Граф коммитов

2006 Коммитов

Автор SHA1 Сообщение Дата
dbradley%netscape.com 371eebbe33 Bug 198668 - [AxPlugin] Ofoto control does not return the value of properties correctly. r=dbradley/adamloc, sr=jst
IDispatch only changes, not part of the regular build
2003-03-26 04:20:33 +00:00
timeless%mozdev.org a4413bbc8a Bug 101198 nsXPCToolsCompiler.cpp > xpctools_private.h incorrectly uses LL_INIT which breaks for !*HAVE_LONG_LONG
removing unused include
r=dbradley sr=dmose
2003-03-25 07:18:15 +00:00
timeless%mozdev.org 072e6da90b Bug 101198 nsXPCToolsCompiler.cpp > xpctools_private.h incorrectly uses LL_INIT which breaks for !*HAVE_LONG_LONG
r=dbradley sr=dmose
2003-03-25 06:57:55 +00:00
bryner%netscape.com 84d98f2f3e Another leak fix from bug 198655 that it turns out we do need. r/sr=brendan. 2003-03-25 00:41:13 +00:00
bryner%netscape.com 3e84c4969c - Fix some JS component loader shutdown leaks.
- Ensure that we use JSPRINCIPALS_DROP() where needed to avoid leaking principals.
- Clear newborn roots when we've finished with a JSContext, so we don't keep objects from being gc'd if there are no other references.

Part of bug 198655.  Thanks to dbaron for much of the component loader patch.  r=shaver, sr=brendan.
2003-03-23 07:22:18 +00:00
dbradley%netscape.com 394011eae7 Minor comment fix 2003-03-23 00:28:40 +00:00
dbaron%dbaron.org 51d4f67690 Add DEBUG_dbaron to the DEBUG_xpc_hacker list. 2003-03-22 17:26:29 +00:00
rogerl%netscape.com c6d91b4c6a Fix bug #122076. r=brendan. Stop [] parse loop from wandering off into the
weeds.
2003-03-22 03:42:22 +00:00
dbradley%netscape.com bfd19c222e bug 197281 - Introduction of XPCConvert::GetNativeInterfaceFromJSObject is causing problems r=jst, sr=alecf 2003-03-22 02:13:38 +00:00
mkaply%us.ibm.com f6bcf1c153 #188249
s/r=brendan
Cleanup #ifdef and add EMX
2003-03-21 22:18:12 +00:00
bolian.yin%sun.com be94514e09 Bug 197088, Netscape 7.02 crashes when my JavaScript accesses a Java string or int. r=av, sr=beard. patch by joshua.xia@sun.com 2003-03-21 03:37:18 +00:00
brendan%mozilla.org cba834a969 Saner ReportReadOnlyScope message names the sealed object, not an undefined property that can't be created in its scope (94693, r=me). 2003-03-20 11:27:33 +00:00
brendan%mozilla.org 86e5524746 Fix JSOP_BINDNAME to have JOF_SET (which is now the same as JOF_ASSIGNING, so use that where it's a clearer name; 94693, r=me). 2003-03-20 11:05:10 +00:00
brendan%mozilla.org 23d950dbcf Fix broken left-associativity and constant folding for + chains involving some string terms (196290, r=shaver by telepathy). 2003-03-19 19:33:33 +00:00
bzbarsky%mit.edu 83e46b1a49 Fix some spelling errors in comments. Bug 106386, patch by Nicholas Bebout
<nb@ipnode.info>, r=doron, sr=dmose
2003-03-19 04:35:08 +00:00
brendan%mozilla.org c0fc066a01 Don't override a prototype property iff it is JSPROP_SHARED (not also if it's a JSPROP_SETTER; 197940, r=rogerl). 2003-03-19 02:25:39 +00:00
brendan%mozilla.org f2c8bb3531 Move readonly pre-ECMA error report down to new place peculiar to sealed
scopes, and share it via downward goto from the old place (which should go
away when a bug on khanson's list whose # I forget is fixed).  This fixes
the hole in sealed scope support mentioned in bug 94693 (r=rogerl/shaver).
2003-03-19 00:53:45 +00:00
seawood%netscape.com 64b0b42708 Fix --enable-xpctools post-minimo bustage 2003-03-17 23:40:16 +00:00
dougt%netscape.com 0d2808c511 Create a new xpcom obsolete library (and component library).
Moved nsFileSpec and related classes into this obsolete library.
  Moved nsRegistry and related libreg functionality into the obsolete library.

