diff --git a/src/MKLSparse.jl b/src/MKLSparse.jl
index 21c211f..6b9dad2 100644
--- a/src/MKLSparse.jl
+++ b/src/MKLSparse.jl
@@ -21,21 +21,18 @@ end
     INTERFACE_GNU
 end
 
-function set_threading_layer(layer::Threading = THREADING_SEQUENTIAL)
+function set_threading_layer(layer::Threading)
     err = @ccall libmkl_rt.MKL_Set_Threading_Layer(layer::Cint)::Cint
     (err == -1) && error("MKL_Set_Threading_Layer() returned -1")
     return nothing
 end
 
-function set_interface_layer(interface::Interface = INTERFACE_ILP64)
+function set_interface_layer(interface::Interface)
     err = @ccall libmkl_rt.MKL_Set_Interface_Layer(interface::Cint)::Cint
     (err == -1) && error("MKL_Set_Interface_Layer() returned -1")
     return nothing
 end
 
-function __init__()
-    set_interface_layer(Base.USE_BLAS64 ? INTERFACE_ILP64 : INTERFACE_LP64)
-end
 
 # Wrappers generated by Clang.jl
 include("libmklsparse.jl")
diff --git a/src/generator.jl b/src/generator.jl
index 68ca14f..f610a1b 100644
--- a/src/generator.jl
+++ b/src/generator.jl
@@ -27,61 +27,130 @@ function _check_mat_mult_matvec(C, A, B, tA)
     end
 end
 
-function cscmv!(transa::Char, α::T, matdescra::String,
-                A::SparseMatrixCSC{T, BlasInt}, x::StridedVector{T},
-                β::T, y::StridedVector{T}) where {T <: BlasFloat}
-    _check_transa(transa)
-    _check_mat_mult_matvec(y, A, x, transa)
-    __counter[] += 1
+for (fname, T) in ((:mkl_scscmv, :Float32   ),
+                   (:mkl_dcscmv, :Float64   ),
+                   (:mkl_ccscmv, :ComplexF32),
+                   (:mkl_zcscmv, :ComplexF64))
+    @eval begin
+        function cscmv!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int32}, x::StridedVector{$T},
+                        β::$T, y::StridedVector{$T})
+            _check_transa(transa)
+            _check_mat_mult_matvec(y, A, x, transa)
+            __counter[] += 1
+            $fname(transa, A.m, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, β, y)
+            return y
+        end
 
-    T == Float32    && (mkl_scscmv(transa, A.m, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, β, y))
-    T == Float64    && (mkl_dcscmv(transa, A.m, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, β, y))
-    T == ComplexF32 && (mkl_ccscmv(transa, A.m, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, β, y))
-    T == ComplexF64 && (mkl_zcscmv(transa, A.m, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, β, y))
-    return y
+        function cscmv!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int64}, x::StridedVector{$T},
+                        β::$T, y::StridedVector{$T})
+            _check_transa(transa)
+            _check_mat_mult_matvec(y, A, x, transa)
+            __counter[] += 1
+            set_interface_layer(INTERFACE_ILP64)
+            $fname(transa, A.m, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, β, y)
+            set_interface_layer(INTERFACE_LP64)
+            return y
+        end
+    end
 end
 
-function cscmm!(transa::Char, α::T, matdescra::String,
-                A::SparseMatrixCSC{T, BlasInt}, B::StridedMatrix{T},
-                β::T, C::StridedMatrix{T}) where {T <: BlasFloat}
-    _check_transa(transa)
-    _check_mat_mult_matvec(C, A, B, transa)
-    mB, nB = size(B)
-    mC, nC = size(C)
-    __counter[] += 1
-    T == Float32    && (mkl_scscmm(transa, A.m, nC, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, β, C, mC))
-    T == Float64    && (mkl_dcscmm(transa, A.m, nC, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, β, C, mC))
-    T == ComplexF32 && (mkl_ccscmm(transa, A.m, nC, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, β, C, mC))
-    T == ComplexF64 && (mkl_zcscmm(transa, A.m, nC, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, β, C, mC))
-    return C
+for (fname, T) in ((:mkl_scscmm, :Float32   ),
+                   (:mkl_dcscmm, :Float64   ),
+                   (:mkl_ccscmm, :ComplexF32),
+                   (:mkl_zcscmm, :ComplexF64))
+    @eval begin
+        function cscmm!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int32}, B::StridedMatrix{$T},
+                        β::$T, C::StridedMatrix{$T})
+            _check_transa(transa)
+            _check_mat_mult_matvec(C, A, B, transa)
+            mB, nB = size(B)
+            mC, nC = size(C)
+            __counter[] += 1
+            $fname(transa, A.m, nC, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, β, C, mC)
+            return C
+        end
+
+        function cscmm!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int64}, B::StridedMatrix{$T},
+                        β::$T, C::StridedMatrix{$T})
+            _check_transa(transa)
+            _check_mat_mult_matvec(C, A, B, transa)
+            mB, nB = size(B)
+            mC, nC = size(C)
+            __counter[] += 1
+            set_interface_layer(INTERFACE_ILP64)
+            $fname(transa, A.m, nC, A.n, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, β, C, mC)
+            set_interface_layer(INTERFACE_LP64)
+            return C
+        end
+    end
 end
 
