I just discovered that they currently go through IsTransparent stuff on Xray wrapper, which is just wrong (because the waive isn't deep). I'm in the middle of some other stuff, so I'll fix this separately. (It's a shame - this bug would let us get rid of the IsTransparent() case on XrayWrapper, except that I'm adding a new use for that case in bug 821850).

My patches for this have broadened a bit, so updating the title here.

Hm, there's a browser-chrome failure that I can't reproduce locally. Let's see what happens on mac: https://tbpl.mozilla.org/?tree=Try&rev=c6fb7ce793bb

Oops, it's "macosx64", not "macosx": https://tbpl.mozilla.org/?tree=Try&rev=ac3e1ce17133

Fixed that bug, should be ready. Uploading patches and flagging blake for review. Aside from general refactoring and cleanup, these patches make nsExpandedPrincipal(foo) have a proper asymmetric relationship with foo, so it blocks bug 821850.

Every time the layout of CompartmentPrivate changes, I forget to rebuild in caps/ and spend half an hour wondering what the heck is going on. :-(

There's a browser-chrome test that does this, which means that _all_ subsequent browser-chrome tests inherit it. So depending on the ordering of cases in WrapperFactory, we might end up using a CrossCompartmentWrapper rather than an XrayWrapper, meaning that stuff like nodePrincipal doesn't work anymore. The semantics of UniversalXPConnect are now entirely dicatated by what makes our test suite go green. So let's not force ourselves to bend over backwards during wrapping to handle this case. And let's fix that stupid test while we're at it.

__scriptOnly__ is unused on mxr and addons-mxr. Morevoer, the current implementation is totally broken, because we check for NNXOW, which only happens when a random content JS object ends up in some other cross-origin scope (via addons, presumably), whereas chrome objects use ChomeObjectWrapper. I'm soon going to replace SCRIPT_ACCESS_ONLY with checked unwrapping, and mark all COWs as unsafe to unwrap (see bug 821573 and bug 658909). So let's just kill this thing here.

I noticed this nonfatal assertion firing, unrelated to my patches. Leaking the holder is not so great. Let's fix this for real.

There's no reason to be doing a dynamic check here, given that the JSClasses will never match. Lets be explicit and safe.

wantXrays means that the sandbox wants Xray wrappers even when accessing same- origin content. The default is true, which Blake says has something to do with GreaseMonkey and days of old. This flag never had an effect for chrome, because the chrome->chrome case always short-circuited to &CrossCompartmentWrapper::singleton. But once we start respecting the flag as a general-purpose indicator that Xrays should be applied same-origin, we need to either add a special case in Rewrap or make the flag reflect reality. The latter seems cleaner and more sane. However, things are complicated by the fact that there's also a completely different, orthogonal usage, whereby setting wantXrays to false implicitly waives Xray on the returned sandbox _and_ on any results returned from evalInSandbox. This is just nuts. The former can be accomplished by callers manually using .wrappedJSObject, and the latter by having EvalInSandbox transitively apply waivers from their sandbox arguments. I've updated the documentation on the MDN page so that it only describes the reasonable usage. The next step is to get rid of the crazy behavior. I think the best path of migration is to have wantXrays: false keep implicitly waiving, but waive return values from EvalInSandbox based on whether the argument was waived. This patch does that.

There's no reason to do this any different than we do for XOWs and such. The only thing this might conceivably support would be certain chrome XPWNs-as-COWs. But that would require that they forced a parent in precreate without being flagged as DOM objects in classinfo. And it's not clear why we'd want to support that. And we're generally moving away from COWs anyway.

This is generally cleaner, and avoids potentially calling these functions multiple times when we start moving this stuff around.

It's pretty orthogonal, and makes the critical block more complicated than it needs to be.

This paves the way for more rule-based selection of wrappers in the common case.

All the casese where we want to waive should now be going through WaiveXrayWrapper.

We tack these onto the tests from bug 812415, adding coverage for nsExpandedPrincipal and making sure that the waivers are deep. We also take the opportunity to check the asymmetric security relationship between a principal and its corresponding nsEP.

Comment on attachment 695882 [details] [diff] [review] Part 13 - Stop checking for Xray waivers in the Xray machinery. v1 Review of attachment 695882 [details] [diff] [review]: ----------------------------------------------------------------- ::: js/xpconnect/wrappers/XrayWrapper.cpp @@ +1239,5 @@ > > bool > IsTransparent(JSContext *cx, JSObject *wrapper) > { > + return false; So, uh, why keep this function?

