JS_THREADSAFE assert triggered in FF 3.6

Running Chromebug on my FF 3.6 debug build I fail on this assert: 105 #if defined(JS_THREADSAFE) 106 #define CHECK_REQUEST(cx) \ 107 JS_ASSERT((cx)->requestDepth || (cx)->thread == (cx)->runtime->gcThread) 108 #else 109 #define CHECK_REQUEST(cx) ((void)0) 110 #endif The assert failure exits (!). Is there a way to say "yea sure whatever, just try your best"? The exact same run on distributed FF 3.6 run with no crash. jjb

Function parameters

hi guys, i am using spider monkey 1.8 to parse javascript source and print the tree. The problem is that i cannot print funtion parameters. For example function product(a,b){return a*b}; parameters a and b are not printed. The code is ////////////////////////////// function = (JSFunction *) JS_GetPrivate(context, (*root->pn_u.func.funpob).object); printf("Function: %s (", JS_GetFunctionName(function)); // Arguments object = JS_GetFunctionObject(function); JSScopeProperty* sp =

closure access to unused variables in closed-over scopes

A firebug user is complaining that Firebug fails in a case like: var www = 'asdasd'; // Can't access 'www' on breakpoint this.closure = function() { rrr = 5; // <<< breakpoint here } http://code.google.com/p/fbug/issues/detail?id=2871 I assume that any objects not referenced by the function in a closure are not 'bound' into the enclosing scope, so there is no bug here. Firebug is just issuing eval('www') in the frame of the function, not much we can get wrong... jjb

Proxies and security wrappers

Sorry this is so long. I have to run immediately to catch a train. There's actually a lot more to say, so please let's discuss... Andreas has been working on an implementation of Mark Miller et al's Proxy proposal, adding the GC support, and so on - which is all good stuff. We're going to explore using Proxy to implement our security wrappers, which is an awesome idea-- but I have misgivings I want to voice now. 1. I think it'll be much easier to write correct security wrappers in C++ than in JS. The main difference is that C++ code doesn't call JS functions implicitly. The

function return type

Hi. Is it possible to know JavaScript function return type, function parameters types, variable type using SpiderMonkey? Example: function doSomething (Date) { try { var Res = check(Date); switch (Res) { case 1 : return (true); break; case 0: return (false); break; case -1: return

Finding array access bounds in SpiderMonkey

Hi, I was wondering if there's any way in SpiderMonkey to figure out the range of accessed elements in an array. (e.g. during parsing or compilation) For example in the following JS code: for (i = 0; i< 100; i++) { myarray[i] = i; } Can I somehow in the parser or compiler detect that elements 0 to 100 in myarray are touched? Thanks, Mojtaba

thiago silva has invited you to join Friendster

You're invited to join thiago silva's network of friends. By joining Friendster, you can reconnect with old friends, meet new friends, start a blog, build a custom profile, keep track of birthdays, and so much more! You can even stay in touch if you move away, switch email addresses, or lose your mobile phone. Click below to join thiago's network. http://www.friendster.com/join.php?inviteref=119553532&amp;invite=vudHYajyOKw87biVGIPQxWcbFQSqmphw5A_2X_uctpJ65MNjwo5ijkFUD9z-GN39&amp;lang=en-US ***************************************************************** If you do not wis

Proposed changes to JS_THREADSAFE rules

This post is about changing the rules about sharing objects in JS_THREADSAFE builds of SpiderMonkey. Feedback is desired, especially from brendan, igor, gwagner, and anyone else interested in taking some of the work. As it stands, in a build with JS_THREADSAFE, JSObjects can theoretically be shared across threads and accessed from any thread. Any locking necessary to avoid crashes supposedly happens internal to the JSAPI. But there are bugs: - Dense arrays are not thread-safe (bug 419537). - Generators, iterators, and E4X objects are not thread-safe (bug 349263).

using /source/js/src/md5.js in /source/toolkit/components/passwordmgr/src/storage-mozStorage.js ?

