1740ef44a7
Respect addProperty failure (148899, r=khanson, sr=shaver).
2002-06-04 18:13:55 +00:00
b20ac26ede
Fixes XPCONNECT_STANDALONE bustage - see bug 121438 for details
2002-05-31 13:51:08 +00:00
32f096fe9c
patch to speed up array.sort of strings by 40% r=brendan, sr=shaver
2002-05-27 05:53:57 +00:00
f981dd90ae
Don't emit JSOP_DEF{VAR,CONST} unless at top-level or in a heavyweight function (146596, r=jband, sr=shaver).
2002-05-26 00:33:23 +00:00
5edfcc1ec0
146210: Add Date.now to mirror JS_Now in script. r=rginda, sr=brendan.
2002-05-24 01:55:15 +00:00
412d54c9d6
Fix Mac bustage from brendan's checkin.
2002-05-22 00:53:12 +00:00
f1f6f1398e
Don't fail to call OBJ_CHECK_ACCESS for o.__proto__ = x, with enough mode bits to say what's happening (143369, r=shaver, sr=jst).
2002-05-22 00:06:40 +00:00
abb8bd42b5
Bug 85267 - clear internal jsdtoa caches when closing down the js engine. Patch by steve.taylor@ndsuk.com r=khanson@netscape.com, sr=brendan@mozilla.org
2002-05-20 20:07:48 +00:00
6778e86241
Followup to checkin for bug 120992. It asserted in an edge case which this fixes. r=mozbot
2002-05-18 17:33:37 +00:00
911958f715
Bug 120992 - avoid expensive locking when converting numbers to strings. Now we look once for the whole operation instead of over and over again any time we access a shared structure. r=khanson@netscape.com, sr=brendan@mozilla.org.
2002-05-18 06:21:49 +00:00
2fdb4e3b04
Checking in jband's fix for bug 143853. Make sure to refresh the cached prototype JSObject when a context is reinitialized and a global object is reused between pages in a window. r=dbradley@netscape.com, sr=jst@netscape.com
2002-05-17 22:36:54 +00:00
c2c188c8cc
Checking in khanson@netscape.com's patch fixing GMT offset day-overflow (65671, r=mccabe, sr=jband).
2002-05-15 19:44:34 +00:00
c683a217ab
Fixes mozilla/strings requiring unfrozen nsCRT class. patch by scc, r=dougt, sr=jag, b=136756
2002-05-15 18:55:21 +00:00
4293400e69
Because js uses static data structures and is linked into other shared libraries, js must be built as a shared library.
...
Thanks to Colin Blake <colin@theblakes.com> for the original patch.
Bug #127442 r=bryner
2002-05-15 01:37:07 +00:00
6fd5862e6e
fixes bug 142870 "nsIFile should use UCS-2 instead of UTF-8"
...
r=dougt sr=alecf
2002-05-07 23:07:19 +00:00
3f518e626f
Avoid toSource bloat in e.stack; fix related crash due to unrooted errObject in js_ErrorToException (139316, r=rginda, sr=shaver).
2002-05-07 08:34:51 +00:00
4d9e715520
Fix old bogus static dollarStr thread-safety bug (142651, r=jband, sr=shaver).
2002-05-07 08:29:39 +00:00
7f766daf07
Adding the functions JS_LockGCThingRT and JS_UnlockGCThingRT, which are variants that only require a JSRuntime, not a JSContext. Converted some internal callers to use the new interface. Bug 141356, patch from shaver, r=bryner, sr=brendan.
2002-05-04 05:28:09 +00:00
9508d71adc
fixes build bustage with --enable-xpctools
...
patch=dmose r/sr=darin
2002-05-03 19:04:14 +00:00
bf8e80d0d1
Fix js_SetProtoOrParent to handle non-native and native-but-incompatible-class-change proto-setting scenarios (140164, r=khanson, sr=jband).
2002-05-02 20:59:35 +00:00
30322b2b64
Don't 'constant-fold' if(false)var foo=42 into nothing (140974, r=khanson, sr=shaver).
2002-04-30 23:28:34 +00:00
e554d83626
fixes bug 129279 "nsIFile unicode/utf8/ascii task"
...
r=dougt sr=alecf
2002-04-27 05:33:09 +00:00
8c580862e8
fix bug 139243. Remove reference count based rooting of JS objects of wrapped natives and replace it with GC-time marking. This is a pretty good DOM performance win. r=dbradley sr=jst.
2002-04-26 06:08:35 +00:00
d4e55c2951
Purge the deflated string cache entry, if any, before finalizing external strings (140070, r=jband, sr=shaver).
2002-04-25 22:55:03 +00:00
dd8c0773df
Reverse of Patch 66477 for bug #121744 . Bug #131348 , r=shaver, sr=brendan.
2002-04-25 20:33:33 +00:00
d817cc069a
fixes memory leak in jsdtoa.c bug 138666, sr=brendan, r=khanson
2002-04-24 23:36:48 +00:00
97531eb92c
Add make -j support to win32 gmake builds.
...
Bug #132848 r=bryner
2002-04-19 02:42:28 +00:00
3f8594ea0e
Forgot to #ifdef JS_THREADSAFE last change (no effect on Mozilla build; fixes broken standalone/single-threaded JS builds, sanity-sr=shaver).
2002-04-04 18:02:59 +00:00
e424e842c6
Followup fix to synchronize with a nt-last-destroy-context racing GC when unpinning pinned atoms (133773, r=shaver, sr=jband, a=rjesup).
2002-04-03 20:38:39 +00:00
8f9516a234
Fixes an installer crash on pre-PCI PowerMacs, bug # 125411. sr=sfraser, r=sdagley, a=asa
2002-04-02 23:27:55 +00:00
090c63d423
More cruft removal. Remove platform.h xp_debug.h xpassert.h & xp_trace.h from the build.
...
Replace currently unused references to XP_ASSERT with one of NS_ASSERTION & friends.
Bug #38061 r=bryner sr=alecf a=tor/drivers a=putterman/adt
2002-04-02 07:57:02 +00:00
0555817eef
Fix next-to-last vs. last context GC race, plus ClaimScope vs. js_DestroyContext race; removed js_ForceGC from the FRIEND JS API (133773, sr=jband&shaver, a=asa).
2002-04-02 04:23:21 +00:00
6cdcd7ef3a
Convert users of nsAReadable[C]String and nsAWritable[C]String typedefs to [const] nsA[C]String. b=131899 r=scc sr=jag a=asa
2002-03-24 00:16:18 +00:00
ac2c08044b
Fix JSOP_DEFFUN to define functions in global code permanently, and not permanently in eval code (131964, r=khanson, sr=shaver, a=scc).
2002-03-23 17:34:24 +00:00
74c88d8460
Cope with OOM creating rt->atomState.table (131246, r=scole@planetweb.com, sr=jband, a=scc, thanks to scole for finding these OOM recovery bugs).
2002-03-23 07:38:08 +00:00
51645833cc
patch #75132 changes for bug#132230 (str_unescape and js_str_escape in jsstr.c ignore OOM errors) Written by Steven Cole, sr=bendan, r=khanson, a=asa. Handles an out of memory conditions more gracefully.
2002-03-22 21:24:46 +00:00
4cf8850876
Updating JS_GetImplementationVersion() to date of latest JS release (67111).
2002-03-22 20:43:47 +00:00
6ce09c1621
fix bug 132713. Use the iid for nsIException to detect exceptions rather than nsIXPCException. This allows for better error reporting of natively implemented exceptions. r=dbradley sr=jst a=scc.
2002-03-22 19:14:57 +00:00
40642fc34c
130405. Makes the xpcom glue library usable. Updates XPCOM Samples. Adds requires lines to a few implict string/category consumers sr=alecf, sr=brendan@mozilla.org a=asa@mozilla.org
2002-03-21 23:43:21 +00:00
16d78379aa
fix bug 125084. Add support for empty arrays in nsIVariant. This is expressed as atype id rather than as a regular array with no elements. Includes changes to xpconnect to support mapping to empty JSArrays, xpconnect tests, and SOAP. Patches are from jcorwin@us.ibm.com and rayw@netscape.com. r=jband sr=brendan a=asa
2002-03-21 23:21:54 +00:00
c2ee1d2afb
Fix js_FinalizeStringRT so it doesn't depend on a dependent string's base, which may already be finalized (132088, r=scole, sr=shaver, a=asa).
2002-03-21 09:06:00 +00:00
5bf172e81d
Patch 74900 for bug #131817 (OOM in AllocSrcNote causes crash in js_NewSrcNote). The patch was contributed by Steven Cole (scole@planetweb.com), sr=brendan, r=khanson, a=scc
2002-03-20 05:45:12 +00:00
94a74d7db5
Defend against early OOM when finishing JSDHashTables (131815, r=scole, sr=jband, a=asa).
2002-03-20 02:22:50 +00:00
a0f36bedac
Fix ChangeScopeTable gross dimensioning bug (131904, r=scole&khanson, sr=jband, a=asa).
