[OpenImageIO] Remove deprecated dependency on AGL

[tophyr]

• Changes comply with the maintainer guide.
• SHA512s are updated for each updated download.
• The "supports" clause reflects platforms that may be fixed by this new version.
• Any fixed CI baseline entries are removed from that file.
• Any patches that are no longer applied are deleted from the port's directory.
• The version database is fixed by rerunning ./vcpkg x-add-version --all and committing the result.
• Only one version is added to each modified port's versions file.

[vicroms]

Please increase the port version in port/openimageio/vcpkg.json and run vcpkg x-add-version openimageio to update the versions database.

[tophyr]

@vicroms done! i was surprised that the git-tree hash changed - i have a hunch that i'm not thinking about git-tree correctly. does that represent the raw git hash of the commit to be downloaded (AcademySoftwareFoundation/OpenImageIO tag v3.0.9.2 in this case) or does it represent some sort of hash after the patches are applied? The former doesn't seem like it would have changed from this PR, but the latter certainly would.

[tophyr]

vcpkg-applied patches causing changes in the git-tree hash would also explain many of the confusing version-database failures i've been having while tinkering on other PRs

[vicroms]

The git-tree is a hash of the contents of the port's folder; it is calculated with git ls-tree and it can be used to recover the contents using git show <git-tree>.

[vicroms]

I suspect there is a missing find_dependency(CpuFeaturesNdkCompat ) introduced by #48311 in opencv4 that is causing the Android failures.

[tophyr]

@vicroms calling mayday here - i'm a bit lost as to what's failing. hitting the build links from this verification run, and then opening up the "artifacts produced" link in the azure pipeline, takes me to a page listing failure artifacts from builds that didn't fail in this run:

this has me doubting the correctness of the logs of the runs that did fail - especially because opencv4:x64-android succeeds locally for me on both windows and linux hosts, on this commit. i'm unsure how to make CI happy, or figure out whatever is actually failing - any advice?

[tophyr]

ah; i take back the "works on my machine" - i always forget that CI builds different flavors than i do by default. building openimageio[core,ffmpeg,freetype,gif,jpegxl,libheif,libraw,opencolorio,opencv,openjpeg,pybind11,tools,webp]:x64-android does indeed repro the (a?) failure.

still fairly mystified by the non-failing failure artifacts in the azure pipeline though.

[dg0yt]

still fairly mystified by the non-failing failure artifacts in the azure pipeline though.

There are expected failures for arm64 windows.

vcpkg/scripts/ci.baseline.txt

Lines 1026 to 1031 in edffab1

	# OpenCV appears to be trying to use __fp16 on a platform that does not support that
	# buildtrees\opencv4\src\4.11.0-46ecfbc8ae.clean\modules\dnn\src\layers\cpu_kernels\conv_winograd_f63.simd.hpp(1371): error C4430: missing type specifier - int assumed. Note: C++ does not support default-int
	#vcpkg-ci-opencv:arm64-windows-static-md=pass
	opencv4:arm64-windows-static-md=fail
	#vcpkg-ci-opencv:arm64-windows=pass
	opencv4:arm64-windows=fail

and then openimageio is a cascading error and not an expected pass.

[dg0yt]

I suspect there is a missing find_dependency(CpuFeaturesNdkCompat ) introduced by #48311 in opencv4 that is causing the Android failures.

#48451 should resolve this.