Updated many callers using the obsolete nsFile spec to use nsIFile and Necko to do file IO.

Combined the following DLLs (source -> dest)
  uriloader -> docshell
  shistory -> docshell
  jsurl -> jsdom
  gkview -> gklayout

Moved nsAdapterEnumerator out of xpcom/ds and into mailnews, since they're the only consumer

Modifed the xpt_link tool so that you can specify a �only include� cid list that can mask CID�s that you are not interested in.

Added build options:
Prevent the building of xpinstall (--disable-xpinstall)
Prevent the building js component loader (--disable-jsloader)
A build option to only build a single profile (--enable-single-profile)
A build flag to only built the required xpfe components (--disable-xpfe-components).

Removal or hiding of unused functions and classes including nsEscape*, nsDequeIterator, nsRecyclingAllocatorImpl, nsDiscriminatedUnion, nsOpaqueKey, nsCRT::strlen, NS_NewCommandLineService

Bug 194240, r/sr = darin, alec.
2003-03-15 01:04:32 +00:00
brendan%mozilla.org 25b813e602 Implement SCOPE_IS_SEALED, JS_SealObject, JS_UnsealObject, etc. to support
sealed standard object graphs, as well as to pave the way for optimizations
to object literals (94693, r=shaver).
2003-03-14 05:24:58 +00:00
beard%netscape.com 8ded18bc1b Fix for bug #53593, sr=brendan, r=beard 2003-03-13 05:37:07 +00:00
brendan%mozilla.org dd545c2cb9 Fix JS_[GS]etContextThread return type; add JSCLASS_NEW_RESOLVE_GETS_START (196966, r=shaver). 2003-03-12 20:29:16 +00:00
dbradley%netscape.com 3b01ed08a0 bug 196241 - Refactors script evaluation pre and post logic into a class. r=timeless, sr=brendan 2003-03-12 20:26:21 +00:00
dbradley%netscape.com f24cadaaaa bug 188229 - [AxPlugin] Security policy should be per-domain configurable r=adamlock, sr=alecf NOTE: these changes are not part of the regular build. 2003-03-12 20:05:54 +00:00
dbradley%netscape.com 612926a718 Bug 194043 - Refactor exception logic between IDispatch and XPConnect. r=adamlock@netscape.com, sr=jst@netscape.com 2003-03-12 15:21:06 +00:00
brendan%mozilla.org 67444da79d Give JSScope a proper flags member, without increasing its size (part of 196097, r=shaver). 2003-03-11 20:30:55 +00:00
cls%seawood.org 5738ef26f0 Adding mingw support to js.
Bug #134113 r=brendan
2003-03-11 03:27:07 +00:00
seawood%netscape.com a115192b1a Remove unneeded local win32 jsinterop.obj target that's causing dependency issuees.
Thanks to John Morrison <jrgm@netscape.com> for the patch.
Bug #196633 r=cls sr=brendan
2003-03-11 00:13:44 +00:00
seawood%netscape.com 06d44dbda8 Fix dependency problems with local targets.
Bug #196633 r=jrgm
2003-03-10 05:16:06 +00:00
brendan%mozilla.org 6ef9f45d83 [Second check-in attempt; last night's was backed out, but I claim it was because of bad dependencies or timestamp problems in the Win32 builds. /be] Add cx->lastAtom, a long overdue addition to the per-context GC root set (196461, r=shaver). 2003-03-09 18:16:27 +00:00
jrgm%netscape.com b14fbf087b backout changes for bug 196461, due to startup crash on win32 2003-03-09 05:57:23 +00:00
brendan%mozilla.org 73a99565bc Add cx->lastAtom, a long overdue addition to the per-context GC root set (196461, r=shaver). 2003-03-09 03:12:56 +00:00
cls%seawood.org d1479589a2 Adding mingw makefile changes.
Bug #134113 r=bryner
2003-03-08 05:57:15 +00:00
brendan%mozilla.org ae0c4d3804 Reorder static and const to match tradition; reomve DEBUG_brendan cruft. 2003-03-08 01:28:02 +00:00
cls%seawood.org d1079607df XP code changes required for mingw landing.
Add NS_STDCALL macros for functions that require stdcall declarations.
Add NS_STDCALL_FUNCPROTO macro for stdcall function prototype typedefs.
Bug #134113 r=dougt sr=dbaron
2003-03-07 06:07:56 +00:00
brendan%mozilla.org dcad0d8960 Infrastructure for bug 92773, r=shaver@mozilla.org:
- call the JSRuntime's checkObjectAccess callback, if configured, for each
  get or set that invokes a user-defined function (a user-defined function is
  a scripted or native function object, not a native JSPropertyOp C function).
  The value passed as an in-out parameter in *vp is the function object, so
  the callback could, e.g., clone function objects and configure them with
  different parent objects, if that helped import them into a trust domain.
- Fix a long-standing bug that broke the deprecated, old-style, and rarely-
  used top-level getter/setter function declaration form:

