[InstCombine] Test for matrix multiplication negation optimisation.

Zain Jaffal · fhahn · commit 56cc49115037 · 2022-09-22T16:49:10.000+01:00
If one of the operands is negated in a multiplication we can optimise the operation by moving the negation to the smallest operand or to the result Reviewed By: fhahn Differential Revision: https://reviews.llvm.org/D133287
diff --git a/llvm/test/Transforms/InstCombine/matrix-multiplication-negation.ll b/llvm/test/Transforms/InstCombine/matrix-multiplication-negation.ll
@@ -0,0 +1,302 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -passes=instcombine -S | FileCheck %s
+
+; The result has the fewest vector elements between the result and the two operands so the negation can be moved there
+define <2 x double> @test_negation_move_to_result(<6 x double> %a, <3 x double> %b) {
+; CHECK-LABEL: @test_negation_move_to_result(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <6 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> [[A_NEG]], <3 x double> [[B:%.*]], i32 2, i32 3, i32 1)
+; CHECK-NEXT:    ret <2 x double> [[RES]]
+;
+  %a.neg = fneg <6 x double> %a
+  %res = tail call <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> %a.neg, <3 x double> %b, i32 2, i32 3, i32 1)
+  ret <2 x double> %res
+}
+
+; The result has the fewest vector elements between the result and the two operands so the negation can be moved there
+; Fast flag should be preserved
+define <2 x double> @test_negation_move_to_result_with_fastflags(<6 x double> %a, <3 x double> %b) {
+; CHECK-LABEL: @test_negation_move_to_result_with_fastflags(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <6 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call fast <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> [[A_NEG]], <3 x double> [[B:%.*]], i32 2, i32 3, i32 1)
+; CHECK-NEXT:    ret <2 x double> [[RES]]
+;
+  %a.neg = fneg <6 x double> %a
+  %res = tail call fast <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> %a.neg, <3 x double> %b, i32 2, i32 3, i32 1)
+  ret <2 x double> %res
+}
+
+; %b has the fewest vector elements between the result and the two operands so the negation can be moved there
+define <9 x double> @test_move_negation_to_second_operand(<27 x double> %a, <3 x double> %b) {
+; CHECK-LABEL: @test_move_negation_to_second_operand(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <27 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> [[A_NEG]], <3 x double> [[B:%.*]], i32 9, i32 3, i32 1)
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %a.neg = fneg <27 x double> %a
+  %res = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> %a.neg, <3 x double> %b, i32 9, i32 3, i32 1)
+  ret <9 x double> %res
+}
+
+; %b has the fewest vector elements between the result and the two operands so the negation can be moved there
+; Fast flag should be preserved
+define <9 x double> @test_move_negation_to_second_operand_with_fast_flags(<27 x double> %a, <3 x double> %b) {
+; CHECK-LABEL: @test_move_negation_to_second_operand_with_fast_flags(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <27 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call fast <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> [[A_NEG]], <3 x double> [[B:%.*]], i32 9, i32 3, i32 1)
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %a.neg = fneg <27 x double> %a
+  %res = tail call fast <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> %a.neg, <3 x double> %b, i32 9, i32 3, i32 1)
+  ret <9 x double> %res
+}
+
+; The result has the fewest vector elements between the result and the two operands so the negation can be moved there
+define <2 x double> @test_negation_move_to_result_from_second_operand(<3 x double> %a, <6 x double> %b){
