[RISCV] Check type size for lax conversions between RVV builtin types…

… and VectorType::RVVFixedLengthDataVector.

This code was copied from SVE and modified for RVV. For SVE, there
is only one size for builtin types so they didn't need to check
the size. For RVV, due to LMUL there are 7 different sizes of builtin
types so we do need to check the size.

I'm not sure we should have lax vector conversions at all for RVV.
That appears to be contributing to #64404

This patch at least fixes the obvious correctness issue.
This should be backported to LLVM 17.

Reviewed By: jacquesguan

Differential Revision: https://reviews.llvm.org/D157130

clang/lib/AST/ASTContext.cpp

@@ -9612,9 +9612,8 @@ bool ASTContext::areLaxCompatibleRVVTypes(QualType FirstType,
       const LangOptions::LaxVectorConversionKind LVCKind =
           getLangOpts().getLaxVectorConversions();
 
-      // If __riscv_v_fixed_vlen != N do not allow GNU vector lax conversion.
-      if (VecTy->getVectorKind() == VectorType::GenericVector &&
-          getTypeSize(SecondType) != getRVVTypeSize(*this, BT))
+      // If __riscv_v_fixed_vlen != N do not allow vector lax conversion.
+      if (getTypeSize(SecondType) != getRVVTypeSize(*this, BT))
         return false;
 
       // If -flax-vector-conversions=all is specified, the types are

clang/test/Sema/riscv-rvv-lax-vector-conversions.c

@@ -2,8 +2,6 @@
 // RUN: %clang_cc1 -triple riscv64-none-linux-gnu -target-feature +f -target-feature +d -target-feature +zve64d -mvscale-min=8 -mvscale-max=8 -flax-vector-conversions=integer -ffreestanding -fsyntax-only -verify=lax-vector-integer %s
 // RUN: %clang_cc1 -triple riscv64-none-linux-gnu -target-feature +f -target-feature +d -target-feature +zve64d -mvscale-min=8 -mvscale-max=8 -flax-vector-conversions=all -ffreestanding -fsyntax-only -verify=lax-vector-all %s
 
-// lax-vector-all-no-diagnostics
-
 // REQUIRES: riscv-registered-target
 
 #define RVV_FIXED_ATTR __attribute__((riscv_rvv_vector_bits(__riscv_v_fixed_vlen)))
@@ -20,6 +18,8 @@ typedef __rvv_uint64m1_t vuint64m1_t;
 typedef __rvv_float32m1_t vfloat32m1_t;
 typedef __rvv_float64m1_t vfloat64m1_t;
 
+typedef __rvv_int64m2_t vint64m2_t;
+
 typedef vfloat32m1_t rvv_fixed_float32m1_t RVV_FIXED_ATTR;
 typedef vint32m1_t rvv_fixed_int32m1_t RVV_FIXED_ATTR;
 typedef float gnu_fixed_float32m1_t GNU_FIXED_ATTR;
@@ -76,3 +76,17 @@ void gnu_allowed_with_all_lax_conversions() {
   // lax-vector-none-error@-1 {{assigning to 'vfloat64m1_t' (aka '__rvv_float64m1_t') from incompatible type}}
   // lax-vector-integer-error@-2 {{assigning to 'vfloat64m1_t' (aka '__rvv_float64m1_t') from incompatible type}}
 }
+
+void not_allowed() {
+  rvv_fixed_int32m1_t fi32m1;
+  vint64m2_t si64m2;
+
+  fi32m1 = si64m2;
+  // lax-vector-none-error@-1 {{assigning to 'rvv_fixed_int32m1_t' (vector of 16 'int' values) from incompatible type}}
+  // lax-vector-integer-error@-2 {{assigning to 'rvv_fixed_int32m1_t' (vector of 16 'int' values) from incompatible type}}
+  // lax-vector-all-error@-3 {{assigning to 'rvv_fixed_int32m1_t' (vector of 16 'int' values) from incompatible type}}
+  si64m2 = fi32m1;
+  // lax-vector-none-error@-1 {{assigning to 'vint64m2_t' (aka '__rvv_int64m2_t') from incompatible type}}
+  // lax-vector-integer-error@-2 {{assigning to 'vint64m2_t' (aka '__rvv_int64m2_t') from incompatible type}}
+  // lax-vector-all-error@-3 {{assigning to 'vint64m2_t' (aka '__rvv_int64m2_t') from incompatible type}}
+}