-function cscsv!(transa::Char, α::T, matdescra::String,
-                A::SparseMatrixCSC{T, BlasInt}, x::StridedVector{T},
-                y::StridedVector{T}) where {T <: BlasFloat}
-    n = checksquare(A)
-    _check_transa(transa)
-    _check_mat_mult_matvec(y, A, x, transa)
-    __counter[] += 1
-    T == Float32    && (mkl_scscsv(transa, A.m, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, y))
-    T == Float64    && (mkl_dcscsv(transa, A.m, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, y))
-    T == ComplexF32 && (mkl_ccscsv(transa, A.m, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, y))
-    T == ComplexF64 && (mkl_zcscsv(transa, A.m, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, y))
-    return y
+for (fname, T) in ((:mkl_scscsv, :Float32   ),
+                   (:mkl_dcscsv, :Float64   ),
+                   (:mkl_ccscsv, :ComplexF32),
+                   (:mkl_zcscsv, :ComplexF64))
+    @eval begin
+        function cscsv!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int32}, x::StridedVector{$T},
+                        y::StridedVector{$T})
+            n = checksquare(A)
+            _check_transa(transa)
+            _check_mat_mult_matvec(y, A, x, transa)
+            __counter[] += 1
+            $fname(transa, A.m, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, y)
+            return y
+        end
+
+        function cscsv!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int64}, x::StridedVector{$T},
+                        y::StridedVector{$T})
+            n = checksquare(A)
+            _check_transa(transa)
+            _check_mat_mult_matvec(y, A, x, transa)
+            __counter[] += 1
+            set_interface_layer(INTERFACE_ILP64)
+            $fname(transa, A.m, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), x, y)
+            set_interface_layer(INTERFACE_LP64)
+            return y
+        end
+    end
 end
 
