Register for HID device notifications on Windows.

Register for HID device notifications on Windows.

Instead of re-enumerating all HID devices whenever asked this change
registers with Windows to receive window messages whenever a HID device
is added or removed. This will allow Chrome to notify apps when devices
are connected so that they do not need to poll.

BUG=376719
Committed: https://crrev.com/08f9e07efe44307b55e5de46721364b3e03b31d6
Cr-Commit-Position: refs/heads/master@{#307124}

[Reilly Grant (use Gerrit), 2014-12-05 22:24:40]

reillyg@chromium.org changed reviewers:
+ rockot@chromium.org

[Reilly Grant (use Gerrit), 2014-12-05 22:24:40]

One more patch for you.

[Ken Rockot(use gerrit already), 2014-12-05 23:27:53]

lgtm

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.cc
File device/hid/hid_service_win.cc (right):

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.c...
device/hid/hid_service_win.cc:74: window_.reset(new base::win::MessageWindow());
nit: Maybe file a bug + add a TODO which expresses the intent to unify device
monitoring into a single service with a single MessageWindow.

[Reilly Grant (use Gerrit), 2014-12-05 23:48:17]

The CQ bit was checked by reillyg@chromium.org

[commit-bot: I haz the power, 2014-12-05 23:48:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/783773002/1

[commit-bot: I haz the power, 2014-12-06 00:44:32]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2014-12-06 00:45:21]

Patchset 1 (id:??) landed as
https://crrev.com/08f9e07efe44307b55e5de46721364b3e03b31d6
Cr-Commit-Position: refs/heads/master@{#307124}

[Nico, 2015-12-09 17:05:45]

thakis@chromium.org changed reviewers:
+ thakis@chromium.org

[Nico, 2015-12-09 17:05:45]

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.cc
File device/hid/hid_service_win.cc (right):

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.c...
device/hid/hid_service_win.cc:294: GetLastError() ==
base::File::FILE_ERROR_ACCESS_DENIED) {
Are you sure this is correct? The CreateFile() MSDN docs say that GetLastError()
will return ERROR_ACCESS_DENIED which has a numerical value of 5.
base::File::FILE_ERROR_ACCESS_DENIED has a numerical value of -5, is a
cross-platform enum, and doesn't mention anywhere that its values are equal to
what the windows function GetLastError() returns.

[Reilly Grant (use Gerrit), 2015-12-09 18:49:30]

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.cc
File device/hid/hid_service_win.cc (right):

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.c...
device/hid/hid_service_win.cc:294: GetLastError() ==
base::File::FILE_ERROR_ACCESS_DENIED) {
On 2015/12/09 17:05:45, Nico wrote:
> Are you sure this is correct? The CreateFile() MSDN docs say that
GetLastError()
> will return ERROR_ACCESS_DENIED which has a numerical value of 5.
> base::File::FILE_ERROR_ACCESS_DENIED has a numerical value of -5, is a
> cross-platform enum, and doesn't mention anywhere that its values are equal to
> what the windows function GetLastError() returns.

Interesting. I wonder if this logic has ever been exercised on Windows. It was
added in https://codereview.chromium.org/240263003.

[Nico, 2015-12-09 19:10:54]

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.cc
File device/hid/hid_service_win.cc (right):

https://codereview.chromium.org/783773002/diff/1/device/hid/hid_service_win.c...
device/hid/hid_service_win.cc:294: GetLastError() ==
base::File::FILE_ERROR_ACCESS_DENIED) {
On 2015/12/09 18:49:29, Reilly Grant wrote:
> On 2015/12/09 17:05:45, Nico wrote:
> > Are you sure this is correct? The CreateFile() MSDN docs say that
> GetLastError()
> > will return ERROR_ACCESS_DENIED which has a numerical value of 5.
> > base::File::FILE_ERROR_ACCESS_DENIED has a numerical value of -5, is a
> > cross-platform enum, and doesn't mention anywhere that its values are equal
to
> > what the windows function GetLastError() returns.
> 
> Interesting. I wonder if this logic has ever been exercised on Windows. It was
> added in https://codereview.chromium.org/240263003.

Deleting this block if it's broken and untested sounds fine to me -- up to you
folks what to do :-)