+; CHECK-LABEL: @test_negation_move_to_result_from_second_operand(
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <6 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <2 x double> @llvm.matrix.multiply.v2f64.v3f64.v6f64(<3 x double> [[A:%.*]], <6 x double> [[B_NEG]], i32 1, i32 3, i32 2)
+; CHECK-NEXT:    ret <2 x double> [[RES]]
+;
+  %b.neg = fneg <6 x double> %b
+  %res = tail call <2 x double> @llvm.matrix.multiply.v2f64.v3f64.v6f64(<3 x double> %a, <6 x double> %b.neg, i32 1, i32 3, i32 2)
+  ret <2 x double> %res
+}
+
+; %a has the fewest vector elements between the result and the two operands so the negation can be moved there
+define <9 x double> @test_move_negation_to_first_operand(<3 x double> %a, <27 x double> %b) {
+; CHECK-LABEL: @test_move_negation_to_first_operand(
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <27 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <9 x double> @llvm.matrix.multiply.v9f64.v3f64.v27f64(<3 x double> [[A:%.*]], <27 x double> [[B_NEG]], i32 1, i32 3, i32 9)
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %b.neg = fneg <27 x double> %b
+  %res = tail call <9 x double> @llvm.matrix.multiply.v9f64.v3f64.v27f64(<3 x double> %a, <27 x double> %b.neg, i32 1, i32 3, i32 9)
+  ret <9 x double> %res
+}
+
+; %a has the fewest vector elements between the result and the two operands so the negation is not moved
+define <15 x double> @test_negation_not_moved(<3 x double> %a, <5 x double> %b) {
+; CHECK-LABEL: @test_negation_not_moved(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <3 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double> [[A_NEG]], <5 x double> [[B:%.*]], i32 3, i32 1, i32 5)
+; CHECK-NEXT:    ret <15 x double> [[RES]]
+;
+  %a.neg = fneg <3 x double> %a
+  %res = tail call <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double> %a.neg, <5 x double> %b, i32 3, i32 1, i32 5)
+  ret <15 x double> %res
+}
+
+; %b as the fewest vector elements between the result and the two operands so the negation is not moved
+define <15 x double> @test_negation_not_moved_second_operand(<5 x double> %a, <3 x double> %b) {
+; CHECK-LABEL: @test_negation_not_moved_second_operand(
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <3 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <15 x double> @llvm.matrix.multiply.v15f64.v5f64.v3f64(<5 x double> [[A:%.*]], <3 x double> [[B_NEG]], i32 5, i32 1, i32 3)
+; CHECK-NEXT:    ret <15 x double> [[RES]]
+;
+  %b.neg = fneg <3 x double> %b
+  %res = tail call <15 x double> @llvm.matrix.multiply.v15f64.v5f64.v3f64(<5 x double> %a, <3 x double> %b.neg, i32 5, i32 1, i32 3)
+  ret <15 x double> %res
+}
+
+; the negation should be moved from the result to operand %a because it has the smallest vector element count
+define <15 x double> @test_negation_on_result(<3 x double> %a, <5 x double> %b) {
+; CHECK-LABEL: @test_negation_on_result(
+; CHECK-NEXT:    [[RES:%.*]] = tail call <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double> [[A:%.*]], <5 x double> [[B:%.*]], i32 3, i32 1, i32 5)
+; CHECK-NEXT:    [[RES_2:%.*]] = fneg <15 x double> [[RES]]
+; CHECK-NEXT:    ret <15 x double> [[RES_2]]
+;
+  %res = tail call <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double> %a, <5 x double> %b, i32 3, i32 1, i32 5)
+  %res.2 = fneg <15 x double> %res
+  ret <15 x double> %res.2
+}
+
+; both negations can be deleted
+define <2 x double> @test_with_two_operands_negated1(<6 x double> %a, <3 x double> %b){
+; CHECK-LABEL: @test_with_two_operands_negated1(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <6 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <3 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> [[A_NEG]], <3 x double> [[B_NEG]], i32 2, i32 3, i32 1)
+; CHECK-NEXT:    ret <2 x double> [[RES]]
