[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation

[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation

APIs and trivial implementation, to unblock Chrome side dev.

BUG=chromium:697028
Review-Url: https://codereview.chromium.org/2763413003
Cr-Commit-Position: refs/heads/master@{#44053}
Committed: https://chromium.googlesource.com/v8/v8/+/78905107d3a6465f346d909fa7f1c22ffec144c2

[Mircea Trofin, 2017-03-22 18:06:45]

The CQ bit was checked by mtrofin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-03-22 18:06:53]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Mircea Trofin, 2017-03-22 18:13:16]

Description was changed from

==========
[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation

APIs and trivial implementation, to unblock Chrome side dev.

BUG=
==========

to

==========
[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation

APIs and trivial implementation, to unblock Chrome side dev.

BUG=chromium:697028
==========

[Mircea Trofin, 2017-03-22 18:13:16]

mtrofin@chromium.org changed reviewers:
+ ahaas@chromium.org, bradnelson@chromium.org

[commit-bot: I haz the power, 2017-03-22 18:13:23]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-22 18:13:24]

Dry run: Try jobs failed on following builders:
  v8_win_compile_dbg on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/3...)

[Mircea Trofin, 2017-03-22 18:16:54]

Added a skeleton API for streamed compilation on the V8 side, to
unblock my work on the Blink side.

[ahaas, 2017-03-22 18:41:58]

lgtm

[Mircea Trofin, 2017-03-22 19:38:16]

The CQ bit was checked by mtrofin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-03-22 19:38:21]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-22 19:45:10]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-22 19:45:11]

Dry run: Try jobs failed on following builders:
  v8_win_compile_dbg on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/3...)
  v8_win_nosnap_shared_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)

[Mircea Trofin, 2017-03-22 20:41:58]

The CQ bit was checked by mtrofin@chromium.org

[Mircea Trofin, 2017-03-22 20:41:58]

The patchset sent to the CQ was uploaded after l-g-t-m from ahaas@chromium.org
Link to the patchset: https://codereview.chromium.org/2763413003/#ps20001
(title: "fix leak")

[commit-bot: I haz the power, 2017-03-22 20:42:04]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-22 20:48:16]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-22 20:48:16]

Try jobs failed on following builders:
  v8_win_compile_dbg on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/3...)
  v8_win_nosnap_shared_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)

[Mircea Trofin, 2017-03-22 21:10:26]

The CQ bit was checked by mtrofin@chromium.org

[Mircea Trofin, 2017-03-22 21:10:26]

The patchset sent to the CQ was uploaded after l-g-t-m from ahaas@chromium.org
Link to the patchset: https://codereview.chromium.org/2763413003/#ps40001
(title: "appease msvc")

[commit-bot: I haz the power, 2017-03-22 21:10:34]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-22 21:15:04]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-22 21:15:05]

Try jobs failed on following builders:
  v8_win_nosnap_shared_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)

[bradnelson, 2017-03-22 22:34:16]

lgtm (sans compile failure on the bots)

[Mircea Trofin, 2017-03-22 23:19:30]

The CQ bit was checked by mtrofin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-03-22 23:19:38]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-22 23:25:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-22 23:25:29]

Dry run: Try jobs failed on following builders:
  v8_win_nosnap_shared_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)

[Mircea Trofin, 2017-03-23 00:53:58]

The CQ bit was checked by mtrofin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-03-23 00:54:02]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Mircea Trofin, 2017-03-23 01:10:55]

The CQ bit was checked by mtrofin@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-03-23 01:11:05]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Mircea Trofin, 2017-03-23 01:18:48]

Patchset #4 (id:60001) has been deleted

[Mircea Trofin, 2017-03-23 01:18:58]

Patchset #3 (id:40001) has been deleted

[Mircea Trofin, 2017-03-23 01:34:11]

Patchset #3 (id:80001) has been deleted

[commit-bot: I haz the power, 2017-03-23 01:37:52]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-23 01:37:52]

Dry run: This issue passed the CQ dry run.

[Mircea Trofin, 2017-03-23 01:46:00]

The CQ bit was checked by mtrofin@chromium.org

[Mircea Trofin, 2017-03-23 01:46:00]

The patchset sent to the CQ was uploaded after l-g-t-m from
bradnelson@chromium.org, ahaas@chromium.org
Link to the patchset: https://codereview.chromium.org/2763413003/#ps100001
(title: "[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation")

[commit-bot: I haz the power, 2017-03-23 01:46:02]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-23 01:47:18]

CQ is committing da patch.

Bot data: {"patchset_id": 100001, "attempt_start_ts": 1490233560205810,
"parent_rev": "06a551ae7c0324a0cdf818144823ecf736353738", "commit_rev":
"78905107d3a6465f346d909fa7f1c22ffec144c2"}

[commit-bot: I haz the power, 2017-03-23 01:47:26]

Description was changed from

==========
[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation

APIs and trivial implementation, to unblock Chrome side dev.

BUG=chromium:697028
==========

to

==========
[wasm] Skeleton WasmModuleObjectBuilder for streamed compilation

APIs and trivial implementation, to unblock Chrome side dev.

