# stack32 out_stack

# int32 out

# stack32 u_stack

# int32 u

# int32 eax

# int32 ebx

# int32 esi

# int32 edi

# int32 ebp

# stack32 ebp_stack

# float80 d0

# float80 d1

# float80 d2

# float80 d3

# float80 d4

# float80 e0

# float80 e1

# float80 e2

# float80 e3

# float80 e4

# float80 f0

# float80 f1

# float80 f2

# float80 f3

# float80 f4

# float80 g0

# float80 g1

# float80 g2

# float80 g3

# float80 g4

# float80 h0

# float80 h1

# float80 h2

# float80 h3

# float80 h4

# float80 i0

# float80 i1

# float80 i2

# float80 i3

# float80 i4

# float80 j0

# float80 j1

# float80 j2

# float80 j3

# float80 j4

# float80 k0

# float80 k1

# float80 k2

# float80 k3

# float80 k4

# enter gaudry_pm_hadamard
.text
.p2align 5
.globl _gaudry_pm_hadamard
.globl gaudry_pm_hadamard
_gaudry_pm_hadamard:
gaudry_pm_hadamard:
mov %esp,%eax
and $31,%eax
add $0,%eax
sub %eax,%esp

# input out_stack

# input u_stack

# caller eax

# caller ebx

# caller esi

# caller edi

# caller ebp

#   u = u_stack
# movl <u_stack=stack32#-2,>u=int32#3
# movl <u_stack=8(%esp,%eax),>u=%edx
movl 8(%esp,%eax),%edx

#   out = out_stack
# movl <out_stack=stack32#-1,>out=int32#2
# movl <out_stack=4(%esp,%eax),>out=%ecx
movl 4(%esp,%eax),%ecx

