Граф коммитов

10280 Коммитов

Автор SHA1 Сообщение Дата
brendan@mozilla.org c18c65386e Try to fix Ts regression (414452, r=shaver). 2008-01-30 00:25:52 -08:00
brendan@mozilla.org 6ac8056119 Move serialized cycle detector for __proto__ and __parent__ into js_GC (414452, r=igor+shaver). 2008-01-29 22:29:49 -08:00
brendan@mozilla.org 7b19c3d0aa Final js1.8 feature: sugar for object destructuring (404734, r=mrbkap). 2008-01-29 22:27:13 -08:00
jst@mozilla.org c0f11e79b8 Landing followup to the fix for bug 412491, this one somehow slipped through, just more of the same. 2008-01-29 21:55:12 -08:00
igor@mir2.org c215b188ed Attachment 300249: better handling of numeric conversions. r,a=brendan 2008-01-29 21:35:34 -08:00
brendan@mozilla.org 6c5c3e2535 Dammit. 2008-01-29 21:14:55 -08:00
mrbkap@gmail.com e19f75b999 Don't wrap chrome objects in SJOWs if we don't have to. bug 414296, r=jst sr=bzbarsky 2008-01-29 21:13:19 -08:00
brendan@mozilla.org db10d95be4 Oops. 2008-01-29 21:12:45 -08:00
brendan@mozilla.org cac0b8dea5 Don't self-lock in the JSGC_BEGIN callback (413097, r=shaver). 2008-01-29 21:10:09 -08:00
crowder@fiverocks.com a1b3b54875 Bug 331043 - Improve hash performance using _rotr intrinsic (js/src hunk), patch by Michael Moy <mmoy@yahoo.com> with updates by me and r=wtc, r=brendan, a=brendan 2008-01-29 18:36:33 -08:00
jst@mozilla.org b5f8c91d5b Fixing bug 413774. Don't suspend/resume requests when making native calls into C++ in XPConnect, do the suspend/resume when pushing/popping JS contexts off of the context stack instead. r=shaver@mozilla.org, sr=brendan@mozilla.org 2008-01-29 18:28:58 -08:00
mrbkap@gmail.com 39cf192e17 Outerize this when we're computing it. bug 394012, r=brendan 2008-01-29 18:16:56 -08:00
crowder@fiverocks.com fa431d439a Bug 408113: allocate stackPools less often, r=brendan, a=blocking1.9 (schrep) 2008-01-29 18:11:30 -08:00
mrbkap@gmail.com 96076d19b0 SJOW's equality hook returns true too much. bug 413902, r+sr=jst a=damons 2008-01-29 18:06:17 -08:00
peterv@propagandism.org 17f5f4406d Fix for bug 407034 (JS_Assert "!rt->gcRunning" unbinding link elements in cycle collector with JS protocol handlers), r/sr=dbaron. 2008-01-29 18:05:43 -08:00
mrbkap@gmail.com cf3152214f Preserve interesting attributes on same-origin XOWs. bug 393269, r+sr=jst a=beltzner 2008-01-29 16:28:30 -08:00
jst@mozilla.org 555d637342 Fixing bug 414658. Make sure to not use a context from a different thread when calling functions on a wrapped JS object. r+sr=peterv@propagandism.org 2008-01-29 15:56:49 -08:00
igor@mir2.org dbb75e97c2 Bug 413744: JS_GCMETER requires to recompile just js/src, not the whole browser. r,a=brendan 2008-01-29 14:17:25 -08:00
peterv@propagandism.org 8affa46361 Try to fix orange 2008-01-29 13:42:05 -08:00
reed@reedloden.com 17dee2a4d9 Bug 411579 - "Optimize read file buffer sizes for faster startup times" [p=jmathies@mozilla.com (Jim Mathies) r=sayrer sr=bsmedberg a=blocking1.9+] 2008-01-29 12:51:46 -08:00
peterv@propagandism.org 55896575d9 Fix for bug 407034 (JS_Assert "!rt->gcRunning" unbinding link elements in cycle collector with JS protocol handlers), r/sr=dbaron. 2008-01-29 12:48:32 -08:00
