8177cabc4e22a9f4f9bdbd241101e020b6bbd62c: Bug 844048 - Track uses of 'var arguments' within eval scripts, always make arguments objects for generators, r=luke.
Brian Hackett <bhackett1024@gmail.com> - Fri, 22 Feb 2013 09:29:28 -0700 - rev 122658
Push 23419 by bhackett@mozilla.com at Fri, 22 Feb 2013 16:29:37 +0000
Bug 844048 - Track uses of 'var arguments' within eval scripts, always make arguments objects for generators, r=luke.
4d301b2bcad047442fa34470893aff52b4df73ee: Bug 836301 - Hoist some assertions, remove a bunch of no-op trap overrides, and add assertions that we've entered our policy. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:34 -0800 - rev 122657
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Hoist some assertions, remove a bunch of no-op trap overrides, and add assertions that we've entered our policy. r=mrbkap
ab07392f24241194e485b42e9c5b598bba5aaf87: Bug 836301 - Hoist enter() calls from {Xray,}Wrapper::foo into Proxy::foo. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:34 -0800 - rev 122656
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Hoist enter() calls from {Xray,}Wrapper::foo into Proxy::foo. r=mrbkap
f2fa2d0bb5e6bb1db92006a8f51eec1f54e927ec: Bug 836301 - Introduce an RAII class for entering policies. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:33 -0800 - rev 122655
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Introduce an RAII class for entering policies. r=mrbkap This will allow us to make some hard assertions that a given policy has been entered exactly once.
b5992ec030e3f3506cb6683d11680c320626f5e5: Bug 836301 - Hoist enter() into BaseProxyHandler. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:33 -0800 - rev 122654
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Hoist enter() into BaseProxyHandler. r=mrbkap
240174280d5258b0601c77398b4e18e48ef8b067: Bug 836301 - Add tracking for whether we have a non-trivial enter() trap. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:33 -0800 - rev 122653
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Add tracking for whether we have a non-trivial enter() trap. r=mrbkap This will allow us to skip the virtual function call for non-security-wrapper proxies, which are the cases where we care most about performance.
804898e09f6e5b58bbc0f759ff9162a455efb58e: Bug 836301 - Stop enter()ing with CALL for nativeCall. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:33 -0800 - rev 122652
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Stop enter()ing with CALL for nativeCall. r=mrbkap This is kind of nonsensical, because CALL means "the wrapped object is being called", whereas nativeCall means "the wrapped object is being unwrapped to have a JSNative invoked on it", which are two very different things. We _could_ add a NATIVECALL enter() trap, but our current policy enforcement around nativeCall involves overriding the trap itself, so we wouldn't use it for anything. So let's just get rid of it.
741efe957058b5f15da8fd649036abab6f065f5e: Bug 836301 - Add Special handling to allow us to call enter() for defineProperty on Xrays. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:33 -0800 - rev 122651
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Add Special handling to allow us to call enter() for defineProperty on Xrays. r=mrbkap
81db941b0b769cdbb57580e5f17b1cecda78fb79: Bug 836301 - Stop using JSRESOLVE_ASSIGNING to determine GET vs SET. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:32 -0800 - rev 122650
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Stop using JSRESOLVE_ASSIGNING to determine GET vs SET. r=mrbkap This is just a heuristic, anyway, and some of the usage is downright broken. There are two cases here: 1 - Deciding what to do for get{Own,}PropertyDescriptor. In these cases, we can just enter with GET and rely on the filtering machinery to filter out dangerous setters for security wrappers. 2 - Custom Xray props. None of these make sense in a |set| context. In fact, they generally have null setters anyway, so we can just assume GET. The policy-entering code in XrayWrapper is super haphazard. We'll get rid of it entirely later in these patches.
e0632e63909741eb1617fc46bce9a67ec4aa6724: Bug 836301 - Assert against JSID_VOID in JSID wrapping. r=mrbkap
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 08:14:32 -0800 - rev 122649
Push 23418 by bobbyholley@gmail.com at Fri, 22 Feb 2013 16:14:54 +0000
Bug 836301 - Assert against JSID_VOID in JSID wrapping. r=mrbkap Luke explained to me that it should never get there.
