Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-10-15 14:25:52 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
381,452 Commits 615 Branches 98 Tags 5.8 GiB
3cf0bd780e
Commit Graph

44569 Commits

Author SHA1 Message Date
Jeff Walden
3cf0bd780e Bug 1064391 - Fix destructuring into a __proto__ pattern. r=shu 
--HG--
extra : rebase_source : dcc08c49d028eeebdcfc1428502fe64e6286b930
 2014-09-08 13:59:47 -07:00
Ehsan Akhgari
37098f6c5e Backed out changeset 9a796d86bae8 (bug 1063728) until people agree what they want to do here 2014-09-08 20:38:21 -04:00
Bobby Holley
87c9e0954f Bug 1062077 - Remove now-no-op error reporter munging in mozJSSubScriptLoader.cpp. r=bz 
In the past this might theoretically have switched us from
NS_ScriptErrorReporter to xpc::SystemErrorReporter, but those have now been
merged.
 2014-09-08 16:30:12 -07:00
Bobby Holley
9ff6cd212d Bug 1062077 - Merge NS_ScriptErrorReporter and xpc::SystemErrorReporter. r=bz 
We keep the latter name because it's referenced in more places. Error reporters
will go away entirely in bug 981187.
 2014-09-08 16:30:11 -07:00
Bobby Holley
0ffae4a42a Bug 1062077 - Switch SystemErrorReporter to xpc::ErrorReport. r=bz 2014-09-08 16:30:11 -07:00
Bobby Holley
7ba663f3ec Bug 1062077 - Hoist console logging duties into xpc::ErrorReport. r=bz 2014-09-08 16:30:11 -07:00
Bobby Holley
e85bd78d5d Bug 1062077 - Introduce xpc::ErrorReport. r=bz 
I've got it! Let's just wrap the JS thing in an XPCOM thing that duplicates
its fields with a different ownership model! We totally need another one of
those, right? And we could even stick it in XPConnect! </sarcasm>

In seriousness - the code to own, format, and display error reports is
currently spread between the JS engine, JSErrorReporters, an async Runnable
abstraction, and elsewhere. We need to condense it somewhere to start chipping
away at this mess.
 2014-09-08 16:30:10 -07:00
Bobby Holley
06cc2ae3c2 Bug 1062631 - Remove exception swallowing code in NS_ScriptErrorReporter. r=bz 
See bug 1062631 comment 0.
 2014-09-08 13:36:22 -07:00
Bobby Holley
0c8cf16e5b Bug 1062631 - Fix up content QI security check and tests. r=bz 
This stuff is broken in a bunch of ways. The test isn't doing the right thing
because it's using SpecialPowers. When we pass the content object to
xpconnectArgument(), it gets wrapped in a SpecialPowers wrapper, so the check
succeeds, only to fail down the line when the QI method is actually invoked
with an IID (which we can't pass to content scopes anyway).

This whole security check is probably useless given the above, but let's be
safe, make it a bit more robust, and fix up the test while we're at it.
 2014-09-08 13:36:21 -07:00
Aleksandar Zlicic
7be7c1afee Bug 1061562 - IonMonkey MIPS: Fix build failures on MIPS caused by SIMD changes r=nbp 2014-09-04 18:11:49 +02:00
Boris Zbarsky
12857de707 Bug 1063521. Remove the WANT_OUTER_OBJECT classinfo flag, since it's now unused. r=bholley 2014-09-08 10:05:11 -04:00
Johannes Schulte
389029b078 Bug 1059241 - Don't mark array as non-packed, when calling pop() . r=bhackett 2014-09-04 14:26:50 +02:00
Brian Hackett
97130a2dda Bug 1063598 - Infer constant properties even when the type property has not yet been instantiated, r=jandem. 2014-09-07 10:27:31 -06:00
Phil Ringnalda
d2bd0b3039 Backed out changeset 7e55ef1c49d1 (bug 1063598) for assertion failures 
CLOSED TREE
 2014-09-06 22:40:45 -07:00
Brian Hackett
2c905451f3 Bug 1063180 - Don't continue analyzing new script properties after the new script has been cleared, r=jandem. 2014-09-06 21:52:31 -06:00
Brian Hackett
8b6aac3a67 Bug 1063598 - Infer constant properties even when the type property has not yet been instantiated, r=jandem. 2014-09-06 21:50:58 -06:00
ziyunfei
9b5d989b59 Bug 1063915 - Make dumpValue() support symbols. r=evilpie 2014-09-06 18:21:20 +08:00
Wes Kocher
cff080e6ec Back out 11 changesets (bug 1063233 and bug 1063247) 
* * *
Backed out changeset 7c7c837585fb (bug 1063233)
* * *
Backed out changeset e584d23b3ce3 (bug 1063233)
* * *
Backed out changeset ee169ca2e00b (bug 1063247)
* * *
Backed out changeset 514c713b2313 (bug 1063247)
* * *
Backed out 3 changesets (bug 1063247)

