This CL makes scaling and cropping lazy in AVFoundationVideoCapturer and
provides optimized paths for SW and HW encoding. For SW encoding, an
efficient NV12 -> I420 cropping and scaling is implemented in
CoreVideoFrameBuffer::NativeToI420. For HW encoding, an efficient NV12 ->
NV12 cropping and scaling is implemented in
CoreVideoFrameBuffer::CropAndScaleTo. The performance improvement over
the existing cropping and scaling is that it is now done in one step
instead of making an intermediary copy of the Y plane.
There might still be room for improvement in the HW path using some HW
support. That will be explored in a future CL.
BUG=b/30939444
Review-Url: https://codereview.webrtc.org/2394483005
Cr-Commit-Position: refs/heads/master@{#14701}
We need to redaclare certain interfaces in the GN build files for the
macOS and iOS framework targets to prevent those symbols being stripped
by the linker.
BUG=webrtc:6434
Review-Url: https://codereview.webrtc.org/2431603002
Cr-Commit-Position: refs/heads/master@{#14691}
The rotation is currently incorrect for the back camera in landscape
mode. The reason is that device rotation needs to be reversed for the
back camera compared to the front camera. The camera sensor can also be
mounted with a specific orientation. So when front camera rotation goes
through: 0->90->180->270, back camera rotation goes in reverse:
180->90->0->270.
BUG=b/31984246,b/30651939
Review-Url: https://codereview.webrtc.org/2401033002
Cr-Commit-Position: refs/heads/master@{#14580}
all methods but a few constructors. And similarly for the
subclass cricket::WebRtcVideoFrame.
TBR=tkchin@webrtc.org # Added an include line
BUG=webrtc:5682
Review-Url: https://codereview.webrtc.org/2315663002
Cr-Commit-Position: refs/heads/master@{#14576}
This makes this target not include the precompiled header, which
is different from GYP.
BUG=webrtc:6412
NOTRY=True
Review-Url: https://codereview.webrtc.org/2402493002
Cr-Commit-Position: refs/heads/master@{#14559}
It eliminates the need for hardcoded enumerations of the supported presets.
Hopefully it enables us to support slightly bigger set of resolutions and make the code slightly cleaner.
BUG=webrtc:6355
Review-Url: https://codereview.webrtc.org/2349223002
Cr-Commit-Position: refs/heads/master@{#14547}
The clang static analyzer seems unable to resolve cpp locks in ObjC code.
As of current time, the clang analyzer has known limitations documented
http://clang.llvm.org/docs/ThreadSafetyAnalysis.html#known-limitations.
From the documentation: "The analysis currently does not do any checking
inside constructors or destructors.
In other words, every constructor and destructor is treated
as if it was annotated with NO_THREAD_SAFETY_ANALYSIS."
This is 'probably' why the analyzer is unable to resolve the lock when
used in ObjC land (the cpp works fine).
The lock can be removed by using atomic property instead.
It's not on performance critical path and we expect updates on just one queue and reads from others. That's why the thread assurance atomic properties bring is enough.
The CL removes rtc_sdk_peerconnection_objc_warnings_config as well as it's no longer needed.
BUG=webrtc:6308
Review-Url: https://codereview.webrtc.org/2372513004
Cr-Commit-Position: refs/heads/master@{#14450}
This CL adds headers that are present in the /Headers directory but not included in the framework target.
BUG=none
NOTRY=true
Review-Url: https://codereview.webrtc.org/2342293002
Cr-Commit-Position: refs/heads/master@{#14413}
This changes most non-test related rtc_source_set targets to be
rtc_static_library instead. Targets without any .cc files are excluded.
This should bring back the build behavior we used to have with GYP
(i.e. same symbols exported in the libjingle_peerconnection.a file, which
are used by some downstream projects).
After doing an Android build with these changes:
$ nm --defined-only -g -C out/Release/lib.unstripped/libjingle_peerconnection_so.so | grep -i createpeerconnectionf
00077c51 T Java_org_webrtc_PeerConnectionFactory_nativeCreatePeerConnectionFactory
$ nm --defined-only -g -C out/Release/obj/webrtc/api/libjingle_peerconnection.a | grep -i createpeerconnectionf
00000001 T webrtc::CreatePeerConnectionFactory(rtc::Thread*, rtc::Thread*, rtc::Thread*, webrtc::AudioDeviceModule*, cricket::WebRtcVideoEncoderFactory*, cricket::WebRtcVideoDecoderFactory*)
00000001 T webrtc::CreatePeerConnectionFactory()
See https://chromium.googlesource.com/chromium/src/+/master/tools/gn/docs/cookbook.md#Note-on-static-libraries
for more details on this.
NOTICE: This should be further cleaned up in the future, to reduce
binary bloat and unnecessary linking time. Right now it's more
important to restore the desired build output though.
BUG=webrtc:6410, chromium:630755
Review-Url: https://codereview.webrtc.org/2361623004
Cr-Commit-Position: refs/heads/master@{#14364}
In the migration to GN templates, some targets got the whole
rtc_common_config removed, which can have unpredicted consequences
in terms of different code behavior due to defines not being set
as expected etc.
It's better to enable this config and only disable the warnings
that fails the build.
BUG=webrtc:6306,webrtc:6307,webrtc:6308
NOTRY=True
Review-Url: https://codereview.webrtc.org/2347263002
Cr-Commit-Position: refs/heads/master@{#14280}
The build artifacts don't look completely identical to the ones generated
by the GYP targets, but manual review shows the same symbols are exported.
On iOS, the version generated by the GN follows convention, including
a "Headers" directory, and the .modulemap file. I think this is preferred
over the gyp version.
BUG=webrtc:6320
NOTRY=True
TESTED=Run AppRTCDemo on iOS + Mac and verified with nm that they export the same symbols.