    getter function f()  { return ++x; }
    setter function f(y) { return x = y; }

  We want js_CheckRedeclaration to complain only if a permanent getter is
  being redefined by another getter, likewise for a setter -- but not when
  (as above) a setter is being added to a top-level object for the same id
  as a pre-existing getter (or vice versa).
2003-03-06 19:45:18 +00:00
dougt%netscape.com c5ac287dab removing NS_MT_SUPPORTED define. b=186494, r=dougt, r=seawood 2003-03-05 23:30:22 +00:00
dougt%netscape.com d707cba238 backing out js changes which may have caused tbox test failures 2003-03-05 05:05:39 +00:00
timeless%mozdev.org 727e7a086b Bug 195289 Enable people to use DumpJSStack and friends from nondebug builds if they list themselves as xpc_hackers
r=dbradley sr=jst
2003-03-05 03:12:04 +00:00
brendan%mozilla.org 25e042cd77 Infrastructure for bug 92773, r=shaver@mozilla.org:
- call the JSRuntime's checkObjectAccess callback, if configured, for each
  get or set that invokes a user-defined function (a user-defined function is
  a scripted or native function object, not a native JSPropertyOp C function).
  The value passed as an in-out parameter in *vp is the function object, so
  the callback could, e.g., clone function objects and configure them with
  different parent objects, if that helped import them into a trust domain.
- Fix a long-standing bug that broke the deprecated, old-style, and rarely-
  used top-level getter/setter function declaration form:

