Looking through old bug tickets... can you still reproduce this issue with the latest version of QEMU? Or could we close this ticket nowadays?
** Changed in: qemu Status: New => Incomplete -- You received this bug notification because you are a member of qemu- devel-ml, which is subscribed to QEMU. https://bugs.launchpad.net/bugs/657329 Title: APIC unusable on QEMU Status in QEMU: Incomplete Bug description: The APIC is unusable with QEMU using x86-64 system emulation. Problem exists in the latest stable QEMU 0.12.5 as well as the latest git head. I am using Mac OS X 10.6, 64-bit version of QEMU. The QEMU binary was configured with: ./configure --target-list=i386-softmmu,x86_64-softmmubck-i-search: conf_ Problem is that the hw/apic.c file (as well as a few other naughty files) rely on the cpu_single_env global - which is set to NULL in cpu-exec.c. Below is a test reading the local APIC version register: Before taking it out: (qemu) xp 0xfee00030 00000000fee00030: 0x00000000 (qemu) After: (qemu) xp 0xfee00030 00000000fee00030: 0x00050011 (qemu) Quick fix below. I don't know if there are any side effects with this, if this is OK maybe we can fix it like this for the stable versions and fix the HEAD to not rely on the cpu_single_env global. diff --git a/cpu-exec.c b/cpu-exec.c index dbdfdcc..3e966d7 100644 --- a/cpu-exec.c +++ b/cpu-exec.c @@ -674,7 +674,17 @@ int cpu_exec(CPUState *env1) env = (void *) saved_env_reg; /* fail safe : never use cpu_single_env outside cpu_exec() */ +#warning fixup devices which rely on this +#if 0 + /* + * Hello. This is wrapped around an #if 0 ... #endif because that's + * what should happen. However, certain naughty devices (like the APIC + * for instance, and a few others), access this global variable. + * + * So this is here for now ... until we fix up those devices. + */ cpu_single_env = NULL; +#endif return ret; } To manage notifications about this bug go to: https://bugs.launchpad.net/qemu/+bug/657329/+subscriptions