2002-03-19 23:25:17 +00:00
cdad9798f3
Fix 'var arguments;' in a function by specializing to JSOP_ARGUMENTS properly (131510, r=khanson, sr=jband, a=asa).
2002-03-19 22:51:48 +00:00
dec943eb10
Bug 106386 rid source of misspellings
...
r=db48x sr=blake a=asa
2002-03-19 04:30:17 +00:00
0c6a6e0115
Forgot to check this in yesterday: fix all engine-defined getters and setters to have no slot [to be JSPROP_SHARED] (130970, r/sr=shaver&jband, a=asa).
2002-03-17 11:01:34 +00:00
9402b61613
Null test to handle recovery from OOM under js_NewContext reported by scole@planetweb.com (r/sr=jband, a=me).
2002-03-16 23:19:35 +00:00
1e126a67d9
Restore non-bogus assertion removed in last rev, and fix the case that tripped it in js_ChangeScopeProperty, by not removing [and maybe freeing sprop->slot] before re-adding sprop; also fix all engine-defined getters and setters to have no slot [to be JSPROP_SHARED] (130970, r/sr=shaver&jband, a=asa).
2002-03-16 04:51:29 +00:00
5ab837c899
bug 129519, "JS_GetPropertyDesc gives up too easily", r=jband, sr=shaver, a=bren
...
dan
Make JS_GetPropertyDesc degrade gracefully instead of fail outright when we run
into a problem fetching the property value.
2002-03-16 01:56:19 +00:00
e67d7ec049
bug #130991 (Out-of-Memory in jsshell causes assertion failure in jsobj.c)
patch by Brendan, sr=jband, r=khanson, a=asa
2002-03-15 06:22:00 +00:00
78b2988012
fix bug 130139. We are iterating an array of pointers not an array of objects. r=dbradley sr=brendan a=asa.
2002-03-15 04:10:25 +00:00
be5d94c4a1
bug #130711 , r=khanson, sr=brendan, a=asa, memory leak in JS_dtoa
2002-03-15 00:11:44 +00:00
40ba9a0a2a
Remove bogus assertion, tripped by js_ChangeScopePropertyAttrs calling js_AddScopeProperty (130970, r=shaver, sr=jband, a=shaver).
2002-03-14 21:59:16 +00:00
6f45f575a4
Fix JS/PL_DHashTableEnumerate to compress or shrink the table after enumeration using the same logic as ADD and REMOVE use (120953, r=dbaron, sr=shaver, a=asa).
2002-03-14 21:55:08 +00:00
0a6eb6a388
Bug 61314 Makefile rules for "jsmath.o" and "jsmathtemp.o" cause stale object files, obscure bugs
...
patch by mang@subcarrier.org r=rogerl sr=brendan a=asa
2002-03-14 21:41:09 +00:00
62bc0d659b
No strict warnings without the strict option, and other js1.5/mozilla1.0 tidying (129972, r=shaver, sr=jband, a=asa).
2002-03-14 00:14:48 +00:00
d3a05a01eb
Fix Array.prototype.sort to follow ECMA and win perf by not (re-)defining length on the 'this' object (130451, r=shaver, sr=jband, a=asa).
2002-03-14 00:10:31 +00:00
05d2d349ce
khanson@netscape.com's patch to switch from QuickSort to heap-sort, plus a crucial ECMA-purity/property-tree-perf fix to InitArrayObject (to SET rather than DEFINE; bug 99120, r=waldemar, sr=shaver&brendan, a=asa).
2002-03-13 01:50:13 +00:00
21978efea8
Missing JSFunction API (129545, r=shaver, sr=jband, a=asa).
2002-03-12 21:58:27 +00:00
5074f6ebd1
Don't forget to lock obj in js_AddNativeProperty and js_ChangeNativePropertyAttrs (130137, r=jband, sr=shaver, a=roc+moz).
2002-03-12 07:17:30 +00:00
cb6ae8ad3e
fix bug 129697. Don't leave a dangling (to crash later) JS root if JS_NewObject fails while creating a wrapped native. r=beard sr=jst a=asa (for trunk and 0.9.9 branch).
2002-03-09 03:01:44 +00:00
75a8745d3d
Avoid MSVC bogo-warning.
2002-03-09 00:55:26 +00:00
efe08cd95c
Fix longstanding bug where watchpoints didn't work with JSPROP_SETTER (127243, r=rginda, sr=shaver, a=dbaron).
2002-03-08 22:46:13 +00:00
cb2b8b58fb
Fix for bug #123920 , prevents reentering the JS engine during Java object JS wrapper finalization, by deferring calls into Java at the end of the garbage collection cycle and storing the wrappers in a linked list, which gets processed by a GC callback.
...
r=jband, sr=brendan, a=asa
2002-03-08 22:37:40 +00:00
d97d6d301f
This patch replaces the bogus USE_AUTOCONF ifdef + platform ifdef tests with a
...
simple HAVE_LOCALTIME_R test. -DHAVE_LOCALTIME_R has been added to the
js/src/config/<platform>.mks as necessary.
bug #128556 r=brendan sr=shaver a=asa
2002-03-08 01:42:32 +00:00
36a0ec70a8
bug 128057, r=brendan, sr=shaver, a=asa
...
check for JS_GetPropertyDesc failure in JS_GetPropertyDescArray
2002-03-06 22:12:38 +00:00
f1a6738b6c
fixes bug 124042 "support internationalized URIs" r=dougt, sr=alecf, a=asa
2002-03-06 07:48:55 +00:00
4047cc7888
Not part of build, for the js testsuite only: add clone shell function to clone a function object (for 127557 regression testing, a=shaver).
2002-03-06 04:45:32 +00:00
e56ba03add
Universal config file for Darwin.
2002-03-05 15:52:37 +00:00
3dbc5b854c
fix for bug #59686 , a=asa@mozilla.org, r=rogerl, sr=shaver.
2002-03-05 15:09:05 +00:00
dac4fdf779
Bug 106864: xptcall support for the ARM platform, and mild jsnum.h arm-define
...
tweaks. Patches from Mark Crichton <crichton@gimp.org>, Russell King
<rmk@arm.linux.org>, Jeroen Dobbelaere <jeroen.dobbelaere@acunia.com>.
r=brendan on jsnum.h, sr=shaver, a=asa for trunk and branch.
2002-03-05 06:35:50 +00:00
8872795d15
Fix for bug 125465. r=jag, sr=jband, a=asa. Add AString, ACString, and AUTF8String support to XPCOM's variant data type. Also, add forward declarations of AString and ACString to nsrootidl.idl.
2002-03-05 02:22:02 +00:00
50d5d72d66
Needed for Mac OS X 10.1.3 / Darwin 5.3
2002-03-04 16:14:41 +00:00
be783ec091
Runaway recursion safety for Exception and js_ErrorToException.
...
The latter no longer calls the former via js_ConstructObject, to
minimize failure opportunities when converting an error into an
exception. Also, the 'stack' property of exception objects
faithfully preserves Unicode in function names and decompiled
actual argument sources.
Also, fun_xdrObject was reversing property order when encoding;
fixed to maximize property tree compression gain.
Also^2, js_ConstructObject was not handling success+not-found
return from FindConstructor.
bug 127136, r=rginda, sr=shaver, a=chofmann
2002-03-03 00:59:16 +00:00
4f6d566527
Fix topcrash regression from bug 62164: keep track of duplicate kids below the root ply of the property tree, so we can fix up each dup's parent link (128258, r=shaver, sr=jband, a=asa).
2002-03-01 23:29:54 +00:00
32d76a128b
Fix call_enumerate to reflect compiler-created properties from function to its call object (127557, r=jband, sr=shaver, a=asa).
2002-03-01 23:25:34 +00:00
8ecde6229e
Fix GC_MARK_DEBUG #ifdef'd code (not part of default build; r=polarbear, sr=shaver, a=dbaron).
2002-03-01 05:22:26 +00:00
72469311ce
part of bug 109652, "Venkman needs to profile"
...
sr=brendan, r=peterv, a=shaver
add JS_Now to the jsapi, and JS_GetScriptVersion to jsdbgapi
2002-02-27 09:17:21 +00:00
2f1de24dbc
Fixing Mac bustage; add a bunch of casts between PRUnichar* and jschar* (why isn't jschar a wchar_t?). r=nisheeth
2002-02-26 03:01:03 +00:00
46db685f9a
Fix for bug 84186. r=dbradley, jag. sr=jband. Added support for CStrings, UTF8String, and AStrings to XPIDL. Also added the -t cmd line flag to XPIDL and XPTLINK that allows us to generate output for a specific XPT version. The XPT version number is being upped from 1.1 to 1.2 with this checkin.
2002-02-26 01:41:42 +00:00
bcc02f397b
Fix stale comment about duplicate formal params (a=drivers, comment-only followup to 62164's big patch).
2002-02-24 00:26:57 +00:00
032c9bde1f
Comment change only, quick followup to last humongous checkin.