+;
+  %a.neg = fneg <6 x double> %a
+  %b.neg = fneg <3 x double> %b
+  %res = tail call <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> %a.neg, <3 x double> %b.neg, i32 2, i32 3, i32 1)
+  ret <2 x double> %res
+}
+
+; both negations will be removed
+define <9 x double> @test_with_two_operands_negated2(<27 x double> %a, <3 x double> %b){
+; CHECK-LABEL: @test_with_two_operands_negated2(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <27 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <3 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> [[A_NEG]], <3 x double> [[B_NEG]], i32 9, i32 3, i32 1)
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %a.neg = fneg <27 x double> %a
+  %b.neg = fneg <3 x double> %b
+  %res = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> %a.neg, <3 x double> %b.neg, i32 9, i32 3, i32 1)
+  ret <9 x double> %res
+}
+
+; both negations will be removed
+define <9 x double> @test_with_two_operands_negated_with_fastflags(<27 x double> %a, <3 x double> %b){
+; CHECK-LABEL: @test_with_two_operands_negated_with_fastflags(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <27 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <3 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call fast <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> [[A_NEG]], <3 x double> [[B_NEG]], i32 9, i32 3, i32 1)
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %a.neg = fneg <27 x double> %a
+  %b.neg = fneg <3 x double> %b
+  %res = tail call fast <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> %a.neg, <3 x double> %b.neg, i32 9, i32 3, i32 1)
+  ret <9 x double> %res
+}
+
+; both negations should be removed
+define <9 x double> @test_with_two_operands_negated2_commute(<3 x double> %a, <27 x double> %b){
+; CHECK-LABEL: @test_with_two_operands_negated2_commute(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <3 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <27 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = call <9 x double> @llvm.matrix.multiply.v9f64.v3f64.v27f64(<3 x double> [[A_NEG]], <27 x double> [[B_NEG]], i32 1, i32 3, i32 9)
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %a.neg = fneg <3 x double> %a
+  %b.neg = fneg <27 x double> %b
+  %res = call <9 x double> @llvm.matrix.multiply.v9f64.v3f64.v27f64(<3 x double> %a.neg, <27 x double> %b.neg, i32 1, i32 3, i32 9)
+  ret <9 x double> %res
+}
+
+define <4 x double> @matrix_multiply_two_operands_negated_with_same_size(<2 x double> %a, <2 x double> %b) {
+; CHECK-LABEL: @matrix_multiply_two_operands_negated_with_same_size(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <2 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <2 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = call <4 x double> @llvm.matrix.multiply.v4f64.v2f64.v2f64(<2 x double> [[A_NEG]], <2 x double> [[B_NEG]], i32 2, i32 1, i32 2)
+; CHECK-NEXT:    ret <4 x double> [[RES]]
+;
+  %a.neg = fneg <2 x double> %a
+  %b.neg = fneg <2 x double> %b
+  %res = call <4 x double> @llvm.matrix.multiply.v4f64.v2f64.v2f64(<2 x double> %a.neg, <2 x double> %b.neg, i32 2, i32 1, i32 2)
+  ret <4 x double> %res
+}
+
+define <2 x double> @matrix_multiply_two_operands_with_multiple_uses(<6 x double> %a, <3 x double> %b) {
+; CHECK-LABEL: @matrix_multiply_two_operands_with_multiple_uses(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <6 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <3 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> [[A_NEG]], <3 x double> [[B_NEG]], i32 2, i32 3, i32 1)
+; CHECK-NEXT:    [[RES_2:%.*]] = shufflevector <6 x double> [[A_NEG]], <6 x double> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[RES_3:%.*]] = fadd <2 x double> [[RES_2]], [[RES]]