#   f0 = *(float64 *) (u + 0)
# fldl 0(<u=int32#3)
# fldl 0(<u=%edx)
fldl 0(%edx)

#   d0 = *(float64 *) (u + 40)
# fldl 40(<u=int32#3)
# fldl 40(<u=%edx)
fldl 40(%edx)

#   g0 = *(float64 *) (u + 80)
# fldl 80(<u=int32#3)
# fldl 80(<u=%edx)
fldl 80(%edx)

#   e0 = *(float64 *) (u + 120)
# fldl 120(<u=int32#3)
# fldl 120(<u=%edx)
fldl 120(%edx)

# stacktop f0
# fxch <f0=float80#4
# fxch <f0=%st(3)
fxch %st(3)

#   d0 += f0
# fadd <f0=float80#1,<d0=float80#3
# fadd <f0=%st(0),<d0=%st(2)
fadd %st(0),%st(2)

# stacktop g0
# fxch <g0=float80#2
# fxch <g0=%st(1)
fxch %st(1)

#   e0 += g0
# fadd <g0=float80#1,<e0=float80#4
# fadd <g0=%st(0),<e0=%st(3)
fadd %st(0),%st(3)

# stacktop f0
# fxch <f0=float80#2
# fxch <f0=%st(1)
fxch %st(1)

#   f0 -= *(float64 *) (u + 40)
# fsubl 40(<u=int32#3)
# fsubl 40(<u=%edx)
fsubl 40(%edx)

# stacktop g0
# fxch <g0=float80#2
# fxch <g0=%st(1)
fxch %st(1)

#   g0 -= *(float64 *) (u + 120)
# fsubl 120(<u=int32#3)
# fsubl 120(<u=%edx)
fsubl 120(%edx)

#     f1 = *(float64 *) (u + 8)
# fldl 8(<u=int32#3)
# fldl 8(<u=%edx)
fldl 8(%edx)

#     d1 = *(float64 *) (u + 48)
# fldl 48(<u=int32#3)
# fldl 48(<u=%edx)
fldl 48(%edx)

#     g1 = *(float64 *) (u + 88)
# fldl 88(<u=int32#3)
# fldl 88(<u=%edx)
fldl 88(%edx)

#     e1 = *(float64 *) (u + 128)
# fldl 128(<u=int32#3)
# fldl 128(<u=%edx)
fldl 128(%edx)

# stacktop f1
# fxch <f1=float80#4
# fxch <f1=%st(3)
fxch %st(3)

#     d1 += f1
# fadd <f1=float80#1,<d1=float80#3
# fadd <f1=%st(0),<d1=%st(2)
fadd %st(0),%st(2)

# stacktop g1
# fxch <g1=float80#2
# fxch <g1=%st(1)
fxch %st(1)

#     e1 += g1
# fadd <g1=float80#1,<e1=float80#4
# fadd <g1=%st(0),<e1=%st(3)
fadd %st(0),%st(3)

# stacktop f1
# fxch <f1=float80#2
# fxch <f1=%st(1)
fxch %st(1)

#     f1 -= *(float64 *) (u + 48)
# fsubl 48(<u=int32#3)
# fsubl 48(<u=%edx)
fsubl 48(%edx)

# stacktop g1
# fxch <g1=float80#2
# fxch <g1=%st(1)
fxch %st(1)

#     g1 -= *(float64 *) (u + 128)
# fsubl 128(<u=int32#3)
# fsubl 128(<u=%edx)
fsubl 128(%edx)

# stacktop d0
# fxch <d0=float80#7
# fxch <d0=%st(6)
fxch %st(6)

#   *(float64 *) (out + 0) = d0
# fstpl 0(<out=int32#2)
# fstpl 0(<out=%ecx)
fstpl 0(%ecx)

# stacktop e0
# fxch <e0=float80#7
# fxch <e0=%st(6)
fxch %st(6)

#   *(float64 *) (out + 40) = e0
# fstpl 40(<out=int32#2)
# fstpl 40(<out=%ecx)
fstpl 40(%ecx)

# stacktop f0
# fxch <f0=float80#4
# fxch <f0=%st(3)
fxch %st(3)

#   *(float64 *) (out + 80) = f0
# fstpl 80(<out=int32#2)
# fstpl 80(<out=%ecx)
fstpl 80(%ecx)

# stacktop g0
# fxch <g0=float80#2
# fxch <g0=%st(1)
fxch %st(1)

#   *(float64 *) (out + 120) = g0
# fstpl 120(<out=int32#2)
# fstpl 120(<out=%ecx)
fstpl 120(%ecx)

# stacktop d1
# fxch <d1=float80#2
# fxch <d1=%st(1)
fxch %st(1)

#     *(float64 *) (out + 8) = d1
# fstpl 8(<out=int32#2)
# fstpl 8(<out=%ecx)
fstpl 8(%ecx)

