Something

void asm_memset(u32 dst, u32 val, u32 count) {  asm volatile (      "      mov     r3,%0\n"      "      mov     r4,%1\n"      "      mov     r5,%1\n"      "      mov     r6,%1\n"      "      mov     r7,%1\n"      "      mov     r8,%1\n"      "      mov     r9,%1\n"      "      mov     r10,%1\n"      "      mov     r11,%1\n"      "      mov     r0,%2\n"      "      add     r0,r0,r3\n"      "Loop: cmp     r0, r3\n"      "      stmneia r3!, {r4-r11}\n"      "      bne     Loop"      : /* no output */       : "r" (dst), "r" (val), "r" (count) // inputs: %0,%1,%2      : "r0", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11" );}