2002-02-23 04:44:40 +00:00
57483a3082
Major footprint/perf win: share property state using a tree whose root represents empty scopes, and whose non-root nodes represent scopes with properties added in order from the root to that node; to use double hashing to map these tree-paths for per-scope property lookup; and avoid locking where possible through immutability and mostly-benign&rare race tolerance (62164, r=shaver, sr=jband).
2002-02-23 03:49:27 +00:00
78dddf0f3b
fix bug 126333. reduce shutdown leakage by releaseing Components object members. r=dbradley sr=alecf
2002-02-20 07:03:27 +00:00
e2486ebfbe
r=thesteve, sr=brendan, bug @119909 - crashing RegExp with > 10 backrefs.
2002-02-20 05:43:26 +00:00
5372835bdb
Bug 119382 - attribute nsIIDPtr IID considered harmful - This patch makes it an error to declare an attribute named IID or a method named getIID. r=jband, sr=jst
2002-02-19 14:48:00 +00:00
db81563416
Bug 121938 - XPCWrappedJS::GetNewORUsed does not look for inherited interfaces when matching interfaces. r=jband, sr=jst
2002-02-19 11:24:21 +00:00
33746feb4c
Fix bad old js_ValueToInt32 to have an int-jsval fast path, to avoid f.p. inaccuracy (126159, r=shaver, sr=jband).
2002-02-19 00:02:29 +00:00
17e692e072
bug 123770. extend xpconnect to use dynamic 'additional' interface info managers to support dynamic type info. Also, abstract out the Components.interfaces object and make the items use a bit less memory. r=dbradley sr=vidur
2002-02-18 23:47:30 +00:00
f2dabe8b2c
Followup fix for ECMA-262 15.10.6.2 step 6, plus warning fixes (76717, r/sr=shaver&jband).
2002-02-18 23:43:28 +00:00
3ae0856e3e
Removing unneeded LIBS setting. LIBS isn't used to link libraries.
2002-02-18 23:19:49 +00:00
1ef661eab1
fix bug 125430. Unbust Components.returnCode by not setting the JSPROP_READONLY bit on the read/write property. r=bzbarsky sr=jst
2002-02-15 03:33:37 +00:00
90a78e1012
Bug 121526 - Need to expose the security info pointer from the nsIXPConnectWrappedNative interface. r=jst, sr=jband
2002-02-15 03:11:15 +00:00
5d58b5146e
bug 124474, "JS_IsNativeFrame should return !fp->script"
...
also includes typo fix from bug 124474
sr=shaver, r=jband
2002-02-14 07:52:02 +00:00
aba15fbd2d
Construct error objects to consolidate their default property setting; this entails adding JS_ConstructObjectWithArguments (123177, r=rginda, sr=shaver).
2002-02-14 07:25:34 +00:00
7a83c28af1
add regression test for bug 125430. NOT PART OF THE BUILD
2002-02-14 06:01:54 +00:00
41da93a53b
Bug 124251 - Fix for some DOMStrings leaking when using -mail. r=dbaron, sr=jst
2002-02-14 00:12:44 +00:00
3cafa8d101
Fixing bug #103351 - using undefined as RegExp value for String.replace.
...
r=khanson, sr=brendan
2002-02-12 19:12:42 +00:00
1e4bb807d7
Hack special frame flags into fp when compiling under JS_Evaluate*InStackFrame, so useless expressions don't get optimized away (92087, r=rginda, sr=jband).
2002-02-12 07:44:48 +00:00
3375f6f6f0
Add exception object stack property, string valued; clean up NULL/VOID mixup, missing OOM test, etc. (123177, r=rginda, sr=shaver).
2002-02-09 06:28:58 +00:00
fa54cd9d8a
Remove bad old regexp object locking across huge, deadlock-prone, so-not-critical sections (123930, r=rogerl, sr=jband&shaver).
2002-02-09 00:14:31 +00:00
a219a25636
Fix recently-uncovered ECMA violation: lastIndex must use double storage (124339, r=rogerl, sr=shaver).
2002-02-08 23:59:49 +00:00
a22a44116b
Make regexp.lastIndex threadsafe by multiplexing it by thread-id in each r.e. (76717, r=rogerl, sr=shaver).
2002-02-08 02:46:03 +00:00
a911b05d03
Avoid all gc <> obj-lock deadlocks by letting the gc run lock-free, safe since all requests are suspended or ended when the gc runs (114712, r=shaver, sr=jband).
2002-02-08 01:55:30 +00:00
2edcffaaed
Quick followup to last checkin -- warning fixes (d'oh!)
2002-02-06 07:56:29 +00:00
2653cdfd3a
Fix call and new expression parsenode beginning line number, where the arglist is on another line from the function or constructor, to avoid a bogus assertion in UPDATE_LINENO_NOTES, which is now replaced by better comments and an avoided gratuitous store to cg->currentLine (123371, r=rginda, sr=shaver).
2002-02-06 07:39:20 +00:00
2c41247d14
Fix failure to check for JS_InitClass failure, reported by Eric Brueggemann <eric_brueggemann@yahoo.com>, r=shaver, sr=mozbot.
2002-02-06 07:01:37 +00:00
dcbd4a1a96
Add better support to xpconnect for converting between variants and arrays. bug 120881. r=dbradley sr=vidur. Trivial fix for case where we were hitting an assert at xpcom shutdown. bug 121536. r=dbradley rs=vidur.
2002-02-05 06:45:08 +00:00
f0115acfab
unbust XPCONNECT_STANDALONE build. No change to default build. bug 121438. rs=jband.
2002-02-03 20:05:22 +00:00
8f2816f49f
Do not place the .res file inthe srcdir
2002-02-01 04:50:32 +00:00
fdf6f6bfee
Fix for(i in null) and for(i in undefined) to error per ECMA, rather than iterating zero times (121744, r=khanson, sr=shaver).
2002-01-31 07:30:05 +00:00
d133d4956f
Making the nsModuleComponentInfo data const; bug 74803. r=dp, sr=waterson
2002-01-30 21:14:20 +00:00
c50ca402ad
1. Converts callers of nsIComponentManagerObsolete to use
...
nsIComponentRegistrar.
2. Converts callers of nsComponentManager::AutoRegister to use
nsIComponentRegistrar's autoRegistrar method.
3. Add nsIComponentRegistrar implmentation to nsComponentManagerImpl.
4. Rearrange nsComponentManager.cpp so that related methods are in the same
place.
5. Added a C-style function NS_GetComponentRegistrar so that getting the
registrar is easier in some places.
6. Added a nsISimpleEnumerator interface on PLDHashTableEnumeratorImpl. in
this way, the same base class can support both old style and new style
enumerations.
7. Fixed a nasty bug where unregistring factories will leave the contract id
hash with a dangling pointer. Now, when unregister is called we search the
contract id hash for entries which have the given doomned cid and remove them.
Bug 115853. r=dp@netscape.com , sr=rpotts@netscape.com
2002-01-29 21:22:13 +00:00
1d1d042dce
Checking in rogerl's fix for double-reports of compile-time errors (120197, r=brendan, sr=shaver).
2002-01-27 04:42:37 +00:00
39ba085fc5
Fix "static" build support for win32 gmake builds.
...
Bug #58981 r=bryner
2002-01-25 03:20:51 +00:00
b8f48f2f9f
Make 'too much recursion' errors be exceptions #if JS_HASH_ERROR_EXCEPTIONS (121658, r=rogerl, sr=shaver).
2002-01-25 00:49:52 +00:00
f303a64843
Checking in patch from Daniel Bratell <bratell@lysator.liu.se>, r=khanson, sr=me.
2002-01-25 00:47:11 +00:00
d8f1df22cb
Fix bug #104375 . Bug in String.replace() with $n where n > # parens.
...
r=khanson, sr=brendan.
2002-01-24 19:27:21 +00:00
755aed19ba
Fixing bug 120901. Eliminating unnecessary QI call in XPCWrappedNative::GetNative(). r=dbradley@netscape.com, sr=jband@netscape.com
2002-01-24 06:03:25 +00:00
cc17fd30b7
Fixing bug 120718. Making XPCReadableJSStringWrapper not allocate non-shared buffer handles, this caused a malloc to happen for every XPCReadableJSStringWrapper that was accessed in mozilla, ~400 allocations at startup. r=dbaron@fas.harvard.edu sr=jaggernaut@netscape.com.
2002-01-24 05:59:41 +00:00
af5d807ca1
Bug 120990 - make conversion from integer to string faster by not using sprintf. r=khanson, sr=brendan
2002-01-24 05:24:39 +00:00
dcbcdb0799
Inline-eliminate js_Emit{Break,Continue} and move SRC_CONTINUE generation to the right place (120571, r=rogerl, sr=jst&jband, a=asa).
2002-01-22 22:06:10 +00:00
0428dc97e3
Fix for bug 120451 -- clean up the sleep Q entry on quit, so that we don't leave a bad entry there and hose the machine some time later. r=sdagley, sr=beard, a=blizzard
2002-01-18 21:40:13 +00:00
5dfeb6c1a2
Configuration to build on Mac OS X 10.1.
