Segfaults From Running Examples - Intel

[ztroop]

Description
I was reporting an issue in bevyengine/bevy#3449 and discovered that the underlying issue I'm experiencing is related to graphics. As suggested, I tried running some of the examples in wgpu and I'm seeing similar errors / segfaults.

System information
Linux 5.4.0-89-generic x86_64 Ubuntu 20.04.3 LTS

ztroop in ~/projects/wgpu/wgpu on master λ sudo dpkg -l | grep intel
ii  intel-media-va-driver:amd64                20.1.1+dfsg1-1                        amd64        VAAPI driver for the Intel GEN8+ Graphics family
ii  intel-microcode                            3.20210608.0ubuntu0.20.04.1           amd64        Processor microcode firmware for Intel CPUs
ii  libdrm-intel1:amd64                        2.4.105-3~20.04.2                     amd64        Userspace interface to intel-specific kernel DRM services -- runtime
ii  libdrm-intel1:i386                         2.4.105-3~20.04.2                     i386         Userspace interface to intel-specific kernel DRM services -- runtime
ii  xserver-xorg-video-intel                   2:2.99.917+git20200226-1              amd64        X.Org X server -- Intel i8xx, i9xx display driver
ztroop in ~/projects/wgpu/wgpu on master λ lspci | grep VGA
00:02.0 VGA compatible controller: Intel Corporation UHD Graphics (rev 02)
ztroop in ~/projects/wgpu/wgpu on master λ sudo lshw -C display
  *-display
       description: VGA compatible controller
       product: UHD Graphics
       vendor: Intel Corporation
       physical id: 2
       bus info: pci@0000:00:02.0
       version: 02
       width: 64 bits
       clock: 33MHz
       capabilities: pciexpress msi pm vga_controller bus_master cap_list rom
       configuration: driver=i915 latency=0
       resources: irq:129 memory:b0000000-b0ffffff memory:a0000000-afffffff ioport:4000(size=64) memory:c0000-dffff

Repro steps
Other examples besides water also produce similar errors.

ztroop in ~/projects/wgpu/wgpu on master λ cargo run --example water
    Finished dev [unoptimized + debuginfo] target(s) in 0.06s
     Running `/home/ztroop/projects/wgpu/target/debug/examples/water`
VUID-VkDebugUtilsMessengerCreateInfoEXT-messageType-parameter(ERROR / SPEC): msgNum: 0 - vkCreateDebugUtilsMessengerEXT: value of pCreateInfo->messageType contains flag bits that are not recognized members of VkDebugUtilsMessageTypeFlagBitsEXT The Vulkan spec states: messageType must be a valid combination of VkDebugUtilsMessageTypeFlagBitsEXT values (https://www.khronos.org/registry/vulkan/specs/1.1-extensions/html/vkspec.html#VUID-VkDebugUtilsMessengerCreateInfoEXT-messageType-parameter)
    Objects: 1
        [0] 0, type: 0, name: NULL
Using Intel(R) UHD Graphics (CML GT2) (Vulkan)
[2021-12-27T19:29:45Z ERROR wgpu_hal::vulkan::instance] VALIDATION [VUID-VkRenderPassBeginInfo-framebuffer-03210 (0x0)]
    	VkRenderPassBeginInfo: Image view #0 created from an image with usage set as 0x16, but image info #0 used to create the framebuffer had usage set as 0x14 The Vulkan spec states: If framebuffer was created with a VkFramebufferCreateInfo::flags value that included VK_FRAMEBUFFER_CREATE_IMAGELESS_BIT, each element of the pAttachments member of a VkRenderPassAttachmentBeginInfo structure included in the pNext chain must be a VkImageView of an image created with a value of VkImageCreateInfo::usage equal to the usage member of the corresponding element of VkFramebufferAttachmentsCreateInfo::pAttachments used to create framebuffer (https://www.khronos.org/registry/vulkan/specs/1.1-extensions/html/vkspec.html#VUID-VkRenderPassBeginInfo-framebuffer-03210)
[2021-12-27T19:29:45Z ERROR wgpu_hal::vulkan::instance] 	objects: (type: RENDER_PASS, hndl: 0x480000000048, name: ?)
[2021-12-27T19:29:45Z ERROR wgpu_hal::vulkan::instance] VALIDATION [VUID-VkRenderPassBeginInfo-framebuffer-03210 (0x0)]
    	VkRenderPassBeginInfo: Image view #1 created from an image with usage set as 0x26, but image info #1 used to create the framebuffer had usage set as 0x24 The Vulkan spec states: If framebuffer was created with a VkFramebufferCreateInfo::flags value that included VK_FRAMEBUFFER_CREATE_IMAGELESS_BIT, each element of the pAttachments member of a VkRenderPassAttachmentBeginInfo structure included in the pNext chain must be a VkImageView of an image created with a value of VkImageCreateInfo::usage equal to the usage member of the corresponding element of VkFramebufferAttachmentsCreateInfo::pAttachments used to create framebuffer (https://www.khronos.org/registry/vulkan/specs/1.1-extensions/html/vkspec.html#VUID-VkRenderPassBeginInfo-framebuffer-03210)
[2021-12-27T19:29:45Z ERROR wgpu_hal::vulkan::instance] 	objects: (type: RENDER_PASS, hndl: 0x480000000048, name: ?)
[1]    63252 segmentation fault (core dumped)  cargo run --example water

[kvark]

If you catch the segfault in gdb, you could check/provide a stack trace. Here, it looks to be vulkan validation that is crashing, not wgpu. I bet if you update the vulkan validation package, it will start running properly (or if you run in release, just for a test).

[ztroop]

I was using the most up to date software available in Ubuntu 20.04.3 LTS. I tried using a newer upstream kernel and mesa components that would help with vulkan functionality and couldn't get this to work. I did a bit more reading around i915, Vulkan and CML GT2 and determined that my best bet was to use a Linux distribution that's closer to upstream changes.

If there's anyone else who happens upon this issue, I ended up upgrading my system to a rolling release Linux distribution. That might sound extreme, but I have all the latest software and I'm able to run all the examples in wgpu without issue.