Add AutoThread types for Windows that initialize COM.

Add AutoThread types for Windows that initialize COM.

This CL adds AutoThread types that create a UI MessageLoop on the
thread and initialize COM for single- or multi-threaded used.

This is a re-land of 

BUG=145856
TBR=alexeypa

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=169595

[Wez, 2012-11-16 08:39:32]

PTAL; this is one of two CLs that try different approaches to adding COM
initialization support to AutoThread.  Let me know what you think, and which you
prefer, with your COM guru hat on. :)

[alexeypa (please no reviews), 2012-11-16 18:16:09]

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
File remoting/base/auto_thread.h (right):

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
remoting/base/auto_thread.h:40: TYPE_UI_COM_MTA
MTA does not need a message pump AFAIR. It is not correct to bind MTA to a UI
message loop. How about providing a mechanism to pass a factory/callback that
will instantiate a scoped object on the thread?

In the worst case I think the approach used in base::Thread is better.

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
remoting/base/auto_thread.h:47: static scoped_refptr<AutoThreadTaskRunner>
CreateWithType(
Add a variant CreateWithTypeAndInitializeCom?

[Sergey Ulanov, 2012-11-16 19:23:32]

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
File remoting/base/auto_thread.h (right):

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
remoting/base/auto_thread.h:34: enum Type {
Do we really need this type? Maybe better to add init_com_with_mta() so that the
API is the same as in base::Thread?

[Wez, 2012-11-16 19:54:54]

Sounds like you both prefer adding init_with_mta() and CreateWithMta() methods;
will follow up with that in a bit.

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
File remoting/base/auto_thread.h (right):

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
remoting/base/auto_thread.h:34: enum Type {
On 2012/11/16 19:23:33, sergeyu wrote:
> Do we really need this type? Maybe better to add init_com_with_mta() so that
the
> API is the same as in base::Thread?

This class is already different enough from base::Thread that I don't think
that's a concern here. But Alex seems to agree; see below. :-/

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
remoting/base/auto_thread.h:40: TYPE_UI_COM_MTA
On 2012/11/16 18:16:09, alexeypa wrote:
> MTA does not need a message pump AFAIR. It is not correct to bind MTA to a UI
> message loop. How about providing a mechanism to pass a factory/callback that
> will instantiate a scoped object on the thread?

Quite right; fixed.

> In the worst case I think the approach used in base::Thread is better.

OK; I'd just hoped to avoid the extra API the base::Thread approach requires.

https://chromiumcodereview.appspot.com/11348087/diff/1/remoting/base/auto_thr...
remoting/base/auto_thread.h:47: static scoped_refptr<AutoThreadTaskRunner>
CreateWithType(
On 2012/11/16 18:16:09, alexeypa wrote:
> Add a variant CreateWithTypeAndInitializeCom?

If I switch the interface to the ugly init_com_with_mta() approach then yes,
I'll add a variant creator function as well.

[Wez, 2012-11-22 21:57:15]

PTAL

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
File remoting/base/auto_thread.cc (right):

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:46: #endif
note-to-self: Remove this.

[alexeypa (please no reviews), 2012-11-26 19:16:13]

lgtm with my comments addressed.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
File remoting/base/auto_thread.cc (right):

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:24: scoped_ptr<base::win::ScopedCOMInitializer>
CreateComInitializerWithType(
nit: CreateComInitializer()

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:95: : startup_data_(NULL),
Initialize |com_init_type_| here.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:101: AutoThread::AutoThread(const char* name,
AutoThreadTaskRunner* joiner)
Initialize |com_init_type_| here.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:151: com_init_type_ = com_init_type;
nit: Add DCHECK_EQ(com_init_type_, COM_INIT_NONE)

[Wez, 2012-11-27 02:22:26]

FYI

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
File remoting/base/auto_thread.cc (right):

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:24: scoped_ptr<base::win::ScopedCOMInitializer>
CreateComInitializerWithType(
On 2012/11/26 19:16:13, alexeypa wrote:
> nit: CreateComInitializer()

Done.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:46: #endif
On 2012/11/22 21:57:15, Wez wrote:
> note-to-self: Remove this.

Done.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:95: : startup_data_(NULL),
On 2012/11/26 19:16:13, alexeypa wrote:
> Initialize |com_init_type_| here.

Done.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:101: AutoThread::AutoThread(const char* name,
AutoThreadTaskRunner* joiner)
On 2012/11/26 19:16:13, alexeypa wrote:
> Initialize |com_init_type_| here.

Done.

https://chromiumcodereview.appspot.com/11348087/diff/2002/remoting/base/auto_...
remoting/base/auto_thread.cc:151: com_init_type_ = com_init_type;
On 2012/11/26 19:16:13, alexeypa wrote:
> nit: Add DCHECK_EQ(com_init_type_, COM_INIT_NONE)

Done.

[commit-bot: I haz the power, 2012-11-27 02:24:36]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/wez@chromium.org/11348087/10001

[commit-bot: I haz the power, 2012-11-27 02:44:09]

Sorry for I got bad news for ya.
Compile failed with a clobber build on linux_rel.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_rel&...
Your code is likely broken or HEAD is junk. Please ensure your
code is not broken then alert the build sheriffs.
Look at the try server FAQ for more details.

[commit-bot: I haz the power, 2012-11-27 03:56:09]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/wez@chromium.org/11348087/9005

[commit-bot: I haz the power, 2012-11-27 07:04:24]

Change committed as 169595