(In reply to :Ms2ger from comment #20) > Comment on attachment 695882 [details] [diff] [review] > Part 13 - Stop checking for Xray waivers in the Xray machinery. v1 > > Review of attachment 695882 [details] [diff] [review]: > ----------------------------------------------------------------- > > ::: js/xpconnect/wrappers/XrayWrapper.cpp > @@ +1239,5 @@ > > > > bool > > IsTransparent(JSContext *cx, JSObject *wrapper) > > { > > + return false; > > So, uh, why keep this function? Because I'm using it again in bug 821850.

I haven't forgot about this -- I've been super busy with b2g stuff and will be until next week.

Comment on attachment 695879 [details] [diff] [review] Part 10 - Move COW prototype remapping out of wrapper selection. v1 Review of attachment 695879 [details] [diff] [review]: ----------------------------------------------------------------- r=me with the question answered. ::: js/xpconnect/wrappers/WrapperFactory.cpp @@ -355,5 @@ > } > } else if (xpc::IsUniversalXPConnectEnabled(target)) { > wrapper = &CrossCompartmentWrapper::singleton; > } else if (originIsChrome) { > - JSFunction *fun = JS_GetObjectFunction(obj); Why bother moving this? @@ +418,5 @@ > + // The prototype chain of COW(obj) looks lke this: > + // > + // COW(obj) => COW(foo) => COW(bar) => contentWin.StandardClass.prototype > + if (wrapper == &ChromeObjectWrapper::singleton) { > + Nit: extra newline here.

Comment on attachment 695880 [details] [diff] [review] Part 11 - Hoist special cases to the top of WrapperFactory::Rewrap. v1 Review of attachment 695880 [details] [diff] [review]: ----------------------------------------------------------------- ::: js/xpconnect/wrappers/WrapperFactory.cpp @@ +351,5 @@ > + > + // If content is accessing a Components object or NAC, we need a special filter, > + // even if the object is same origin. > + } else if (IsComponentsObject(obj) && !AccessCheck::isChrome(target)) { > + wrapper = &FilteringWrapper<CrossCompartmentSecurityWrapper, This branch and the one below it are over-indented.

(In reply to Blake Kaplan (:mrbkap) from comment #24) > Comment on attachment 695879 [details] [diff] [review] > Part 10 - Move COW prototype remapping out of wrapper selection. v1 > > Review of attachment 695879 [details] [diff] [review]: > ----------------------------------------------------------------- > > r=me with the question answered. > > ::: js/xpconnect/wrappers/WrapperFactory.cpp > @@ -355,5 @@ > > } > > } else if (xpc::IsUniversalXPConnectEnabled(target)) { > > wrapper = &CrossCompartmentWrapper::singleton; > > } else if (originIsChrome) { > > - JSFunction *fun = JS_GetObjectFunction(obj); > > Why bother moving this? Because I want wrapper computation to be as succinct as possible. I don't consider that check to be very important to someone skimming over our wrapper behavior, so I wanted to move it to the "Junk we do for COWs" section, below.

Comment on attachment 695881 [details] [diff] [review] Part 12 - Replace security wrapper enumeration with a more rule-based approach. v1 Review of attachment 695881 [details] [diff] [review]: ----------------------------------------------------------------- Very nice patch. I'm not entirely convinced that this is easier to understand or to tweak but I see why it's useful.

Pushed to inbound: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?changeset=c356aef54656 This was green on try (comment 2, comment 5), and I've also pushed other try pushes on top of it, such as bug 821850 comment 12.

https://hg.mozilla.org/mozilla-central/rev/7198eb5958a2 https://hg.mozilla.org/mozilla-central/rev/a6059bb2bc6c https://hg.mozilla.org/mozilla-central/rev/1e063f6c170a https://hg.mozilla.org/mozilla-central/rev/d8c8bfd7f3e4 https://hg.mozilla.org/mozilla-central/rev/1aa557c6712b https://hg.mozilla.org/mozilla-central/rev/228181ec5626 https://hg.mozilla.org/mozilla-central/rev/c5d438850b86 https://hg.mozilla.org/mozilla-central/rev/29eae99e14f7 https://hg.mozilla.org/mozilla-central/rev/96b499530382 https://hg.mozilla.org/mozilla-central/rev/588adb5ab729 https://hg.mozilla.org/mozilla-central/rev/43be433d5832 https://hg.mozilla.org/mozilla-central/rev/309a0c60b871 https://hg.mozilla.org/mozilla-central/rev/16afc05bc402 https://hg.mozilla.org/mozilla-central/rev/c356aef54656