#     *(float64 *) (out + 48) = e1
# fstpl 48(<out=int32#2)
# fstpl 48(<out=%ecx)
fstpl 48(%ecx)

# stacktop f1
# fxch <f1=float80#2
# fxch <f1=%st(1)
fxch %st(1)

#     *(float64 *) (out + 88) = f1
# fstpl 88(<out=int32#2)
# fstpl 88(<out=%ecx)
fstpl 88(%ecx)

#     *(float64 *) (out + 128) = g1
# fstpl 128(<out=int32#2)
# fstpl 128(<out=%ecx)
fstpl 128(%ecx)

#       f2 = *(float64 *) (u + 16)
# fldl 16(<u=int32#3)
# fldl 16(<u=%edx)
fldl 16(%edx)

#       d2 = *(float64 *) (u + 56)
# fldl 56(<u=int32#3)
# fldl 56(<u=%edx)
fldl 56(%edx)

#       g2 = *(float64 *) (u + 96)
# fldl 96(<u=int32#3)
# fldl 96(<u=%edx)
fldl 96(%edx)

#       e2 = *(float64 *) (u + 136)
# fldl 136(<u=int32#3)
# fldl 136(<u=%edx)
fldl 136(%edx)

# stacktop f2
# fxch <f2=float80#4
# fxch <f2=%st(3)
fxch %st(3)

#       d2 += f2
# fadd <f2=float80#1,<d2=float80#3
# fadd <f2=%st(0),<d2=%st(2)
fadd %st(0),%st(2)

# stacktop g2
# fxch <g2=float80#2
# fxch <g2=%st(1)
fxch %st(1)

#       e2 += g2
# fadd <g2=float80#1,<e2=float80#4
# fadd <g2=%st(0),<e2=%st(3)
fadd %st(0),%st(3)

# stacktop f2
# fxch <f2=float80#2
# fxch <f2=%st(1)
fxch %st(1)

#       f2 -= *(float64 *) (u + 56)
# fsubl 56(<u=int32#3)
# fsubl 56(<u=%edx)
fsubl 56(%edx)

# stacktop g2
# fxch <g2=float80#2
# fxch <g2=%st(1)
fxch %st(1)

#       g2 -= *(float64 *) (u + 136)
# fsubl 136(<u=int32#3)
# fsubl 136(<u=%edx)
fsubl 136(%edx)

#         f3 = *(float64 *) (u + 24)
# fldl 24(<u=int32#3)
# fldl 24(<u=%edx)
fldl 24(%edx)

#         d3 = *(float64 *) (u + 64)
# fldl 64(<u=int32#3)
# fldl 64(<u=%edx)
fldl 64(%edx)

#         g3 = *(float64 *) (u + 104)
# fldl 104(<u=int32#3)
# fldl 104(<u=%edx)
fldl 104(%edx)

#         e3 = *(float64 *) (u + 144)
# fldl 144(<u=int32#3)
# fldl 144(<u=%edx)
fldl 144(%edx)

# stacktop f3
# fxch <f3=float80#4
# fxch <f3=%st(3)
fxch %st(3)

#         d3 += f3
# fadd <f3=float80#1,<d3=float80#3
# fadd <f3=%st(0),<d3=%st(2)
fadd %st(0),%st(2)

# stacktop g3
# fxch <g3=float80#2
# fxch <g3=%st(1)
fxch %st(1)

#         e3 += g3
# fadd <g3=float80#1,<e3=float80#4
# fadd <g3=%st(0),<e3=%st(3)
fadd %st(0),%st(3)

# stacktop f3
# fxch <f3=float80#2
# fxch <f3=%st(1)
fxch %st(1)

#         f3 -= *(float64 *) (u + 64)
# fsubl 64(<u=int32#3)
# fsubl 64(<u=%edx)
fsubl 64(%edx)

# stacktop g3
# fxch <g3=float80#2
# fxch <g3=%st(1)
fxch %st(1)

#         g3 -= *(float64 *) (u + 144)
# fsubl 144(<u=int32#3)
# fsubl 144(<u=%edx)
fsubl 144(%edx)

# stacktop d2
# fxch <d2=float80#7
# fxch <d2=%st(6)
fxch %st(6)

#       *(float64 *) (out + 16) = d2
# fstpl 16(<out=int32#2)
# fstpl 16(<out=%ecx)
fstpl 16(%ecx)

# stacktop e2
# fxch <e2=float80#7
# fxch <e2=%st(6)
fxch %st(6)

