diff --git a/llvm/test/Transforms/InstCombine/matrix-multiplication-negation.ll b/llvm/test/Transforms/InstCombine/matrix-multiplication-negation.ll new file mode 100644 --- /dev/null +++ b/llvm/test/Transforms/InstCombine/matrix-multiplication-negation.ll @@ -0,0 +1,120 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --force-update +; RUN: opt < %s -passes=instcombine -S | FileCheck %s + +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 +} + +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 +} + +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 +} + +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 +} + +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 +} + +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: [[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 + %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, i32 9, i32 3, i32 1) + 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> +; 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> + %res.3 = fadd <12 x double> %res.2, %res + ret <12 x double> %res.3 +} + +define <15 x double> @chain_of_matrix_mutliplies(<3 x double> %a, <5 x double> %b, <35 x double> %c) { +; CHECK-LABEL: @chain_of_matrix_mutliplies( +; 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) + %res.2 = tail call <21 x double> @llvm.matrix.multiply.v21f64.v15f64.v35f64(<15 x double> %res, <35 x double> %c, i32 3, i32 5, i32 7) + ret <15 x double> %res +} + +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: [[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: [[RES_2:%.*]] = tail call <6 x double> @llvm.matrix.multiply.v6f64.v15f64.v10f64(<15 x double> [[RES]], <10 x double> [[C:%.*]], i32 3, i32 5, i32 2) +; CHECK-NEXT: ret <6 x double> [[RES_2]] +; + %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) + %res.neg = fneg <15 x double> %res + %res.2 = tail call <6 x double> @llvm.matrix.multiply.v6f64.v15f64.v10f64(<15 x double> %res, <10 x double> %c, i32 3, i32 5, 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 <9 x double> @llvm.matrix.multiply.v9f64.v27f64.v3f64(<27 x double>, <3 x double>, i32 immarg, i32 immarg, i32 immarg) #1 +declare <15 x double> @llvm.matrix.multiply.v15f64.v3f64.v5f64(<3 x double>, <5 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