NVIDIA driver bug? #4
Description
Just found your new renderer project, and decided to give it a spin on my Arch Linux system. It compiles cleanly, but when running sk_renderer_test my whole desktop environment locks up for a few seconds (including mouse cursor), gets back going after that, but the render window stays empty (non-updated). I also don't see any errors from your code, but do see those nasty Xid lines in my system logs. This screams NVIDIA driver bug to me, where I'm currently on 580.95.05 on a TITAN V. Just curious if you have been testing in a similar environment, i.e. Linux + NVIDIA GPU + NVIDIA binary blob driver? I run X11, btw, no Wayland.
melis@blackbox 10:10:~/c/sk_renderer/build$ ./sk_renderer_test
[warning] Extension 'VK_EXT_present_mode_fifo_latest_ready' not available, skipping
[warning] Layer 'VK_LAYER_KHRONOS_validation' not available, skipping
[info] Selected discrete GPU: NVIDIA TITAN V
[info] Using GPU: NVIDIA TITAN V
[info] Available device extensions (220):
[info] VK_KHR_16bit_storage
[info] VK_KHR_8bit_storage
[info] VK_KHR_acceleration_structure
[info] VK_KHR_bind_memory2
[info] VK_KHR_buffer_device_address
[info] VK_KHR_calibrated_timestamps
[info] VK_KHR_copy_commands2
[info] VK_KHR_create_renderpass2
[info] VK_KHR_dedicated_allocation
[info] VK_KHR_deferred_host_operations
[info] ... and 210 more
[info] Using transfer queue (family 1)
[info] Thread slot #0 initialized
[info] Selected surface format: 50 (colorSpace: 0)
[info] Swapchain created: requested 4, actual 4 images
[info] sk_renderer initialized successfully!
[info] Thread slot #1 initialized
[info] GLTF: Loading started
[info] GLTF: Ready (1 meshes, 3 textures)
[info] Thread #1 shutdown, marking as non-alive for reuse
[info] Switched to scene: GLTF Model
[info] Application created successfully!
[info] Available scenes: 10 (use arrow keys to switch)
[info] Render targets created: 1280x720, color=1, depth=1, scene=0
^C[info] Completed 7 frames, shutting down...
[info] Application destroyed