    getter function f()  { return ++x; }
    setter function f(y) { return x = y; }

  We want js_CheckRedeclaration to complain only if a permanent getter is
  being redefined by another getter, likewise for a setter -- but not when
  (as above) a setter is being added to a top-level object for the same id
  as a pre-existing getter (or vice versa).
2003-03-05 01:33:19 +00:00
brendan%mozilla.org cf97227e83 Quick followup to patch for 174341. 2003-03-04 00:01:21 +00:00
brendan%mozilla.org 020aef7d75 My take on a patch and great idea from igor@icesoft.com (Igor Bukanov) for bug 195385, r=shaver. 2003-03-01 06:39:51 +00:00
timeless%mozdev.org df64a519af Bug 136264 HTMLCollection Object wrapped by XPConnect has members whose names shadow each other
r=jst sr=jst
2003-02-28 19:07:51 +00:00
brendan%mozilla.org 68cef095fc Fix typo in old comment. 2003-02-28 07:18:40 +00:00
brendan%mozilla.org 6e0c7cdde4 Constipation of (JS|PL)DHashTableOps (195298, r/sr=shaver/alecf). 2003-02-28 07:17:59 +00:00
timeless%mozdev.org d11119926a Bug 136260 ChromeWindow Object wrapped by XPConnect has members whose names shadow each other
r=jst sr=jst
2003-02-28 02:34:01 +00:00
brendan%mozilla.org 4f5a98cd5c - Move left-associative binary tree flattening from the post-order position
in js_FoldConstants where it was added (suboptimally: it worked, but it ran
  so late that js_FoldConstants recursion was not reduced, only js_EmitTree
  recursion), to NewBinary, where it avoids JSParseNode allocations up front
  and reduces recursion in all parse-tree walking.
- This change enables js_FoldConstants to see a very long concatenation of
  string literals as a PN_LIST node, so it can quickly concatenate without
  running afoul of O(n^2) problems inherent in js_ConcatStrings applied to
  two atomized strings (the old way of folding string concatenations, still
  used for any pairwise string literal concatenation).
- Further optimize the first change for the second by having NewBinary set a
  new pn_strcat flag (overlaying the pn_extra field) of the PN_LIST arm of
  the JSParseNode.pn_u union, whenever it sees at least one string literal in
  a concatenation that might be folded (whose operands might all be constants
  of string or number type).
- Notes:
  - only string and number constants are folded (not boolean or null
  constants);
  - only all-constant left-associated binary expression chains are folded,
    so 2 * foo * 3 is not folded using commutativity of * over numbers, nor
    is "hi" + " there" + foo folded to "hi there" + foo.
  - gcc3.2 -O and objdump -x say I added 708 bytes of instructions with this
    change.  I tried to keep it down to what was necessary for common script;
    I don't think JS needs an optimizing-compiler-strength constant folder,
    and I don't think this 1K bloat is too high a price to pay for this fix.
    But I'll certainly work on reducing footprint elsewhere, if I can.
- Bug 174341, r=shaver.
2003-02-27 23:04:46 +00:00
pschwartau%netscape.com 2f97a8c92b Get JS standalone building with Makefile.ref on Mac OSX (195134, r=cls). Not part of browser build. 2003-02-27 21:52:13 +00:00
brendan%mozilla.org 1d7a5aa298 Checking in heap-sort speedup from Igor Bukanov <igor@icesoft.no> (bug 181828, r=me). 2003-02-27 01:43:44 +00:00
mkaply%us.ibm.com 7b82e8d993 #156505
r=dbradley, sr=brendan
If you ask for input, you should fflush after the fprintf
2003-02-26 15:22:03 +00:00
brendan%mozilla.org 7915c3d008 Optimize the foo = '' + bar; case in js_ConcatStrings to do foo = bar -- patch from Tim Lundeen <tlundeen@webcrossing.com>, r=me. 2003-02-26 01:49:32 +00:00
brendan%mozilla.org c68f6d3980 Fix misspelling in comment. 2003-02-26 01:29:55 +00:00
mkaply%us.ibm.com 907ef1cd66 Overzealous -removed a paren 2003-02-25 05:59:50 +00:00
mkaply%us.ibm.com 5e0e0118e6 IRIX bustage - for anyone blaming this code in the future, checkin was Bug 146873 by mstoltz 2003-02-25 05:58:21 +00:00
timeless%mozdev.org 2f675849fa Bug 131275 JS_AddRoot could provide a default root name if you define NAME_ALL_GC_ROOTS