-function cscsm!(transa::Char, α::T, matdescra::String,
-                A::SparseMatrixCSC{T, BlasInt}, B::StridedMatrix{T},
-                C::StridedMatrix{T}) where {T <: BlasFloat}
-    mB, nB = size(B)
-    mC, nC = size(C)
-    n = checksquare(A)
-    _check_transa(transa)
-    _check_mat_mult_matvec(C, A, B, transa)
-    __counter[] += 1
-    T == Float32    && (mkl_scscsm(transa, A.n, nC, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, C, mC))
-    T == Float64    && (mkl_dcscsm(transa, A.n, nC, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, C, mC))
-    T == ComplexF32 && (mkl_ccscsm(transa, A.n, nC, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, C, mC))
-    T == ComplexF64 && (mkl_zcscsm(transa, A.n, nC, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, C, mC))
-    return C
+for (fname, T) in ((:mkl_scscsm, :Float32   ),
+                   (:mkl_dcscsm, :Float64   ),
+                   (:mkl_ccscsm, :ComplexF32),
+                   (:mkl_zcscsm, :ComplexF64))
+    @eval begin
+        function cscsm!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int32}, B::StridedMatrix{$T},
+                        C::StridedMatrix{$T})
+            mB, nB = size(B)
+            mC, nC = size(C)
+            n = checksquare(A)
+            _check_transa(transa)
+            _check_mat_mult_matvec(C, A, B, transa)
+            __counter[] += 1
+            $fname(transa, A.n, nC, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, C, mC)
+            return C
+        end
+
+        function cscsm!(transa::Char, α::$T, matdescra::String,
+                        A::SparseMatrixCSC{$T, Int64}, B::StridedMatrix{$T},
+                        C::StridedMatrix{$T})
+            mB, nB = size(B)
+            mC, nC = size(C)
+            n = checksquare(A)
+            _check_transa(transa)
+            _check_mat_mult_matvec(C, A, B, transa)
+            __counter[] += 1
+            set_interface_layer(INTERFACE_ILP64)
+            $fname(transa, A.n, nC, α, matdescra, A.nzval, A.rowval, A.colptr, pointer(A.colptr, 2), B, mB, C, mC)
+            set_interface_layer(INTERFACE_LP64)
+            return C
+        end
+    end
 end
diff --git a/src/matdescra.jl b/src/matdescra.jl
index 247275a..6dc25ad 100644
--- a/src/matdescra.jl
+++ b/src/matdescra.jl
@@ -5,4 +5,4 @@ matdescra(A::UnitLowerTriangular) = "TLUF"
 matdescra(A::UnitUpperTriangular) = "TUUF"
 matdescra(A::Symmetric) = string('S', A.uplo, 'N', 'F')
 matdescra(A::Hermitian) = string('H', A.uplo, 'N', 'F')
-matdescra(A::SparseMatrixCSC) = "GUUF"
+matdescra(A::SparseMatrixCSC) = "GFNF"
diff --git a/src/matmul.jl b/src/matmul.jl
index 6baa8a3..ddb2e50 100644
--- a/src/matmul.jl
+++ b/src/matmul.jl
@@ -6,23 +6,18 @@ _get_data(A::UpperTriangular) = triu(A.data)
 _get_data(A::UnitLowerTriangular) = tril(A.data)
 _get_data(A::UnitUpperTriangular) = triu(A.data)
 _get_data(A::Symmetric) = A.data
+_get_data(A::Hermitian) = A.data
 
 _unwrap_adj(x::Union{Adjoint,Transpose}) = parent(x)
 _unwrap_adj(x) = x
 
-const SparseMatrices{T} = Union{SparseMatrixCSC{T,BlasInt},
-                        Symmetric{T,SparseMatrixCSC{T,BlasInt}},
-                        LowerTriangular{T, SparseMatrixCSC{T,BlasInt}},
-                        UnitLowerTriangular{T, SparseMatrixCSC{T,BlasInt}},
-                        UpperTriangular{T, SparseMatrixCSC{T,BlasInt}},
-                        UnitUpperTriangular{T, SparseMatrixCSC{T,BlasInt}}}
-
-for T in [Complex{Float32}, Complex{Float64}, Float32, Float64]
+for T in [Float32, Float64, Complex{Float32}, Complex{Float64}]
+for INT in [Int32, Int64]
 for mat in (:StridedVector, :StridedMatrix)
 for (tchar, ttype) in (('N', :()),
                        ('C', :Adjoint),
                        ('T', :Transpose))
-    AT = tchar == 'N' ? :(SparseMatrixCSC{$T,BlasInt}) : :($ttype{$T,SparseMatrixCSC{$T,BlasInt}})
+    AT = tchar == 'N' ? :(SparseMatrixCSC{$T,$INT}) : :($ttype{$T,SparseMatrixCSC{$T,$INT}})
     @eval begin
         function mul!(C::$mat{$T}, adjA::$AT, B::$mat{$T}, α::Number, β::Number)
             A = _unwrap_adj(adjA)
@@ -45,10 +40,10 @@ for (tchar, ttype) in (('N', :()),
         end
     end
 
-    for w in (:Symmetric, :LowerTriangular, :UnitLowerTriangular, :UpperTriangular, :UnitUpperTriangular)
+    for w in (:Symmetric, :Hermitian, :LowerTriangular, :UnitLowerTriangular, :UpperTriangular, :UnitUpperTriangular)
         AT = tchar == 'N' ?
-            :($w{$T,SparseMatrixCSC{$T,BlasInt}}) :
-            :($ttype{$T,$w{$T,SparseMatrixCSC{$T,BlasInt}}})
+            :($w{$T,SparseMatrixCSC{$T,$INT}}) :
+            :($ttype{$T,$w{$T,SparseMatrixCSC{$T,$INT}}})
         @eval begin
             function mul!(C::$mat{$T}, adjA::$AT, B::$mat{$T}, α::Number, β::Number)
                 A = _unwrap_adj(adjA)
@@ -71,7 +66,7 @@ for (tchar, ttype) in (('N', :()),
             end
         end
 