BUG=chromium:697028

Review-Url: https://codereview.chromium.org/2763413003
Cr-Commit-Position: refs/heads/master@{#44053}
Committed:
https://chromium.googlesource.com/v8/v8/+/78905107d3a6465f346d909fa7f1c22ffec...
==========

[commit-bot: I haz the power, 2017-03-23 01:47:27]

Committed patchset #3 (id:100001) as
https://chromium.googlesource.com/v8/v8/+/78905107d3a6465f346d909fa7f1c22ffec...

[clemensh, 2017-03-27 07:39:19]

clemensh@google.com changed reviewers:
+ clemensh@google.com

[clemensh, 2017-03-27 07:39:20]

https://codereview.chromium.org/2763413003/diff/100001/include/v8.h
File include/v8.h (right):

https://codereview.chromium.org/2763413003/diff/100001/include/v8.h#newcode3917
include/v8.h:3917: // TODO(mtrofin): rename WasmCompiledModule to
WasmModuleObject, for
FYI: Ben once had the plan to rename this to WasmSpecializedCode.
The problem with the current name (and also "WasmModuleObject") is that this
suggests that only one such object exists per module, which is not the case.

[Mircea Trofin, 2017-03-27 07:58:30]

https://codereview.chromium.org/2763413003/diff/100001/include/v8.h
File include/v8.h (right):

https://codereview.chromium.org/2763413003/diff/100001/include/v8.h#newcode3917
include/v8.h:3917: // TODO(mtrofin): rename WasmCompiledModule to
WasmModuleObject, for
On 2017/03/27 07:39:20, clemensh wrote:
> FYI: Ben once had the plan to rename this to WasmSpecializedCode.
> The problem with the current name (and also "WasmModuleObject") is that this
> suggests that only one such object exists per module, which is not the case.

By "module" you mean "wire bytes"?

An instance of WasmCompiledModule is the v8 wrapping of an instance of the JS
type "WebAssembly.Module".  Not sure where the problem is: the user knows they
can create multiple WebAssembly.Module objects from the same instance of wire
bytes. Same here.

[Clemens Hammacher, 2017-03-27 08:05:34]

On 2017/03/27 at 07:58:30, mtrofin wrote:
> https://codereview.chromium.org/2763413003/diff/100001/include/v8.h
> File include/v8.h (right):
> 
>
https://codereview.chromium.org/2763413003/diff/100001/include/v8.h#newcode3917
> include/v8.h:3917: // TODO(mtrofin): rename WasmCompiledModule to
WasmModuleObject, for
> On 2017/03/27 07:39:20, clemensh wrote:
> > FYI: Ben once had the plan to rename this to WasmSpecializedCode.
> > The problem with the current name (and also "WasmModuleObject") is that this
> > suggests that only one such object exists per module, which is not the case.
> 
> By "module" you mean "wire bytes"?
> 
> An instance of WasmCompiledModule is the v8 wrapping of an instance of the JS
type "WebAssembly.Module".  Not sure where the problem is: the user knows they
can create multiple WebAssembly.Module objects from the same instance of wire
bytes. Same here.

That correspondance does not hold (and that's exactly the problem).
The WasmCompiledModule object is cloned for each (live) instance. Hence it
corresponds more to an instance than a module.
And it's purpose is mainly to hold the code objects that were specialized to
that instance. If the last instance of a "WebAssembly.Module" is GCed, then the
WasmCompiledModule object of that instance is reset and reused for another
instantiation of the same module.

[Mircea Trofin, 2017-03-27 08:09:33]

On 2017/03/27 08:05:34, Clemens Hammacher wrote:
> On 2017/03/27 at 07:58:30, mtrofin wrote:
> > https://codereview.chromium.org/2763413003/diff/100001/include/v8.h
> > File include/v8.h (right):
> > 
> >
>
https://codereview.chromium.org/2763413003/diff/100001/include/v8.h#newcode3917
> > include/v8.h:3917: // TODO(mtrofin): rename WasmCompiledModule to
> WasmModuleObject, for
> > On 2017/03/27 07:39:20, clemensh wrote:
> > > FYI: Ben once had the plan to rename this to WasmSpecializedCode.
> > > The problem with the current name (and also "WasmModuleObject") is that
this
> > > suggests that only one such object exists per module, which is not the
case.
> > 
> > By "module" you mean "wire bytes"?
> > 
> > An instance of WasmCompiledModule is the v8 wrapping of an instance of the
JS
> type "WebAssembly.Module".  Not sure where the problem is: the user knows they
> can create multiple WebAssembly.Module objects from the same instance of wire
> bytes. Same here.
> 
> That correspondance does not hold (and that's exactly the problem).
> The WasmCompiledModule object is cloned for each (live) instance. Hence it
> corresponds more to an instance than a module.
> And it's purpose is mainly to hold the code objects that were specialized to
> that instance. If the last instance of a "WebAssembly.Module" is GCed, then
the
> WasmCompiledModule object of that instance is reset and reused for another
> instantiation of the same module.

Ah, there is a problem, but that's the name: this here is actually the module
object, not
just the compiled functions. It's the "m" and "n", separately, in:

var m = new WebAssembly.Module(some_bytes);
var n = new WebAssembly.Module(some_bytes);

Hence my wanting to rename the external concept to match the JS API and also the
internal
API (i::WasmModuleObject), which is different from the internal notion of
"compiled part" and all that.