2019-10-16 J.C. Jones <jjones@mozilla.com>
* lib/softoken/pkcs11c.c:
Bug 1459141 - Backed out changeset 474d62c9d0db for PK11_Wrap/Unwrap
issues r=me
[f10c3e0757b7] [NSS_3_47_BETA3]
2019-10-15 J.C. Jones <jjones@mozilla.com>
* .hgtags:
Added tag NSS_3_47_BETA2 for changeset f657d65428c6
[3ca8b20b24ee]
* cmd/addbuiltin/addbuiltin.c:
Bug 1465613 - Fixup clang format a=bustage
[f657d65428c6] [NSS_3_47_BETA2]
2019-10-11 Marcus Burghardt <mburghardt@mozilla.com>
* automation/abi-check/expected-report-libnss3.so.txt, automation/abi-
check/expected-report-libsmime3.so.txt, automation/abi-check
/expected-report-libssl3.so.txt, cmd/addbuiltin/addbuiltin.c,
cmd/lib/secutil.c, gtests/softoken_gtest/manifest.mn,
gtests/softoken_gtest/softoken_gtest.gyp,
gtests/softoken_gtest/softoken_nssckbi_testlib_gtest.cc,
lib/certdb/certdb.c, lib/certdb/certt.h, lib/ckfw/builtins/README,
lib/ckfw/builtins/certdata.txt, lib/ckfw/builtins/manifest.mn,
lib/ckfw/builtins/nssckbi.h, lib/ckfw/builtins/testlib/Makefile,
lib/ckfw/builtins/testlib/builtins-testlib.gyp,
lib/ckfw/builtins/testlib/certdata-testlib.txt,
lib/ckfw/builtins/testlib/config.mk,
lib/ckfw/builtins/testlib/manifest.mn, lib/ckfw/builtins/testlib
/nssckbi-testlib.rc,
lib/ckfw/builtins/testlib/testcert_err_distrust.txt,
lib/ckfw/builtins/testlib/testcert_no_distrust.txt,
lib/ckfw/builtins/testlib/testcert_ok_distrust.txt,
lib/ckfw/manifest.mn, lib/nss/nss.def, lib/pki/pki3hack.c,
lib/softoken/sdb.c, lib/util/pkcs11n.h, nss.gyp, tests/cert/cert.sh:
Bug 1465613 - Created two new fields for scheduled distrust from
builtins and updated support commands. r=jcj,kjacobs,mt
Added two new fields do scheduled distrust of CAs in
nssckbi/builtins. Also, created a testlib to validate these fields
with gtests.
[52024949df95]
2019-10-14 Martin Thomson <martin.thomson@gmail.com>
* lib/ssl/tls13con.c:
Bug 1588557 - Fix debug statement, r=jcj
[0f563a2571c3]
2019-10-15 Dana Keeler <dkeeler@mozilla.com>
* gtests/mozpkix_gtest/pkixder_universal_types_tests.cpp,
lib/mozpkix/include/pkix/pkixder.h, lib/mozpkix/lib/pkixcert.cpp:
bug 1579060 - fix handling of issuerUniqueID and subjectUniqueID in
mozilla::pkix::BackCert r=jcj
According to RFC 5280, the definitions of issuerUniqueID and
subjectUniqueID in TBSCertificate are as follows:
issuerUniqueID [1] IMPLICIT UniqueIdentifier OPTIONAL,
subjectUniqueID [2] IMPLICIT UniqueIdentifier OPTIONAL,
where UniqueIdentifier is a BIT STRING.
IMPLICIT tags replace the tag of the underlying type. For these
fields, there is no specified class (just a tag number within the
class), and the underlying type of BIT STRING is "primitive" (i.e.
not constructed). Thus, the tags should be of the form CONTEXT
SPECIFIC | [number in class], which comes out to 0x81 and 0x82,
respectively.
When originally implemented, mozilla::pkix incorrectly required that
the CONSTRUCTED bit also be set for these fields. Consequently, the
library would reject any certificate that actually contained these
fields. Evidently such certificates are rare.
[c50f933d37a5]
2019-10-14 Deian Stefan <deian@cs.ucsd.edu>
* lib/softoken/pkcs11c.c:
Bug 1459141 - Rewrite softoken CBC pad check to be constant time.
r=kjacobs,jcj
[474d62c9d0db]
2019-10-11 J.C. Jones <jjones@mozilla.com>
* .hgtags:
Added tag NSS_3_47_BETA1 for changeset 93245f5733b3
[f60dbafbc182]
Differential Revision: https://phabricator.services.mozilla.com/D49470
--HG--
extra : moz-landing-system : lando
A Debugger.Frame referring to a generator or async call must track all
resumptions of that call. For the debugger to detect resumptions, the
generator/async function's script must be compiled with instrumentation on
JSOP_AFTERYIELD bytecodes. DebuggerFrame::setGenerator calls DebugScript::
incrementGeneratorObserverCount to request this.