Dear beloved newsgroup :-) Maybe you can help me. I don't know exactly who in xpcom interpretes the javascript xpcom components before registering them (does spidermonkey do that, too?), but there's my problem: I need to be able to do hashes in passwordmanager's storage-mozStorage.js. I'm looking for the simplest way to use code from somewhere else, like a "library". (If possible, I don't want to call another xpcom component to do the hashes for me - but if I need to, which one would you use? It doesn't matter which algorithm, could also be SHA-something or other...). I

AST Dump

Hi, all, I am trying to add a new functionality in spiderMonkey, to dump the Abstract Syntax Tree of a javaScript program in dot format. This is something like clang does, in LLVM, for instance, and it helps debugging. I have a tree here: http://dcc.ufmg.br/~rsol/docs/ast.pdf, that my procedure has produced to this js http://dcc.ufmg.br/~rsol/docs/t4.js. The whole procedure is almost done. But I am missing two things: 1 - What is the best way to integrate this feature? Using a new argument in the command line or creating a new environment variable? Should the name of the dot f

Interpretor parking in JS

Hey folks, does spidermonkey provide a way to suspend/resume execution of scripts? NOTE: while JS_SuspendRequest or JS_YieldRequest does allow GC to be unlocked it is not what I am looking for. I am running JS in an embedded system (non threaded mode) where I will have to make multiple networking and other potentially long lived requests in a single interpretor. Lets say I am executing the script shown below and my_custom_request is a *native call* implemented in C that carries out a networking request. The native that would normally block for the IO completion and instead of blo

Regular Expressions -- use NFAs when not using backreferences for better performance?

Hi there, I've recently been reading through http://swtch.com/~rsc/regexp/regexp1.html, which describes a different approach to evaluating regular expressions. Have a read, but the upshot of the article is that there exists an alternative method of evaluating regexes (using non- deterministic finite automata, or NFAs) that provides O(n^2) performance, as opposed to the Perl algorithm's O(2^n). The only caveat is that using NFAs, it's not possible to use backreferences. Is it perhaps worthwhile investigating using NFAs when presented with 'simple' regexes to improve performance? Lo

Serializing the global object

Hi I am experimenting with serializing the global object via XDR. I came across the function js_XDRBlockObject in jsobj.cpp and was wondering what this function does. Can it be used to serialize an arbitrary Javascript object or do I have to write my own code for that? Thank you

out of memory error from spidermonkey 1.7 ( not seen on 1.5 )

I am seeing an issue with 1.7 where it is giving out of memory error. Spidermonkey 1.5 works fine. We have this script, for( i = 1; i < 126000; i ++ ) { myExposedMethod("myclass.name = " + myclass.name ); } where myExposedMethod is a C function exposed to the script by embedding host. Similarly myclass is also an class exposed by embeddnig host. myclass.name is a string which remains rooted ( JS_AddRoot ). I even made myExposedMethod to just do nothing and return JS_TRUE, still I am seeing the issue. On spidermonkey 1.5, this works fine even with loop count increased to 10

Is spider monkey 1.8 is released.

Hi, We would like to upgrade JS Engine version from 1.6 to 1.8 as one of the bugfix which we are waiting for is provided in 1.8 But when I visited the spider-monkey site (https://developer.mozilla.org/En/ SpiderMonkey/1.8), it still shows coming soon. The site was last updated in March 2009. Can anyone please tell me whether its released or when the release 1.8 would be available. The site mentions that its candidate release. Thanks and With Warm Regards, -RV

Authority of plugins

ActionScript provides the ExternalInterface class, which allows a Flash developer to expose a subset of the methods in his program to JavaScript code. This limits the authority of the JavaScript code to affect the internal state of the Flash app, which is appropriate; it's rather like a firewall. The web page, on the other hand, has only a very crude way of limiting the authority of the Flash app to affect the internal state of the browser: all or nothing, as specified in the allowScriptAccess or allowNetworking parameters. JavaScript rewriting projects like Facebook's FBJS, Googl

Removing JS_strtod and JS_dtostr

I'm fixing bug 509857 ("get rid of LOCK_DTOA"). This avoids entering and leaving a lock in some cases where we convert a string to a number or vice versa. However, with my changes it's no longer convenient to implement JS_strtod and JS_dtostr. Note that these functions have never been in jsapi.h. I'm not sure why they are named with the uppercase JS_ prefix. Is anyone using these? Speak up! -j

Blancpain Villeret Mens Watch 4082-3642-55B

Blancpain Villeret Mens Watch 4082-3642-55B http://www.watchestrade.net/Blancpain-Villeret.html http://www.watchestrade.net/Blancpain-Watches.html http://www.watchestrade.net/Blancpain-Villeret-Mens-Watch-4082-3642-55B.html http://www.watchestrade.net/ <br />Blancpains self-winding chronographs are among the smallest and the slimmest chronograph movements in the world owing to a totally new construction concept: The coupling-clutch wheel also functions as a brake; no additional coupling wheel or blocker is required. And only one finger drives the minute and hour coun

