New issue
Advanced search Search tips

Issue 883496 link

Starred by 2 users
Status: Started
Owner:
scottmg@chromium.org
Cc:
w...@chromium.org
seantopping@chromium.org
haraken@chromium.org
ianloic@google.com
Components:
EstimatedDays: ----
NextAction: ----
OS: Fuchsia
Pri: 3
Type: Bug

Blocked on:
issue 902969
issue 893230


Sign in to add a comment

fuchsia: JS FIDL bindings

Project Member Reported by scottmg@chromium.org, Sep 12 
To facilitate exposing things to other things.

Comment 1 by scottmg@chromium.org, Sep 12

Owner: scottmg@chromium.org
Status: Started (was: Untriaged)

Comment 2 by scottmg@chromium.org, Sep 12

Cc: seantopping@chromium.org w...@chromium.org ianloic@google.com
Very (very) rough prototyping/hacking here: https://chromium-review.googlesource.com/c/chromium/src/+/1222697

This is targeting a Chromium-friendly JS output following: https://chromium.googlesource.com/chromium/src/+/master/styleguide/web/web.md#JavaScript.

Some implications:
- Annotating types with Closure annotations
- No TS
- Assuming a recent-ish JS engine

Starting with the Dart fidlgen as a base, as well as design and tests that ianloic@ has done on a snazzy in-Fuchsia generator written in TS.

Also TBD, I forked a few useful files from Garnet's fidl compiler which seemed helpful to unmarshal/resolve the json IR.

As of this writing it only emits some borked hybrid mess of JS (enum, const, struct) and Dart (the rest). i.e. not much to look at yet.

I also have no idea how it's actually going to do any calling, or get access to services, or hook into the build system, or really do much of anything in Chromium yet. So I'm not sure at all that this is the solution to any problem yet.

Comment 3 by scottmg@chromium.org, Sep 13 

Yeah, I dunno. Doing this to target Closure instead of TypeScript seems pretty silly and awkward. Hrm.

Comment 4 by haraken@chromium.org, Oct 1 

Very native question: What's a difference between FIDL and Web IDL? In other words, why can't Fuchsia simply use Web IDL to expose things?

Comment 5 by w...@chromium.org, Oct 1 

Re #4: FIDL is Fuchsia IDL, the platform-native IPC format.

For various applications it will be useful to be able to have web-based content that is aware of platform/application specific (non-web-standard) APIs, equivalent to what we have today on various platforms with WebView.

The FIDL/JS bindings described here will allow custom APIs to be injected for these kinds of use-cases, similar to the special Mojo/JS API bindings that e.g. Chrome's Media Router provides to its component extension.

Perhaps I'm misunderstanding what you're suggesting with "simply use WebIDL", though?

Comment 6 by haraken@chromium.org, Oct 3

Cc: haraken@chromium.org
My question is: Blink already has a Web IDL compiler. Why can't you reuse the IDL compiler for FIDL? In other words, does FIDL have a different syntax from Web IDL?

For example, it looks like this CL (https://chromium-review.googlesource.com/c/chromium/src/+/1222697) is implementing a parser of IDL. Can you probably reuse the existing IDL parser (https://cs.chromium.org/chromium/src/tools/idl_parser/?q=idl_parser&sq=package:chromium&dr)?

Comment 7 by scottmg@chromium.org, Oct 3 

Hi, thanks for taking a look.

There's no parser here, just a generator. Yes, FIDL has a different syntax than Web IDL. The FIDL language is specified by https://fuchsia.googlesource.com/zircon/+/master/docs/fidl/language.md and is used to specify communication protocols between services on Fuchsia at the OS level.

`fidlc` which is provided by the Fuchsia SDK parses the .fidl files and generates a type-resolved (sort of flattened) .json file representing the DOM the fidl represents. The compiler in the CL you linked then converts that json representation to .js code that can encode and decode messages in the format specified by https://fuchsia.googlesource.com/zircon/+/master/docs/fidl/wire-format/index.md.