2002-01-17 06:39:04 +00:00
50295b2702
Fixing part of bug 118933. Speeding up the global resolve hook in the DOM JS helpers by eliminating the use of nsHashtable and nsStringKey in favor of pldhash. Adding global HashString(const nsA[C]String&) functions that calculate a hash code from a nsA[C]String, code by dbaron@fas.harward.edu. Also checking in brendan@mozilla.org's fix for a type error in the jsdhash/pldhash API found while working on this. sr=brendan@mozilla.org, r=peterv@netscape.com
2002-01-17 04:08:14 +00:00
362b589445
Missing part of fix for bug 117153, fixing threadsafety problems with shared empty buffer handles. b=120300 r=jag sr=alecf a=blizzard
2002-01-17 00:07:15 +00:00
be480d0e2d
Fixed RegExp bug #72964 , r=khanson, sr=brendan
2002-01-16 06:47:06 +00:00
08c2912aac
Fixed RegExp bugs #105972 & #87231 . r=khanson, sr=brendan.
2002-01-16 06:06:34 +00:00
71be933776
Simplify #ifdefs in JS_Assert to default to fprintf (45673, debug only, r=rogerl, sr=me, patch from jband).
2002-01-16 05:17:58 +00:00
958c05b6cc
Checking in patch for rob1@rekl.yi.org, r=khanson, sr=brendan (but neither file is part of the build).
2002-01-16 05:12:53 +00:00
e56e20b1dc
Use dynamic library & xpidl_module names from nmake build for win32 gmake builds for compatibility.
...
bug #58981 sr=jag
2002-01-16 05:10:09 +00:00
007db84d18
Checking in patch for Steven Cole <scole@planetweb.com> to fix js_DoubleToECMAInt32 (120083, r=khanson, sr=brendan).
2002-01-16 00:17:16 +00:00
81bbc17978
Bug 111068 - SyncXPCContextLists uses of XPCContext::Mark is not thread safe. patch by jband, r=dbradley, sr=jst
2002-01-15 22:49:46 +00:00
66c9536fcc
fix bug 119387. Add a non-allocating IID version of the frequently used getIIDForParam method. r=dbradley sr=jst
2002-01-14 20:02:14 +00:00
cfc984cc6c
Bug 119743 - autocomplete doesn't work in gcc 2.96 -O2 builds. r=hixie, sr=brendan.
2002-01-14 06:45:21 +00:00
d24a0fe0ca
Updating JS_GetImplementationVersion() to date of latest JS release (67111).
2002-01-11 23:18:39 +00:00
28d27d47f8
Remove Call constructor (118732) and fix old bug in Function where it failed to convert args to strings (118849; both r=timeless, sr=shaver).
2002-01-10 08:51:01 +00:00
7bdffe5186
Bug 117276 - regression on gcc 2.95.x debug builds. Since we don't
...
understand exactly what the problem is here, and it does not seem to
happen on gcc 2.96, and 2.95 doesn't have the alias optimizations that
called for the new code in the first place, just make 2.95 use the old
code. r=timeless, sr=brendan.
2002-01-09 19:23:37 +00:00
85a52892dc
Don't use C++ comments in C files.
2002-01-08 02:29:11 +00:00
5965318acb
C++ side of fix for bug 100212 - move consumers of nsIFile.URL into nsIIOService. r=dougt, sr=darin
2002-01-08 01:31:25 +00:00
8d24cf46e4
Checking in patch written by balleysson@bigfoot.com based on tenthumbs@cybernex.net's sketch, to avoid mallocs in js_strtod (115627, r=khanson, sr=me).
2002-01-07 23:42:25 +00:00
56d43516cb
Group date and time fields in toString's result so as to work with PR_ParseTimeString (118266, r=khanson, sr=shaver).
2002-01-07 11:09:04 +00:00
2ba4c7d57c
Null pointer defense in FUN_CALLER fun_getProperty case (117307, r/sr skipped, trivial).
2002-01-01 21:12:21 +00:00
8962b0ec01
Fix old bug introduced by rev 3.2, where this() crashes due to uninitialized obj register (116228, r=timeless, sr=shaver).
2001-12-27 08:47:33 +00:00
61a97fd94a
Bugzilla Bug 106386 rid source of these misspellings: persistant persistance priviledge protocal editting editted targetted targetting
...
r='s from many people. sr=jst
2001-12-23 23:23:41 +00:00
3a73c9be8d
fixing dumbass error. This could should have returned null if obsoleteManager was null, not the inverse. r=jband@netscape.com
2001-12-20 00:26:28 +00:00
de1592438e
Don't double-drop a property when strict-warning about duplicate formals (115436, r=khanson, sr=jband).
2001-12-19 02:10:26 +00:00
21b53b29a8
fix for bug #115545 . r=sdagley, sr=sfraser
2001-12-19 00:23:05 +00:00
849d297364
nsIComponentManager API Changes (bug 98553)
...
a) create a new nsIComponentManager with only four functions on it:
CreateInstance CreateInstanceByContractID GetClassInfo GetClassInfoByContractID.
b) rename the old nsIComponentManager to nsIComponentManagerObsolete.
c) fixes callers which use to access the nsIComponentManager for component
registration functionality. These callers will temporary use the
nsIComponentManagerObsolete interface.
d) Create a new API NS_GetComponentManager() which mirrors the
NS_GetServiceManager()
e) Perserves the old NS_GetGlobalComponentManager(). Note the cast usage.
r/sr = rpotts@netscape.com alecf@netscape.com brendan@mozilla.org
2001-12-19 00:12:41 +00:00
45bfbf0658
Landing the rest of the win32 gmake changes:
...
* Adds Makefile.ins to win32 specific dirs
* Adds WINNT ifdefs to Makefile.ins
* Causes NSPR to be compiled with --with-mozilla
* Misc general Makefile.in cleanup
Bug #58981 r=mcafee
2001-12-18 09:14:29 +00:00
ee12112b29
Remove timer from REQUIRES, the makefile.win edition
2001-12-16 18:08:20 +00:00
15273ea50a
bug 78611. New XP threadsafe timer implementation. r=dougt sr=brendan
2001-12-16 06:13:17 +00:00
51fa852613
Make OS/2 use the win32 build changes and finish LIB_PREFIX substitution.
...
Thanks to Javier Pedemonte (pedemont@us.ibm.com ) for the patch.
Bug #58981 r=cls
2001-12-14 23:49:04 +00:00
74151505f6
Long ago, mccabe copied the principals setting code in Function from late to early in the function (should've moved, but no one noticed); last month, I fixed bug 90546 in the original site but not in the copy; this change eliminates the duplication, moving the fix up to the copy's site (90546, r=jst, sr=jband, a=chofmann).
2001-12-13 00:27:20 +00:00
7b2b1b44ad
Fix try/finally and try/guarded-catches/finally hard cases (104077, r=jband, sr=shaver).
2001-12-12 05:00:29 +00:00
140631bce2
Checking in rogerl@netscape.com's patch for bug 100199 (r=rginda, sr=me).
2001-12-12 04:27:50 +00:00
f7b755c4e0
Prevent XPCReadableJSStringWrapper from having a null buffer handle. b=112496 r=shaver, jag sr=jst
2001-12-08 22:49:05 +00:00
944c82b990
Bug104843 - Added requires to makefile
2001-12-07 13:34:48 +00:00
16e4f2d4d9
Drop optimization level for select files on Irix when compiling with MIPSpro to workaround suspected compiler bug.
...
Thanks to Nick Blievers <nickb@adacel.com.au> for the patch.
Bug #79562 r=brendan sr=cls
2001-12-07 09:12:16 +00:00
5586bf8bdc
Bugs #100199 & 78156. r=khanson, sr=brendan. Fixing multiline matching and
...
pathological character class cases in regular expression handling.
2001-12-05 23:58:51 +00:00
eb9c1aae74
not part of default build
...
bug 111365, JS standalone doesn't build on solaris.
point AS to /usr/ccs/bin/as
2001-12-05 19:32:16 +00:00
c59566e1a8
not part of default build
...
bug 111365, JS standalone doesn't build on solaris.
point AS to /usr/ccs/bin/as
2001-12-05 00:31:55 +00:00
c8994a7ac0
Initial landing of win32 support for autoconf builds.
...
Bug #58981 sr=alecf
2001-12-04 21:17:48 +00:00
824e7d22ca
Make nsSharable[C]String writable by overriding appropriate methods. Ensure that it doesn't store null buffer handle pointer internally, and override get() on nsXPIDL[C]String for backwards-compatibility. Collapse nsSharedBufferHandle and nsFlexBufferHandle into nsSharedBufferHandle which knows its length but can't have a storage start distinct from its data start. Convert string users who were working around broken nsSharableString::Assign back to using it. b=104663, 100751, 74726 r=jag sr=scc
2001-11-28 05:22:57 +00:00
25f3ad4fb3
Update all pointers to an arena that moved when it was reallocated (111557, r=shaver, sr=jband).