-        if w != :Symmetric
+        if w != :Symmetric && w != :Hermitian
             @eval begin
                 function ldiv!(α::Number, adjA::$AT,
                                B::$mat{$T}, C::$mat{$T})
@@ -98,4 +93,5 @@ for (tchar, ttype) in (('N', :()),
     end
 end
 end # mat
+end # INT
 end # T
diff --git a/test/test_BLAS.jl b/test/test_BLAS.jl
index c569eee..b6a64aa 100644
--- a/test/test_BLAS.jl
+++ b/test/test_BLAS.jl
@@ -2,94 +2,126 @@ using MKLSparse
 using Test, SparseArrays, LinearAlgebra
 
 sA = sprand(5, 5, 0.01)
-sS = sA'sA
+sS = sA' * sA
 sTl = tril(sS)
 sTu = triu(sS)
 
 @test MKLSparse.matdescra(Symmetric(sTl,:L)) == "SLNF"
 @test MKLSparse.matdescra(Symmetric(sTu,:U)) == "SUNF"
+@test MKLSparse.matdescra(Hermitian(sTl,:L)) == "HLNF"
+@test MKLSparse.matdescra(Hermitian(sTu,:U)) == "HUNF"
 @test MKLSparse.matdescra(LowerTriangular(sTl)) == "TLNF"
 @test MKLSparse.matdescra(UpperTriangular(sTu)) == "TUNF"
 @test MKLSparse.matdescra(UnitLowerTriangular(sTl)) == "TLUF"
 @test MKLSparse.matdescra(UnitUpperTriangular(sTu)) == "TUUF"
-@test MKLSparse.matdescra(sA) == "GUUF"
+@test MKLSparse.matdescra(sA) == "GFNF"
 
-macro test_blas(ex)
+#=macro test(ex)
     return quote
         MKLSparse.__counter[] = 0
         @test $(esc(ex))
         @test MKLSparse.__counter[] == 1
     end
-end
+end=#
 
 @testset "matrix-vector multiplication (non-square)" begin
-    for i = 1:5
-        a = sprand(10, 5, 0.5)
-        b = rand(5)
-        @test_blas maximum(abs.(a*b - Array(a)*b)) < 100*eps()
-        b = rand(5, 5)
-        @test_blas maximum(abs.(a*b - Array(a)*b)) < 100*eps()
-        b = rand(10)
-        @test_blas maximum(abs.(a'*b - Array(a)'*b)) < 100*eps()
-        @test_blas maximum(abs.(transpose(a)*b - Array(a)'*b)) < 100*eps()
-        b = rand(10,10)
-        @test_blas maximum(abs.(a'*b - Array(a)'*b)) < 100*eps()
-        @test_blas maximum(abs.(transpose(a)*b - Array(a)'*b)) < 100*eps()
+    T = Float64
+    @testset "interface = $interface" for interface in ("LP64", "ILP64")
+        S = interface == "LP64" ? Int32 : Int64
+        for i = 1:5
+            a = sprand(T, 10, 5, 0.5)
+            a = SparseMatrixCSC{T, S}(a)
+            b = rand(T, 5)
+            @test maximum(abs.(a*b - Array(a)*b)) < 100*eps(T)
+            b = rand(T, 5, 5)
+            @test maximum(abs.(a*b - Array(a)*b)) < 100*eps(T)
+            b = rand(T, 10)
+            @test maximum(abs.(a'*b - Array(a)'*b)) < 100*eps(T)
+            @test maximum(abs.(transpose(a)*b - Array(a)'*b)) < 100*eps(T)
+            b = rand(T, 10, 10)
+            @test maximum(abs.(a'*b - Array(a)'*b)) < 100*eps()
+            @test maximum(abs.(transpose(a)*b - Array(a)'*b)) < 100*eps(T)
+        end
     end
 end
 
