[TIMOB-23903] Build against V8 5.7.492.71 #8382
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sgtcoolguy
changed the title
sgtcoolguy
force-pushed
the
v8-5.3.332.41
branch
from
8bcb990
to
d53225b
Compare
sgtcoolguy
changed the title
sgtcoolguy
force-pushed
the
v8-5.3.332.41
branch
from
4b86202
to
0369fb4
Compare
garymathews
suggested changes
Apr 27, 2017
| Local<Function> constructor; | ||
| MaybeLocal<Function> maybeConstructor = pt->GetFunction(context); | ||
| if (!maybeConstructor.ToLocal(&constructor)) { | ||
| V8Util::fatalException(isolate, tryCatch); |
There was a problem hiding this comment.
titanium::V8Util::fatalException(isolate, tryCatch);| MaybeLocal<Object> maybeInstance = constructor->NewInstance(context); | ||
| Local<Object> moduleInstance; | ||
| if (!maybeInstance.ToLocal(&moduleInstance)) { | ||
| V8Util::fatalException(isolate, tryCatch); |
There was a problem hiding this comment.
titanium::V8Util::fatalException(isolate, tryCatch);
…lates, so the Set*Method calls needed to set the generated FunctionTemplate (and not it's function). Also need to hack around hanging an External wrapping a jclass on the __javaClass_ property by using a natiev data property getter callback that gets the External passed as Data and returning that.
…gher version/moduleApiVersion against a build of the SDK
…mplate must be throwing some uncaught excpetion, so the toLocalChecked() on GetFunction is failing.
…with a proxy that we then go back and kill the proxy/JS object on the C++ side. Tehre's still an issue here where we're not cleaning up jclass references somehow that I need to fix...
…undary to Proxy instances
…hey don't *all* barf/crash
sgtcoolguy
force-pushed
the
v8-5.3.332.41
branch
from
131d63f
to
4310252
Compare
garymathews
approved these changes
May 2, 2017
|
@garymathews Given that they've moved away from the older debug protocol, I think e should look into simply updating studio's debugger for Android to work with the new inspector protocol so we can update V8 to latest. It shouldn't be too bad since I already had to write the inspector protocol plus extra layers for iOS/JSCore. Here's the studio ticket for updating the Android debugger (again!): https://jira.appcelerator.org/browse/TISTUD-8750 |
|
@sgtcoolguy So it's no breaking change? Otherwise, we cannot cause all native modules on Github and our marketplace to be recompiled again, especially not in a non-major release version :(. |
sgtcoolguy
added a commit
that referenced
this pull request
Jul 21, 2017
This reverts commit 193e13e.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
https://jira.appcelerator.org/browse/TIMOB-23903
This updates Android's runtime to build against V8 5.7.492.71
V8 introduced yet more breaking API changes and I was unable to keep that from propagating out to our native modules. So this is a BREAKING NATIVE MODULE CHANGE!
Specifically, V8 removed WeakCallbackData and moved to WeakCallbackInfo. That was a minor update and I think could have avoided breaking module compatibility.
The larger change they made was to restrict values you can set as properties on a Template. We were setting actual Function objects on the prototype and instance templates for the types - we needed to change to setting the FunctionTemplate itself, not the function it generated. Secondly, it meant we could set an External wrapping the class (java class in JNI) that the proxy represented. This was used to instantiate instances of the Java class itself when constructing a proxy object in JS-world. To workaround this, instead of explicitly setting that External as the value for javaClass, I now set a "native data property" for that property name with a getter callback. This means it's a property that always goes straight to the native callback (can't be overridden in JS). And the data value we set with it gets passed to every callback invocation, so we use that as a sneaky "hack" to pass along the External wrapping the java class. The callback method simply grabs that value and returns it.
Relates to tidev/v8_titanium#12