Other similar tools include one written in Go, which generates bindings for
C++, Go, Rust, and Syzkaller: https://fuchsia.googlesource.com/garnet/+/master/go/src/fidl/compiler/backend (which is maintained by the Fuchsia team), and a Dart bindings generator https://fuchsia.googlesource.com/topaz/+/master/bin/fidlgen_dart/. The Fuchsia team does not maintain bindings for JavaScript though.

So, I'm attempting to make an equivalent JavaScript version to those generators. For some APIs (for example in Cast) we would like to allow the .js to send and receive FIDL messages directly rather than writing send/receive in C++ for all of the APIs and then adding wrappers to expose that functionality to JS.

In https://chromium-review.googlesource.com/c/chromium/src/+/1222697, I included a unittest binary to exercise the bindings. I'd like to run those tests by using v8 of course, hence the new introduced dependency.

Comment 8 by haraken@chromium.org, Oct 3 

Thanks for the details!

(I'm wondering why FIDL needs to have a different syntax than Web IDL but that's not something I should have a strong opinion about. I understand the situation :-)

Comment 9 by scottmg@chromium.org, Oct 3 

No idea! :) It's a reasonable question, but I don't really know much of the background on how the team settled on fidl other than it originally evolved from Mojo and .mojom files. I suspect some of the differences come from differing design requirements. For example, fidl is used to communicate with quite low-level services, so messages must be encode/decodable in plain C without any dynamic memory allocation.
Project Member

Comment 10 by bugdroid1@chromium.org, Oct 3 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/4411204b3a414c296efa418d9cbe56d52119e1b5

commit 4411204b3a414c296efa418d9cbe56d52119e1b5
Author: Scott Graham <scottmg@chromium.org>
Date: Wed Oct 03 19:46:26 2018

fuchsia: Beginnings of FIDL JS bindings

This is the very initial stages of writing a bindings generator to allow
interop to/from JS with FIDL.

The main pieces are:
- the build time generator which consumes the JSON IR from `fidlc` and
  outputs a fidl.js
- integration with gn in fidl_library.gni
- a JS runtime support library (mostly helps with message encoding)
- new unittest binary (which also currently holds some Zircon
  integration points that will eventually be moved to a C++ runtime
  support library.)

There are many, many things it does not yet handle, e.g. responses
(either sync or async), events, passing structs, passing unions, ...

For reference, test/simple.fidl generates this
https://gist.github.com/sgraham/3935c64d5bc1b67eea1f6e3b38fef6f1


Bug: 883496
Change-Id: I58e5d2b81213f20a3198b68cbd2bc01660ba58f7
Reviewed-on: https://chromium-review.googlesource.com/c/1222697
Reviewed-by: Jeremy Roman <jbroman@chromium.org>
Reviewed-by: Adam Klein <adamk@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#596329}
[modify] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/BUILD.gn
[modify] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/build/config/fuchsia/fidl_library.gni
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/BUILD.gn
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/DEPS
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/fidl.py
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/gen.py
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/runtime/fidl.mjs
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/runtime/zircon.cc
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/runtime/zircon.h
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/test/simple.fidl
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/third_party/__init__.py
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/third_party/enum34/LICENSE
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/third_party/enum34/README.chromium
[add] https://crrev.com/4411204b3a414c296efa418d9cbe56d52119e1b5/tools/fuchsia/fidlgen_js/third_party/enum34/__init__.py

Comment 11 by w...@chromium.org, Oct 14

Blockedon: 893230

Comment 12 by w...@chromium.org, Oct 16

Blocking: 893229
Project Member

Comment 13 by bugdroid1@chromium.org, Oct 17 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e

commit d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e
Author: Scott Graham <scottmg@chromium.org>
Date: Wed Oct 17 01:57:27 2018

fuchsia: Add support for FIDL/JS responses

Responses are always returned asynchronously as JS promises (the
newly added zx.objectWaitOne() helper also returns a promise).