 #?
 @testset "complex matrix-vector multiplication" begin
-    for i = 1:5
-        a = I + im * 0.1*sprandn(5, 5, 0.2)
-        b = randn(5,3) + im*randn(5,3)
-        c = randn(5) + im*randn(5)
-        d = randn(5) + im*randn(5)
-        α = rand(ComplexF64)
-        β = rand(ComplexF64)
-        @test_blas (maximum(abs.(a*b - Array(a)*b)) < 100*eps())
-        @test_blas (maximum(abs.(a'*b - Array(a)'*b)) < 100*eps())
-        @test_blas (maximum(abs.(transpose(a)*b - transpose(Array(a))*b)) < 100*eps())
-        @test_blas (maximum(abs.(mul!(similar(b), a, b) - Array(a)*b)) < 100*eps())
-        @test_blas (maximum(abs.(mul!(similar(c), a, c) - Array(a)*c)) < 100*eps())
-        @test_blas (maximum(abs.(mul!(similar(b), transpose(a), b) - transpose(Array(a))*b)) < 100*eps())
-        @test_blas (maximum(abs.(mul!(similar(c), transpose(a), c) - transpose(Array(a))*c)) < 100*eps())
-        @test_blas (maximum(abs.(mul!(copy(b), a, b, α, β) - (α*(Array(a)*b) + β*b))) < 100*eps())
-        @test_blas (maximum(abs.(mul!(copy(b), transpose(a), b, α, β) - (α*(transpose(Array(a))*b) + β*b))) < 100*eps())
-        @test_blas (maximum(abs.(mul!(copy(c), transpose(a), c, α, β) - (α*(transpose(Array(a))*c) + β*c))) < 100*eps())
-        α = β = 1 # test conversion to float
-        @test_blas (maximum(abs.(mul!(copy(b), a, b, α, β) - (α*(Array(a)*b) + β*b))) < 100*eps())
-        @test_blas (maximum(abs.(mul!(copy(b), transpose(a), b, α, β) - (α*(transpose(Array(a))*b) + β*b))) < 100*eps())
-        @test_blas (maximum(abs.(mul!(copy(c), transpose(a), c, α, β) - (α*(transpose(Array(a))*c) + β*c))) < 100*eps())
+    T = ComplexF64
+    R = Float64
+    @testset "interface = $interface" for interface in ("LP64", "ILP64")
+        S = interface == "LP64" ? Int32 : Int64
+        for i = 1:5
+            a = I + im * 0.1 * sprandn(T, 5, 5, 0.2)
+            a = SparseMatrixCSC{T,S}(a)
+            b = randn(T, 5, 3) + im * randn(T, 5, 3)
+            c = randn(T, 5) + im * randn(T, 5)
+            d = randn(T, 5) + im * randn(T, 5)
+            α = rand(T)
+            β = rand(T)
+            @test (maximum(abs.(a*b - Array(a)*b)) < 100*eps(R))
+            @test (maximum(abs.(a'*b - Array(a)'*b)) < 100*eps(R))
+            @test (maximum(abs.(transpose(a)*b - transpose(Array(a))*b)) < 100*eps(R))
+            @test (maximum(abs.(mul!(similar(b), a, b) - Array(a)*b)) < 100*eps(R))
+            @test (maximum(abs.(mul!(similar(c), a, c) - Array(a)*c)) < 100*eps(R))
+            @test (maximum(abs.(mul!(similar(b), transpose(a), b) - transpose(Array(a))*b)) < 100*eps(R))
+            @test (maximum(abs.(mul!(similar(c), transpose(a), c) - transpose(Array(a))*c)) < 100*eps(R))
+            @test (maximum(abs.(mul!(copy(b), a, b, α, β) - (α*(Array(a)*b) + β*b))) < 100*eps(R))
+            @test (maximum(abs.(mul!(copy(b), transpose(a), b, α, β) - (α*(transpose(Array(a))*b) + β*b))) < 100*eps(R))
+            @test (maximum(abs.(mul!(copy(c), transpose(a), c, α, β) - (α*(transpose(Array(a))*c) + β*c))) < 100*eps(R))
+            α = β = 1 # test conversion to float
+            @test (maximum(abs.(mul!(copy(b), a, b, α, β) - (α*(Array(a)*b) + β*b))) < 100*eps(R))
+            @test (maximum(abs.(mul!(copy(b), transpose(a), b, α, β) - (α*(transpose(Array(a))*b) + β*b))) < 100*eps(R))
+            @test (maximum(abs.(mul!(copy(c), transpose(a), c, α, β) - (α*(transpose(Array(a))*c) + β*c))) < 100*eps(R))
 