However, incrementing the generator observer count marks the script as a
debuggee script. It is an invariant of the debugger that all frames running
debuggee scripts must themselves be marked as debuggee frames. So
DebuggerFrame::setGenerator needs to also call
Debugger::ensureExecutionObservabilityOfScript, which walks the stack and marks
other frames as debuggees where necessary.
Differential Revision: https://phabricator.services.mozilla.com/D46528
--HG--
extra : moz-landing-system : lando
The extra heading causes some exposure problems with the way
the sidebar menu collapses docs under a certain hierarchy depth.
Since the user documentation is more important than the developer
documentation, it makes sense not to collapse it.
DONTBUILD
Differential Revision: https://phabricator.services.mozilla.com/D49461
--HG--
extra : moz-landing-system : lando
All but browser_bug744745.js seem to pass even without the fixes I
made, which seems odd.
browser_bug1058164.js is a little odd because it passes in {} instead
of a boolean for the useCapture argument. I think this ends up calling
addEventListener(..., {}, false), which should be the equivalent of
addEventListener(..., {}).
Differential Revision: https://phabricator.services.mozilla.com/D49453
--HG--
extra : moz-landing-system : lando
Check the return code of write() to determine if the output line has actually been written completely; if not, write the remainder.
Tests suggest that incomplete writes are possible when the buffer exceeds a few thousand bytes. Very long log lines are unusual but this is important for cases like reftest failures where an encoded screenshot is dumped to the log.
Differential Revision: https://phabricator.services.mozilla.com/D48218
--HG--
extra : moz-landing-system : lando
Not a fix. This patch adds some assertions that we don't ever set the realm of an ObjectGroup such that it points to a gray global object. We have these asserts already for other types of edge but this one is indirect because we store a pointer to the Realm, and get the global from that.
Differential Revision: https://phabricator.services.mozilla.com/D49302
--HG--
extra : moz-landing-system : lando
When remapping wrappers the embedding can call this to create a new proxy object while in a realm whose global is gray. This breaks the JS API invariants and can cause black to gray GC edges to be creaated. Fix this by triggering the global's read barrier here.
Differential Revision: https://phabricator.services.mozilla.com/D49430
--HG--
extra : moz-landing-system : lando
"disjunction" and "unit" types aren't yet supported, because ICU doesn't
provide a C-API for this functionality. "short" and "narrow" styles aren't
supported for the same reason.
Differential Revision: https://phabricator.services.mozilla.com/D40437
--HG--
extra : moz-landing-system : lando
RelativeTimeFormat.cpp doesn't call `getInternals()` with an unresolved
internals object, so this didn't blow up in the past.
Drive-by fixes:
- Remove the unused "methodName" argument from `getRelativeTimeFormatInternals`.
Differential Revision: https://phabricator.services.mozilla.com/D40436
--HG--
extra : moz-landing-system : lando
By getting the right screen size we can shrink the popup menus which overflows
the screen size under Wayland. The ScreenManager does not help us, because we
can't get absolute window position, but we can use gdk_display_get_monitor_at_window
and gdk_monitor_get_workarea to get the correct screen rectangle.
Differential Revision: https://phabricator.services.mozilla.com/D49289
--HG--
extra : moz-landing-system : lando
Many `GlobalObject::getOrCreateXYZ()` methods explicitly perform casts to `NativeObject*`
even though the callers directly assign to `JSObject*`. So remove all explicit casts to
`NativeObject` which aren't needed in any callers to reduce unnecessary code and to make
the code a bit leaner.
Differential Revision: https://phabricator.services.mozilla.com/D42887
--HG--
extra : moz-landing-system : lando
AsyncGenerator initialisation isn't required to also always initialise AsyncFromSyncIteratorPrototype,
so split AsyncFromSyncIteratorPrototype, and its prototype, AsyncFromIteratorPrototype, from the async
generator initialisation.
Differential Revision: https://phabricator.services.mozilla.com/D42886
--HG--
extra : moz-landing-system : lando
In a follow-up bug this will be changed to use `ClassSpec`, along with the rest
of the JSProtoKey classes still using non-ClassSpec initialisation.
Differential Revision: https://phabricator.services.mozilla.com/D42883
--HG--
extra : moz-landing-system : lando
Jan-Ivar Bruaroey
jaril
J.C. Jones
Jim Blandy
Andreas Tolfsen
Wes Kocher
Gabriel Luong
David Teller
Andrew McCreight
Ehsan Akhgari
Emma Malysz
Geoff Brown
Jon Coppeard
Mats Palmgren
shindli
Jorg K
Tomislav Jovanovic
Boris Zbarsky
André Bargull
Cosmin Sabou
Christian Holler
Erica Wright
Mike Conley
Harry Twyford
Jan de Mooij
Florin Strugariu
David Walsh
Jan Horak
Gijs Kruitbosch
Jared Wein