Tag Heuer Link Blue Mother-of-Pearl Ladies Watch WJF131D.BA0572

Tag Heuer Link Blue Mother-of-Pearl Ladies Watch WJF131D.BA0572 http://www.watchestrade.net/TAG-Heuer-Link.html http://www.watchestrade.net/TAG-Heuer-Watches.html http://www.watchestrade.net/Tag-Heuer-Link-Blue-Mother-of-Pearl-Ladies-Watch-WJF131D.BA0572.html http://www.watchestrade.net/ <br />Stainless steel case and bracelet. Blue mother-of-pearl dial with 11 Top Wesselton (VVS) diamond hour markers. Screw-in steel crown. Double safety clasp with folding buckle. Curved scratch resistant sapphire crystal. Quartz movement. Water resistant at 200 meters (660 feet). Ta

Audemars Piguet Jules Audemars Equation of Time Mens Watch 26003BC.OO.D002CR.01

Audemars Piguet Jules Audemars Equation of Time Mens Watch 26003BC.OO.D002CR.01 http://www.watchestrade.net/Audemars-Piguet-Jules-Audemars.html http://www.watchestrade.net/Audemars-Piguet-Watches.html http://www.watchestrade.net/Audemars-Piguet-Jules-Audemars-Equation-of-Time-Mens-Watch-26003BC.OO.D002CR.01.html http://www.watchestrade.net/ <br /> <br> Jules Audemars and Edward Piguet were both in their early twenties when they founded Audemars Piguet in 1875. In 1972 AP launched a line of mechanical watches with hexagonal screws that fastened a porthole-like bezel.

New harness for shell JSTests

I just landed to the tracemonkey repo a new harness for running the shell JSTests that we've been working on for a while. My primary goal was to make the tests run faster on multiple cores and to be easier to run day-to-day for devs, but it also supports Tinderbox-style reporting and a bunch of other new features. The basic usage is: # in directory js/src/tests python jstests.py PATH_TO_JS This will run with a progress bar and report results in a compact way. The 3 numbers on the left are mochitest-style "passed/failed/known fails". There is no need to do a base

How to build JS 1.8.0 using MSVC 6.0 (win32)

Hi, How can we build Spidermonkey JS 1.8.0 in Windows (32-bit) using Microsoft Visual C++ 6.0 compiler (even using the command line build methods)? I tried to follow the same approach as that for JS 1.5, but came across multiple issues: Issue1: js.mak not available. I tried to use the js.mak provided at http://blog.endflow.net/?p=55 <http://blog.endflow.net/?p=55&lang=en> &lang=en Issue2: jsparse.c compilation issues related to static asserts: jsparse.c(93) : error C2057 : expected constant expression jsparse.c(93) : error C2466 : cannot allocate an

Lock contention and performance problems

I am also running into lock contention performance problems mentioned in this thread <http://groups.google.com/group/mozilla.dev.tech.js- engine/browse_thread/thread/d70153f89d8d0736/f259b8abdd767642? lnk=gst&q=mutex#f259b8abdd767642>. Instantiating interpretor per thread seems to work great and I am running with JS_THREADSAFE option. The only catch now seem to be that each of the runtime is creating a back ground thread. I have some pretty strict requirements on the ability to create threads per-core and was wondering if there was any way to prevent the back ground thread (if at al

64 bit porting of JS

Hi, I am using JS 1.5 RC 6 codebase released in 2005. In recent discussions, I got that Spidermonkey supports 64 bit platform ( http://groups.google.com/group/mozilla.dev.tech.js-engine/browse_thread/thread/d108545385160c52). Even though I didn't get any specific document talking about support of 64 bit platform architecture. Now, I have tried compiling same code base over x86_64 architecture. Though, I am able to build it but with lot of compilation warnings questioning its successful execution over a 64 bit platform. I am listing some of these warnings, which are specific t

js 1.8.1 gc doesn't automatically run

Hi everyone, I've been playing with the latest 1.8.1 source, and it appears that something is wrong with the garbage collection threshold specified in JS_NewRuntime(). Here's some quick example code: for ( var i = 0 ; i < 1000 ; i++ ) { print(i); //gc(); obj = new Array(); for ( var j = 0 ; j < 100000 ; j++ ) obj[j] = "1234567890"; } I'm using the latest source updated today (11/12/09) compiled in Linux, and executed with the js shell. If I let this run to completion with 1.8.1 it consumes around 1GB of memory, and the js sh

Deprecating JSCLASS_SHARE_ALL_PROPERTIES?