+; CHECK-NEXT:    ret <2 x double> [[RES_3]]
+;
+  %a.neg = fneg <6 x double> %a
+  %b.neg = fneg <3 x double> %b
+  %res = tail call <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double> %a.neg, <3 x double> %b.neg, i32 2, i32 3, i32 1)
+  %res.2 = shufflevector <6 x double> %a.neg, <6 x double> undef,
+  <2 x i32>  <i32 0, i32 1>
+  %res.3 = fadd <2 x double> %res.2, %res
+  ret <2 x double> %res.3
+}
+
+define <9 x double> @matrix_multiply_two_operands_with_multiple_uses2(<27 x double> %a, <3 x double> %b, ptr %a_loc, ptr %b_loc){
+; CHECK-LABEL: @matrix_multiply_two_operands_with_multiple_uses2(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <27 x double> [[A:%.*]]
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <3 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> [[A_NEG]], <3 x double> [[B_NEG]], i32 9, i32 3, i32 1)
+; CHECK-NEXT:    store <27 x double> [[A_NEG]], ptr [[A_LOC:%.*]], align 256
+; CHECK-NEXT:    store <3 x double> [[B_NEG]], ptr [[B_LOC:%.*]], align 32
+; CHECK-NEXT:    ret <9 x double> [[RES]]
+;
+  %a.neg = fneg <27 x double> %a
+  %b.neg = fneg <3 x double> %b
+  %res = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> %a.neg, <3 x double> %b.neg, i32 9, i32 3, i32 1)
+  store <27 x double> %a.neg, ptr %a_loc
+  store <3 x double> %b.neg, ptr %b_loc
+  ret <9 x double> %res
+}
+
+define <12 x double> @fneg_with_multiple_uses(<15 x double> %a, <20 x double> %b){
+; CHECK-LABEL: @fneg_with_multiple_uses(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <15 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double> [[A_NEG]], <20 x double> [[B:%.*]], i32 3, i32 5, i32 4)
+; CHECK-NEXT:    [[RES_2:%.*]] = shufflevector <15 x double> [[A_NEG]], <15 x double> undef, <12 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11>
+; CHECK-NEXT:    [[RES_3:%.*]] = fadd <12 x double> [[RES_2]], [[RES]]
+; CHECK-NEXT:    ret <12 x double> [[RES_3]]
+;
+  %a.neg = fneg <15 x double> %a
+  %res = tail call <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double> %a.neg, <20 x double> %b, i32 3, i32 5, i32 4)
+  %res.2 = shufflevector <15 x double> %a.neg, <15 x double> undef,
+  <12 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11>
+  %res.3 = fadd <12 x double> %res.2, %res
+  ret <12 x double> %res.3
+}
+
+define <12 x double> @fneg_with_multiple_uses_2(<15 x double> %a, <20 x double> %b, ptr %a_loc){
+; CHECK-LABEL: @fneg_with_multiple_uses_2(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <15 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double> [[A_NEG]], <20 x double> [[B:%.*]], i32 3, i32 5, i32 4)
+; CHECK-NEXT:    store <15 x double> [[A_NEG]], ptr [[A_LOC:%.*]], align 128
+; CHECK-NEXT:    ret <12 x double> [[RES]]
+;
+  %a.neg = fneg <15 x double> %a
+  %res = tail call <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double> %a.neg, <20 x double> %b, i32 3, i32 5, i32 4)
+  store <15 x double> %a.neg, ptr %a_loc
+  ret <12 x double> %res
+}
+; negation should be moved to the second operand given it has the smallest operand count
+define <72 x double> @chain_of_matrix_mutliplies(<27 x double> %a, <3 x double> %b, <8 x double> %c) {
+; CHECK-LABEL: @chain_of_matrix_mutliplies(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <27 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> [[A_NEG]], <3 x double> [[B:%.*]], i32 9, i32 3, i32 1)
+; CHECK-NEXT:    [[RES_2:%.*]] = tail call <72 x double> @llvm.matrix.multiply.v72f64.v9f64.v8f64(<9 x double> [[RES]], <8 x double> [[C:%.*]], i32 9, i32 1, i32 8)
+; CHECK-NEXT:    ret <72 x double> [[RES_2]]
+;
+  %a.neg = fneg <27 x double> %a
+  %res = tail call <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double> %a.neg, <3 x double> %b, i32 9, i32 3, i32 1)
