Skip to content

Further expanded vector operations to force scalarization#3

Open
delvinbuckley-ops wants to merge 1 commit into
vortexgpgpu:vortex_3.xfrom
delvinbuckley-ops:local-bug-fixes
Open

Further expanded vector operations to force scalarization#3
delvinbuckley-ops wants to merge 1 commit into
vortexgpgpu:vortex_3.xfrom
delvinbuckley-ops:local-bug-fixes

Conversation

@delvinbuckley-ops

Copy link
Copy Markdown

Further expanded vector operations to force scalarization. This fixes Clang compiler crashes (like "Cannot select: v4f32 = fadd") caused by frameworks that request explicit float4 vector math. By marking these operations as Expand, high-level vectored types will be converted into 32-bit scalar instructions.

… Clang compiler crashes (like "Cannot select: v4f32 = fadd") caused by frameworks that request explicit float4 vector math. By marking these operations as Expand, high-level vectored types will be converted into 32-bit scalar instructions.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant