# Copyright (c) 2021-2022 Huawei Device Co., Ltd.
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

.record Math <external>
.function f64 Math.sqrt(f64 a0) <external>
.function f64 Math.absF64(f64 a0) <external>

.function u1 main(){
    movi v0, 6
    movi v1, 480
    fmovi.64 v2, 8.909361304388094
    call test, v0, v1, v2, v0
    return
}

.function u1 test(i32 a0, i32 a1, f64 a2){
    fmovi.64 v3, 0.0              #total
    mov v4, a0          #loop counter
    lda a0
loop:
    jgt a1, loop_exit
    call.short spectralnorm, v4, v4
    fadd2.64 v3
    sta.64 v3
    lda v4
    muli 2
    sta v4
    jmp loop
loop_exit:
    lda.64 v3
    fsub2.64 a2
    sta.64 v5
    call.short Math.absF64, v5, v5
    fldai.64 1e-13
    fcmpl.64 v5
    jltz exit_failure
    ldai 0
    return
exit_failure:
    ldai 1
    return
}

.function f64 spectralnorm(i32 a0){
    mov v0, a0
    newarr v1, v0, f64[]              #u
    newarr v2, v0, f64[]              #w
    newarr v3, v0, f64[]              #v
    fmovi.64 v4, 0.0       #vBv
    fmovi.64 v5, 0.0       #vv
    movi.64 v6, 0          #loop_counter
    movi v20, 10
    ldai 0
loop:
    jeq a0, loop_exit
    fldai.64 1.0
    fstarr.64 v1, v6
    fldai.64 0.0
    fstarr.64 v2, v6
    fldai.64 0.0
    fstarr.64 v3, v6
    inci v6, 1
    lda v6
    jmp loop
loop_exit:
    movi v6, 0
    ldai 0
loop2:
    jeq v20, loop2_exit
    call AtAu, v1, v3, v2, v1
    call AtAu, v3, v1, v2, v1
    inci v6, 1
    lda v6
    jmp loop2
loop2_exit:
    movi v6, 0
    ldai 0
loop3:
    jeq a0, loop3_exit
    lda v6
    fldarr.64 v1
    sta.64 v7
    lda.64 v6
    fldarr.64 v3
    fmul2.64 v7
    fadd2.64 v4
    sta.64 v4
    lda v6
    fldarr.64 v3
    sta.64 v7
    fmul2.64 v7
    fadd2.64 v5
    sta.64 v5
    inci v6, 1
    lda v6
    jmp loop3
loop3_exit:
    lda.64 v4
    fdiv2.64 v5
    sta.64 v9
    call.short Math.sqrt, v9, v9
    return.64
}

.function void AtAu(f64[] a0, f64[] a1, f64[] a2){
    call.short Au, a0, a2
    call.short Atu, a2, a1
    return.void
}

.function void Atu(f64[] a0, f64[] a1){
    movi v2, 0                      #loop_counter
    lenarr a0
    sta v0
    lenarr a1
    sta v1
    ldai 0
loop:
    jeq v1, loop_exit
    fmovi.64 v3, 0.0                  #t
    movi v4, 0                      #loop2_counter
    ldai 0
loop2:
    jeq v0, loop2_exit
    call.short A, v4, v2
    sta.64 v5
    lda v4
    fldarr.64 a0
    fmul2.64 v5
    fadd2.64 v3
    sta.64 v3
    inci v4, 1
    lda v4
    jmp loop2
loop2_exit:
    lda.64 v3
    fstarr.64 a1, v2
    inci v2, 1
    lda v2
    jmp loop
loop_exit:
    return.void
}

.function void Au(f64[] a0, f64[] a1){
    movi v2, 0                      #loop_counter
    lenarr a0
    sta v0
    lenarr a1
    sta v1
    ldai 0
loop:
    jeq v1, loop_exit
    fmovi.64 v3, 0.0                  #t
    movi v4, 0                      #loop2_counter
    ldai 0
loop2:
    jeq v0, loop2_exit
    call.short A, v2, v4
    sta.64 v5
    lda v4
    fldarr.64 a0
    fmul2.64 v5
    fadd2.64 v3
    sta.64 v3
    inci v4, 1
    lda v4
    jmp loop2
loop2_exit:
    lda.64 v3
    fstarr.64 a1, v2
    inci v2, 1
    lda v2
    jmp loop
loop_exit:
    return.void
}

.function f64 A(i32 a0, i32 a1){
    fmovi.64 v10, 1.0
    fmovi.64 v11, 2.0
    lda a0
    i32tof64
    sta.64 a0
    lda a1
    i32tof64
    sta.64 a1
    fadd2.64 a0
    fadd2.64 v10
    sta.64 v2
    lda.64 a0
    fadd2.64 a1
    fmul2.64 v2
    fdiv2.64 v11
    fadd2.64 a0
    fadd2.64 v10
    sta.64 v2
    lda.64 v10
    fdiv2.64 v2
    return.64
}