bee4a533e23ef69a5a44883d987727a94d6e504c: Bug 839647 - Remove crash test workaround r=jesup
Adam Roach [:abr] <adam@nostrum.com> - Fri, 22 Feb 2013 10:08:54 -0600 - rev 122648
Push 23417 by adam@nostrum.com at Fri, 22 Feb 2013 16:10:48 +0000
Bug 839647 - Remove crash test workaround r=jesup
4d6b042ba9fa41cca882fb22145b0a39a7e3de44: Bug 839647 - Synchronize CCApp thread start-up r=ehugg
Adam Roach [:abr] <adam@nostrum.com> - Fri, 22 Feb 2013 10:08:47 -0600 - rev 122647
Push 23417 by adam@nostrum.com at Fri, 22 Feb 2013 16:10:48 +0000
Bug 839647 - Synchronize CCApp thread start-up r=ehugg
b332047abd6c27d470839c450540373d2f753dd5: Bug 843231 - Compute whether to use an XBL scope exactly once. r=bz
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 07:56:03 -0800 - rev 122646
Push 23416 by bobbyholley@gmail.com at Fri, 22 Feb 2013 15:56:18 +0000
Bug 843231 - Compute whether to use an XBL scope exactly once. r=bz
e177d717382fbfd2afd9f1c28f2286bbcf178798: Bug 843231 - Alter the XBL scope test to make it compatible with reading the dom.xbl_scopes only once per global. r=bz
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 07:56:03 -0800 - rev 122645
Push 23416 by bobbyholley@gmail.com at Fri, 22 Feb 2013 15:56:18 +0000
Bug 843231 - Alter the XBL scope test to make it compatible with reading the dom.xbl_scopes only once per global. r=bz
fe3729c8396d5b3e3f28093bbe4b064ff2faec01: Bug 843231 - Remove SetGlobal. r=bz
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 07:56:03 -0800 - rev 122644
Push 23416 by bobbyholley@gmail.com at Fri, 22 Feb 2013 15:56:18 +0000
Bug 843231 - Remove SetGlobal. r=bz This stuff has all been simplified significantly with CPG, and this is no longer necessary. We always have a global when we construct a scope, and it never changes.
35d293a7541d972478ebcb39b64ef041cf29733b: Bug 843231 - Move the nulling out of mPrototypeNoHelper from SetGlobal to RemoveWrappedNativeProtos. r=bz
Bobby Holley <bobbyholley@gmail.com> - Fri, 22 Feb 2013 07:56:02 -0800 - rev 122643
Push 23416 by bobbyholley@gmail.com at Fri, 22 Feb 2013 15:56:18 +0000
Bug 843231 - Move the nulling out of mPrototypeNoHelper from SetGlobal to RemoveWrappedNativeProtos. r=bz The only call-site for both of these is JSD->nsXPConnect::InitClasses, so this is equivalent. All this will go away soon anyhow when JSD goes away.
21d3ff83095f8b7ac17fae4ced3e76f3fc7ac7ac: Bug 842089. MediaStream doesn't implement EventTarget yet. r=bz
Robert O'Callahan <robert@ocallahan.org> - Fri, 22 Feb 2013 10:33:30 -0500 - rev 122642
Push 23415 by bzbarsky@mozilla.com at Fri, 22 Feb 2013 15:34:52 +0000
Bug 842089. MediaStream doesn't implement EventTarget yet. r=bz
05eeb90bbc31dee7082dae9fa5ac9f33f8ee856a: Bug 629801 Implement HTML5 <time> element. r=smaug, peterv, hsivonen
David Humphrey (:humph) <david.humphrey@senecacollege.ca> - Fri, 22 Feb 2013 10:07:42 -0500 - rev 122641
Push 23414 by bzbarsky@mozilla.com at Fri, 22 Feb 2013 15:14:08 +0000
Bug 629801 Implement HTML5 <time> element. r=smaug, peterv, hsivonen
b9b189c8951bacea8609b612dac61cbc1a678310: Bug 842561. Assert that our inheritance chain is correct for WebIDL objects. r=peterv
Boris Zbarsky <bzbarsky@mit.edu> - Fri, 22 Feb 2013 09:56:29 -0500 - rev 122640
Push 23414 by bzbarsky@mozilla.com at Fri, 22 Feb 2013 15:14:08 +0000
Bug 842561. Assert that our inheritance chain is correct for WebIDL objects. r=peterv We could drop the descriptor for Text if we changed nsIDocument::CreateTextNode to return an already_AddRefed<Text>, but then we'd need more casting in nsDocument.cpp for the XPCOM CreateTextNode. Not sure which way is better, really.
4c8080f73b606866f0a0e55a35714ab33ebecaf9: Bug 842726. Kill off GetterShim and the classinfo support for chrome-only stuff on Document now that we can do that via WebIDL. r=peterv
Boris Zbarsky <bzbarsky@mit.edu> - Fri, 22 Feb 2013 09:56:29 -0500 - rev 122639
Push 23414 by bzbarsky@mozilla.com at Fri, 22 Feb 2013 15:14:08 +0000
Bug 842726. Kill off GetterShim and the classinfo support for chrome-only stuff on Document now that we can do that via WebIDL. r=peterv
(0) -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 +300000 tip