Remove the requirement of D3D_FEATURE_LEVEL_11_0, but export the D3D_FEATURE_LEVEL through APIs

Remove the requirement of D3D_FEATURE_LEVEL_11_0, but export the D3D_FEATURE_LEVEL through APIs

D3D_FEATURE_LEVEL_11_0 is not offically documented on MSDN to be a requirement
of DXGI duplicator APIs. So instead of using D3D_FEATURE_LEVEL_11_0 as a
requirement in D3dDevice::Initialize(), this change adds a
ScreenCapturerWinDirectx::SupportedFeatureLevel() function to retrieves minimum
and maximium for further reference (in HostTraits to control the experiment).

BUG=314516
Committed: https://crrev.com/556f49d6b6009f4110872f06510e734fdde74a79
Cr-Commit-Position: refs/heads/master@{#14962}

[Hzj_jie, 2016-11-01 22:34:27]

Patchset #1 (id:1) has been deleted

[Hzj_jie, 2016-11-01 22:43:46]

Patchset #1 (id:20001) has been deleted

[Hzj_jie, 2016-11-01 23:59:40]

Patchset #3 (id:80001) has been deleted

[Hzj_jie, 2016-11-01 23:59:47]

Patchset #2 (id:60001) has been deleted

[Hzj_jie, 2016-11-01 23:59:54]

Patchset #1 (id:40001) has been deleted

[Hzj_jie, 2016-11-02 00:00:14]

zijiehe@chromium.org changed reviewers:
+ sergeyu@chromium.org

[Sergey Ulanov, 2016-11-07 19:37:16]

LGTM, but please see my suggesions

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
File webrtc/modules/desktop_capture/win/d3d_device.cc (right):

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/d3d_device.cc:46: if (feature_level <
D3D_FEATURE_LEVEL_11_0) {
Do we need this check at all? Maybe just log feature_level below, when we fail
to get IDXGIDevice?

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
File webrtc/modules/desktop_capture/win/dxgi_duplicator_controller.cc (right):

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/dxgi_duplicator_controller.cc:54: if
(min_feature_level != nullptr) {
nit: remove "!=nullptr" part.

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/dxgi_duplicator_controller.cc:143:
max_feature_level_ = D3D_FEATURE_LEVEL_9_1;
Maybe set these to -1 for initialization. Otherwise you have min > max, which
just doesn't look right.
Also this code won't work correctly if any of the devices has feature level
below 9.1 or above 11.

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
File webrtc/modules/desktop_capture/win/screen_capturer_win_directx.h (right):

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/screen_capturer_win_directx.h:42: static bool
SupportedFeatureLevels(D3D_FEATURE_LEVEL* min_feature_level,
Maybe we should expose some other hardware info here? E.g. manufacturer name,
driver version, etc? If so then it's probably better to do it with a single
method.

[Hzj_jie, 2016-11-08 00:24:48]

Patchset #2 (id:120001) has been deleted

[Hzj_jie, 2016-11-08 01:03:32]

Patchset #2 (id:140001) has been deleted

[Hzj_jie, 2016-11-08 01:25:40]

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
File webrtc/modules/desktop_capture/win/d3d_device.cc (right):

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/d3d_device.cc:46: if (feature_level <
D3D_FEATURE_LEVEL_11_0) {
On 2016/11/07 19:37:16, Sergey Ulanov wrote:
> Do we need this check at all? Maybe just log feature_level below, when we fail
> to get IDXGIDevice?

Indeed I cannot quite tell whether the following d3d_device_.As() statement or
DxgiOutputDuplicator::DuplicateOutput() will fail. Unfortunately, nothing has
been mentioned in MSDN. That's the reason I removed the "return false" here, and
export the version in API.
I can totally remove this check once the API version has been exported and
logged by client.

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
File webrtc/modules/desktop_capture/win/dxgi_duplicator_controller.cc (right):

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/dxgi_duplicator_controller.cc:54: if
(min_feature_level != nullptr) {
On 2016/11/07 19:37:16, Sergey Ulanov wrote:
> nit: remove "!=nullptr" part.