nboyd@atg.com e83839494a Missing ECMA conformance test, see https://bugzilla.mozilla.org/show_bug.cgi?id=392593 2008-01-29 07:02:12 -08:00
brendan@mozilla.org 65dd6882bf Revert last checkin (414452). 2008-01-29 00:45:14 -08:00
brendan@mozilla.org a786eebf13 Move guts of js_SetProtoOrParent to jsgc.c, unifying serialization and deadlock avoidance machinery (414452, r=igor). 2008-01-28 23:18:21 -08:00
jst@mozilla.org 0886c37f01 Fixing windows bustage. 2008-01-28 15:54:56 -08:00
mrbkap@gmail.com eab5666aef Backing out to see if this is the cause for apparent random crashes. 2008-01-28 15:52:54 -08:00
jst@mozilla.org e6a7385bbe Fixing bug 413559. Make XPConnect use faster accessors for JS class/parent/private/proto. r=mrbkap@gmail.com, sr=brendan@mozilla.org 2008-01-28 15:19:51 -08:00
mrbkap@gmail.com ede517d088 There might be a pushed context but no running code. bug 413200, r+sr=jst a=blocking-1.9+ 2008-01-28 15:10:30 -08:00
mrbkap@gmail.com e127b20c84 Create a more valid pseudo-frame for liveconnect to avoid null ptr dereferences. bug 409720, r=brendan sr=jst a=brendan 2008-01-28 15:07:29 -08:00
mrbkap@gmail.com 10c6e0c9d1 Don't allocate links on the heap. bug 410223, r+sr=brendan a=schrep 2008-01-28 15:03:42 -08:00
igor@mir2.org c195802bc0 Bug 413793: fixing JS compilation. r=outside of the browser three 2008-01-28 14:36:59 -08:00
brendan@mozilla.org 4759d85bef Must lock newborn block object (shared) scope before calling js_GetMutableScope (413850, r=mrbkap). 2008-01-27 18:13:00 -08:00
mrbkap@gmail.com bb45b18132 Fix think-o causing valgrind errors during evalcx(). bug 414219, r=shaver NPOTB 2008-01-27 09:27:01 -08:00
brendan@mozilla.org 932fb674a1 Fix #if-related bug that broken compilation of pre-1.7 versions (bug 412894, r=timeless). 2008-01-26 23:58:06 -08:00
jst@mozilla.org 69a56c708b Landing fix for bug 412462. Do the right thing when converting objects wrapped in XOWs. Patch by mrbkap@gmail.com, r=jst@mozilla.org, sr=brendan@mozilla.org 2008-01-25 16:36:44 -08:00
jst@mozilla.org a116be76ca Landing fix for bug 411093. Make sure functions get the right filename. Patch by mrbkap@gmail.com, r=brendan@mozilla.org 2008-01-25 16:28:08 -08:00
jst@mozilla.org 275f61abc4 Fixing bug 397791. Prevent document principals from ever changing, and make us not use XOWs for same origin document objects. r=jonas@sickin.cc, sr=bzbarsky@mit.edu 2008-01-25 13:49:11 -08:00
jst@mozilla.org 33bb88f086 Fixing bug 413569. Make XPCWrappedNative::CallMethod() initialize the nsAutoString buffer used for [out] string param handling only when needed. r+sr=jonas@sicking.cc 2008-01-23 17:17:13 -08:00
jst@mozilla.org 2133a00277 Fixing bug 413045. Make some JS API functions faster by accessing obj->fslots[] directly when possible. r=brendan@mozilla.org, r=igor@mir2.org. 2008-01-23 14:05:43 -08:00
igor@mir2.org 57c91f16d7 Bug 413556: Patch from Jason Orendorff to fix JS_ConvertArguments. r,a=brendan 2008-01-23 06:20:38 -08:00
