…call

…appers, update examples_tests submodule

…pping into vertex shader - TODO: fill clipping plane distances

…sl and spirv_intrinsics/core.hlsl because of microsoft/DirectXShaderCompiler#6751

…e it

…ipeline barier to change the font image's layout after the buffer update from TRANSFER_DST_OPTIMAL to READ_ONLY_OPTIMAL

…ngine for this purpose with full test suite. Enter runtime issues, I think it's because of how I currently share resources across modules and the fact ImGUI context is static

…into gpgpuimgui

…ate memory corruption issues. Make NITE work, now we just need to take care of imguizmo which doesn't like the configuration

…fig header on fly with CMake and share across imgui projects. TODO: add NBL_IMGUI_WITH_TEST_ENGINE enabled by default which makes all imgui projects use our configuration with embedding the test engine and enables the nite tool, otherwise disables stuff and limits the imgui build to minimum

…e and use running modes

…d and executed immediately after startup or should let user browse GUI and execute stuff by hand, use queued mode with CTest

…bla into gpgpuimgui

…nt atlas as 2D array texture, use BDA for requesting texture index. Add comments regarding our base instance indexing as replacement for gl_DrawID and a few more comments + minor changes

…define, refactor current legacy mouse handle & add key events handle

…into gpgpuimgui

…tb_truetype

…texture use request + no hardcoded ds layout etc

…into gpgpuimgui and resolve conflicts

…into gpgpuimgui

…into gpgpuimgui

…xtension, ask user for descriptor set layout at create time & let they pass descriptor set within render call

… errors!) - turns out it must be passed to another stage

… submodule - turns out I have issues with blend equation, add a temporary solution which turns alpha to 1 when non-atlas font textures are sampled

…2 to export/import it's constructor in dll mode

…ions instead of std::vector objects, update examples_tests submodule. TODO: Instead of allocating at render call I need to start with preallocated like 2 MB buffer for which multi alloc requests will be performed & handle overflows

…alloc but now I have white screen + swapchain present fails terminating my exe, I need to take care of intended submit + sync things first, also my queue submit seems to take much more time than before

…aken ImGUI Vertex with Indices types [typo] resulting in incorrect memory write. I see the scene nicely again, but something is decreasing my ref counted counter for mdi buffer which is proxy passed with dont_grab_t{}! (wut, inspecting it)

… keep it alive and not let it decrease reference counter to 0

…g multi_deallocate

…call, add more validation & logs, multi_deallocate at then end of the render call with future scratch semaphore (we treating it as wait semaphore), temporary terminate application on scratch semaphore-spotted counter overflow, update examples_tests submodule

…ave issues with multi alloc hence I guess something off with my wait semaphore

…r, add more logs (+ tmp logs) - I have issue with deferred offset memory deallocation, it doesn't execute (maybe because of my wait semaphore?)

…t work (Erfan thank you for help), update examples_tests submodule

…some typos, add comments & update examples_tests submodule

…ferred memory free anyway

…ded at update stage, should never be kept & passed with constructor. Let user pass subpassIx while creating the UI for internal pipeline creation purposes, update examples_tests submodule, clean some code & add comments

…proper extended instruction set for our opcodes ( `ext_instruction(opcode[, set])` )

…buffer themselves & validate its creation params, memory, allocate + mapping flags, update examples_tests submodule