[llvm] [AArch64] Add FeatureZCRegMoveFPR128 subtarget feature (PR #152906)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-aarch64

Author: Tomer Shafir (tomershafir)

<details>
<summary>Changes</summary>

Adds a subtarget feature called FeatureZCRegMoveFPR128 that enables to query wether the target supports zero cycle reg move for FPR128 NEON registers, and embeds it into the appropriate processors.

Its a preparation for future optimizations.

---
Full diff: https://github.com/llvm/llvm-project/pull/152906.diff


2 Files Affected:

- (modified) llvm/lib/Target/AArch64/AArch64Features.td (+3) 
- (modified) llvm/lib/Target/AArch64/AArch64Processors.td (+10) 


``````````diff
diff --git a/llvm/lib/Target/AArch64/AArch64Features.td b/llvm/lib/Target/AArch64/AArch64Features.td
index c1c1f0a1024d0..55aea17d29f55 100644
--- a/llvm/lib/Target/AArch64/AArch64Features.td
+++ b/llvm/lib/Target/AArch64/AArch64Features.td
@@ -621,6 +621,9 @@ def FeatureZCRegMoveGPR64 : SubtargetFeature<"zcm-gpr64", "HasZeroCycleRegMoveGP
 def FeatureZCRegMoveGPR32 : SubtargetFeature<"zcm-gpr32", "HasZeroCycleRegMoveGPR32", "true",
                                         "Has zero-cycle register moves for GPR32 registers">;
 
+def FeatureZCRegMoveFPR128 : SubtargetFeature<"zcm-fpr128", "HasZeroCycleRegMoveFPR128", "true",
+                                        "Has zero-cycle register moves for FPR128 registers">;
+
 def FeatureZCRegMoveFPR64 : SubtargetFeature<"zcm-fpr64", "HasZeroCycleRegMoveFPR64", "true",
                                         "Has zero-cycle register moves for FPR64 registers">;
 
diff --git a/llvm/lib/Target/AArch64/AArch64Processors.td b/llvm/lib/Target/AArch64/AArch64Processors.td
index 1bc1d98a6f65b..1f5a6ff1328e7 100644
--- a/llvm/lib/Target/AArch64/AArch64Processors.td
+++ b/llvm/lib/Target/AArch64/AArch64Processors.td
@@ -321,6 +321,7 @@ def TuneAppleA7  : SubtargetFeature<"apple-a7", "ARMProcFamily", "AppleA7",
                                     FeatureFuseAES, FeatureFuseCryptoEOR,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing,
                                     FeatureZCZeroingFPWorkaround]>;
@@ -335,6 +336,7 @@ def TuneAppleA10 : SubtargetFeature<"apple-a10", "ARMProcFamily", "AppleA10",
                                     FeatureFuseCryptoEOR,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -348,6 +350,7 @@ def TuneAppleA11 : SubtargetFeature<"apple-a11", "ARMProcFamily", "AppleA11",
                                     FeatureFuseCryptoEOR,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -361,6 +364,7 @@ def TuneAppleA12 : SubtargetFeature<"apple-a12", "ARMProcFamily", "AppleA12",
                                     FeatureFuseCryptoEOR,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -374,6 +378,7 @@ def TuneAppleA13 : SubtargetFeature<"apple-a13", "ARMProcFamily", "AppleA13",
                                     FeatureFuseCryptoEOR,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -392,6 +397,7 @@ def TuneAppleA14 : SubtargetFeature<"apple-a14", "ARMProcFamily", "AppleA14",
                                     FeatureFuseLiterals,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -410,6 +416,7 @@ def TuneAppleA15 : SubtargetFeature<"apple-a15", "ARMProcFamily", "AppleA15",
                                     FeatureFuseLiterals,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -428,6 +435,7 @@ def TuneAppleA16 : SubtargetFeature<"apple-a16", "ARMProcFamily", "AppleA16",
                                     FeatureFuseLiterals,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -446,6 +454,7 @@ def TuneAppleA17 : SubtargetFeature<"apple-a17", "ARMProcFamily", "AppleA17",
                                     FeatureFuseLiterals,
                                     FeatureStorePairSuppress,
                                     FeatureZCRegMoveGPR64,
+                                    FeatureZCRegMoveFPR128,
                                     FeatureZCRegMoveFPR64,
                                     FeatureZCZeroing]>;
 
@@ -463,6 +472,7 @@ def TuneAppleM4 : SubtargetFeature<"apple-m4", "ARMProcFamily", "AppleM4",
                                      FeatureFuseCryptoEOR,
                                      FeatureFuseLiterals,
                                      FeatureZCRegMoveGPR64,
+                                     FeatureZCRegMoveFPR128,
                                      FeatureZCRegMoveFPR64,
                                      FeatureZCZeroing
                                      ]>;

``````````

</details>


https://github.com/llvm/llvm-project/pull/152906