r=brendan
2003-02-25 03:29:24 +00:00
mstoltz%netscape.com db7833ec52 Bug 146873 - Check JS Enabled status before calling from java into JS. r=jst, sr=heikki. 2003-02-25 01:37:47 +00:00
timeless%mozdev.org 0acb5ea8ca Bug 140031 Change jstypes.h to use __declspec instead of _declspec
r=rogerl
2003-02-24 21:37:54 +00:00
timeless%mozdev.org 0e996b7cf4 Bug 186937 JSLL_UCMP missing from !JS_HAVE_LONG_LONG builds
patch by scole@planetweb.com r=brendan
2003-02-24 21:37:23 +00:00
timeless%mozdev.org 8facc552af Bug 188398 JavaScript Engine doesn't work well for dates after 2038 on Windows
r=rogerl
2003-02-24 19:20:24 +00:00
timeless%mozdev.org 4d48249ffe Bug 134728 mozJSComponentLoader::UnloadAll => JSCLAutoContext::JSCLAutoContext Getting service @mozilla.org/js/xpc/ContextStack;1 on shutdown.
r=dbradley sr=brendan
2003-02-23 07:03:11 +00:00
seawood%netscape.com 7f752bf63b Use va_copy if available for VARARGS_ASSIGN
Bug #187180 r=blizzard sr=brendan
2003-02-23 06:59:39 +00:00
brendan%mozilla.org 9c387e3aa5 Cleanup GC_CALLBACK blunder (190813 followup, r=shaver). 2003-02-22 22:56:57 +00:00
brendan%mozilla.org ed7dbec5c1 Fix from pthomas@suse.de (Philipp Thomas) for 64-bit conflict between ptrdiff_t and JS_BIT/JSUint32 (r=shaver, a=dbaron). 2003-02-20 22:23:03 +00:00
mkaply%us.ibm.com a77ccce713 #193611
r=rginda, sr=brendan, a=dbaron
Set the MAX_INTERP_LEVEL for OS/2 to something lower than 1000 (250)
2003-02-19 21:19:54 +00:00
dbradley%netscape.com 74be079a50 Bug 192281 - Error calling IDispatch.setCallback on active-x control. This patch turns on wrapping of JS objects by an IDispatch interface. All changes are outside of Mozilla build. r=adamlock, sr=jst, a=dbaron 2003-02-19 15:19:22 +00:00
brendan%mozilla.org 9fd257d039 Hack around OBJ_ENUMERATE design flaw that requires the property iterator object to root its parent slot (193393, r=shaver, a=asa). 2003-02-18 20:08:21 +00:00
dbradley%netscape.com 92aa9a24b8 Bug 140852 (160602) - String(819187200000) == '8191871:0000' in xpcshell, browser. r=seawood, a=brendan 2003-02-17 13:02:08 +00:00
locka%iol.ie b864237a15 Fix blacklist/whitelist registry code in COM connect. b=193256 r=dbradley sr=alecf a=dbaron 2003-02-16 19:50:14 +00:00
locka%iol.ie 9108ceef0c Fix use of IObjectSafety, cleanup nsAxSecurityPolicy.js & subscribe to hosting flag pref changes, using medium settings as the default. b=191758 r=dbradley@netscape.com sr=alecf@netscape.com a=asa@mozilla.org 2003-02-13 12:17:21 +00:00
timeless%mozdev.org fdd98cf2ed Bug 191675 JS Shell (1.3.1+) Quit() calls exit() without cleaning up.
r=rogerl
2003-02-07 03:51:59 +00:00
timeless%mozdev.org 89f3a5e056 Bug 107002 Makefile.ref doesn't build on rh7.1 with readline
r=rogerl
2003-02-07 03:10:59 +00:00
dbradley%netscape.com e858034c7a Bug 190813 - Browser hangs indefinitely on above URL. r=shaver, a=dbaron 2003-02-06 14:54:59 +00:00
locka%iol.ie 1023aaea93 Fix ClassIsListed which returns wrong value when registry key is missing. b=191131 r=dbradley@netscape.com sr=darin@netscape.com a=asa@mozilla.org 2003-01-30 13:22:01 +00:00
darin%netscape.com 8632233c45 final patch for bug 190326 "crash in nsJARChannel.cpp when launching calendar"
r=dougt sr=alecf a=asa
2003-01-29 06:47:55 +00:00
brendan%mozilla.org b8d97d8b8c Defend against bad EndMuxedDocument calls, and assert in the underlying {JS,PL}_DHashTableRawRemove that entry is live (188744, r=jrgm, sr=ben). 2003-01-17 10:01:50 +00:00
dougt%netscape.com 6508e0713e move nsTraceMalloc out of XPCOM. Factory out stackwalking code into new files. 189221 r=chak, 186585 r=blythe sr=dbaron. Most of this patch was contributed by seawood. 2003-01-15 20:40:06 +00:00
dbaron%dbaron.org 88c249d811 Most of the changes needed for Mozilla to build with gcc's -pedantic-errors option. b=175423 r=seawood sr=bryner 2003-01-13 23:15:42 +00:00
pschwartau%netscape.com e4c62ce9fd Updating JS_GetImplementationVersion() to date of latest JS release (67111). 2003-01-11 02:05:00 +00:00