The wait and response are serviced by libasync waits, and so must be
cancelled when the v8 context is torn down. The (default) dispatcher
must outlive any v8 contexts with which it is used.

Bug: 883496
Change-Id: Ib8735c6ec4196d36b53929341d57e843abaeea3b
Reviewed-on: https://chromium-review.googlesource.com/c/1240204
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Reviewed-by: Jeremy Roman <jbroman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#600238}
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/gin/converter.cc
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/gin/converter.h
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/gin/public/gin_embedders.h
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/BUILD.gn
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/runtime/fidl.mjs
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/runtime/zircon.cc
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/runtime/zircon.h
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/d2f4ed4be95bf6d9fb5445a85baeba40ef51db5e/tools/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 16 by bugdroid1@chromium.org, Oct 17 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/24f11a7b6de9857295bef2b1ebbb512aedd828c0

commit 24f11a7b6de9857295bef2b1ebbb512aedd828c0
Author: Scott Graham <scottmg@chromium.org>
Date: Wed Oct 17 21:28:53 2018

fuchsia: FIDL JS: format gen.py with yapf --style=chromium

Also, don't import * from fidl to avoid pylint complaining on Gerrit.

With a couple tweaks to avoid 80col being exceeded for presubmit. No
intended functionality change (and the generated fidl.js is unchanged).

Bug: 883496
Change-Id: If4335f92d7630dd4b543dc686d2baf6aca5d2b6e
Reviewed-on: https://chromium-review.googlesource.com/c/1258087
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#600562}
[modify] https://crrev.com/24f11a7b6de9857295bef2b1ebbb512aedd828c0/tools/fuchsia/fidlgen_js/gen.py
Project Member

Comment 18 by bugdroid1@chromium.org, Oct 23 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2d91b4aa42886d07d770be6df218c014e967f2f1

commit 2d91b4aa42886d07d770be6df218c014e967f2f1
Author: Scott Graham <scottmg@chromium.org>
Date: Tue Oct 23 17:21:18 2018

fuchsia: Simplify exposed bindings names

Also, remove channelCreate and handleClose since they aren't yet useful,
and we'll replace them with something better soon (MessagePort,
Transferrable, etc.)

Bug: 883496
Change-Id: I3861398035e79029cdbde01cc9a4ce4ea872ab0b
Reviewed-on: https://chromium-review.googlesource.com/c/1292721
Reviewed-by: Kevin Marshall <kmarshall@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#601999}
[modify] https://crrev.com/2d91b4aa42886d07d770be6df218c014e967f2f1/tools/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/2d91b4aa42886d07d770be6df218c014e967f2f1/tools/fuchsia/fidlgen_js/runtime/fidl.mjs
[modify] https://crrev.com/2d91b4aa42886d07d770be6df218c014e967f2f1/tools/fuchsia/fidlgen_js/runtime/zircon.cc
[modify] https://crrev.com/2d91b4aa42886d07d770be6df218c014e967f2f1/tools/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
Project Member

Comment 20 by bugdroid1@chromium.org, Nov 5 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/bd60a551e79c6a8bf31e0eadb7abafe11c81b147

commit bd60a551e79c6a8bf31e0eadb7abafe11c81b147
Author: Scott Graham <scottmg@chromium.org>
Date: Mon Nov 05 20:51:41 2018

Fuchsia: Implement handling of arrays in FIDL/JS

Bug: 883496
Change-Id: I4981cf288566aed5f95e3ca3b8632e5462203ee3
Reviewed-on: https://chromium-review.googlesource.com/c/1311262
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#605455}
[modify] https://crrev.com/bd60a551e79c6a8bf31e0eadb7abafe11c81b147/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/bd60a551e79c6a8bf31e0eadb7abafe11c81b147/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/bd60a551e79c6a8bf31e0eadb7abafe11c81b147/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 22 by bugdroid1@chromium.org, Nov 6 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a0ca0340f2abe5d8d0d6427c8d3bb47d6b7f7665

commit a0ca0340f2abe5d8d0d6427c8d3bb47d6b7f7665
Author: Scott Graham <scottmg@chromium.org>
Date: Tue Nov 06 00:22:09 2018

