projects / platform / upstream / VK-GL-CTS.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 5655067) | patch
Use memcpy to copy floats for vertex input
authorMark Adams <marka@nvidia.com>
Sat, 16 Jan 2016 16:37:31 +0000 (11:37 -0500)
committerMark Adams <marka@nvidia.com>
Sat, 16 Jan 2016 16:37:31 +0000 (11:37 -0500)
commitb1d85fb2842920a4f2eff81f4dd70df286dbddf3
treec3b0ea068f20ebe27b2de6f909d1ed66d83b6a35tree | snapshot
parent56550672b1b4838ff0bfc3d9b1a96c8e127103bdcommit | diff
Use memcpy to copy floats for vertex input

The destination buffer may not be aligned as required to
cast as a float* or deFloat16*. This was causing crashes
due to bus errors on 32-bit ARM systems in several of the
dEQP-VK.pipeline.vertex_input.multiple_attributes.* tests.
external/vulkancts/modules/vulkan/pipeline/vktPipelineVertexInputTests.cpp diff | blob | history
Domain: Graphics System / GL;