[AArch64] Add option -msve-streaming-vector-bits= . (#144611)

This is similar to -msve-vector-bits, but for streaming mode: it
constrains the legal values of "vscale", allowing optimizations based on
that constraint.

This also fixes conversions between SVE vectors and fixed-width vectors
in streaming functions with -msve-vector-bits and
-msve-streaming-vector-bits.

This rejects any use of arm_sve_vector_bits types in streaming
functions; if it becomes relevant, we could add
arm_sve_streaming_vector_bits types in the future.

This doesn't touch the __ARM_FEATURE_SVE_BITS define.

1 files changed, 8 insertions, 2 deletions

diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
index 70a0979..776a646 100644
--- a/clang/lib/CodeGen/CodeGenFunction.cpp
+++ b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -1109,10 +1109,16 @@ void CodeGenFunction::StartFunction(GlobalDecl GD, QualType RetTy,
 
   // Add vscale_range attribute if appropriate.
   llvm::StringMap<bool> FeatureMap;
-  bool IsArmStreaming = false;
+  auto IsArmStreaming = TargetInfo::ArmStreamingKind::NotStreaming;
   if (FD) {
     getContext().getFunctionFeatureMap(FeatureMap, FD);
-    IsArmStreaming = IsArmStreamingFunction(FD, true);
+    if (const auto *T = FD->getType()->getAs<FunctionProtoType>())
+      if (T->getAArch64SMEAttributes() &
+          FunctionType::SME_PStateSMCompatibleMask)
+        IsArmStreaming = TargetInfo::ArmStreamingKind::StreamingCompatible;
+
+    if (IsArmStreamingFunction(FD, true))
+      IsArmStreaming = TargetInfo::ArmStreamingKind::Streaming;
   }
   std::optional<std::pair<unsigned, unsigned>> VScaleRange =
       getContext().getTargetInfo().getVScaleRange(getLangOpts(), IsArmStreaming,