#       *(float64 *) (out + 56) = e2
# fstpl 56(<out=int32#2)
# fstpl 56(<out=%ecx)
fstpl 56(%ecx)

# stacktop f2
# fxch <f2=float80#4
# fxch <f2=%st(3)
fxch %st(3)

#       *(float64 *) (out + 96) = f2
# fstpl 96(<out=int32#2)
# fstpl 96(<out=%ecx)
fstpl 96(%ecx)

# stacktop g2
# fxch <g2=float80#2
# fxch <g2=%st(1)
fxch %st(1)

#       *(float64 *) (out + 136) = g2
# fstpl 136(<out=int32#2)
# fstpl 136(<out=%ecx)
fstpl 136(%ecx)

# stacktop d3
# fxch <d3=float80#2
# fxch <d3=%st(1)
fxch %st(1)

#         *(float64 *) (out + 24) = d3
# fstpl 24(<out=int32#2)
# fstpl 24(<out=%ecx)
fstpl 24(%ecx)

#         *(float64 *) (out + 64) = e3
# fstpl 64(<out=int32#2)
# fstpl 64(<out=%ecx)
fstpl 64(%ecx)

# stacktop f3
# fxch <f3=float80#2
# fxch <f3=%st(1)
fxch %st(1)

#         *(float64 *) (out + 104) = f3
# fstpl 104(<out=int32#2)
# fstpl 104(<out=%ecx)
fstpl 104(%ecx)

#         *(float64 *) (out + 144) = g3
# fstpl 144(<out=int32#2)
# fstpl 144(<out=%ecx)
fstpl 144(%ecx)

#           f4 = *(float64 *) (u + 32)
# fldl 32(<u=int32#3)
# fldl 32(<u=%edx)
fldl 32(%edx)

#           d4 = *(float64 *) (u + 72)
# fldl 72(<u=int32#3)
# fldl 72(<u=%edx)
fldl 72(%edx)

#           g4 = *(float64 *) (u + 112)
# fldl 112(<u=int32#3)
# fldl 112(<u=%edx)
fldl 112(%edx)

#           e4 = *(float64 *) (u + 152)
# fldl 152(<u=int32#3)
# fldl 152(<u=%edx)
fldl 152(%edx)

# stacktop f4
# fxch <f4=float80#4
# fxch <f4=%st(3)
fxch %st(3)

#           d4 += f4
# fadd <f4=float80#1,<d4=float80#3
# fadd <f4=%st(0),<d4=%st(2)
fadd %st(0),%st(2)

# stacktop g4
# fxch <g4=float80#2
# fxch <g4=%st(1)
fxch %st(1)

#           e4 += g4
# fadd <g4=float80#1,<e4=float80#4
# fadd <g4=%st(0),<e4=%st(3)
fadd %st(0),%st(3)

# stacktop f4
# fxch <f4=float80#2
# fxch <f4=%st(1)
fxch %st(1)

#           f4 -= *(float64 *) (u + 72)
# fsubl 72(<u=int32#3)
# fsubl 72(<u=%edx)
fsubl 72(%edx)

# stacktop g4
# fxch <g4=float80#2
# fxch <g4=%st(1)
fxch %st(1)

#           g4 -= *(float64 *) (u + 152)
# fsubl 152(<u=int32#3)
# fsubl 152(<u=%edx)
fsubl 152(%edx)

#   j0 = *(float64 *) (out + 0)
# fldl 0(<out=int32#2)
# fldl 0(<out=%ecx)
fldl 0(%ecx)

#   h0 = *(float64 *) (out + 40)
# fldl 40(<out=int32#2)
# fldl 40(<out=%ecx)
fldl 40(%ecx)

#   k0 = *(float64 *) (out + 80)
# fldl 80(<out=int32#2)
# fldl 80(<out=%ecx)
fldl 80(%ecx)

#   i0 = *(float64 *) (out + 120)
# fldl 120(<out=int32#2)
# fldl 120(<out=%ecx)
fldl 120(%ecx)

# stacktop j0
# fxch <j0=float80#4
# fxch <j0=%st(3)
fxch %st(3)

#   h0 += j0
# fadd <j0=float80#1,<h0=float80#3
# fadd <j0=%st(0),<h0=%st(2)
fadd %st(0),%st(2)

# stacktop k0
# fxch <k0=float80#2
# fxch <k0=%st(1)
fxch %st(1)

#   i0 += k0
# fadd <k0=float80#1,<i0=float80#4
# fadd <k0=%st(0),<i0=%st(3)
fadd %st(0),%st(3)

# stacktop j0
# fxch <j0=float80#2
# fxch <j0=%st(1)
fxch %st(1)