2001-11-27 21:27:59 +00:00
51a4520c4a
Bugzilla Bug 107823 change nsIFile
...
to use 'Modified' instead of 'Modification'
r=akkana sr=brendan
2001-11-26 08:05:05 +00:00
106186f720
Thanks to scole@planetweb.com (Steven Cole) for fixing date_format to respect PRMJ_FormatTime's return value (102159, r=beard, sr=brendan).
2001-11-21 23:53:54 +00:00
3350a394b5
Don't eagerly allocate fatlock freelist chunks; don't preallocate so much, and use a smaller global lock hashtable (109218, r=shaver/sr=jband).
2001-11-21 23:28:52 +00:00
7733357d51
Rename 'install' build phase to 'libs'.
...
Bug #56601 r=pavlov
2001-11-21 09:49:41 +00:00
9b46da979c
Thanks to scole@planetweb.com, no more un-ECMA error for /* in a /**/ comment (110286, r=khanson, sr=brendan).
2001-11-20 23:35:45 +00:00
d2eb2bf046
Better fix than 0.9.6 minimal fix for NUL-termination assumption in replace (108832, r=shaver, sr=jband).
2001-11-20 21:26:33 +00:00
5eaba7585d
Don't warn or 'correct' = to == in a condition unless the right operand has greater precedence than == (106244, r=shaver, sr=jband).
2001-11-20 05:20:57 +00:00
43f2ce47ff
bug 110903, patch=brendan, sr=shaver, r=me
...
Call new script hook for scripts created via xdr.
2001-11-20 02:47:41 +00:00
024ffbebe7
Don't ass-u-me fp->argv => fp->fun (109635, sr=shaver/jband).
2001-11-20 02:37:01 +00:00
060425ec59
Bugzilla Bug 110080 change !NS_SUCCEEDED(status) to NS_FAILED(status)
...
s/!NS_SUCCEEDED/NS_FAILED/g;s/!NS_FAILED/NS_SUCCEEDED/g;
r=dougt sr=hewitt
2001-11-19 06:45:00 +00:00
051d085765
Remove __call__ and add a callback used to check f.caller access for any function object f (109113, r=jst, sr=jband).
2001-11-16 02:56:49 +00:00
c7ca889121
Build libmozjs during the install phase.
...
Bug #104307 sr=jband
2001-11-14 04:52:55 +00:00
f2ee8260a0
Use NS_LITERAL_STRING instead of NS_ConvertASCIItoUCS2 where possible
...
(i.e. where the string is a literal). Bug 104159, r=jag, sr=alecf
2001-11-14 01:33:42 +00:00
8b9fb46499
Avoid ClaimScope (lock-free to -full transition) in JS_GetClass and js_FinalizeObject called from the GC (109183, r=shaver, sr=jband).
2001-11-11 21:25:13 +00:00
cd8dc8ee2c
not part of build. Now that null DOM strings are correctly supported we can fix the test case
2001-11-10 22:35:26 +00:00
91397f1e55
For now (0.9.6), str_replace requires an independent replacement string (108832, r=shaver, sr=jband, a=asa).
2001-11-08 23:53:33 +00:00
63cb2c1077
Fix Function.call(this, 'alert(42)')() so the anonymous function finds its nearest scripted caller's principals (90546, sr=shaver&jband).
2001-11-07 06:34:04 +00:00
8cd8d91750
Ensure that string literals are used as |const char*| rather than |char*|. r=jag sr=brendan b=107052
2001-11-07 06:24:10 +00:00
5991fb004d
#105476
...
r=os2team, sr=brendan@mozilla.org
Add ASM implementation of js_CompareAndSwap for OS/2 VACPP
2001-11-07 05:20:33 +00:00
bfb2f49916
Add assertions that users of nsDependentC?String meet the requirements of nsAFlatC?String. Add nsDependentSingleFragmentC?Substring for users that don't. Additional cleanup after bug 100214. b=104651 r=jag sr=scc
2001-11-07 04:12:20 +00:00
c220d0c707
Avoid recursive join/toString death (108440, r=shaver, sr=jband).
2001-11-07 00:15:44 +00:00
be20b21443
Move the oversized arena back-pointer to a well-aligned header extension lying just behind a->base (108013, sr=waterson&jband).
2001-11-07 00:11:22 +00:00
0c81cb1768
better error message when zero passed where object or null expected. bug 108284. r=dbradley sr=shaver. Also, fix uninitilized counter var in debug only code rs=jband
2001-11-06 00:40:54 +00:00
be10c3b1a8
Backing out fix for remote mach-o builds as it left mach-o builds fragile.
...
Add TK_CFLAGS back to default CFLAGS/CXXFLAGS.
Bug #107696
2001-11-03 03:29:05 +00:00
2085b262f0
bug 108257, "source notes broken for functions compiled via JS_CompileUCFunctionForPrincipals", sr=brendan, r=shaver
...
no need to emit the function body after calling FunctionBody
2001-11-03 00:10:47 +00:00
5cb0f9371f
Bug 83426 - Remove DOM exception hack. r=jst, r=dbradley, sr=jband
2001-11-01 15:52:53 +00:00
3bcdaf198c
fixing warnings from recent landing and uncommenting test cases I did not mean to comment out for the checkin
2001-11-01 05:18:35 +00:00
3c38914b93
Fix nul-termination assumption hidden in lastIndexOf, broken by dependent strings fix for bug 56940 (107771, r=jag, sr=waterson).
2001-11-01 03:19:54 +00:00
a2511b96a8
fix Mac bustage. I can't believe the Mac can't do this conversion without a cast!
2001-11-01 02:36:19 +00:00
3ff792dd5d
the HP compiler says that anonymous unions can't even have structs. So the union must be named. Fixing that bustage
2001-11-01 01:45:48 +00:00
c052cc3bf0
fix Mac bustage (I hope). I think these names conflicted with other names
2001-11-01 01:15:53 +00:00
8a2a775382
Since '-framework Carbon' causes binaries to require console access to run, do not link using that flag by default. Otherwise, remote builds & non-console tinderboxes will break. Mac OSX Mach-O only.
...
Bug #107696 r=mozbot
2001-11-01 00:54:48 +00:00
b52992027b
b=107823, r=ccarlen, sr=brendan
...
--pete
2001-11-01 00:11:15 +00:00
84bfc894a4
add support for nsIVariant and nsIPropertyBag. bug 44675 and bug 98209. sr=vidur r=dbradley
2001-10-31 23:35:07 +00:00
7491994718
bug 99601, don't install xpconnect debugger keyword handler if one is already there, r=jband, sr=brendan
2001-10-31 23:13:17 +00:00
2caef4a183
Don't try to correct = to == if strict, do it only if downrev (106244, r=jband, sr=shaver).
2001-10-31 03:17:00 +00:00
7000513580
Ensure atom strings are immutable -- duh! (107138, r=jag, sr=jband).
2001-10-30 02:33:41 +00:00
671178cbdb
Consolidate property cache flush under JS_ClearScope (107265, r=shaver, sr=jband).
2001-10-29 19:30:03 +00:00
d27c4cb86f
Followup to bug 56940's big patch, tighten GC-mark case analysis, assume atom strings are immutable.
2001-10-28 22:37:25 +00:00
114a9d244c
Ensure proper pointer alignment in oversized arena footers (105726, r/sr=jband,shaver).
2001-10-28 20:41:53 +00:00
467cafd25f
Avoid stack overflow in js_EmitTree on long left-associative operator chains (98901, r=jband, sr=shaver).
2001-10-27 18:38:16 +00:00
9b5ae05e33
Landing the rest of the mach-o OSX changes.
...
Bug #75653 r=cls sr=waterson
2001-10-27 07:38:18 +00:00
41da6c554a
Initial checkin of build file for Darwin1.4.mk. It just includes Darwin1.3.mk. Because I happened to be fooling around with an OS X system.
2001-10-26 09:38:43 +00:00
a6790acff4
other part of bug 106130, remove unicharutil dependency and linking (just requires updates)
2001-10-26 03:06:04 +00:00
63754279fd
epstein@tellme.com's patch implementing JS_ClearNewbornRoots (106790, r=brendan, sr=jband).
2001-10-26 02:38:10 +00:00
862a59da87
khanson@netscape.com's fix for silent ValueIsLength failure (101488, r=rogerl, sr=me).
2001-10-26 02:35:01 +00:00
05bb290330
Fixing bug 106823. Pass the length to nsAString::Assign() to make sure we don't loose embedded null characters in strings, and also to speed things up a bit. Also trunkate strings before 'voiding' them to make sure old string data is deleted from the string before it's reused. r=bzbarsky@mit.edu, rs=jband@netscape.com
2001-10-26 02:18:16 +00:00
900bb81362
Bug 106130 - Remove dependancy of xpconnect on intl. r=alecf, sr=shaver, patch=jband
2001-10-25 12:07:23 +00:00
894b19f8c3
Add mutable (growable or dependent, two subtypes) strings to solve O(n^2) and O(n^3) growth rates (56940, r=rogerl, sr=jband&shaver).