Fuchsia: Implement receive of unions to JS from C++

Bug: 883496
Change-Id: I402ef87f893488bcd15ec171e18add9918022a95
Reviewed-on: https://chromium-review.googlesource.com/c/1313570
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#605528}
[modify] https://crrev.com/a0ca0340f2abe5d8d0d6427c8d3bb47d6b7f7665/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/a0ca0340f2abe5d8d0d6427c8d3bb47d6b7f7665/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/a0ca0340f2abe5d8d0d6427c8d3bb47d6b7f7665/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 23 by bugdroid1@chromium.org, Nov 6 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/6de6f090852edc44b9c97fb80bcf22c0593786a2

commit 6de6f090852edc44b9c97fb80bcf22c0593786a2
Author: Scott Graham <scottmg@chromium.org>
Date: Tue Nov 06 05:40:41 2018

Fuchsia: Handling of more cases for default initialization in FIDL/JS

Initializing a struct member with the value of an enum would previously
fail to compile; implement this. This is slightly awkward because the
front end does not scope the enum value to the enum like the C++ (and
JS) bindings but rather only to the library. So the compiler needs to
know the type to which it's being assigned and use that to determine how
the enum's value is intended to be qualified.

Bug: 883496
Change-Id: Id8a9f75d07bb0acb11963ecbd37d4915b36c28e0
Reviewed-on: https://chromium-review.googlesource.com/c/1313286
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#605601}
[modify] https://crrev.com/6de6f090852edc44b9c97fb80bcf22c0593786a2/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/6de6f090852edc44b9c97fb80bcf22c0593786a2/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/6de6f090852edc44b9c97fb80bcf22c0593786a2/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 24 by bugdroid1@chromium.org, Nov 6 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d32c33c2032b0d72d0831272184fa267d3813d43

commit d32c33c2032b0d72d0831272184fa267d3813d43
Author: Scott Graham <scottmg@chromium.org>
Date: Tue Nov 06 18:19:15 2018

Fuchsia: Implement vector<string>, fix traversal order

Previously the body of structs was traversed in-order, but out-of-line
items were deferred to a queue, making traversal mostly breadth-first.
Instead, fix encoding by emitting any out-of-line items by allocating
(and filling) directly following the current item.

Bug: 883496
Change-Id: If157930521e4cf5f89f702207defe91beb862def
Reviewed-on: https://chromium-review.googlesource.com/c/1316429
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#605745}
[modify] https://crrev.com/d32c33c2032b0d72d0831272184fa267d3813d43/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/d32c33c2032b0d72d0831272184fa267d3813d43/build/fuchsia/fidlgen_js/runtime/fidl.mjs
[modify] https://crrev.com/d32c33c2032b0d72d0831272184fa267d3813d43/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/d32c33c2032b0d72d0831272184fa267d3813d43/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 25 by bugdroid1@chromium.org, Nov 6 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cb745a3edd5eb1f3e26721e21ec43b8f1e5bcff4

commit cb745a3edd5eb1f3e26721e21ec43b8f1e5bcff4
Author: Scott Graham <scottmg@chromium.org>
Date: Tue Nov 06 21:49:23 2018

Fuchsia: Implement vectors-of-more-things in FIDL/JS

Vectors of primitive types and strings were previously implemented; this
extends support to vectors of unions and structs.