+  %res.2 = tail call <72 x double> @llvm.matrix.multiply.v72f64.v9f64.v8f64(<9 x double> %res, <8 x double> %c, i32 9, i32 1, i32 8)
+  ret <72 x double> %res.2
+}
+
+; first negation should be moved to %a
+; second negation should be moved to the result of the second multipication
+define <6 x double> @chain_of_matrix_mutliplies_with_two_negations(<3 x double> %a, <5 x double> %b, <10 x double> %c) {
+; CHECK-LABEL: @chain_of_matrix_mutliplies_with_two_negations(
+; CHECK-NEXT:    [[B_NEG:%.*]] = fneg <5 x double> [[B:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double> [[A:%.*]], <5 x double> [[B_NEG]], i32 3, i32 1, i32 5)
+; CHECK-NEXT:    [[RES_NEG:%.*]] = fneg <15 x double> [[RES]]
+; CHECK-NEXT:    [[RES_2:%.*]] = tail call <6 x double> @llvm.matrix.multiply.v6f64.v15f64.v10f64(<15 x double> [[RES_NEG]], <10 x double> [[C:%.*]], i32 3, i32 5, i32 2)
+; CHECK-NEXT:    ret <6 x double> [[RES_2]]
+;
+  %b.neg = fneg <5 x double> %b
+  %res = tail call <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double> %a, <5 x double> %b.neg, i32 3, i32 1, i32 5)
+  %res.neg = fneg <15 x double> %res
+  %res.2 = tail call <6 x double> @llvm.matrix.multiply.v6f64.v15f64.v10f64(<15 x double> %res.neg, <10 x double> %c, i32 3, i32 5, i32 2)
+  ret <6 x double> %res.2
+}
+
+; negation should be propagated to the result of the second matrix multiplication
+define <6 x double> @chain_of_matrix_mutliplies_propagation(<15 x double> %a, <20 x double> %b, <8 x double> %c){
+; CHECK-LABEL: @chain_of_matrix_mutliplies_propagation(
+; CHECK-NEXT:    [[A_NEG:%.*]] = fneg <15 x double> [[A:%.*]]
+; CHECK-NEXT:    [[RES:%.*]] = tail call <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double> [[A_NEG]], <20 x double> [[B:%.*]], i32 3, i32 5, i32 4)
+; CHECK-NEXT:    [[RES_2:%.*]] = tail call <6 x double> @llvm.matrix.multiply.v6f64.v12f64.v8f64(<12 x double> [[RES]], <8 x double> [[C:%.*]], i32 3, i32 4, i32 2)
+; CHECK-NEXT:    ret <6 x double> [[RES_2]]
+;
+  %a.neg = fneg <15 x double> %a
+  %res = tail call <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double> %a.neg, <20 x double> %b, i32 3, i32 5, i32 4)
+  %res.2 = tail call <6 x double> @llvm.matrix.multiply.v6f64.v12f64.v8f64(<12 x double> %res, <8 x double> %c, i32 3, i32 4, i32 2)
+  ret <6 x double> %res.2
+}
+
+declare <2 x double> @llvm.matrix.multiply.v2f64.v6f64.v3f64(<6 x double>, <3 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <4 x double> @llvm.matrix.multiply.v4f64.v2f64.v2f64(<2 x double>, <2 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <2 x double> @llvm.matrix.multiply.v2f64.v3f64.v6f64(<3 x double>, <6 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double>, <3 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <9 x double> @llvm.matrix.multiply.v9f64.v3f64.v27f64(<3 x double>, <27 x double>, i32 immarg, i32 immarg, i32 immarg)
+declare <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double>, <5 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <15 x double> @llvm.matrix.multiply.v15f64.v5f64.v3f64(<5 x double>, <3 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <72 x double> @llvm.matrix.multiply.v72f64.v9f64.v8f64(<9 x double>, <8 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <12 x double> @llvm.matrix.multiply.v12f64.v15f64.v20f64(<15 x double>, <20 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <21 x double> @llvm.matrix.multiply.v21f64.v15f64.v35f64(<15 x double>, <35 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <6 x double> @llvm.matrix.multiply.v6f64.v15f64.v10f64(<15 x double>, <10 x double>, i32 immarg, i32 immarg, i32 immarg) #1
+declare <6 x double> @llvm.matrix.multiply.v6f64.v12f64.v8f64(<12 x double>, <8 x double>, i32 immarg, i32 immarg, i32 immarg) #1
