Enable C++ 20 standard globally

[SaschaWillems]

Description

Note: This is work-in-progress and not ready for review yet. I didn't set it to "draft", as I wanted to test CI to see where things break with C++ 20.

Refs #1201

General Checklist:

Please ensure the following points are checked:

• My code follows the coding style
• I have reviewed file licenses
• I have commented any added functions (in line with Doxygen)
• I have commented any code that could be hard to understand
• My changes do not add any new compiler warnings
• My changes do not add any new validation layer errors or warnings
• I have used existing framework/helper functions where possible
• My changes do not add any regressions
• I have tested every sample to ensure everything runs correctly
• This PR describes the scope and expected impact of the changes I am making

Note: The Samples CI runs a number of checks including:

• I have updated the header Copyright to reflect the current year (CI build will fail if Copyright is out of date)
• My changes build on Windows, Linux, macOS and Android. Otherwise I have documented any exceptions

If this PR contains framework changes:

• I did a full batch run using the batch command line argument to make sure all samples still work properly

[SaschaWillems]

Once we merge this, I think we can remove one or two external dependencies (submodules). Right now. printing formatted messages (e.g. errors) is knda complicated and goes through our framework and two externeal dependencies. One of those is "fmt", and it seems that it's functionality is also present in the STL with C++20. So we could get rid of "fmt" and maybe also "spdlog". I don't think we need to have such a complicated setup just for logging.

[SaschaWillems]

@jackk18912525: Can you please remove your spam comment and also remove yourself from the reviewers?

[SaschaWillems]

A blocker is the HCWPipe third party dependency. I can't get it to compile after updating to C++20, and updating the submodule doesn't help either (esp.a s the library is now differently named). I need help on this from someone from arm, maybe @JoseEmilio-ARM?

Submitted accidentally

[SaschaWillems]

@JoseEmilio-ARM : Any update on this? I'm stuck until someone with proper knowledge updates HCWPipe.

[JoseEmilio-ARM]

@JoseEmilio-ARM : Any update on this? I'm stuck until someone with proper knowledge updates HCWPipe.

Hi @SaschaWillems , please try again with the changes in #1273

[SaschaWillems]

Awesome. Thank you very much :)

[CLAassistant]

All committers have signed the CLA.

[SaschaWillems]

Builds are still failing for me on Android. Seems the HCWStatsProvider needs to be updated to work with C++20:

/home/runner/work/Vulkan-Samples/Vulkan-Samples/framework/stats/hwcpipe_stats_provider.cpp:47:14: error: no matching constructor for initialization of 'StatDataMap' (aka 'unordered_map<StatIndex, StatData, StatIndexHash>')
2025-02-25T16:05:45.9780418Z      47 |         StatDataMap hwcpipe_stats = {
2025-02-25T16:05:45.9781083Z         |                     ^               ~
2025-02-25T16:05:45.9781771Z      48 |             {StatIndex::gpu_cycles,            {hwcpipe_counter::MaliGPUActiveCy}},
2025-02-25T16:05:45.9782517Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9783593Z      49 |             {StatIndex::gpu_vertex_cycles,     {hwcpipe_counter::MaliNonFragQueueActiveCy, {MaliNonFragActiveCy, MaliBinningQueueActiveCy} }},
2025-02-25T16:05:45.9785185Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9785969Z      50 |             {StatIndex::gpu_load_store_cycles, {hwcpipe_counter::MaliLSIssueCy}},
2025-02-25T16:05:45.9786661Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9787351Z      51 |             {StatIndex::gpu_tiles,             {hwcpipe_counter::MaliFragTile}},
2025-02-25T16:05:45.9787997Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9788683Z      52 |             {StatIndex::gpu_killed_tiles,      {hwcpipe_counter::MaliFragTileKill}},
2025-02-25T16:05:45.9789578Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9790788Z      53 |             {StatIndex::gpu_fragment_cycles,   {hwcpipe_counter::MaliFragQueueActiveCy, {MaliFragActiveCy, MaliMainQueueActiveCy}}},
2025-02-25T16:05:45.9791865Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9792759Z      54 |             {StatIndex::gpu_fragment_jobs,     {hwcpipe_counter::MaliFragQueueJob, {MaliMainQueueJob}}},
2025-02-25T16:05:45.9793575Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9794289Z      55 |             {StatIndex::gpu_ext_reads,         {hwcpipe_counter::MaliExtBusRdBt}},
2025-02-25T16:05:45.9794949Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9795639Z      56 |             {StatIndex::gpu_ext_writes,        {hwcpipe_counter::MaliExtBusWrBt}},
2025-02-25T16:05:45.9796294Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9796994Z      57 |             {StatIndex::gpu_ext_read_stalls,   {hwcpipe_counter::MaliExtBusRdStallCy}},
2025-02-25T16:05:45.9797691Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9798399Z      58 |             {StatIndex::gpu_ext_write_stalls,  {hwcpipe_counter::MaliExtBusWrStallCy}},
2025-02-25T16:05:45.9799103Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9799785Z      59 |             {StatIndex::gpu_ext_read_bytes,    {hwcpipe_counter::MaliExtBusRdBy}},
2025-02-25T16:05:45.9800443Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9801325Z      60 |             {StatIndex::gpu_ext_write_bytes,   {hwcpipe_counter::MaliExtBusWrBy}},
2025-02-25T16:05:45.9802002Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2025-02-25T16:05:45.9811329Z      61 |             {StatIndex::gpu_tex_cycles,        {hwcpipe_counter::MaliTexIssueCy}}};
2025-02-25T16:05:45.9812044Z         |             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Any idea on how to fix this?