Bug 527805 <https://bugzilla.mozilla.org/show_bug.cgi?id=527805> indicates that classes with the JSCLASS_SHARE_ALL_PROPERTIES flag are seriously broken with the JIT enabled. It has been busted for quite some time. Brendan suggests just deleting the feature. Mozilla isn't using it. Does anyone need it? -j

tracemonkey tests

Hello! Tracemonkey has tests directory in it's distribution, and special perl script jsDriver.pl that perform testing. However, some tests fail, and it is normal - that tells us wiki (some tests are obsolete, etc.). Is there any way to run subset of tests that always must pass (even in case of active development), just to verify that current verdion of tracemonkey is working correctly? I using tracemonkey from hg.mozilla.org/tracemonkey .

Mozilla memory usage

**Please let me know if I need to post to a different group.** Hi Our product is built on top of Mozilla Firefox 3.0. We use it differently from a regular user since ours is an automated solution and in a given time span it issues lot more requests. One of the problems we are facing is that Mozilla is using lot of memory, if it is run for a long time. In some cases, long time can be few hours and in other cases it is more than 18 hours. What I noticed was that mozilla is holding on to lot of memory. VM size is getting more than 1GB and eventually it is crashing. Then I did so

Treehydra as JS test suite

Taras says JS development has been regularly busting Treehydra lately. A recent example would be https://bugzilla.mozilla.org/show_bug.cgi?id=526299 Taras suggested adding Treehydra's unit tests to our regular JS testing. What do JS devs think of that? Currently, it would require building Treehydra (including building a patched gcc) and then running 'make check' in Treehydra. -- Dave

how to define my own error.errCode and error.errMsg?

hi, guys, i am not sure how to throw the exception so that in js code i can using print(error.errCode + error.errMsg). here is the example: //c++ code void my_own_func( ) { try { // some c++ code here }catch(exception&error) { // ------what to do here???? using JS_ReportErrorNumber ?i have tested it but not run well. } } // filename: test.js try{ // some js code; } catch(error) { print(error.errCode + error.errMsg); //how can make this run well???= ? } thx =A3=BA=A3=A9 --=20 best regards

Reg 64-bit architecture support for Spidermonkey 1.5 and NSPR 4.4.1 components

Hi, I would like to get some information regarding 64-bit platform architecture support for Spidermonkey and NSPR components. For a server application, we are using Spidermonkey JS Engine 1.5 version and NSPR 4.4.1 version components. Till now we have been using 32-bit libraries of these components in different UNIX-based platforms. We are evaluating technical aspects of migrating the application to 64-bit platforms (including x86_64 and IA64 models). Towards this, I have a few queries which I am listing down below: a) Are Spidermonkey 1.5 and NSPR

Nanojit merge complete: FLAG DAY

Hi, Please read this email in full if you work on nanojit. You're likely to lose work in the near future if you do not. Today is our "nanojit merge completion" flag day. The merge is done; we finished at 7pm pacific tonight, revision 9df70d3d88ea in nanojit-central. This corresponds to the following two revisions in tamarin and tracemonkey: http://hg.mozilla.org/tamarin-redux/rev/6e89a44f87de http://hg.mozilla.org/tracemonkey/rev/a658f130ad09 This means several things for anyone who works on tamarin or tracemonkey, and has occasion to make changes to the nanojit subdir

Use of 'bool' in SpiderMonkey code

If we're migrating towards bool/true/false over JSBool/JS_TRUE/JS_FALSE, what's our policy towards SpiderMonkey headers #included into C code? Should they continue to use JSBool? #include <stdbool.h>, even though that introduces a definition for 'bool' into client code? C99 has a _Bool built-in type, which allows new headers to use the type without introducing definitions for 'bool' into old code, because _Bool isn't defined in C++; defining it ourselves seems like asking for more trouble. My personal preference would be to simply #include <stdbool.h> in jstypes.h if !defi

how to disassemble JS for an entire program?

In the documentation of spidermonkey's JS shell, it says the following about dissrc command: dissrc([function]) Disassembles the JavaScript bytecode for the entire program, or for the specified function, showing the source lines. This function only works with programs loaded from files, either using the -f flag on launching the shell, or by using the load() function. But I am not very clear how can I use dissrc to disassemble the entire program. The example given only demos how to disassemble a function object. How could I achieve similar things on entire programs? Or is this a m

Gecko object key interpretation specific: a bug or a feature?