#   j0 -= *(float64 *) (out + 40)
# fsubl 40(<out=int32#2)
# fsubl 40(<out=%ecx)
fsubl 40(%ecx)

# stacktop k0
# fxch <k0=float80#2
# fxch <k0=%st(1)
fxch %st(1)

#   k0 -= *(float64 *) (out + 120)
# fsubl 120(<out=int32#2)
# fsubl 120(<out=%ecx)
fsubl 120(%ecx)

# stacktop d4
# fxch <d4=float80#7
# fxch <d4=%st(6)
fxch %st(6)

#           *(float64 *) (out + 32) = d4
# fstpl 32(<out=int32#2)
# fstpl 32(<out=%ecx)
fstpl 32(%ecx)

# stacktop e4
# fxch <e4=float80#7
# fxch <e4=%st(6)
fxch %st(6)

#           *(float64 *) (out + 72) = e4
# fstpl 72(<out=int32#2)
# fstpl 72(<out=%ecx)
fstpl 72(%ecx)

# stacktop f4
# fxch <f4=float80#4
# fxch <f4=%st(3)
fxch %st(3)

#           *(float64 *) (out + 112) = f4
# fstpl 112(<out=int32#2)
# fstpl 112(<out=%ecx)
fstpl 112(%ecx)

# stacktop g4
# fxch <g4=float80#2
# fxch <g4=%st(1)
fxch %st(1)

#           *(float64 *) (out + 152) = g4
# fstpl 152(<out=int32#2)
# fstpl 152(<out=%ecx)
fstpl 152(%ecx)

# stacktop h0
# fxch <h0=float80#2
# fxch <h0=%st(1)
fxch %st(1)

#   *(float64 *) (out + 0) = h0
# fstpl 0(<out=int32#2)
# fstpl 0(<out=%ecx)
fstpl 0(%ecx)

# stacktop j0
# fxch <j0=float80#3
# fxch <j0=%st(2)
fxch %st(2)

#   *(float64 *) (out + 40) = j0
# fstpl 40(<out=int32#2)
# fstpl 40(<out=%ecx)
fstpl 40(%ecx)

# stacktop i0
# fxch <i0=float80#2
# fxch <i0=%st(1)
fxch %st(1)

#   *(float64 *) (out + 80) = i0
# fstpl 80(<out=int32#2)
# fstpl 80(<out=%ecx)
fstpl 80(%ecx)

#   *(float64 *) (out + 120) = k0
# fstpl 120(<out=int32#2)
# fstpl 120(<out=%ecx)
fstpl 120(%ecx)

#     j1 = *(float64 *) (out + 8)
# fldl 8(<out=int32#2)
# fldl 8(<out=%ecx)
fldl 8(%ecx)

#     h1 = *(float64 *) (out + 48)
# fldl 48(<out=int32#2)
# fldl 48(<out=%ecx)
fldl 48(%ecx)

#     k1 = *(float64 *) (out + 88)
# fldl 88(<out=int32#2)
# fldl 88(<out=%ecx)
fldl 88(%ecx)

#     i1 = *(float64 *) (out + 128)
# fldl 128(<out=int32#2)
# fldl 128(<out=%ecx)
fldl 128(%ecx)

# stacktop j1
# fxch <j1=float80#4
# fxch <j1=%st(3)
fxch %st(3)

#     h1 += j1
# fadd <j1=float80#1,<h1=float80#3
# fadd <j1=%st(0),<h1=%st(2)
fadd %st(0),%st(2)

# stacktop k1
# fxch <k1=float80#2
# fxch <k1=%st(1)
fxch %st(1)

#     i1 += k1
# fadd <k1=float80#1,<i1=float80#4
# fadd <k1=%st(0),<i1=%st(3)
fadd %st(0),%st(3)

# stacktop j1
# fxch <j1=float80#2
# fxch <j1=%st(1)
fxch %st(1)

#     j1 -= *(float64 *) (out + 48)
# fsubl 48(<out=int32#2)
# fsubl 48(<out=%ecx)
fsubl 48(%ecx)

# stacktop k1
# fxch <k1=float80#2
# fxch <k1=%st(1)
fxch %st(1)

#     k1 -= *(float64 *) (out + 128)
# fsubl 128(<out=int32#2)
# fsubl 128(<out=%ecx)
fsubl 128(%ecx)

#       j2 = *(float64 *) (out + 16)
# fldl 16(<out=int32#2)
# fldl 16(<out=%ecx)
fldl 16(%ecx)

#       h2 = *(float64 *) (out + 56)
# fldl 56(<out=int32#2)
# fldl 56(<out=%ecx)
fldl 56(%ecx)