2001-10-25 00:26:38 +00:00
1541469c6a
Bug 53057: Fixing up users of implicit |CharT*| conversion operators for nsCString to use |.get()| instead, rr=dbaron, rs=scc
2001-10-24 09:43:59 +00:00
ec6e6aa5da
Fixing bug 69468. Finally adding support for string nullness to XPConnect. r=jaggernaut@netscape.com, shaver@mozilla.org, sr=jband@netscape.com
2001-10-24 07:15:06 +00:00
b41bf3571c
Since the purpose of --enable-ultrasparc is to just flip on the ultrasparc-specific js compare & swap code, move the asflags/defines to turn it on back into js/src/Makefile.in rather than causing the entire project to be built with those flags.
2001-10-23 08:17:16 +00:00
0a62d4e8d5
99163 - removing nsObserverService.h includes. Also fixing 106090, extra methods on nsObserverService. r=shaver@mozilla.org, sr=jband@netscape.com
2001-10-22 22:01:27 +00:00
c8469b7ac5
bug 100214 - add libunicharutil_s to all unix projects that require it, r=cls
2001-10-22 20:44:02 +00:00
3e278e4913
Bug 105973: s/Observe/observe/
2001-10-22 04:16:33 +00:00
d18d7e2b17
nsIObserver and nsIObserverService API freeze. r=rpotts@netscape.com, alecf@netscape.com. bug 99163
2001-10-19 20:52:59 +00:00
ef987373e8
Avoid converting lock-free to lock-full scope from GC mark call-outs (105571, r=jst, sr=jband).
2001-10-19 03:02:41 +00:00
19f0e89f2b
bug 103649 - Crash when rendering a page with hundreds of iframes - This reduces the number of calls to JS_ContextIterator from XPCJSRuntime::SyncXPCContextList. It is only the first step in addressing this bug. r=jband, sr=jst
2001-10-17 11:52:55 +00:00
38b3c5bb4b
Checking in scole@planetweb.com's patch to prefix str_escape with js_ (96084, r=rogerl, sr=brendan).
2001-10-17 08:16:53 +00:00
43a911aeb6
Fixes for bug 80981 (``Need extended jump bytecode to avoid "script too large"
...
errors, etc.''):
We now ReportStatementTooLarge only if
- a jump offset overflows 32 bits, signed;
- there are 2**32 or more span dependencies in a script;
- a backpatch chain link is more than (2**30 - 1) bytecodes long;
- a source note's distance from the last note, or from script main entry
point, is > 0x7fffff bytes.
Narrative of the patch, by file:
- js.c
The js_SrcNoteName array of const char * is now a js_SrcNoteSpec array of
"specifiers", structs that include a const char *name member. Also, due to
span-dependent jumps at the ends of basic blocks where the decompiler knows
the basic block length, but not the jump format, we need an offset operand
for SRC_COND, SRC_IF_ELSE, and SRC_WHILE (to tell the distance from the
branch bytecode after the condition expression to the span-dependent jump).
- jsarena.[ch]
JS arenas are used mainly for last-in-first-out allocation with _en masse_
release to the malloc pool (or, optionally, to a private freelist). But
the code generator needs to allocate and grow (by doubling, to avoid O(n^2)
growth) allocations that hold bytecode, source notes, and span-dependency
records. This exception to LIFO allocation works by claiming an entire
arena from the pool and realloc'ing it, as soon as the allocation size
reaches the pool's default arena size. Call such an allocation a "large
single allocation".
This patch adds a new arena API, JS_ArenaFreeAllocation, which can be used
to free a large single allocation. If called with an allocation that's not
a large single allocation, it will nevertheless attempt to retract the arena
containing that allocation, if the allocation is last within its arena.
Thus JS_ArenaFreeAllocation adds a non-LIFO "free" special case to match the
non-LIFO "grow" special case already implemented under JS_ARENA_GROW for
large single allocations.
The code generator still benefits via this extension to arenas, over purely
manual malloc/realloc/free, by virtue of _en masse_ free (JS_ARENA_RELEASE
after code generation has completed, successfully or not).
To avoid searching for the previous arena, in order to update its next
member upon reallocation of the arena containing a large single allocation,
the oversized arena has a back-pointer to that next member stored (but not
as allocable space within the arena) in a (JSArena **) footer at its end.
- jscntxt.c
I've observed for many scripts that the bytes of source notes and bytecode
are of comparable lengths, but only now am I fixing the default arena size
for cx->notePool to match the size for cx->codePool (1024 instead of 256).
- jsemit.c
Span-dependent instructions in JS bytecode consist of the jump (JOF_JUMP)
and switch (JOF_LOOKUPSWITCH, JOF_TABLESWITCH) format opcodes, subdivided
into unconditional (gotos and gosubs), and conditional jumps or branches
(which pop a value, test it, and jump depending on its value). Most jumps
have just one immediate operand, a signed offset from the jump opcode's pc
to the target bytecode. The lookup and table switch opcodes may contain
many jump offsets.
This patch adds "X" counterparts to the opcodes/formats (X is suffixed, btw,
to prefer JSOP_ORX and thereby to avoid colliding on the JSOP_XOR name for
the extended form of the JSOP_OR branch opcode). The unextended or short
formats have 16-bit signed immediate offset operands, the extended or long
formats have 32-bit signed immediates. The span-dependency problem consists
of selecting as few long instructions as possible, or about as few -- since
jumps can span other jumps, extending one jump may cause another to need to
be extended.
Most JS scripts are short, so need no extended jumps. We optimize for this
case by generating short jumps until we know a long jump is needed. After
that point, we keep generating short jumps, but each jump's 16-bit immediate
offset operand is actually an unsigned index into cg->spanDeps, an array of
JSSpanDep structs. Each struct tells the top offset in the script of the
opcode, the "before" offset of the jump (which will be the same as top for
simplex jumps, but which will index further into the bytecode array for a
non-initial jump offset in a lookup or table switch), the after "offset"
adjusted during span-dependent instruction selection (initially the same
value as the "before" offset), and the jump target (more below).
Since we generate cg->spanDeps lazily, from within js_SetJumpOffset, we must
ensure that all bytecode generated so far can be inspected to discover where
the jump offset immediate operands lie within CG_CODE(cg). But the bonus is
that we generate span-dependency records sorted by their offsets, so we can
binary-search when trying to find a JSSpanDep for a given bytecode offset,
or the nearest JSSpanDep at or above a given pc.
To avoid limiting scripts to 64K jumps, if the cg->spanDeps index overflows
65534, we store SPANDEP_INDEX_HUGE in the jump's immediate operand. This
tells us that we need to binary-search for the cg->spanDeps entry by the
jump opcode's bytecode offset (sd->before).
Jump targets need to be maintained in a data structure that lets us look
up an already-known target by its address (jumps may have a common target),
and that also lets us update the addresses (script-relative, a.k.a. absolute
offsets) of targets that come after a jump target (for when a jump below
that target needs to be extended). We use an AVL tree, implemented using
recursion, but with some tricky optimizations to its height-balancing code
(see http://www.enteract.com/~bradapp/ftp/src/libs/C++/AvlTrees.html ).
A final wrinkle: backpatch chains are linked by jump-to-jump offsets with
positive sign, even though they link "backward" (i.e., toward lower bytecode
address). We don't want to waste space and search time in the AVL tree for
such temporary backpatch deltas, so we use a single-bit wildcard scheme to
tag true JSJumpTarget pointers and encode untagged, signed (positive) deltas
in JSSpanDep.target pointers, depending on whether the JSSpanDep has a known
target, or is still awaiting backpatching.
Note that backpatch chains would present a problem for BuildSpanDepTable,
which inspects bytecode to build cg->spanDeps on demand, when the first
short jump offset overflows. To solve this temporary problem, we emit a
proxy bytecode (JSOP_BACKPATCH; JSOP_BACKPATCH_PUSH for jumps that push a
result on the interpreter's stack, namely JSOP_GOSUB; or JSOP_BACKPATCH_POP
for branch ops) whose nuses/ndefs counts help keep the stack balanced, but
whose opcode format distinguishes its backpatch delta immediate operand from
a normal jump offset.
The cg->spanDeps array and JSJumpTarget structs are allocated from the
cx->tempPool arena-pool. This created a LIFO vs. non-LIFO conflict: there
were two places under the TOK_SWITCH case in js_EmitTree that used tempPool
to allocate and release a chunk of memory, during whose lifetime JSSpanDep
and/or JSJumpTarget structs might also be allocated from tempPool -- the
ensuing release would prove disastrous. These bitmap and table temporaries
are now allocated from the malloc heap.
- jsinterp.c
Straightforward cloning and JUMP => JUMPX mutating of the jump and switch
format bytecode cases.
- jsobj.c
Silence warnings about %p used without (void *) casts.
- jsopcode.c
Massive and scary decompiler whackage to cope with extended jumps, using
source note offsets to help find jumps whose format (short or long) can't
be discovered from properties of prior instructions in the script.
One cute hack here: long || and && expressions are broken up to wrap before
the 80th column, with the operator at the end of each non-terminal line.