Having a statement such as: var obj = { foo: 'bar', default: 'default_value' } Firefox 3.5.3 creates obj with properties foo and default. IE reports syntax error "Expected identifier, string or number", same or similar syntax error occurs on Safari, Chrome and Opera. By placing default into quotes we are making the code valid for all browsers in question: var obj = { foo: 'bar', 'default': 'default_value' } My question is if it is a Gecko bug, a convenience extension or the proper implementation by ECMA with others being wrong on that?

User property of a <form> element is not enumerable

If I create a <form> element and set a new property on it, it's not visible when enumerating properties. As far as I know it's because <form> elements implement an enumerate hook to access form fields. Is there a way how this hook can be bypassed to get user defined properties? (I am a Firefox extension - Firebug) Example: var form= document.createElement('form'); form.myProp = {ID:'test', Value:'someValue'}; // Following cycle doesn't see the "myProp" property for (var p in form) { // ... } Thanks! Honza

SpiderMonkey 1.8.1 now throwing TypeError when setting a property that has only a getter

Hi all, According to https://bugzilla.mozilla.org/show_bug.cgi?id=478047, newer versions of SpiderMonkey now throw a TypeError: setting a property that has only a getter instead of silently ignoring the assignment as before. var o = { get p() { return "a"; } }; o.p = "b"; >>TypeError: setting a property that has only a getter Although I understand the reasons behind the change, isn't Mozilla worried about introducing a change that breaks backwards compatibility with already existing JavaScript code? What's your take on this issue? Thanks! Pedro

Custom JSMarkOp callback not triggered in GC

Hi All, I have added a custom object to SpiderMonkey and am testing it with the browser integrated. Everything works fine except the Garbage Collector. My custom object has several JSObject* as its private data members. I have also declared the JSClass flags as JSCLASS_MARK_IS_TRACE and defined the JSMarkOp callback as JS_CLASS_TRACE(custom_trace). In this custom_trace, i walk over all the private JSObject* and invoke JS_CallTracer after checking their traceability. The strange problem is that this custom_trace callback is never triggered. My logs confirm that it is not invoked -

undefined symbol: __udivsi3 ?

Hi All : I wang to compile js-1.8.0-rc1.tar.gz for at91sam9261 arm board with = compiler cross-3.4.1.=20 this is my steps. cd src make -f Makefile.ref LD=3Darm-linux-ld CC=3Darm-linux-gcc = CCC=3Darm-linux-g++ cp jscpucfg to my board and run , cp file jsautocfg.h to = src/Linux_All_DBG.OBJ/ make -f Makefile.ref LD=3Darm-linux-ld CC=3Darm-linux-gcc = CCC=3Darm-linux-g++ cp jskwgen to my board and run , cp file jsautokw.h to = src/Linux_All_DBG.OBJ/ make -f Makefile.ref LD=3Darm-linux-ld CC=3Darm-linux-gcc = CCC=3Darm-linux-g++ at last , it'

detect useless return value

Hello, I am trying to optimize some of my native functions. My aim is to know in advance if the returned value will be used by the caller. For example my JSClass Socket has a Write() method that send data through a socket and returns the data that has not been sent (overflow). This feature exists but is rarely used, this is why I wondering if it is possible to detect (by looking at the next opcode) if the creation of the returned value is optional. I noticed that the bytecode of functions whose the return value is not used look like this: 00046: 13 callname "MyFct" 00049: 13

State of Tamarin

Hi, i just wanted to ask: What is the state of tamarin? Will it be integrated in future firefox js engines - or is it cancelled and will only be used in flash player engines? Kind Regards, moonglum

MAC OSX 10.6 - Static linking issue

Hi, Im linking spidermonkey (1.8.1) as static library against a dynamic library : gcc -std=c99 -D_GNU_SOURC -Wall -g -bundle -undefined suppress -flat_namespace -o lib/libmod_spidermonkey.so libape-spidermonkey.c -I../libs/js1.8/src/dist/include/ -L../libs/js1.8/src/ -ljs_static -Xlinker -rpath But when Im trying to load the generated library (with a dlopen), I get this : dyld: lazy symbol binding failed: Symbol not found: __Znwm Referenced from: /Users/para/dev/ape_1.0/APE_Server/modules/lib/libmod_spidermonkey.so Expected in: flat namespace I don't have any