#       k2 = *(float64 *) (out + 96)
# fldl 96(<out=int32#2)
# fldl 96(<out=%ecx)
fldl 96(%ecx)

#       i2 = *(float64 *) (out + 136)
# fldl 136(<out=int32#2)
# fldl 136(<out=%ecx)
fldl 136(%ecx)

# stacktop j2
# fxch <j2=float80#4
# fxch <j2=%st(3)
fxch %st(3)

#       h2 += j2
# fadd <j2=float80#1,<h2=float80#3
# fadd <j2=%st(0),<h2=%st(2)
fadd %st(0),%st(2)

# stacktop k2
# fxch <k2=float80#2
# fxch <k2=%st(1)
fxch %st(1)

#       i2 += k2
# fadd <k2=float80#1,<i2=float80#4
# fadd <k2=%st(0),<i2=%st(3)
fadd %st(0),%st(3)

# stacktop j2
# fxch <j2=float80#2
# fxch <j2=%st(1)
fxch %st(1)

#       j2 -= *(float64 *) (out + 56)
# fsubl 56(<out=int32#2)
# fsubl 56(<out=%ecx)
fsubl 56(%ecx)

# stacktop k2
# fxch <k2=float80#2
# fxch <k2=%st(1)
fxch %st(1)

#       k2 -= *(float64 *) (out + 136)
# fsubl 136(<out=int32#2)
# fsubl 136(<out=%ecx)
fsubl 136(%ecx)

# stacktop h1
# fxch <h1=float80#7
# fxch <h1=%st(6)
fxch %st(6)

#     *(float64 *) (out + 8) = h1
# fstpl 8(<out=int32#2)
# fstpl 8(<out=%ecx)
fstpl 8(%ecx)

# stacktop j1
# fxch <j1=float80#5
# fxch <j1=%st(4)
fxch %st(4)

#     *(float64 *) (out + 48) = j1
# fstpl 48(<out=int32#2)
# fstpl 48(<out=%ecx)
fstpl 48(%ecx)

# stacktop i1
# fxch <i1=float80#6
# fxch <i1=%st(5)
fxch %st(5)

#     *(float64 *) (out + 88) = i1
# fstpl 88(<out=int32#2)
# fstpl 88(<out=%ecx)
fstpl 88(%ecx)

# stacktop k1
# fxch <k1=float80#2
# fxch <k1=%st(1)
fxch %st(1)

#     *(float64 *) (out + 128) = k1
# fstpl 128(<out=int32#2)
# fstpl 128(<out=%ecx)
fstpl 128(%ecx)

# stacktop h2
# fxch <h2=float80#4
# fxch <h2=%st(3)
fxch %st(3)

#       *(float64 *) (out + 16) = h2
# fstpl 16(<out=int32#2)
# fstpl 16(<out=%ecx)
fstpl 16(%ecx)

#       *(float64 *) (out + 56) = j2
# fstpl 56(<out=int32#2)
# fstpl 56(<out=%ecx)
fstpl 56(%ecx)

# stacktop i2
# fxch <i2=float80#2
# fxch <i2=%st(1)
fxch %st(1)

#       *(float64 *) (out + 96) = i2
# fstpl 96(<out=int32#2)
# fstpl 96(<out=%ecx)
fstpl 96(%ecx)

#       *(float64 *) (out + 136) = k2
# fstpl 136(<out=int32#2)
# fstpl 136(<out=%ecx)
fstpl 136(%ecx)

#         j3 = *(float64 *) (out + 24)
# fldl 24(<out=int32#2)
# fldl 24(<out=%ecx)
fldl 24(%ecx)

#         h3 = *(float64 *) (out + 64)
# fldl 64(<out=int32#2)
# fldl 64(<out=%ecx)
fldl 64(%ecx)

#         k3 = *(float64 *) (out + 104)
# fldl 104(<out=int32#2)
# fldl 104(<out=%ecx)
fldl 104(%ecx)

#         i3 = *(float64 *) (out + 144)
# fldl 144(<out=int32#2)
# fldl 144(<out=%ecx)
fldl 144(%ecx)

# stacktop j3
# fxch <j3=float80#4
# fxch <j3=%st(3)
fxch %st(3)

#         h3 += j3
# fadd <j3=float80#1,<h3=float80#3
# fadd <j3=%st(0),<h3=%st(2)
fadd %st(0),%st(2)

# stacktop k3
# fxch <k3=float80#2
# fxch <k3=%st(1)
fxch %st(1)