Backed out changeset a6033944c4c1 (bug 1063247)
Backed out changeset c7ffa64956c6 (bug 1063247)
Backed out changeset dd80b2984fe2 (bug 1063247)
* * *
Backed out 4 changesets (bug 1063233)

Backed out changeset 1f6db27e6240 (bug 1063233)
Backed out changeset b8e3d887faa2 (bug 1063233)
Backed out changeset 8e1913c07322 (bug 1063233)
Backed out changeset 04d42b23a489 (bug 1063233)
 2014-09-05 17:42:45 -07:00
Jim Blandy
eeb67c06a0 Bug 1063233: Re-include js/src/jit/JitOptions.cpp in UNIFIED_SOURCES. r=none 2014-09-05 15:05:40 -07:00
Ehsan Akhgari
bf881ddde5 Bug 1063728 - Run check_spidermonkey_style.py before starting to build; r=jandem 2014-09-05 18:00:24 -04:00
Jim Blandy
c46d5a6268 Bug 1063233: Remove use of UbiNode.h from IonCode.h, with consequent changes; Windows non-unified builds seem to be unhappy with it. CLOSED TREE r=bustage 2014-09-05 14:57:42 -07:00
Tom Schuster
dcb538e74a Bug 1061270 - Fix crash with cross-compartment Set cloning. r=luke 2014-09-05 23:17:39 +02:00
Ehsan Akhgari
37ea61cf99 Bustage fix for bug 1060976, irc-r=efaust, landed on a CLOSED TREE 2014-09-05 15:47:00 -04:00
Ehsan Akhgari
21a9749477 Bug 1060976 - Fix more bad implicit constructors in the JS engine; r=jorendorff 2014-09-05 15:11:39 -04:00
Boris Zbarsky
b0b2acb195 Bug 1017424 part 9. Remove redundant xpconnect code looking for ChromeWindow/ModalContentWindow classes. r=bholley 2014-09-05 14:28:46 -04:00
Boris Zbarsky
7dd6d51191 Bug 1017424 part 7. Remove code to handle XPConnect window wrappers from WindowNamedPropertiesHandler and XPCJSRuntime. r=peterv 2014-09-05 14:28:45 -04:00
Boris Zbarsky
4532962c88 Bug 1017424 part 5. Remove classinfo for Window. r=peterv 2014-09-05 14:28:45 -04:00
Boris Zbarsky
b197b61892 Bug 1017424 part 4. Remove Web IDL quickstubs for Window. r=peterv 2014-09-05 14:28:44 -04:00
Jim Blandy
bad296b02c Bug 1063247: Include declaration for JS::dbg::SetDebuggerMallocSizeOf. CLOSED TREE r=bustage 2014-09-05 11:24:51 -07:00
Jim Blandy
fd3073d0be Bug 1063247: Stop checking for malloc_size; we need to get the header file righ. CLOSED TREE r=bustage 2014-09-05 11:11:35 -07:00
Jim Blandy
71484f95ca Bug 1063247: Let embeddings tell Debugger how to find the size of a malloc'd block of memory. r=terrence 2014-09-05 10:43:36 -07:00
Jim Blandy
796704a305 Bug 1063247: Amend JS::ubi::Node::size and its implementations to expect a mozilla::MallocSizeOf function. r=terrence 
Note that JS::ubi::Node::size has no callers at present, so we can change its
type without changing any callers.
 2014-09-05 10:43:35 -07:00
Jim Blandy
96c87e3727 Bug 1063247: Allow implicit construction of JS::ubi::Node from JS::HandleValue. r=terrence 
This seems reasonable, since a ubi::Node is really just a pointer with some
introspection support, so the value and the Node are pretty similar. And I don't
see cases where you'd want an error.
 2014-09-05 10:43:35 -07:00
Jim Blandy
2987a288ea Bug 1063233: Move JS::ubi::Node support for js::types::TypeObject next to where it is declared/defined. r=terrence 2014-09-05 10:37:19 -07:00
Jim Blandy
5d4c33121b Bug 1063233: Move JS::ubi::Node support for js::jit::JitCode next to where it is declared/defined. r=terrence 2014-09-05 10:37:19 -07:00
Jim Blandy
f45cc6900e Bug 1063233: Move JS::ubi::Node support for js::Shape and js::BaseShape next to their declarations and definitions. r=terrence 2014-09-05 10:36:07 -07:00
Jim Blandy
2f06da2369 Bug 1063233: Move JS::ubi::Node support for js::LazyScript to sit next to the rest of LazyScript's implementation. r=terrence 2014-09-05 10:34:03 -07:00
Jim Blandy
724dc1950b Bug 1062089: Rename js/public/DebugAPI.h to js/public/Debug.h. It's a public header file; *obviously* it's an API. r=jorendorff 
--HG--
rename : js/public/DebugAPI.h => js/public/Debug.h
 2014-09-05 10:32:29 -07:00
Hannes Verschore
6724407033 Bug 1061646 - Inline CheckForInterrupt on windows, just like we do on linux, r=jorendorff 2014-09-05 18:28:15 +02:00
Brian Hackett
a299798558 Bug 1061318 - Improve type information for RegExp.exec result objects, r=h4writer. 2014-09-05 09:27:21 -06:00
Brian Hackett
a08965c116 Bug 1062648 - Treat scripts which have been ion compiled or inlined before as hot enough to inline, r=h4writer. 2014-09-05 09:20:39 -06:00
Chris Peterson
7bea2be20a Bug 1036781 - Part 12: Replace MOZ_ASSUME_UNREACHABLE with MOZ_CRASH in js/src/jit. r=jandem 2014-08-21 21:48:01 -07:00
Nicholas Nethercote
94bc3e8dde Bug 1040593 (attempt 2) - Partly allocate elements for |new Array(N)| when N > 2048. r=jandem,bhackett. 
--HG--
extra : rebase_source : 2bcccb3aa1543649eb30cd8c43680fc09ec6df91
 2014-08-27 23:35:57 -07:00
Wes Kocher
edce5ffd3c Backed out changeset 94eed55e5a5b (bug 1040593) on a CLOSED TREE 2014-09-04 18:00:59 -07:00
Nicholas Nethercote
10873906cf Bug 1040593 - Partly allocate elements for |new Array(N)| when N > 2048. r=jandem,bhackett. 
This patch makes |new Array(N)| always allocate at least
|EagerAllocationsMaxLength| elements up front.

