[llvm] [LV] Transform to handle exits in the scalar loop (PR #148626)

[Graham Hunter via llvm-commits]

================
@@ -355,6 +451,70 @@ define void @inner_loop_trip_count_depends_on_outer_iv(ptr align 8 dereferenceab
 ; CHECK:       exit:
 ; CHECK-NEXT:    ret void
 ;
+; EE-SCALAR-LABEL: define void @inner_loop_trip_count_depends_on_outer_iv(
+; EE-SCALAR-SAME: ptr align 8 dereferenceable(1792) [[THIS:%.*]], ptr [[DST:%.*]]) {
+; EE-SCALAR-NEXT:  entry:
+; EE-SCALAR-NEXT:    [[GEP_SRC:%.*]] = getelementptr i8, ptr [[THIS]], i64 1000
+; EE-SCALAR-NEXT:    br label [[OUTER_HEADER:%.*]]
+; EE-SCALAR:       outer.header:
+; EE-SCALAR-NEXT:    [[OUTER_IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[OUTER_IV_NEXT:%.*]], [[OUTER_LATCH:%.*]] ]
+; EE-SCALAR-NEXT:    [[C_1:%.*]] = icmp eq i64 [[OUTER_IV]], 0
+; EE-SCALAR-NEXT:    br i1 [[C_1]], label [[THEN:%.*]], label [[INNER_HEADER_PREHEADER:%.*]]
+; EE-SCALAR:       inner.header.preheader:
+; EE-SCALAR-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[OUTER_IV]], 4
+; EE-SCALAR-NEXT:    br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
+; EE-SCALAR:       vector.ph:
+; EE-SCALAR-NEXT:    [[N_MOD_VF:%.*]] = urem i64 [[OUTER_IV]], 4
+; EE-SCALAR-NEXT:    [[N_VEC:%.*]] = sub i64 [[OUTER_IV]], [[N_MOD_VF]]
+; EE-SCALAR-NEXT:    [[WIDE_LOAD:%.*]] = load <4 x ptr>, ptr [[GEP_SRC]], align 8
----------------
huntergr-arm wrote:

No. The block above this performs a min_iters_check, ensuring we have at least <VF> lanes to process, so no masking for the condition load is required until we enter the vector body. (And if we reduced the vector trip count, we wouldn't need a masked load there either).

https://github.com/llvm/llvm-project/pull/148626