vvl: Test a new .clang-format #9172
Conversation
|
CI Vulkan-ValidationLayers build queued with queue ID 340475. |
|
CI Vulkan-ValidationLayers build # 18574 running. |
|
So for instance I like that for GPU-AV specifically because it turns this: into this:
So parsing the captured variables list in the lambda is way easier |
|
CI Vulkan-ValidationLayers build # 18574 failed. |
layers/best_practices/bp_state.h
Outdated
| namespace bp_state { | ||
| class Image : public vvl::Image { | ||
| namespace bp_state | ||
| { |
There was a problem hiding this comment.
moving brackets around seems very unnecessary to me
| CALL_STATE vkGetPhysicalDeviceSurfacePresentModesKHRState = UNCALLED; | ||
| CALL_STATE vkGetPhysicalDeviceSurfaceFormatsKHRState = UNCALLED; | ||
| uint32_t surface_formats_count = 0; | ||
| CALL_STATE vkGetPhysicalDeviceLayerPropertiesState = UNCALLED; // Currently unused |
There was a problem hiding this comment.
not a huge fan of lining things up in vertical columns. Fortan called and wants it format back
| uint32_t numDrawCallsDepthOnly = 0; | ||
| struct RenderPassState | ||
| { | ||
| bool depthAttachment = false; |
There was a problem hiding this comment.
Is the issue the blanks between the variable name and its default value? I should be able to remove that
There was a problem hiding this comment.
All the extra spaces on lines like this. trying to make vertical columns will lead to lots of churn whenever someone adds or removes a declaration that causes the columns to change. And inevitably someone who isn't fully up to speed on git clang-format will waste a bunch of time trying to fix the columns manually.
| const VkMemoryHeap& memory_heap, | ||
| std::optional<vvl::DedicatedBinding>&& dedicated_binding, | ||
| uint32_t physical_device_count) : | ||
| vvl::DeviceMemory(handle, |
There was a problem hiding this comment.
this part is mostly nicer
There was a problem hiding this comment.
well until you have a 13-inch laptop screen and this takes up half of it 🤷♂️
arno-lunarg
force-pushed
the
arno-new-clang-format
branch
from
5230e4a to
997f2ba
Compare
|
CI Vulkan-ValidationLayers build queued with queue ID 340999. |
|
CI Vulkan-ValidationLayers build # 18582 running. |
|
CI Vulkan-ValidationLayers build # 18582 failed. |
| return ValidateBuildAccelerationStructure(commandBuffer, error_obj.location); | ||
| } | ||
|
|
||
| bool BestPractices::ValidateBuildAccelerationStructure(VkCommandBuffer commandBuffer, const Location& loc) const { | ||
| bool skip = false; | ||
| bool skip = false; |
There was a problem hiding this comment.
I don't get why we need this white space like this... I'm gonna just go crazy and end up trying to add more newlines to satisfy my OCD to not have this like this
| uint64_t num_submits = 0; | ||
| bool uses_vertex_buffer = false; | ||
| uint32_t small_indexed_draw_call_count = 0; | ||
| uint64_t num_submits = 0; |
There was a problem hiding this comment.
sorry, not of fan have to highlight over a line to make sure I know which type it is once these get many lines stack on it
|
just want to say if we needed to revert something like #9200 that would have been a nightmare/impossible if we touch every file like this.... I just don't see what we gain from this, the whole idea of clang-format is we picked something and stick with it, we plan to update this clang-format again in 2 years when someone else suggests it is better, if we change it now, what justification is there to not change it again in the future |
No description provided.