[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v2 1/2] xc_cpuid_x86.c: Simplify masking conditions and remove redundant work

On Tue, Sep 9, 2014 at 11:13 PM, z <alfred.z.song@xxxxxxxxx> wrote:

On Tue, Sep 9, 2014 at 10:49 PM, Jan Beulich <JBeulich@xxxxxxxx> wrote:
>>> On 09.09.14 at 16:43, <alfred.z.song@xxxxxxxxx> wrote:
> And I agree with you is_pae does not mean the guest must be long mode,
> instead, it is long mode _or_ pae mode.

Not even that - is_pae only allows the guest to enter PAE mode,
it doesn't require it to.


Sorry. Yep. If so, is_pae is not proper here for SYSCALL which only requires long mode for Intel. There might be two ways (one of them):

1. make a new flag (not is_64bit, since it is "a property of hypervisor") to indicate the guest is long mode (not pae) and keep the work at tools level
2. simply leave the work to vmexit handler if guest run CPUID



Hi, Jan

It seems that vmx_cpuid_intercept() has covered SYSCALL masking when vmexit. So we do not need is_pae here again and it could be dropped from my point of view.

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.