Lines Matching full:neon

30 // Floating Point, HW Division and Neon Support
99 def FeatureNEON           : SubtargetFeature<"neon", "HasNEON", "true",
100                                              "Enable NEON instructions",
257 // Some targets (e.g. Cortex-A9) prefer VMOVSR to VMOVDRR even when using NEON
269 // Some targets (e.g. Cortex-A9) have muxed AGU and NEON/FPU.
270 // True if the AGU and NEON/FPU units are multiplexed.
273                                              "Has muxed AGU and NEON/FPU">;
302 // True if splat a register between VFP and NEON instructions.
303 def FeatureSplatVFPToNeon : SubtargetFeature<"splat-vfp-neon",
305                                              "Splat register from VFP to NEON",
308 // Whether or not it is profitable to expand VFP/NEON MLA/MLS instructions.
312                                              "Expand VFP/NEON MLA/MLS instructions">;
314 // Some targets have special RAW hazards for VFP/NEON VMLA/VMLS.
315 // True if VFP/NEON VMLA/VMLS have special RAW hazards.
320 // VFP to NEON, as an execution domain optimization.
321 // True if VMOVRS, VMOVSR and VMOVS will be converted from VFP to NEON.
322 def FeatureNEONForFPMovs  : SubtargetFeature<"neon-fpmovs",
325                                              "VMOVS to NEON">;
327 // Some processors benefit from using NEON instructions for scalar
330 // Use the method useNEONForSinglePrecisionFP() to determine if NEON should actually be used.
334                                              "Use NEON for single precision FP">;
350 // play nicely with other VFP / NEON instructions, and it's generally better
352 // If the VFP2 / NEON instructions are available, indicates
355                                              "Disable VFP / NEON MAC instructions">;
358 // If the VFP4 / NEON instructions are available, indicates
361                                              "Disable VFP / NEON FMA instructions">;
364 /// True if NEON has special multiplier accumulator