Done.

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/dxgi_duplicator_controller.cc:143:
max_feature_level_ = D3D_FEATURE_LEVEL_9_1;
On 2016/11/07 19:37:16, Sergey Ulanov wrote:
> Maybe set these to -1 for initialization. Otherwise you have min > max, which
> just doesn't look right.
> Also this code won't work correctly if any of the devices has feature level
> below 9.1 or above 11.

In D3dDevice::EnumDevices(), since we do not provide a D3D_FEATURE_LEVEL*, the
default feature levels is 9.1 to 11.0 (https://goo.gl/lcMhss). So here 11.0 and
9.1 is the largest and smallest possibilities.

But I do not think -1 is a safe value, considering Microsoft uses top 20 bits
for major version, middle 4 bits for minor version, low 8 bits as preserved
(https://goo.gl/tU4DoL). -1 means version 1048575.255.*(?). It looks like 0 is
not really possible, since there should be no 0.0.0(?).

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
File webrtc/modules/desktop_capture/win/screen_capturer_win_directx.h (right):

https://codereview.webrtc.org/2468083002/diff/100001/webrtc/modules/desktop_c...
webrtc/modules/desktop_capture/win/screen_capturer_win_directx.h:42: static bool
SupportedFeatureLevels(D3D_FEATURE_LEVEL* min_feature_level,
On 2016/11/07 19:37:16, Sergey Ulanov wrote:
> Maybe we should expose some other hardware info here? E.g. manufacturer name,
> driver version, etc? If so then it's probably better to do it with a single
> method.

Yes, I have added a D3dInfo structure for further extension.

[Hzj_jie, 2016-11-08 02:14:42]

The CQ bit was checked by zijiehe@chromium.org

[Hzj_jie, 2016-11-08 02:14:43]

The patchset sent to the CQ was uploaded after l-g-t-m from sergeyu@chromium.org
Link to the patchset: https://codereview.webrtc.org/2468083002/#ps160001 (title:
"Resolve review comments")

[commit-bot: I haz the power, 2016-11-08 02:14:52]

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

[commit-bot: I haz the power, 2016-11-08 02:35:19]

Committed patchset #2 (id:160001)

[commit-bot: I haz the power, 2016-11-08 02:35:22]

Description was changed from

==========
Remove the requirement of D3D_FEATURE_LEVEL_11_0, but export the
D3D_FEATURE_LEVEL through APIs

D3D_FEATURE_LEVEL_11_0 is not offically documented on MSDN to be a requirement
of DXGI duplicator APIs. So instead of using D3D_FEATURE_LEVEL_11_0 as a
requirement in D3dDevice::Initialize(), this change adds a
ScreenCapturerWinDirectx::SupportedFeatureLevel() function to retrieves minimum
and maximium for further reference (in HostTraits to control the experiment).

BUG=314516
==========

to

==========
Remove the requirement of D3D_FEATURE_LEVEL_11_0, but export the
D3D_FEATURE_LEVEL through APIs

D3D_FEATURE_LEVEL_11_0 is not offically documented on MSDN to be a requirement
of DXGI duplicator APIs. So instead of using D3D_FEATURE_LEVEL_11_0 as a
requirement in D3dDevice::Initialize(), this change adds a
ScreenCapturerWinDirectx::SupportedFeatureLevel() function to retrieves minimum
and maximium for further reference (in HostTraits to control the experiment).

BUG=314516

Committed: https://crrev.com/556f49d6b6009f4110872f06510e734fdde74a79
Cr-Commit-Position: refs/heads/master@{#14962}
==========

[commit-bot: I haz the power, 2016-11-08 02:35:23]

Patchset 2 (id:??) landed as
https://crrev.com/556f49d6b6009f4110872f06510e734fdde74a79
Cr-Commit-Position: refs/heads/master@{#14962}