- jsopcode.h, jsopcode.tbl
The new extended jump opcodes, formats, and fundamental parameterization
macros. Also, more comments.
- jsparse.c
Random and probably only aesthetic fix to avoid decorating a foo[i]++ or
--foo[i] parse tree node with JSOP_SETCALL, wrongly (only foo(i)++ or
--foo(i), or the other post- or prefix form operator, should have such an
opcode decoration on its parse tree).
- jsscript.h
Random macro naming sanity: use trailing _ rather than leading _ for macro
local variables in order to avoid invading the standard C global namespace.
2001-10-17 03:16:48 +00:00
e6e171574e
trivial warning fix. rs=jband
2001-10-16 05:48:25 +00:00
eb9671e246
Add SetAlphaBounds, TABLE_SIZE, and MIN_ALPHA APIs (103990, r=dbaron, sr=waterson).
2001-10-16 05:40:27 +00:00
4f3a44d63b
API Freeze for nsIServiceManager r=shaver@mozilla.org, sr=rpotts@netscape.com bug 99147
2001-10-16 03:35:52 +00:00
12d650dc85
Fix js_GC to mark max(#actuals, #formals) slots at fp->argv (104584, r=shaver, sr=jband).
2001-10-16 02:21:02 +00:00
7e89125660
Bug 102725 -- problems converting numbers to strings when using gcc -O2. r=bbaetz, sr=brendan.
2001-10-15 22:31:39 +00:00
e21b56807d
not yet part of build. support tracking the iid in array of interfaces
2001-10-12 20:29:19 +00:00
064c4cae19
Avoid O(n**2) growth rate of token buffer arena-based reallocation (103042, r=jband, sr=shaver).
2001-10-10 20:19:12 +00:00
900a5bfba9
NOT PART OF BUILD. bug 44675. bug 98209. New files
2001-10-09 23:42:22 +00:00
6907d52b27
bug 102750 - Can't call JSIID.equals from web content. r=mstoltz, sr=jband
2001-10-09 18:54:37 +00:00
5992cc2572
fix bug 103259. Avoid crasher by correctly marking sharedscriptable object so that JSClasses don't get collected prematurely even when wrapper is 'stillborn' and not reflected into JS. r=dbradley sr=vidur.
2001-10-08 19:56:25 +00:00
0c2b3bc449
Bug 81444: eliminate |null_nsCOMPtr|. Eliminating all callers, r=syd, rs=mozbot.
2001-10-07 10:58:39 +00:00
c39b98b2a4
Clean up REQUIRES lines.
2001-10-07 10:35:41 +00:00
95df849df8
Fix bogus return JS_FALSE without error report for empty flat string regexp match/replace (83293, r=rogerl, sr=shaver).
2001-10-06 08:07:43 +00:00
9324b81dee
Fix for bug 99663 (for loop resolves properties of the object being enumerated
...
with JSRESOLVE_ASSIGNING, wrongly), plus a few miscellaneous bugfixes.
- Combine the JSStackFrame members constructing, special, overrides, and
reserved into a uint32 flags member.
- Separate JOF_ASSIGNING from the JOF_SET bytecode format flag, and impute
JSRESOLVE_ASSIGNING from the presence of JOF_ASSIGNING among the current
opcode's format flags. To handle the for-in loop opcodes, which do more
than simply assign -- in particular, they do property lookups whose resolve
hook outcalls should not be flagged with JSRESOLVE_ASSIGNING -- a new frame
flag, JSFRAME_ASSIGNING, has been added.
- Fix interpreter version selection to respect JS_SetVersion, whose effect on
cx->version is "sticky".
- Fix js_DecompileValueGenerator to deal with JSOP_ENUMELEM -- it never had,
as this testcase shows (it crashes without this patch):
version(120);
eval("function fe(s) { for (it[s] in this); }");
try { fe('rdonly'); } catch (e) { print(e); }
2001-10-03 06:39:30 +00:00
aa050fb3a4
bug 100665 - Maybe xpconnect should not init the XPCOutArg JSClass. r=jband, sr=shaver
2001-10-03 02:57:42 +00:00
9bb2766908
Bug 102294 - printf has an extra parameter. r=jband
2001-10-03 02:48:40 +00:00
9a4acbab7a
Bug 45797: replace remaining uses of deprecated macros NS_IMPL_QUERY_INTERFACE and friends with NS_IMPL_QUERY_INTERFACE1 and friends and remove the deprecated macros. r=jag sr=alecf
2001-10-02 03:18:08 +00:00
8cf1697439
Check in patch for bug 94580, thanks to Bernard Alleysson <balleysson@bigfoot.com> for researching and writing it (r=me, sr=shaver).
2001-09-29 21:30:26 +00:00
ca6197295f
Bug 100476: Convert uses of member functions ToNewUnicode, ToNewCString and ToNewUTF8String to their global versions and remove support from nsCString and nsString. r=dbaron, rs=scc
2001-09-29 08:28:41 +00:00
28d15740ef
Fixing xpctools build on win32. r=cls.
2001-09-29 06:24:41 +00:00
2dca78dc34
Items moved to bugzilla or have already been addressed
2001-09-29 00:13:34 +00:00
3305314761
Bug 97967 - leak of nsLocalFile on failure. Patch by dbaron, r=dbradley, sr=jband
2001-09-28 21:49:00 +00:00
4e12e44b2f
Relicensing Round 1, Take 2. Most C-like NPL files -> NPL/GPL/LGPL. Bug 98089.
2001-09-28 20:14:13 +00:00
77d6e73e42
Bug 94506: parameter named "arguments" is not accessible in JavaScript method.
...
Now we check for the magic "arguments" name after var-and-arg slot
optimizations. r=rogerl, sr=brendan
2001-09-27 14:52:14 +00:00
06c740491e
Don't report uncaught exceptions unless returning to native code (97646, r=jband, sr=shaver).
2001-09-27 05:04:45 +00:00
0b484536e8
Bug 94859 silence compiler warning about unused variable r=rginda, sr=jband
2001-09-26 04:11:54 +00:00
3558787cfd
Report uncaught exceptions from all JS_Compile* APIs (97646, sr=jband&shaver).
2001-09-25 06:43:33 +00:00
a47dae7c99
patch from bug 99838 - Changes to remove xpconnects dependency on DOM r=jband, sr=jst
2001-09-24 22:57:47 +00:00
1f85e703a9
patch from bug 99838 - Changes to remove xpconnects dependency on DOM r=jband, sr=jst
2001-09-24 22:48:35 +00:00
69b6808ca6
Use default EXPORTS rules instead of installing headers by hand.
...
Bug #98371 sr=alecf
2001-09-23 03:34:32 +00:00
801cc884a1
I'm tired of this warning: '*/' found outside of comment. This was much discussed. But no one ever bothered to actually check in the trivial fix. r=rginda rs=jband.
2001-09-22 17:42:36 +00:00
d2a5676d57
Bug 100873 JS Hard Codes JS_HAVE_LONG_LONG to the detriment of FreeBSD
...
r=cls sr=shaver
2001-09-21 14:09:54 +00:00
afa3e0b204
fix bug 100222. Fix crash by only releasing other objects in xpcwrappedjs dtor if xpconnect has not yet shutdown. r=dbradley sr=jst.
2001-09-20 00:51:38 +00:00
1856815ff1
Oops.
2001-09-20 00:02:59 +00:00
bc444f528a
bug #98089 : ripped new license
2001-09-19 22:39:41 +00:00
229409f54e
fix bug 100260. xpconnect should not hold a weak reference to a classinfo object after the xpcwrappednativeproto that was holding a strong ref is destroyed. sr=jst, r=dbradley,markh.
2001-09-19 06:22:17 +00:00
30680afe44
Fix bug #98306 , ParseAtom crash for "/[/]/" r=jst, r/sr=be.
2001-09-18 22:36:14 +00:00
d1a7bfa843
One module per line for REQUIRES. r=/sr=alecf
2001-09-18 22:01:13 +00:00
dc40187223
Bug 73353: splitting the modules on the REQUIRES lines in Makefile.in across multiple lines to more clearly show the changes made. sr=alecf
2001-09-18 13:41:47 +00:00
c18c814935
Use the maximum of fp->argc and fp->fun->nargs to bound fp->argv (97921, sr=jband&shaver).
2001-09-18 01:47:41 +00:00
62424ee7b2
Final REQUIRES changes for bug 98371, adding support for MOZ_TRACK_MODULE_DEPS to win32. r=cls (and tested by cls!)
2001-09-17 19:06:35 +00:00
cf80f7240c
Fix unsafe JS_DHashTableRawRemove usage; unoverload cx->resolving into a counter and a JSDHashTable pointer (94422, r/sr=jband,waterson with dbaron sanity checks).
2001-09-14 03:37:15 +00:00
3efd0e519f
1. Fix jsdhash/pldhash.[ch] so that add/remove cycles for a given k do not
...
create a long chain of removed sentinels. Also, when adding k to a table
where k is not mapped, but where k hashes to a chain that includes removed
sentinels, recycle the first removed sentinel in the chain for k's entry.