Review-Url: https://codereview.webrtc.org/2340633003
Cr-Commit-Position: refs/heads/master@{#14228}
This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.
Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.
BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True
Review-Url: https://codereview.webrtc.org/2313943002
Cr-Commit-Position: refs/heads/master@{#14150}
Reason for revert:
Downstream build is fixed.
Original issue's description:
> Revert of Ignore Camera and Flip bits in CVO when parsing video rotation (patchset #3 id:80001 of https://codereview.webrtc.org/2280703002/ )
>
> Reason for revert:
> Breaks downstream build.
>
> Original issue's description:
> > Ignore Camera and Flip bits in CVO when parsing video rotation
> >
> > Currently, if WebRTC receives a CVO byte where the Camera or Flip bit is
> > set, then rotation is incorrectly parsed as 0. This CL fixes that issue.
> > The Camera and Flip bit is still unimplemented and will just be ignored
> > though.
> >
> > BUG=webrtc:6120
> > R=danilchap@webrtc.org, pthatcher@webrtc.org, tommi@webrtc.org
> >
> > Committed: f9e1b922ef
>
> TBR=pthatcher@webrtc.org,danilchap@webrtc.org,tommi@webrtc.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=webrtc:6120
>
> Committed: https://crrev.com/97667c7746282704acccd896e26175decee349c0
> Cr-Commit-Position: refs/heads/master@{#14035}
TBR=pthatcher@webrtc.org,danilchap@webrtc.org,tommi@webrtc.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=webrtc:6120
Review-Url: https://codereview.webrtc.org/2320913003
Cr-Commit-Position: refs/heads/master@{#14124}
There is no clear reason to have them in build_overrides, and
webrtc/build seems to be a better place.
Also, delete build_overrides/webrtc.gni
NOTRY=True
BUG=webrtc:5949
Review-Url: https://codereview.webrtc.org/2309253004
Cr-Commit-Position: refs/heads/master@{#14108}
Remove common_inherited_config from the targets and add it to the
template instead.
BUG=webrtc:6187
NOTRY=True
Review-Url: https://codereview.webrtc.org/2311843002
Cr-Commit-Position: refs/heads/master@{#14069}
Remove common_config from the targets' config and add
it to the template instead.
BUG=webrtc:6187
NOTRY=True
Review-Url: https://codereview.webrtc.org/2300413002
Cr-Commit-Position: refs/heads/master@{#14063}
Defines the rtc_executable, rtc_source_set, rtc_test and
rtc_static_library templates.
These templates provide no functionality yet, but will enable common
configuration to be introduced, avoiding repetition in every target
Changes summary:
- Prepend rtc_ to test, source_set, executable and static_library targets
- Change "configs -= [" to "suppressed_configs += ["
- Include webrtc/build/webrtc.gni where it wasn't included yet
- Delete import("//testing/test.gni"), since rtc_test makes it unnecessary.
BUG=webrtc:6187
TBR=henrik.lundin@webrtc.org,tommi@webrtc.org
NOTRY=True
Review-Url: https://codereview.webrtc.org/2301053002
Cr-Commit-Position: refs/heads/master@{#14043}
Reason for revert:
Breaks downstream build.
Original issue's description:
> Ignore Camera and Flip bits in CVO when parsing video rotation
>
> Currently, if WebRTC receives a CVO byte where the Camera or Flip bit is
> set, then rotation is incorrectly parsed as 0. This CL fixes that issue.
> The Camera and Flip bit is still unimplemented and will just be ignored
> though.
>
> BUG=webrtc:6120
> R=danilchap@webrtc.org, pthatcher@webrtc.org, tommi@webrtc.org
>
> Committed: f9e1b922efTBR=pthatcher@webrtc.org,danilchap@webrtc.org,tommi@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6120
Review-Url: https://codereview.webrtc.org/2300323002
Cr-Commit-Position: refs/heads/master@{#14035}
Currently, if WebRTC receives a CVO byte where the Camera or Flip bit is
set, then rotation is incorrectly parsed as 0. This CL fixes that issue.
The Camera and Flip bit is still unimplemented and will just be ignored
though.
BUG=webrtc:6120
R=danilchap@webrtc.org, pthatcher@webrtc.org, tommi@webrtc.org
Review URL: https://codereview.webrtc.org/2280703002 .
Cr-Commit-Position: refs/heads/master@{#14027}
WebRTC no longer has any restriction on what thread frames should be
delivered on. One possible problem with this CL is that NV21->I420
conversion and scaling is done on the thread that delivers frames, which
might cause fps regressions.
R=nisse@webrtc.org, perkj@webrtc.org, tkchin@webrtc.org
Review URL: https://codereview.webrtc.org/2137503003 .
Cr-Commit-Position: refs/heads/master@{#14021}
The rotation is currently always applied by AVFoundation by
reconfiguring the capture connection video orientation. This CL sets the
rotation field in the frame instead. This avoids the current flash in
the video when the device is rotated, and also avoids reconfiguring the
local encoder and remote decoder when the device is rotated.
BUG=b/30651939
Review-Url: https://codereview.webrtc.org/2271583003
Cr-Commit-Position: refs/heads/master@{#13916}
We've seen some cases of nonrecoverable runtime error when entering the foreground. This is a theoretical fix to see if we can restart after willEnterForeground in didBecomeActive instead.
NOTRY=True
BUG=
Review-Url: https://codereview.webrtc.org/2258583004
Cr-Commit-Position: refs/heads/master@{#13903}
Theoretical fix to address some EAGLContext issues from other UIImageViews that could be active.
NOTRY=True
BUG=
Review-Url: https://codereview.webrtc.org/2259513002
Cr-Commit-Position: refs/heads/master@{#13806}
