Интересный листинг.

00000000004000b0 <.text>:
  4000b0:    55                       push   %rbp
  4000b1:    48 89 e5                 mov    %rsp,%rbp
  4000b4:    6a 04                    pushq  $0x4           ; #define __NR_stat 4
  4000b6:    58                       pop    %rax
  4000b7:    6a 09                    pushq  $0x9           ; строка Hello World занимает 11 байт. 
  4000b9:    5a                       pop    %rdx
  4000ba:    be cc 00 40 00           mov    $0x4000cc,%esi
  4000bf:    31 ff                    xor    %edi,%edi
  4000c1:    0f 05                    syscall 
  4000c3:    6a 01                    pushq  $0x1           ; #define __NR_write 1
  4000c5:    58                       pop    %rax
  4000c6:    31 ff                    xor    %edi,%edi
  4000c8:    0f 05                    syscall 
  4000ca:    eb fe                    jmp    0x4000ca

#define __NR_exit 60

Вот как выглядит корректный и с претензией минимальность

; fasm demonstration of writing 64-bit ELF executable
; (thanks to Franti.ek G.bri.)
; syscall numbers: /usr/src/linux/include/asm-x86_64/unistd.h
; parameters order:
; r9    ; 6th param
; r8    ; 5th param
; r10   ; 4th param
; rdx   ; 3rd param
; rsi   ; 2nd param
; rdi   ; 1st param
; eax   ; syscall_number
; syscall
format ELF64 executable 3
segment readable executable
entry $
           mov     edx,msg_size    ; CPU zero extends 32-bit operation to 64-bit
                                   ; we can use less bytes than in case mov rdx,...
           lea     rsi,[msg]
           mov     edi,1           ; STDOUT
           mov     eax,1           ; sys_write
           syscall
           xor     edi,edi         ; exit code 0
           mov     eax,60          ; sys_exit
           syscall
segment readable writeable
msg db 'Hello 64-bit world!',0xA
msg_size = $-msg