x86_64 cmpxchg behavior in qemu tcg does not match the real CPU
Copied from https://bugs.launchpad.net/qemu/+bug/1915327
Still broken in f2da205c
Consider the following little program:
$ cat 1.c
#include <stdio.h>
int main() {
int mem = 0x12345678;
register long rax asm("rax") = 0x1234567812345678;
register int edi asm("edi") = 0x77777777;
asm("cmpxchg %[edi],%[mem]"
: [ mem ] "+m"(mem), [ rax ] "+r"(rax)
: [ edi ] "r"(edi));
long rax2 = rax;
printf("rax2 = %lx\n", rax2);
}
According to the Intel Manual, cmpxchg should not touch the accumulator in case the values are equal, which is indeed the case on the real CPU:
$ gcc 1.c
$ ./a.out
rax2 = 1234567812345678
However, QEMU appears to zero extend EAX to RAX:
$ qemu-x86_64 ./a.out
rax2 = 12345678
This is also the case for lock cmpxchg.
Edited by Ilya Leoshkevich