Bug: 883496
Change-Id: Id00174b05d04c3f432fae39112afde658e533806
Reviewed-on: https://chromium-review.googlesource.com/c/1318676
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#605826}
[modify] https://crrev.com/cb745a3edd5eb1f3e26721e21ec43b8f1e5bcff4/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/cb745a3edd5eb1f3e26721e21ec43b8f1e5bcff4/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/cb745a3edd5eb1f3e26721e21ec43b8f1e5bcff4/build/fuchsia/fidlgen_js/test/simple.fidl

Comment 26 by scottmg@chromium.org, Nov 8

Blockedon: 902969
Project Member

Comment 27 by bugdroid1@chromium.org, Nov 8 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/ee2630a9c5f0d653041c8d05ce475afc161b23bd

commit ee2630a9c5f0d653041c8d05ce475afc161b23bd
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Nov 08 00:59:44 2018

Fuchsia: Add missing FIDL/JS vector cases for bool, float32, float64

This also required implementing the type tables for int64/uint64 as
BigInt. Additionally, some test support code to pull BigInt out of V8 as
64-bit types because gin assumes always-Number.

Bug: 883496
Change-Id: I10d07124e59f0fdde8f053007eca30508751b211
Reviewed-on: https://chromium-review.googlesource.com/c/1318729
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606261}
[modify] https://crrev.com/ee2630a9c5f0d653041c8d05ce475afc161b23bd/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/ee2630a9c5f0d653041c8d05ce475afc161b23bd/build/fuchsia/fidlgen_js/runtime/fidl.mjs
[modify] https://crrev.com/ee2630a9c5f0d653041c8d05ce475afc161b23bd/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/ee2630a9c5f0d653041c8d05ce475afc161b23bd/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 28 by bugdroid1@chromium.org, Nov 8 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c438e9912ea8b48536b977598f908162b7f6b295

commit c438e9912ea8b48536b977598f908162b7f6b295
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Nov 08 02:37:40 2018

Fuchsia: Implement passing of vector<handle<X>> in FIDL/JS

Bug: 883496
Change-Id: I2a769b4779ff0f56622ed62099ee0b620aefb16e
Reviewed-on: https://chromium-review.googlesource.com/c/1318745
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606295}
[modify] https://crrev.com/c438e9912ea8b48536b977598f908162b7f6b295/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/c438e9912ea8b48536b977598f908162b7f6b295/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/c438e9912ea8b48536b977598f908162b7f6b295/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 29 by bugdroid1@chromium.org, Nov 8 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/987097ee69f35bdc8b5ac93c489c944b30eb580c

commit 987097ee69f35bdc8b5ac93c489c944b30eb580c
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Nov 08 05:25:07 2018

Fuchsia: Fix handling of nullable vectors in FIDL/JS

Bug: 883496
Change-Id: I656ef4172089f65b6febf0ba3075a510bb4e9e11
Reviewed-on: https://chromium-review.googlesource.com/c/1318835
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606354}
[modify] https://crrev.com/987097ee69f35bdc8b5ac93c489c944b30eb580c/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/987097ee69f35bdc8b5ac93c489c944b30eb580c/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/987097ee69f35bdc8b5ac93c489c944b30eb580c/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 30 by bugdroid1@chromium.org, Nov 8 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/8be79151580abce8c361e17f74f61b2c961e3a00

commit 8be79151580abce8c361e17f74f61b2c961e3a00
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Nov 08 17:55:41 2018

Fuchsia: Improve handling of constant values in FIDL/JS

- Properly suffixes BigInt constants with 'n'
- Fixes stringification of constants into valid .js syntax
- Improve default values handling in structs which was only partially
  implemented previously

Bug: 883496
Change-Id: Id270d2f34dd40fa970d58448c1002c691db35a87
Reviewed-on: https://chromium-review.googlesource.com/c/1321577
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606529}
[modify] https://crrev.com/8be79151580abce8c361e17f74f61b2c961e3a00/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/8be79151580abce8c361e17f74f61b2c961e3a00/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/8be79151580abce8c361e17f74f61b2c961e3a00/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 31 by bugdroid1@chromium.org, Nov 8 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1453d9de047aedd6b084f3abdacb858109c0f576

commit 1453d9de047aedd6b084f3abdacb858109c0f576
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Nov 08 19:40:26 2018