dbaron%dbaron.org 3ea0dc57d4 Bug 178643: Remove uses of NS_INIT_ISUPPORTS, since it is no longer needed. r=timeless sr=jag 2003-01-08 21:07:51 +00:00
brendan%mozilla.org 4131c7c96c Don't hold the GC lock when calling the GC callback (186592, r=shaver). 2003-01-07 21:56:52 +00:00
dbradley%netscape.com 311c9ba0ac Bug 184491 - Security Manager blocks invoking methods on some IDispatch based objects. r=adamlock, sr=jst 2003-01-03 15:48:20 +00:00
seawood%netscape.com 16e7da2837 nobrainer changes while thinking about mingw support
Bug #134113 r=dmose
2003-01-03 08:58:59 +00:00
seawood%netscape.com d5efcdfb6d Start installing GRE libraries & components into a separate dist/gre directory as part of the default build.
Bug #186241 r=dougt
2002-12-28 01:15:07 +00:00
brendan%mozilla.org e0e5f44e58 Fixed so overriding Function.prototype.toSource doesn't mess up F.p.toString output, which should decompile the function, including nested function expressions (178389, r=shaver). 2002-12-21 02:14:53 +00:00
colin%theblakes.com 0824e190c0 OpenVMS build changes. b=180290. r=cls 2002-12-12 00:08:59 +00:00
locka%iol.ie 73d7ed36cb Share pref settings between com connect and plugin. b=179573 r=dbradley@netscape.com sr=brendan@mozilla.org a=asa@mozilla.org 2002-12-09 21:04:22 +00:00
locka%iol.ie 6838389e7d Share pref settings between com connect and plugin. b=179573 r=dbradley@netscape.com sr=brendan@mozilla.org a=asa@mozilla.org 2002-12-09 20:48:39 +00:00
dbradley%netscape.com f9cb7f63fa Bug 173146 - add support to XPConnect for IDispatch interfaces. Final implementation, this is the result of the code walk through with jband, jst, adam, and rpots. sr=jst, r=adam, a=asa 2002-12-09 19:41:54 +00:00
brendan%mozilla.org eaeb21dee8 Optimize s.match(r) into r.test(s) if possible (no bug, r=shaver). 2002-11-30 06:51:27 +00:00
dbradley%netscape.com 03dbe4ab1c Bug 180182 - crash[@ 0x10101010 - js_GetSlotThreadSafe - JS_GetPrivate]
This was caused by GC occuring within JS_CloneFunctionObject, and the jsval passed in wasn't protected. r=dbradley, sr=jst, patch=jband
2002-11-27 03:55:17 +00:00
brendan%mozilla.org 48668108fd Fix rt->checkObjectAccess call in fun_getProperty to avoid non-object-accessing calls (r=waterson). 2002-11-27 02:11:53 +00:00
brendan%mozilla.org 9f86c8f220 Comment nit-picks (comment changes only). 2002-11-26 00:33:08 +00:00
brendan%mozilla.org 5380b54b0e Relax comments restricting Operate calls from enumerator to allow lookups (no bug, comment change only). 2002-11-26 00:32:07 +00:00
sicking%bigfoot.com 8d87d5e285 Bug 8929: Kill NS_COMFALSE
r=peterv sr=bz
2002-11-25 11:21:22 +00:00
dbradley%netscape.com 86ca2cd40a Bug 173146 - This is the third patch in the series for IDispatch support. The bulk of the changes are comments and formatting, and some better error handling. r=adamlock, sr=jst and the code review group. NOTE: The bulk of this code is still not part of the build and the reviews reflect the XPConnect portions only. 2002-11-22 23:04:53 +00:00
pschwartau%netscape.com 6d450f7126 One-line fix for bug 62187, "JS shell producing too much console output on Mac" (r=waldemar, r=rginda). Without this fix, running the test driver is impossible on Mac 9. [NOT PART OF BROWSER BUILD]. 2002-11-19 20:53:17 +00:00
pschwartau%netscape.com 7ae2d05ef4 Bringing Windows makefile for JS shell in sync with recent changes made to the browser build and to the js.mak(VC++) build of the JS shell; see bug 160592. [DOES NOT AFFECT BROWSER BUILD]. 2002-11-19 20:29:38 +00:00
timeless%mozdev.org b6b14bf40e Bug 180783 nsDependentCString filename(PR_GetEnv("MOZILLA_JS_PROFILER_OUTPUT")) => ASSERTION: nsDependentCString must wrap a non-NULL buffer: 'aPtr'
r=dbradley sr=bz
2002-11-19 05:06:27 +00:00
timeless%mozdev.org 384066e98e Bug 58221 don't use strlen to check if a string is of length 0
patch by aaronl@vitelus.com r=timeless sr=bz
2002-11-17 05:16:49 +00:00
dbradley%netscape.com 1621e27d5b Bug 180098 - "--enable-activex-scripting" option seems to break composition window. Error was in enumeration when IDispatch logic was turned on. r=adamlock, sr=jst 2002-11-15 13:57:16 +00:00