summaryrefslogtreecommitdiff
path: root/mlir/lib/Conversion/MathToROCDL
AgeCommit message (Expand)Author
2025-11-02[mlir] Remove redundant typename (NFC) (#166108)Kazu Hirata
2025-10-17[MLIR][ROCDL] Add math.clampf -> rocdl.fmed3 conversion (#163520)Keshav Vinayak Jha
2025-10-14Revert "[MLIR][ROCDL] Add math.clampf -> rocdl.fmed3 conversion" (#163447)Fabian Mora
2025-10-14[MLIR][ROCDL] Add math.clampf -> rocdl.fmed3 conversion (#163259)Keshav Vinayak Jha
2025-07-30[MLIR] Migrate some conversion passes and dialects to LDBG() macro (NFC) (#15...Mehdi Amini
2025-07-04[mlir] Remove unused includes (NFC) (#147101)Kazu Hirata
2025-02-27[MLIR][ROCDL] Add conversion of math.erfc to AMD GPU library calls (#128899)Jan Leyonberg
2025-01-13[MLIR][ROCDL] Convert `math::fpowi` to ROCDL call (#122640)lialan
2024-10-05[mlir][NFC] Mark type converter in `populate...` functions as `const` (#111250)Matthias Springer
2024-09-23[mlir][AMDGPU] Add support for AMD f16 math library calls (#108809)Daniel Hernandez-Juarez
2024-09-04[MLIR][ROCDL] Remove patterns for ops supported as intrinsics in the AMDGPU b...Jan Leyonberg
2024-07-25[MLIR][GPUToNVVM] support fastMath and other non-supported mathOp (#99890)runseny
2024-07-17[MLIR][ROCDL] Refactor conversion of math operations to ROCDL calls to a sepa...Jan Leyonberg