Upcoming nanojit-central, import logic and graph structure

Hi, This is another super-long email about the nanojit merge. Sorry. We're Approaching The End of the nanojit merge, in the sense that there are no (or very few) cross-nanojit features remaining on either fork of the nanojit-central synthetic history, and the only substantial diffs are in the ARM backend, that several contributors on both teams are now looking at. In light of this, I've been making some preparations for a "final" transition to working out of a single nanojit-central branch, with a "reversed flow" of changes from nanojit-central (NC) to tamarin-redux (TR) and tr

Debugging object identity test

There is a bug somewhere in jsd that causes the following code to fail: For |frame| an instance of jsdIStackFrame: --- var functionObject = frame.script.functionObject.getWrappedValue(); var frameThis = frame.thisValue.getWrappedValue(); for (var p in frameThis) { if (frameThis[p] === functionObject) // never true .... } --- I'd like to investigate to see I can find a hack-around. So I'd like to look at the code for operator=== when these objects come in. Can anyone give me a hint on which file to start looking? Thanks, jjb

Tail Call Optimization coming to SpiderMonkey

Background: http://en.wikipedia.org/wiki/Tail_call http://en.wikipedia.org/wiki/Tail_recursion Please see: https://bugzilla.mozilla.org/show_bug.cgi?id=445363 In particular, the proposed insertion of some kind of ellipsis in the Error object's stack property value, to indicate that a tail call reused a frame, overwriting its record of the caller with the callee's activation information, see this comment: https://bugzilla.mozilla.org/show_bug.cgi?id=445363#c31 I hope to get Proper Tail Calls included as part of the guaranteed semantics of ECMAScript some day, but we can't

SpiderMonkey 1.8.1 release

Hi, I'd like to get some clarification what is the version of SpiderMonkey (TraceMonkey) shipped with Firefox 3.5.3? According to the string returned by JSAPI it is "JavaScript-C 1.8.0 pre-release 1 2007-10-03". Is it correct? Can anyone tell me when 1.8.1 release will come out? In case it exists how can I take it? Best regards, alerma

var, window, and object.watch()

I asked this over in .platform and Boris suggested that folks reading here may know more about the .watch() issue. For Firebug 1.5 we are adding break on DOM object property change (along with break on HTML, XHR, and Error). But one important case fails: properties declared with 'var' at window level. Is there something different about a property declared with var vs one just attached to the window? This code: -------- function dumpIt(prop, oldval, newval) { window.dump("changed: "+prop+" from "+oldval+" to "+newval+"\n"); return newval; } var avar = "avar";

Operator Overloading

Hi, I'm embedding SpiderMonkey to script a 3D application. I would like to expose my C++ 3D-math classes (Vector, Matrix etc) to javascript, which is fine. However the nicest thing would be to allow the use of operator syntax with these, so you do, eg, 'a+b' to add two 3D vectors together (as opposed to 'a.add(b)' or 'add(a,b)' ). Can this be done at all? If not from Javascript, then is there any way from the C end of things? It will be _much_ simpler for users to use the standard operator syntax, rather than have to worry about treating every operation as a function call. Thanks,

jsdIStackFrame.line changes?

When running on the FF 3.7 nightly build, Firebug 1.5a25 fails this test case https://getfirebug.com/tests/content/script/singleStepping/index.html If you follow the instructions on the page using Firebug 1.5a25 and FF 3.5.3 you end up on line 14 at step 3. If you follow the instructions on the page using Firebug 1.5a25 and FF 3.7 you end up on line 13 at step 3. 13 function clicker() { 14 dataDiv = document.getElementById("data"); 15 dataDiv.innerHTML += fibonacci(iter++) + "<br/>"; 16 } It looks like the line table for 'clicker' is slightly different with pc=0 b

bug 469718 - JavaScript Browser jsreftests

https://bugzilla.mozilla.org/show_bug.cgi?id=469718 landed on tracemonkey yesterday which enables the ability to run the JavaScript tests in the browser using the reftest framework. To run the tests, do make -C $(OBJDIR) jstestbrowser https://bugzilla.mozilla.org/show_bug.cgi?id=509629 tracks dmandelin's new shell test framework which will use the same manifests. One change you will see right away is the JavaScript tests have moved from js/tests to js/src/tests. All new tests should be placed in js/src/tests. The other change is the existence of the reftest-style manif