TestNetworkLinkIdHashingWindows has two issues that, frighteningly, cancelled each other out under most conditions.
First is that the `sscanf` format string uses curly braces when the test data does not have braces.
Second is that the order of `nwGUIDS` does not match the sorting behavior of `nsNotifyAddrListener::HashSortedNetworkIds`.
What ended up happening was that the sscanf failed, and left the entire GUID uninitialized. Then we used that uninitialized data over and over in `nwGUIDS` so the order didn't matter. However, under AddressSanitizer, the failure became evident, because (I think, haven't verified) that ASan's instrumentation messes with the contents of the stack between the four GUID parses, so we no longer use the same uninitialized data four times. And for bonus fun, this wasn't noticed in CI because we don't (yet) run ASan on GTests for Windows.
Debugging this was... quite an adventure.
Differential Revision: https://phabricator.services.mozilla.com/D63532
--HG--
extra : moz-landing-system : lando
When it's first starting up, when mConfirmationState != CONFIRM_OK
the TRR service is not ready to use.
For TRR_FIRST requests we need to fallback to DNS while the service starts up.
Differential Revision: https://phabricator.services.mozilla.com/D61218
--HG--
extra : moz-landing-system : lando
When it's first starting up, when mConfirmationState != CONFIRM_OK
the TRR service is not ready to use.
For TRR_FIRST requests we need to fallback to DNS while the service starts up.
Differential Revision: https://phabricator.services.mozilla.com/D61218
--HG--
extra : moz-landing-system : lando
When it's first starting up, when mConfirmationState != CONFIRM_OK
the TRR service is not ready to use.
For TRR_FIRST requests we need to fallback to DNS while the service starts up.
Differential Revision: https://phabricator.services.mozilla.com/D61218
--HG--
extra : moz-landing-system : lando
This change is backwards incompatible with the older cookies.sqlite
files, which means files saved from newer versions of Firefox will
no longer be possible to open in older versions of Firefox.
Differential Revision: https://phabricator.services.mozilla.com/D60464
--HG--
rename : netwerk/test/unit/test_schema_3_migration.js => netwerk/test/unit/test_schema_10_migration.js
extra : moz-landing-system : lando
This change is backwards incompatible with the older cookies.sqlite
files, which means files saved from newer versions of Firefox will
no longer be possible to open in older versions of Firefox.
Differential Revision: https://phabricator.services.mozilla.com/D60464
--HG--
rename : netwerk/test/unit/test_schema_3_migration.js => netwerk/test/unit/test_schema_10_migration.js
extra : moz-landing-system : lando
Fetch, configure, and run node for Android on the test host, just like Linux tests do.
Make the node/HTTP/2 environment variables available to the tests on the device, and
use adb port forwarding to connect sockets. Finally, enable tests skipped for node.
Differential Revision: https://phabricator.services.mozilla.com/D60204
--HG--
extra : moz-landing-system : lando
Fetch, configure, and run node for Android on the test host, just like Linux tests do.
Make the node/HTTP/2 environment variables available to the tests on the device, and
use adb port forwarding to connect sockets. Finally, enable tests skipped for node.
Differential Revision: https://phabricator.services.mozilla.com/D60204
--HG--
extra : moz-landing-system : lando
Previously we had no way from excluding just one channel from TRR mode3.
The solution was to add the captive portal domain to the exclusion list.
Now the captive portal channel is marked with nsIRequest.DISABLE_TRR_MODE so
the exclusion is not necessary anymore.
Differential Revision: https://phabricator.services.mozilla.com/D48820
--HG--
extra : moz-landing-system : lando
The forced garbage collection has been added a decade ago for a test that
exhibited a memory increase over time in Firefox. Now that the test is no
longer present in the tree, and the garbage collector got a lot of improvements
over the last years, there is no compelling reason to keep the call to
"forceGC()" in the closing handler of httpd.js.
Differential Revision: https://phabricator.services.mozilla.com/D58550
--HG--
extra : moz-landing-system : lando
On windows 10-64 asan, we get a failure on the network io test that
checks the sqlite DB is touched. According to the logs, this failure
happens because when the test occurs the write happens in the wal
file. Adapted the test to also check for that location.
Differential Revision: https://phabricator.services.mozilla.com/D58908
--HG--
extra : moz-landing-system : lando
Andrea Marchesini
Cosmin Sabou
Valentin Gosu
Coroiu Cristina
Arthur Iakab
Noemi Erli
Andreea Pavel
David Major
Neil Deakin
Masatoshi Kimura
Christian Holler
Honza Bambas
Dragana Damjanovic
Razvan Maries
Kershaw Chang
Ehsan Akhgari
Bogdan Tara
Geoff Brown
Narcis Beleuzu
Ting-Yu Lin
Dorel Luca
Matt Woodrow
Magnus Melin
Sylvestre Ledru
Henrik Skupin
Tarek Ziadé