-        c = randn(6) + im*randn(6)
-        @test_throws DimensionMismatch transpose(a)*c
-        @test_throws DimensionMismatch a.*c
-        @test_throws DimensionMismatch a.*c
+            c = randn(T, 6) + im * randn(T, 6)
+            @test_throws DimensionMismatch transpose(a)*c
+            @test_throws DimensionMismatch a.*c
+            @test_throws DimensionMismatch a.*c
+        end
     end
 end
 
 @testset "triangular" begin
-    n = 100
-    A = sprandn(n, n, 0.5) + sqrt(n)*I
-    b = rand(n)
-    symA = A + transpose(A)
-    trilA = tril(A)
-    triuA = triu(A)
-    trilUA = tril(A, -1) + I
-    triuUA = triu(A, 1)  + I
+    @testset "T = $T" for T in (Float32, Float64, ComplexF32, ComplexF64)
+        @testset "interface = $interface" for interface in ("LP64", "ILP64")
+            S = interface == "LP64" ? Int32 : Int64
+            n = 100
+            A = sprandn(T, n, n, 0.5) + sqrt(n)*I
+            A = SparseMatrixCSC{T,S}(A)
+            b = rand(T, n)
+            trilA = tril(A)
+            triuA = triu(A)
+            trilUA = tril(A, -1) + I
+            triuUA = triu(A, 1)  + I
 
-    @test_blas LowerTriangular(trilA) \ b ≈ Array(LowerTriangular(trilA)) \ b
-    @test_blas LowerTriangular(trilA) * b ≈ Array(LowerTriangular(trilA)) * b
+            @test LowerTriangular(trilA) \ b ≈ Array(LowerTriangular(trilA)) \ b
+            @test LowerTriangular(trilA) * b ≈ Array(LowerTriangular(trilA)) * b
 
-    @test_blas UpperTriangular(triuA) \ b ≈ Array(UpperTriangular(triuA)) \ b
-    @test_blas UpperTriangular(triuA) * b ≈ Array(UpperTriangular(triuA)) * b
+            @test UpperTriangular(triuA) \ b ≈ Array(UpperTriangular(triuA)) \ b
+            @test UpperTriangular(triuA) * b ≈ Array(UpperTriangular(triuA)) * b
 
-    @test_blas UnitLowerTriangular(trilUA) \ b ≈ Array(UnitLowerTriangular(trilUA)) \ b
-    @test_blas UnitLowerTriangular(trilUA) * b ≈ Array(UnitLowerTriangular(trilUA)) * b
+            @test UnitLowerTriangular(trilUA) \ b ≈ Array(UnitLowerTriangular(trilUA)) \ b
+            @test UnitLowerTriangular(trilUA) * b ≈ Array(UnitLowerTriangular(trilUA)) * b
 
-    @test_blas UnitUpperTriangular(triuUA) \ b ≈ Array(UnitUpperTriangular(triuUA)) \ b
-    @test_blas UnitUpperTriangular(triuUA) * b ≈ Array(UnitUpperTriangular(triuUA)) * b
+            @test UnitUpperTriangular(triuUA) \ b ≈ Array(UnitUpperTriangular(triuUA)) \ b
+            @test UnitUpperTriangular(triuUA) * b ≈ Array(UnitUpperTriangular(triuUA)) * b
+        end
+    end
+end
 
-    @test_blas Symmetric(symA) * b ≈ Array(Symmetric(symA)) * b
+@testset "Symmetric -- Hermitian" begin
+    @testset "T = $T" for T in (Float32, Float64, ComplexF32, ComplexF64)
+        @testset "interface = $interface" for interface in ("LP64", "ILP64")
+            S = interface == "LP64" ? Int32 : Int64
+            n = 100
+            A = sprandn(T, n, n, 0.5) + sqrt(n)*I
+            b = rand(T, n)
+            symA = A + transpose(A)
+            hermA = A + adjoint(A)
+
+            @test Symmetric(symA) * b ≈ Array(Symmetric(symA)) * b
+            @test Hermitian(hermA) * b ≈ Array(Hermitian(hermA)) * b
+        end
+    end
 end