2. Cache cx->resolving till js_DestroyContext, to avoid high JSDHashTable
new/destroy overhead in js_LookupProperty.
3. Add NS_TraceStack to nsTraceMalloc.[ch] and clean the .c file up a bit.
2001-09-12 06:39:31 +00:00
1db0a058f3
Make generated header use common install rules.
...
Bug #59454 r=bryner sr=alecf
2001-09-12 06:35:50 +00:00
3b2127bd59
Checking in for Ulrich Drepper. Bug #99246 . r=beard, sr=brendan
2001-09-12 03:19:08 +00:00
8bf5905937
add REQUIRES support to jpeg, sun-java, db, etc.. b=98371 r=cls
2001-09-10 20:48:38 +00:00
911df23fb2
bug 95534, r=brendan, sr=jband
...
fix semantics of JSTRAP_CONTINUE to prevent a debugger client from telling the engine to continue when we all know it'll proobably just crash.
2001-09-06 21:46:18 +00:00
02986411d1
fix bug 96511. Meant #ifdef not #if. sr=brendan r=jband
2001-09-06 19:01:17 +00:00
0a93bb62fa
fix bug 98450. We need to be calling OBJ_DROP_PROPERTY after OBJ_LOOKUP_PROPERTY. Also, this was refactored on dbradley's suggestion. sr=brendan r=dbradley
2001-09-06 06:29:19 +00:00
23d7dc717d
fix bug 97444. It is not good to patch a different fun into the frame. Let's safely shunt aside the callee frame instead. r=rogerl sr=brendan
2001-09-05 21:25:09 +00:00
622522e134
Ignore property found in non-native prototype (12367, sr=jband&shaver, a=asa).
2001-09-03 22:29:12 +00:00
0cf73f1847
Fix toSource on exception objects so it uses toSource to generate properly-quoted, embedded string literals for message and filename (bug 96284, r=jband, sr=shaver, a=asa).
2001-09-03 19:36:24 +00:00
f14e08526e
bug 88130, patch=jband, r=me,dbradley, sr=jst, a=brendan
...
clear exception state before and after calling out via xpconnect.
remove redundant exception clear from CallQueryInterfaceOnJSObject.
2001-09-01 22:47:47 +00:00
56d03ab63d
fix bug 97555. Need to get the correct value of staep into the iterator object when we fail during enumeration init. r=jst sr=brendan a=asa
2001-08-31 21:48:18 +00:00
43c254feee
Fix 97540, r=rginda, sr=jband, a=asa:
...
- The most significant fix, to keep JSStackFrame.spbase, the operand stack base pointer for an active frame, null except when there is an operand stack allocated and in use by js_Interpret. Previously, spbase would point after args and local vars (if any), then advance upon allocation of the (possibly discontiguous) operand stack space. This made for a fatal ambiguity: js_AllocStack, called by XPConnect, could not tell when there was allocated operand stack space above the frame's sp, which needs to be set to a known (JSVAL_VOID) state for exact GC to work. Now, the GC doesn't have to mark any operand stack space for a frame whose spbase is null, and js_AllocStack doesn't need to void any unused space for such a frame.
- Fixes to reload the JSRuntime's callHook or executeHook after calling or executing, in case the debugger removes the hook. In which case, it must clean up any dynamic memory held by hookData, but in any event, in which case the engine must not call the post-call or post-execute hook.
- While debugging with rginda, I was horrified to see his trivial testcase function, expressed as a lambda, fail to be invoked using the "inline_call" machinery in js_Interpret (which avoids js_Interpret recursion through js_Invoke for most JS functions). The problem was a test of fun->flags == 0 conditioning the /* inline_call: */ code. Since that test was written, at least one JSFUN_* flag (JSFUN_LAMBDA, used only for pretty-printing or accurate decompilation) has been added. But all along, that test was an over-optimization (testing against 0 without &'ing certain flags), making for an accident waiting to happen -- which did happen. The relevant flags are JSFUN_HEAVYWEIGHT (set by the compiler when a function calls eval, uses with, or otherwise needs an activation object for its scope; if lightweight, the compiler can see the function's scope and eliminate it via specialized bytecodes) and JSFUN_BOUND_METHOD (for Java method calls, where |this| binds statically to the instance, not dynamically to the calling expression reference's base object, as in JS).
2001-08-31 21:25:26 +00:00
f95e84e0e7
fix bug 96725. Avoid infinite recursion in call to QI'd xbl binding by avoiding the mutation of the interface set when the object implementing the interface is a wrappedJS around our wrappedNative's own JSObject. r=dbradley sr=hyatt a=brendan,hyatt,jband
2001-08-28 21:52:10 +00:00
6f3e4115eb
Bug 94243: VoidArray usage patch. a=asa sr=brendan r=waterson r=bienvenu
...
r=pierre r=jband/dbradley Also fixes some whitespace issues.
2001-08-27 06:15:54 +00:00
f319c1d822
Check keyword version when scanning, not by installing different keywords on version selection (when starting to compile; bug 96562, r=rogerl, sr=shaver&jband, a=dbaron).
2001-08-24 03:32:31 +00:00
f81e32a193
Avoid corrupting backpatch chains by checking for jump offset operand overflow (90445, khanson's patch with minor changes, sr=shaver&me, a=asa).
2001-08-23 02:07:07 +00:00
8cc94e0ba1
Checking in khanson's patch for bug 85112, based on joemansh@interl.net's algorithm (r=beard, sr=me).
2001-08-22 05:42:16 +00:00
834611cda0
Forget to include jsopcode.h.
2001-08-22 05:33:32 +00:00
2b5a1171f6
Support ref()++ etc. for native ref (75688) and fix exn_toSource (96284); sr=shaver&jband.
2001-08-22 04:59:28 +00:00
68f889e9bd
Checking in shaver's patch for bug 78100, adding JS_GetExternalStringGCType (sr=brendan&jband).
2001-08-21 02:53:19 +00:00
113ad642ae
Replace uses of deprecated NS_IMPL_ISUPPORTS and NS_IMPL_QUERYINTERFACE macros with NS_IMPL_{ISUPPORTS,QUERYINTERFACE}{1,0}. r=jag rs=brendan b=45797
2001-08-21 01:48:11 +00:00
b0e3e4ab62
Don't resolve standard classes if JSRESOLVE_ASSIGNING (NOT PART OF BUILD).
2001-08-17 10:40:15 +00:00
6843e21c2b
Checking in khanson's patch with whitespace cleanups for bug 89443, to avoid stack overflow on too many || operators in one expression (r=rogerl, sr=me).
2001-08-17 10:29:59 +00:00
4e1703cf54
Shrinking realloc can fail (95517, basic patch from epstein@tellme.com, r=brendan, sr=jband).
2001-08-17 09:56:48 +00:00
2bad71c89b
fix bug 94752. Let xpconnect decide when to call JS_DestroyContext. It sometimes needs to defer the call until JS code running on the context is finished. r=dbradley sr=jst
2001-08-15 04:49:09 +00:00
e1d4f83235
Context typemapping changed
2001-08-14 13:09:43 +00:00
a3c1dd1b09
Use NS_PTR_TO_INT32 macros to do 64-bit safe pointer conversions.
...
Bug #20860 r=Roland.Mainz@informatik.med.uni-giessen.de sr=brendan@mozilla.org
2001-08-14 07:59:59 +00:00
c04a3d6833
Attempt to build on Unix
2001-08-11 04:57:50 +00:00
9abbf226e2
JSOP_BINDNAME needs JOF_SET flag for JSRESOLVE_ASSIGNING to be imputed (94631, r=shaver, sr=jband).
2001-08-10 23:57:07 +00:00
bc3efa22e4
Bug 88130: XPConnect disturbs exception state of the call-context. This is jband's patch to reorder calls and fix whitspace. r=rginda, sr=jst
2001-08-10 19:23:39 +00:00
466a26d77c
Bug 93790: xpconnect does not correctly reflect nsIClassInfo interface. r=dbradley, sr=jst, patch by jband
2001-08-10 14:13:17 +00:00
brendan%mozilla.org
dbradley%netscape.com
khanson%netscape.com
shaver%mozilla.org
dbaron%fas.harvard.edu
bratell%lysator.liu.se
jst%netscape.com
dougt%netscape.com
seawood%netscape.com
darin%netscape.com
bryner%netscape.com
jband%netscape.com
beard%netscape.com
pschwartau%netscape.com
timeless%mac.com
rginda%netscape.com
nisheeth%netscape.com
sfraser%netscape.com
rogerl%netscape.com
alecf%netscape.com
jaggernaut%netscape.com
pavlov%netscape.com
bzbarsky%mit.edu
mkaply%us.ibm.com
pete%alphanumerica.com
mike+mozilla%meer.net
cls%seawood.org
gerv%gerv.net
scc%mozilla.org
blizzard%redhat.com
rjesup%wgate.com
pavel%gingerall.cz
val4%cornell.edu