Fixes to CMake and version upgrade with CMake

[weirdbeardgame]

FindWayland.cmake: Changed incorrect use of FIND_PACKAGE_HANDLE_STANDARD_ARGS which is meant to link and set the main package to found rather then link individual libs

Added CMake 3.22.0 as max version. Added package version in CMakeLists as per standard CMP0048

Description of Changes

Changes incorrect FIND_PACKAGE_HANDLE_STANDARD_ARGS to link wayland libs as imported targes. Increases the minimum Cmake version to include 3.11 and more modern 3.22 and sets "package version" as per https://cmake.org/cmake/help/latest/policy/CMP0048.html#cmp0048

Rationale behind Changes

To use more modern CMAKE features for those with 3.22 and clear warnings related to that. To fix several incorrect linkages with wayland clearing warnings like

  The package name passed to `find_package_handle_standard_args` (WAYLAND)
  does not match the name of the calling package (Wayland).  This can lead to
  problems in calling code that expects `find_package` result variables
  (e.g., `_FOUND`) to follow a certain pattern.
Call Stack (most recent call first):
  cmake/FindWayland.cmake:56 (FIND_PACKAGE_HANDLE_STANDARD_ARGS)
  cmake/SearchForStuff.cmake:184 (find_package)
  CMakeLists.txt:31 (include)
This warning is for project developers.  Use -Wno-dev to suppress it.

Suggested Testing Steps

See if it builds without those warnings

[F0bes]

Gave it a spin with the Wayland API On.

• Builds with no cmake warnings
• Runs a game

Didn't notice anything immediately incorrect with the FindWayland script but I'm no cmake wizard.

[TellowKrinkle]

Since you're adding imported targets, could you also edit pcsx2/CMakeLists.txt to stop looking at things like WAYLAND_CLIENT_FOUND and ${WAYLAND_CLIENT_LIBRARIES} and just link against Wayland::Client?

[TellowKrinkle]

This enables many years worth of CMake policies so we should get some testing on Linux to make sure nothing broke

[weirdbeardgame]

Tested between Xorg and Wayland. Package built correctly and on basic use with loading a game seemed functional as it should be. Wayland was active as noted by the pause and resume tested resulting in the current black window glitch speed seemed fast enough for the earlier fixes to be active nearing full-speed at times. But sluggish on Nvidia compared to Xorg still.

[F0bes]

gt3regression.mp4

Doesn't happen on master, both renderers, SW mode settings do not matter.
Happens with or without -DWAYLAND_API=on passed to cmake.

Ok, hold on a second. Reproducing this is a pain because savestates do have an affect.

[TellowKrinkle]

Doesn't happen on master, both renderers, SW mode settings do not matter. Happens with or without -DWAYLAND_API=on passed to cmake.

Can you check which commit does it? I would expect it's the CMake minimum version one but just to make sure

[F0bes]

Doesn't happen on master, both renderers, SW mode settings do not matter. Happens with or without -DWAYLAND_API=on passed to cmake.

Can you check which commit does it? I would expect it's the CMake minimum version one but just to make sure

Nevermind, I just got unlucky when testing this PR. After many attempts you can reproduce it on master.

Edit:
Yeah sorry I forgot to mention that it builds and plays a game for me.

[weirdbeardgame]

As a side note. This works in the AUR environment as well

[weirdbeardgame]

Anything else needed to merge this?

[Mrlinkwii]

runing X11 and cmake 3.16.3 , the PR complied fine and worked like master