#         i3 += k3
# fadd <k3=float80#1,<i3=float80#4
# fadd <k3=%st(0),<i3=%st(3)
fadd %st(0),%st(3)

# stacktop j3
# fxch <j3=float80#2
# fxch <j3=%st(1)
fxch %st(1)

#         j3 -= *(float64 *) (out + 64)
# fsubl 64(<out=int32#2)
# fsubl 64(<out=%ecx)
fsubl 64(%ecx)

# stacktop k3
# fxch <k3=float80#2
# fxch <k3=%st(1)
fxch %st(1)

#         k3 -= *(float64 *) (out + 144)
# fsubl 144(<out=int32#2)
# fsubl 144(<out=%ecx)
fsubl 144(%ecx)

#           j4 = *(float64 *) (out + 32)
# fldl 32(<out=int32#2)
# fldl 32(<out=%ecx)
fldl 32(%ecx)

#           h4 = *(float64 *) (out + 72)
# fldl 72(<out=int32#2)
# fldl 72(<out=%ecx)
fldl 72(%ecx)

#           k4 = *(float64 *) (out + 112)
# fldl 112(<out=int32#2)
# fldl 112(<out=%ecx)
fldl 112(%ecx)

#           i4 = *(float64 *) (out + 152)
# fldl 152(<out=int32#2)
# fldl 152(<out=%ecx)
fldl 152(%ecx)

# stacktop j4
# fxch <j4=float80#4
# fxch <j4=%st(3)
fxch %st(3)

#           h4 += j4
# fadd <j4=float80#1,<h4=float80#3
# fadd <j4=%st(0),<h4=%st(2)
fadd %st(0),%st(2)

# stacktop k4
# fxch <k4=float80#2
# fxch <k4=%st(1)
fxch %st(1)

#           i4 += k4
# fadd <k4=float80#1,<i4=float80#4
# fadd <k4=%st(0),<i4=%st(3)
fadd %st(0),%st(3)

# stacktop j4
# fxch <j4=float80#2
# fxch <j4=%st(1)
fxch %st(1)

#           j4 -= *(float64 *) (out + 72)
# fsubl 72(<out=int32#2)
# fsubl 72(<out=%ecx)
fsubl 72(%ecx)

# stacktop k4
# fxch <k4=float80#2
# fxch <k4=%st(1)
fxch %st(1)

#           k4 -= *(float64 *) (out + 152)
# fsubl 152(<out=int32#2)
# fsubl 152(<out=%ecx)
fsubl 152(%ecx)

# stacktop h3
# fxch <h3=float80#7
# fxch <h3=%st(6)
fxch %st(6)

#         *(float64 *) (out + 24) = h3
# fstpl 24(<out=int32#2)
# fstpl 24(<out=%ecx)
fstpl 24(%ecx)

# stacktop j3
# fxch <j3=float80#5
# fxch <j3=%st(4)
fxch %st(4)

#         *(float64 *) (out + 64) = j3
# fstpl 64(<out=int32#2)
# fstpl 64(<out=%ecx)
fstpl 64(%ecx)

# stacktop i3
# fxch <i3=float80#6
# fxch <i3=%st(5)
fxch %st(5)

#         *(float64 *) (out + 104) = i3
# fstpl 104(<out=int32#2)
# fstpl 104(<out=%ecx)
fstpl 104(%ecx)

# stacktop k3
# fxch <k3=float80#2
# fxch <k3=%st(1)
fxch %st(1)

#         *(float64 *) (out + 144) = k3
# fstpl 144(<out=int32#2)
# fstpl 144(<out=%ecx)
fstpl 144(%ecx)

# stacktop h4
# fxch <h4=float80#4
# fxch <h4=%st(3)
fxch %st(3)

#           *(float64 *) (out + 32) = h4
# fstpl 32(<out=int32#2)
# fstpl 32(<out=%ecx)
fstpl 32(%ecx)

#           *(float64 *) (out + 72) = j4
# fstpl 72(<out=int32#2)
# fstpl 72(<out=%ecx)
fstpl 72(%ecx)

# stacktop i4
# fxch <i4=float80#2
# fxch <i4=%st(1)
fxch %st(1)

#           *(float64 *) (out + 112) = i4
# fstpl 112(<out=int32#2)
# fstpl 112(<out=%ecx)
fstpl 112(%ecx)

#           *(float64 *) (out + 152) = k4
# fstpl 152(<out=int32#2)
# fstpl 152(<out=%ecx)
fstpl 152(%ecx)

# leave
add %eax,%esp
ret