How to fix ASM Sigmentation fault when calling a macro?

Question

Qt

I am developing a Qt C++ application. Some methods call functions written in assembler in a separate file. I am compiling the application for a 64-bit processor, so 64-bit registers are used in asma. A segmentation fault is thrown when trying to run a simple macro from a function.

.macro my_asm_macro src, so, dest, do
    movl    \so+60(\src), %r8d
    movl    \so+44(\src), %r9d
    movl    \so+28(\src), %r10d
    movl    \so+12(\src), %r11d
    movl    %r8d, \do+12(\dest)
    movl    %r9d, \do+28(\dest)
    movl    %r10d, \do+44(\dest)
    movl    %r11d, \do+60(\dest)
    movl    \so+40(\src), %r8d
    movl    \so+8(\src), %r9d
    movl    \so+48(\src), %r10d
    movl    \so+16(\src), %r11d
    movl    %r8d, \do+8(\dest)
    movl    %r9d, \do+40(\dest)
    movl    %r10d, \do+16(\dest)
    movl    %r11d, \do+48(\dest)
    movl    \so+20(\src), %r8d
    movl    \so+4(\src), %r9d
    movl    \so+52(\src), %r10d
    movl    \so+36(\src), %r11d
    movl    %r8d, \do+4(\dest)
    movl    %r9d, \do+20(\dest)
    movl    %r10d, \do+36(\dest)
    movl    %r11d, \do+52(\dest)
    movl    \so+0(\src), %r8d
    movl    \so+24(\src), %r9d
    movl    \so+32(\src), %r10d
    movl    \so+56(\src), %r11d
    movl    %r8d, \do+0(\dest)
    movl    %r9d, \do+24(\dest)
    movl    %r10d, \do+32(\dest)
    movl    %r11d, \do+56(\dest)
.endm

my_asm_function:
    my_asm_macro %rdi, 0, %rsp, 0           # sigfault here
    my_asm_macro %rdi, 64, %rsp, 64
    my_asm_macro %rdi, 128, %rsp, 128
    my_asm_macro %rdi, 192, %rsp, 192
    movdqa  64(%rsp), %xmm0
    movdqa  80(%rsp), %xmm1
    movdqa  96(%rsp), %xmm2
    movdqa  112(%rsp), %xmm3
    movdqa  128+64(%rsp), %xmm8
    movdqa  128+80(%rsp), %xmm9
    movdqa  128+96(%rsp), %xmm10
    movdqa  128+112(%rsp), %xmm11
    movdqa  256+64(%rsp), %xmm12
    movdqa  256+80(%rsp), %xmm13
    movdqa  256+96(%rsp), %xmm14
    movdqa  256+112(%rsp), %xmm15

    movq    %rsi, %rbx
    leaq    3*131072(%rsi), %rax

Answer 1

What does the declaration and call of this function look like from C++? What platform (linux/...)?