It also changes |EagerAllocationsMaxLength| from 2048 to 2046, so that when the
two extra slots needed for |ObjectElements| are included, the resulting
allocation size is a power-of-two.

--HG--
extra : rebase_source : 17e069bc9d39c5d5ee957a99e11ab189f7493540
 2014-08-27 23:35:57 -07:00
Jim Blandy
48ad8c1cc3 Bug 1061961: Fully initialize ScriptSourceObjects built by XDR and for Function.prototype. r=shu 2014-09-04 15:40:10 -07:00
Jeff Walden
0aeca94121 Bug 929642 - Make .arguments and .caller throw a TypeError when accessed on builtin functions. r=till 
--HG--
extra : rebase_source : 94f0c50b6828e8cdcf35002dfa41543d2370fb26
 2014-09-03 10:12:42 -07:00
Jeff Walden
7f5f3f7054 Bug 1061853 - Remove a stale bit of code dealing with the parsing of shorthands in destructuring assignment object patterns as not being done by identifierName() (they are now). r=shu 
--HG--
extra : rebase_source : b8de5cacd646f0f4ca70179c9fa673291157807d
 2014-09-03 17:28:10 -07:00
Jeff Walden
2c8a9f2eca Bug 1061853 - Rename a variable from |member| to |propdef| (referring to the items in an object literal during bytecode emission). r=shu 
--HG--
extra : rebase_source : aa5351b3ac87b93bfc16d64780635a92805b5c64
 2014-09-03 17:26:50 -07:00
Jeff Walden
874117b538 Bug 1061853 - Make duplicate __proto__ in an object literal a syntax error. Also add a test for __proto__ in object literals in various forms (normal, shorthand, computed, method, generator versions of all of these, getter, setter), exhaustively combined in literals with one to three elements. r=shu 
--HG--
extra : rebase_source : b269d45c8f2eaff9dec519ebcaa4b659a6fdf042
 2014-09-02 11:14:31 -07:00