Fuchsia: Add support for vector of enum in FIDL/JS

Bug: 883496
Change-Id: Ib611c2d195b401475bb343099c8b12a644e942a7
Reviewed-on: https://chromium-review.googlesource.com/c/1324095
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606569}
[modify] https://crrev.com/1453d9de047aedd6b084f3abdacb858109c0f576/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/1453d9de047aedd6b084f3abdacb858109c0f576/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/1453d9de047aedd6b084f3abdacb858109c0f576/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 32 by bugdroid1@chromium.org, Nov 8 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a3e345885f6e31a236ec75efabf8ac2e61c93041

commit a3e345885f6e31a236ec75efabf8ac2e61c93041
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Nov 08 21:56:06 2018

Fuchsia: Handle references to types later in file in FIDL/JS

Bug: 883496
Change-Id: Ic8973eb46a68ecf09a57572b1d694cb4812c897f
Reviewed-on: https://chromium-review.googlesource.com/c/1324209
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606613}
[modify] https://crrev.com/a3e345885f6e31a236ec75efabf8ac2e61c93041/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/a3e345885f6e31a236ec75efabf8ac2e61c93041/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 33 by bugdroid1@chromium.org, Nov 9 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/6c65375765dac98249a8b3bfe596845bf0bae3d7

commit 6c65375765dac98249a8b3bfe596845bf0bae3d7
Author: Scott Graham <scottmg@chromium.org>
Date: Fri Nov 09 00:37:32 2018

Fuchsia: Implement request<Interface> in FIDL/JS

Fairly basic and the connection must be manually torn down for now, but
allows requesting an interface implemention from the JS side.

Bug: 883496
Change-Id: I8367fd50097403ee2f9911657afbabdd8d461a41
Reviewed-on: https://chromium-review.googlesource.com/c/1324929
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606663}
[modify] https://crrev.com/6c65375765dac98249a8b3bfe596845bf0bae3d7/build/fuchsia/fidlgen_js/gen.py
[modify] https://crrev.com/6c65375765dac98249a8b3bfe596845bf0bae3d7/build/fuchsia/fidlgen_js/runtime/zircon.cc
[modify] https://crrev.com/6c65375765dac98249a8b3bfe596845bf0bae3d7/build/fuchsia/fidlgen_js/test/fidlgen_js_unittest.cc
[modify] https://crrev.com/6c65375765dac98249a8b3bfe596845bf0bae3d7/build/fuchsia/fidlgen_js/test/simple.fidl
Project Member

Comment 34 by bugdroid1@chromium.org, Nov 27 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/52f0d9110df3466dc3452a44b0b136d20b181534

commit 52f0d9110df3466dc3452a44b0b136d20b181534
Author: Dan Elphick <delphick@chromium.org>
Date: Tue Nov 27 20:03:27 2018

[cleanup] Fix deprecated V8 function usage in zircon.cc

Fixes uses of deprecated Object::Set and FunctionTemplate::GetFunction
to use Context version that return Maybe values.

Bug: v8:7283, v8:7295, v8:8238, chromium:883496
Change-Id: Ia400ec8b36d138b9f6b87c4a91f7ccb12c2eeefa
Reviewed-on: https://chromium-review.googlesource.com/c/1352314
Commit-Queue: Dan Elphick <delphick@chromium.org>
Reviewed-by: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#611276}
[modify] https://crrev.com/52f0d9110df3466dc3452a44b0b136d20b181534/build/fuchsia/fidlgen_js/runtime/zircon.cc

Comment 35 by w...@chromium.org, Jan 19 (4 days ago)

Blocking: -893229
►

Sign in to add a comment