[stubs] Using template magic to get the call interface descriptor type from callable.

[stubs] Using template magic to get the call interface descriptor type from callable.

BUG=v8:5236

[Igor Sheludko, 2016-07-27 11:50:26]

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

[commit-bot: I haz the power, 2016-07-27 11:50:28]

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

[Igor Sheludko, 2016-07-27 11:53:30]

Patchset #1 (id:1) has been deleted

[Igor Sheludko, 2016-07-27 12:46:40]

Patchset #1 (id:20001) has been deleted

[Igor Sheludko, 2016-07-27 12:46:48]

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

[commit-bot: I haz the power, 2016-07-27 12:46:51]

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

[Igor Sheludko, 2016-07-27 12:49:38]

ishell@chromium.org changed reviewers:
+ mstarzinger@chromium.org

[Igor Sheludko, 2016-07-27 12:49:39]

PTAL

[Michael Starzinger, 2016-07-27 12:57:18]

mstarzinger@chromium.org changed reviewers:
+ mvstanton@chromium.org

[Michael Starzinger, 2016-07-27 12:57:18]

mstarzinger@chromium.org changed required reviewers:
+ mvstanton@chromium.org

[Michael Starzinger, 2016-07-27 12:57:19]

Not a huge fan of this template magic. What is the overall goal that this CL is
aiming at? Also, mvstanton@, WDYT?

[commit-bot: I haz the power, 2016-07-27 13:08:35]

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

[commit-bot: I haz the power, 2016-07-27 13:08:35]

Dry run: This issue passed the CQ dry run.

[Igor Sheludko, 2016-07-27 13:09:19]

On 2016/07/27 12:57:19, Michael Starzinger wrote:
> Not a huge fan of this template magic. What is the overall goal that this CL
is
> aiming at? Also, mvstanton@, WDYT?

The goal is to make sure that we always use the right call interface descriptor
type (and therefore parameter indices)
which is defined in the stub. Whenever something is changed somewhere the C++
compiler should show us an error.

We abstracted the code stubs by CodeFactory interface and clients do not have
access to the right descriptor type anymore.
So CallableD<> is a way to export this information from the CodeFactory.

I'm not sure about the names. Probably, in the end Callable should become
CallableBase and CallableD<> - Callable<>.

[Michael Starzinger, 2016-07-27 13:14:36]

On 2016/07/27 13:09:19, Igor Sheludko wrote:
> On 2016/07/27 12:57:19, Michael Starzinger wrote:
> > Not a huge fan of this template magic. What is the overall goal that this CL
> is
> > aiming at? Also, mvstanton@, WDYT?
> 
> The goal is to make sure that we always use the right call interface
descriptor
> type (and therefore parameter indices)
> which is defined in the stub. Whenever something is changed somewhere the C++
> compiler should show us an error.
> 
> We abstracted the code stubs by CodeFactory interface and clients do not have
> access to the right descriptor type anymore.
> So CallableD<> is a way to export this information from the CodeFactory.
> 
> I'm not sure about the names. Probably, in the end Callable should become
> CallableBase and CallableD<> - Callable<>.

IIUC then it still requires all call sites to be adapted in case the interface
is changed. So all that one gains is static checking from the C++ compiler.
Clients still need access to the actual CallInterfaceDescriptor. It is just that
the specific type of the CallInterfaceDescriptor can now be derived from the
signature of the various factory methods (instead of naming the type directly).
IMHO this is not worth the additional verbosity. Just my two cents.

[Igor Sheludko, 2016-09-24 14:57:18]

Description was changed from

==========
[stubs] Using template magic to get the call interface descriptor type from
callable.

BUG=v8:5236
==========

to

==========
[stubs] Using template magic to get the call interface descriptor type from
callable.

BUG=v8:5236
==========