[gpx1000]

Sorry Sascha, I meant to do that change on my branch and offer it as a PR. my troubles with CLion and Android Studio's github plugin continue to annoy. The change to the constructor in that class seems to allow Android to build locally. I don't want to step on your toes, so please back it out if you don't like it. I'm happy to fix the clang format and copyright header checks, but again, don't want to step on your toes; just wanted to make a suggestion for help.

[SaschaWillems]

No worries. That's perfectly fine. And thank you very much for taking care of this, I couldn't figure out how to fix this 👍🏻

[gpx1000]

I'm always happy to bang my head against a problem till it works ;-)

That said, I've been playing around locally with your branch on adding modules... that's... well it'll take a bit from what I can tell. At least CMake has most of the support. It should help a LOT with compile times. But the changes are getting gnarly. Anyway, I think this PR has all changes necessary for C++20, so that's most of the problem.

[SaschaWillems]

So with Steven's fix this now builds on all platforms and is ready for review :)

[gary-sweet]

This does seem to run ok for me, but I am seeing some build warnings from the third party code with this change (using gcc-12.3-1.0 on aarch64):

In file included from third_party/catch2/src/catch2/benchmark/detail/catch_stats.cpp:10:
third_party/catch2/src/catch2/../catch2/benchmark/detail/catch_stats.hpp: In instantiation of 'Catch::Benchmark::Estimate<double> Catch::Benchmark::Detail::bootstrap(double, Iterator, Iterator, const sample&, Estimator&&) [with Iterator = __gnu_cxx::__normal_iterator<double*, std::vector<double> >; Estimator = double (*&)(__gnu_cxx::__normal_iterator<double*, std::vector<double> >, __gnu_cxx::__normal_iterator<double*, std::vector<double> >); sample = std::vector<double>]':
third_party/catch2/src/catch2/benchmark/detail/catch_stats.cpp:244:37:   required from here
third_party/catch2/src/catch2/../catch2/benchmark/detail/catch_stats.hpp:99:118: note: parameter passing for argument of type 'std::pair<double, double>' when C++17 is enabled changed to match C++14 in GCC 10.1
   99 |                 std::tie(sum_squares, sum_cubes) = std::accumulate(jack.begin(), jack.end(), std::make_pair(0., 0.), [jack_mean](std::pair<double, double> sqcb, double x) -> std::pair<double, double> {
      |                                                                                                                      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  100 |                     auto d = jack_mean - x;
      |                     ~~~~~~~~~~~~~~~~~~~~~~~                                                                           
  101 |                     auto d2 = d * d;
      |                     ~~~~~~~~~~~~~~~~                                                                                  
  102 |                     auto d3 = d2 * d;
      |                     ~~~~~~~~~~~~~~~~~                                                                                 
  103 |                     return { sqcb.first + d2, sqcb.second + d3 };
      |                     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~                                                     
  104 |                 });
      |                 ~     

This one pops up 15 times:

In file included from third_party/ktx/other_include/glm/detail/type_half.hpp:19,
                 from third_party/ktx/other_include/glm/detail/func_packing.inl:5,
                 from third_party/ktx/other_include/glm/detail/func_packing.hpp:168,
                 from third_party/ktx/other_include/glm/packing.hpp:6,
                 from third_party/ktx/other_include/glm/glm.hpp:84,
                 from framework/common/glm_common.h:26,
                 from framework/common/helpers.h:39,
                 from framework/core/descriptor_set_layout.h:20,
                 from framework/core/hpp_descriptor_set_layout.h:20,
                 from framework/core/command_buffer.h:23,
                 from framework/gui.h:29,
                 from framework/platform/plugins/plugin.h:28,
                 from app/plugins/plugins.h:20,
                 from app/plugins/plugins.cpp:20:
third_party/ktx/other_include/glm/detail/type_half.inl: In function 'float glm::detail::overflow()':
third_party/ktx/other_include/glm/detail/type_half.inl:12:27: warning: compound assignment with 'volatile'-qualified left operand is deprecated [-Wvolatile]
   12 |                         f *= f; // this will overflow before the for loop terminates
      |                         ~~^~~~

[SaschaWillems]

Thanks for bringing that up. Catch2 is something we want to get rid of anyway (see #1303). As for KTX: we may want to update that in another PR, I guess it's using it's own outdated version of glm.