igor@mir2.org 8c529adb1c Bug 412340: avois rehashing of alreday atomized strings. r,a=brendan 2008-01-23 05:17:47 -08:00
igor@mir2.org 812a206f33 Bug 413565: GC now put all free cells to free lists. r,a=brendan 2008-01-23 05:05:48 -08:00
igor@mir2.org ad3ec80c7b Bug 411722: backing out again to fix incorrect error recovery 2008-01-23 01:56:28 -08:00
igor@mir2.org d9c8e2ea86 Bug 411722: faster implementation of js_GetLocalNameArray. r,a=brendan 2008-01-23 01:38:32 -08:00
igor@mir2.org 236f0f10fb Bug 411722: the new version of the patch shows the same bad Ts regression, backing out again. 2008-01-22 01:06:06 -08:00
brendan@mozilla.org 38e6d23318 Fix #ifdef NARCISSUS code to track patch for bug 409476 (NPOTB). 2008-01-22 00:56:26 -08:00
igor@mir2.org c60b21490b Bug 411722: faster implementation of js_GetLocalNames. r,a=brendan 2008-01-22 00:21:54 -08:00
sayrer@gmail.com 88dbacb40b Bug 411365. Shark functions. r/sr=brendan 2008-01-21 22:37:53 -08:00
mrbkap@gmail.com 7b788f7834 Use already running code's context when executing JS code from C++. bug 412598, r+sr=jst 2008-01-21 16:27:13 -08:00
jst@mozilla.org 4ad4aba591 Fixing bug 412698. Make XPConnect cache the per thread data for the main thread and use the cache when running on the main thread to save on time getting at the per thread data. r=dbradley@gmail.com, sr=brendan@mozilla.org 2008-01-20 20:51:55 -08:00
jst@mozilla.org b028d9847a Fixing bug 408301. Make more XPConnect wrappers share their JS object maps. r=peterv@propagandism.org, sr=brendan@mozilla.org 2008-01-20 13:03:52 -08:00
igor@mir2.org afa3a88d13 Bug 413241: fixing bogus assertion in Statements() from jsparse.c. r=brendan,mrbkap a=brendan 2008-01-20 12:55:42 -08:00
igor@mir2.org 587c01d442 Bug 411722: Backing out due to bad Ts regression. 2008-01-20 11:11:59 -08:00
igor@mir2.org 950a93d935 Bug 411722: faster implementation of js_GetLocalNames. r=brendan a=blocking1.9+ 2008-01-20 09:42:38 -08:00
peterv@propagandism.org cc0c8428bf Part of fix for bug 412491 (function objects cloned by XPConnect still keep hidden window alive late into shutdown). r=igor, sr=jst. 2008-01-20 09:29:35 -08:00
timeless@mozdev.org 6ebd0d5466 Bug 334195 XPCWrappedNative::HandlePossibleNameCaseError dereferences an unitialized value if !set r=dbradley sr=jst a=mtschrep 2008-01-20 03:21:36 -08:00
igor@mir2.org db688a1770 Bug 410649: function statement and destructuring parameter name clash now favours the function. r,a=brendan 2008-01-20 02:34:06 -08:00
reed@reedloden.com 5eaa89d176 I'm dumb. Please forgive me. Yes, that includes you, shaver. 2008-01-20 01:54:09 -08:00
reed@reedloden.com c031ce7f17 Bug 331545 - "Use _InterlockedCompareExchange for js_CompareAndSwap" [p=mmoy@yahoo.com (Michael Moy) r+a1.9=brendan a=blocking1.9+] 2008-01-19 20:39:10 -08:00
jwalden@mit.edu dae8a6b9a4 Fix typo noted by igor, bug 409476 2008-01-19 18:52:23 -08:00
sayrer@gmail.com 95af9eef8d Bug 412874. Need Shark functions in JS component global scope. r/sr=brendan 2008-01-19 18:31:10 -08:00
shaver@mozilla.org fb0b46ab3c Config file for Solaris 10 from Wesley Garland (wes@page.ca), r=shaver. 2008-01-19 18:12:59 -08:00
