Re: Intel XNU bug report
Re: Intel XNU bug report
- Subject: Re: Intel XNU bug report
- From: Michael Smith <email@hidden>
- Date: Tue, 8 Aug 2006 12:27:20 -0700
On Aug 8, 2006, at 12:22 PM, email@hidden wrote:
Now that the source is released, we might as well get to work.
That's the point of releasing the source in the first place, right?
File: xnu-792.10.96/osfmk/i386/acpi.c
Line: 151
if (cpu_mode_is64bit()) {
cpu_IA32e_enable(current_cpu_datap());
cpu_desc_load64(current_cpu_datap());
kprintf("acpi_sleep_kernel 64-bit mode re-enabled\n");
fast_syscall_init64();
} else {
fast_syscall_init()
}
Bug: The 64 bit mode should be restored if the cpu is *NOT* currently
in 64 bit mode.
Without double-checking the source, I think you are making an
erroneous assumption here.
The first line should read:
if (!cpu_mode_is64bit()) {
I believe the applicable logic is that, returning from sleep the CPU
will either be in 32-bit mode
on a 32-bit system, or 64-bit mode on a 64-bit system.
In the former case, nothing need be done as that is the mode in which
the kernel operates. On
non-64-bit systems, attempting to restore 64-bit mode operation would
be fatal.
In the latter case, however, the kernel is run in IA32e mode and thus
it's necessary to change.
The diagnostic there is perhaps confusing the matter.
= Mike
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden