Implement bmsk.clamp.b32

[zluda-violet]

Uses __ockl_bfm_u32.

[Copilot]

Pull request overview

This pull request implements support for the PTX bmsk.clamp.b32 instruction, which creates bit field masks with clamping behavior. The implementation uses AMD's __ockl_bfm_u32 built-in function to generate the appropriate bit masks.

Key Changes:

• Added parser support for the bmsk.clamp.b32 instruction with proper AST representation
• Implemented bmsk_clamp_b32 function using __ockl_bfm_u32 OCKL built-in
• Added test case verifying clamp behavior (position=24, width=12 → 0xFF000000)

Reviewed changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
ptx_parser/src/lib.rs	Adds parser definition for bmsk instruction with clamp mode
ptx_parser/src/ast.rs	Defines Bmsk instruction AST structure and imports BmskMode type
ptx/src/test/spirv_run/mod.rs	Registers new test case for bmsk_clamp_b32
ptx/src/test/spirv_run/bmsk_clamp_b32.ptx	PTX test file demonstrating bmsk.clamp.b32 usage
ptx/src/test/ll/bmsk_clamp_b32.ll	Expected LLVM IR output showing function call to implementation
ptx/src/pass/replace_instructions_with_functions.rs	Routes Bmsk instruction to bmsk_clamp_b32 function
ptx/src/pass/llvm/emit.rs	Marks Bmsk as a function-replaced instruction in LLVM emission
ptx/src/pass/instruction_mode_to_global_mode/mod.rs	Excludes Bmsk from mode transformation processing
ptx/src/pass/insert_post_saturation.rs	Marks Bmsk as not requiring post-saturation handling
ptx/lib/zluda_ptx_impl.cpp	Implements bmsk_clamp_b32 using __ockl_bfm_u32 OCKL built-in
ptx/lib/zluda_ptx_impl.bc	Updated compiled bitcode with new implementation

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.