jwalden@mit.edu 24f7096762 Bug 409476 - js_ValueToBoolean is pure, remove cx and out parameters and just return a boolean value for better perf. r+a=brendan 2008-01-19 16:56:08 -08:00
igor@mir2.org 6b9e1e7766 Bug 411025: temp rooting implemenation cleanup, r,a=brendan 2008-01-19 16:46:37 -08:00
sayrer@gmail.com ab17fc5d6c Bug 411365. JS Shark renames per brendan. r/a=brendan 2008-01-19 15:33:08 -08:00
igor@mir2.org c526c4b477 Bug 413104: access to JSString is hidden behind macros even for a flat string. 2008-01-19 13:15:08 -08:00
reed@reedloden.com d10363409b Back out Igor's patch from bug 412340 due to consistent orange on fxdbug-win32-tb. 2008-01-19 00:06:45 -08:00
igor@mir2.org bb8f471f9d Bug 412340: JString stores a flag to indicate that it was atomized. r=brendan a=blocking1.9+ 2008-01-18 18:56:51 -08:00
peterv@propagandism.org 09001414ad Back out remainder of patch for bug 412491 to try to fix orange. 2008-01-18 07:36:28 -08:00
peterv@propagandism.org 733973188f Back out to try to fix orange. 2008-01-18 05:59:07 -08:00
peterv@propagandism.org 442e717108 Fix for bug 412491 (function objects cloned by XPConnect still keep hidden window alive late into shutdown). r=igor, sr=jst. 2008-01-18 05:29:06 -08:00
reed@reedloden.com f015d06d01 Back out patch from bug 411579, as there may be a Ts perf regression. 2008-01-18 02:36:40 -08:00
reed@reedloden.com e2f4447ded Bug 412811 - "xpconnect-tools build broken" [p=armin76@gentoo.org (Raúl Porcel) r+sr=bsmedberg (NPODB)] 2008-01-18 01:09:18 -08:00
reed@reedloden.com cd2af3de01 Bug 411579 - "Optimize read file buffer sizes for faster startup times" [p=jmathies@mozilla.com (Jim Mathies) r=sayrer sr=bsmedberg a=blocking1.9+] 2008-01-18 01:08:09 -08:00
reed@reedloden.com 4a77dda5cf Bug 411968 - "Finishing JS_AlreadyHasOwnProperty" (CHECK_REQUEST(cx) in JS_AlreadyHasOwn{,UC}Property) [p=jorendorff@mozilla.com (Jason Orendorff) r+a1.9=brendan] 2008-01-17 02:28:57 -08:00
sayrer@gmail.com f2a9049b13 Bug 411365. Start and stop Shark from JS. r=crowder, sr=jst 2008-01-16 12:42:50 -08:00
igor@mir2.org e5294ff020 Bug 412047: optimizing regexp upper/lower. r=crowder,a=blocking1.9 2008-01-16 00:31:15 -08:00
reed@reedloden.com 48b624cd11 Bug 411285 - "js_BoyerMooreHorspool is sometimes called for very short texts" [p=Seno.Aiko@gmail.com r=crowder a1.9=schrep] 2008-01-15 22:37:13 -08:00
bclary@bclary.com 70124a60a3 JavaScript Tests - test iterator value scope in array comprehension, by Norris Boyd, bug 412467, not part of the build 2008-01-15 20:47:31 -08:00
mrbkap@gmail.com 83e127e0b0 Don't call js_PopStatement if there was a parse error. bug 410852, r=brendan 2008-01-15 14:24:54 -08:00
bent.mozilla@gmail.com 55194785a8 Bug 352791 - "Permission denied to get property XULElement.ownerDocument". Patch by Blake Kaplan <mrbkap@gmail.com>. r=jst,bzbarsky. sr=brendan. a=blocking1.9. 2008-01-15 10:23:46 -08:00
benjamin@smedbergs.us c1214a52cd Bug 411327 - nsIXPCNativeCallContext should not inherit from nsISupports, r=mrbkap, a=schrep 2008-01-15 07:57:43 -08:00
