListener.

Differential Revision: https://phabricator.services.mozilla.com/D78488
 2020-06-09 07:24:15 +00:00
Jean-Yves Avenard
3fce8c1366 Bug 1641737 - P5. Remove unnecessary method. r=mattwoodrow 
Differential Revision: https://phabricator.services.mozilla.com/D78210
 2020-06-09 07:24:04 +00:00
Jean-Yves Avenard
f5611e2936 Bug 1641737 - P3. Streamline cleanup so it's all done in the same place. r=mattwoodrow 
Differential Revision: https://phabricator.services.mozilla.com/D78180
 2020-06-09 07:23:49 +00:00
Jean-Yves Avenard
fdc7c5c703 Bug 1641737 - P2. Use MozPromise with DocumentLoadListener and remove cycle. r=mattwoodrow. 
Differential Revision: https://phabricator.services.mozilla.com/D78179
 2020-06-09 10:11:11 +00:00
Matt Woodrow
4ec2ae121c Bug 1638711 - Report correct error code in DocumentLoadListener if AsyncOpen fails. r=jya,necko-reviewers,valentin 
Differential Revision: https://phabricator.services.mozilla.com/D77954
 2020-06-08 23:17:52 +00:00