benjamin@smedbergs.us b3e87aa63b Bug 411327 - nsIXPCNativeCallContext should not inherit from nsISupports, r=mrbkap, a=schrep 2008-01-15 07:50:57 -08:00
bclary@bclary.com a2cb4b68cf JavaScript Test - fix thinko in ecma_3/Number/15.7.4.3-01.js, bug 412068 2008-01-14 23:09:46 -08:00
jwalden@mit.edu c9b4d9d6f4 Bug 412068 - Number.prototype.toLocaleString incorrectly accesses the memory where its first argument should be, even if no first argument is actually given. Tag-team r=igor, r+a=brendan By the way, this checkin occurred on an orange tree. Just like every other patch in the last day. Yay checkin policy. 2008-01-14 20:16:08 -08:00
bclary@bclary.com 12fa39509d Sisyphus|JavaScript Tests - fix runtests.sh -e extra argument processing, bug 412317, not part of the build 2008-01-14 11:07:01 -08:00
sayrer@gmail.com 52d1056980 Bug 412210. SpiderMonkey faster at -O2 with msvc. r/a=brendan 2008-01-13 20:45:42 -08:00
bclary@bclary.com 4cd27a119e Sisyphus|JavaScript Tests - TinderboxPrint of test counts contain leading whitespace, bug 412183, not part of the build 2008-01-13 10:44:48 -08:00
jag@tty.nl 348b5a2de0 Bug 406800: HP's OA crash [@js_FinalizeObject][@ RtlpDeCommitFreeBlock] when loading blade enclosure info. r/sr=peterv GC was clearing mJSVal without updating mData, so XPCTraceableVariant's destructor would get confused and call Cleanup() on mData when it shouldn't (the buffer pointed to wasn't ours to free, you see). Instead of avoiding Cleanup(), make it be harmless by nulling out the pointer to the buffer. 2008-01-13 10:21:08 -08:00
bclary@bclary.com b081323f55 JavaScript Tests - update public failures, bug 314874 verified fixed 2008-01-13 09:40:04 -08:00
brendan@mozilla.org a83ad1b4dd * Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases). 2008-01-12 16:31:31 -08:00
reed@reedloden.com 84015e0baf Change to ISO C90 comment style to fix warning in sayrer's last patch. 2008-01-11 23:37:32 -08:00
sayrer@gmail.com 83a63cb101 Bug 353962. Use OSAtomic on Darwin for JS thinlocks. r/a=brendan 2008-01-11 23:16:14 -08:00
dwitte@stanford.edu 3f33f45d2a thoroughly whack mallocfest in nsID/nsJSID and friends. b=410250, r+sr=jst, a=blocking1.9+ 2008-01-11 20:30:42 -08:00
bclary@bclary.com b3ff48cc38 JavaScript Tests - regression test for bug 411893, by Jeff Walden, not part of the build 2008-01-11 19:54:40 -08:00
bclary@bclary.com ba4c806ad1 JavaScript Tests - regression test for bug 411889, by Jeff Walden, not part of the build 2008-01-11 19:48:16 -08:00
myk@mozilla.org 3db3cf881d backing out jst's fix for bug 408301 because it seems to have turned some tinderboxen red, others orange, and caused web content regressions 2008-01-11 17:56:21 -08:00
jwalden@mit.edu 01c28c6669 Fix dumb mistake in the last toString patch that causes proto-acid3 to crash. 2008-01-11 16:12:36 -08:00
jwalden@mit.edu bb236cb5c9 Bug 411893 - num.toPrecision(undefined) !== num.toString(). r=crowder, a=brendan 2008-01-11 15:12:15 -08:00
jwalden@mit.edu e958678360 Bug 411889 - Given a = num.toString(), b = num.toString(undefined), c = num.toString(10), expect (a === b && b === c). r=crowder, a=brendan 2008-01-11 15:07:13 -08:00