Re: [Xen-devel] [PATCH 1/2] xen: Drop XEN_DOMCTL_suppress_spurious_page_faults

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 27/08/2019 16:39, Jan Beulich wrote:
> On 13.08.2019 12:53, Andrew Cooper wrote:
>> This functionality is obsolete.  It was introduced by c/s 39407bed9c0
>> into
>> Xend, but never exposed in libxl.
>
> This is good enough a reason I think (hope), while ...
>
>> While not explicitly limited to PV guests, this is PV-only by virtue
>> of its
>> position in the pagefault handler.
>>
>> Looking though the XenServer templates, this was used to work around
>> bugs in
>> the 32bit RHEL/CentOS 4.{5..7} kernels (fixed in 4.8).  RHEL 4 as a
>> major
>> version when out if support in 2017.
>
> ... I wouldn't accept this as justification,

I'd be approaching this differently if support were in libxl, but...

> since people are
> free to use out-of-support software for as long as they wish.

... there are a lot of very good reasons not to keep on using
out-of-support software.  A recent example is the WannaCry outbreak
which spread in part due to the still-prolific use of Windows XP.

Furthermore, people running kernels which are that obsolete will also be
running similarly-obsolete versions of Xen, so this change isn't going
to affect them.

>
>> Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
>
> Hypervisor parts
> Reviewed-by: Jan Beulich <jbeulich@xxxxxxxx>

Thanks,

> with one further suggestion:
>
>> --- a/xen/arch/x86/traps.c
>> +++ b/xen/arch/x86/traps.c
>> @@ -1452,20 +1452,6 @@ void do_page_fault(struct cpu_user_regs *regs)
>>                 error_code, _p(addr));
>>       }
>>   -    if (
>> unlikely(current->domain->arch.suppress_spurious_page_faults) )
>> -    {
>> -        pf_type = spurious_page_fault(addr, regs);
>> -        if ( (pf_type == smep_fault) || (pf_type == smap_fault))
>> -        {
>> -            printk(XENLOG_G_ERR "%pv fatal SM%cP violation\n",
>> -                   current, (pf_type == smep_fault) ? 'E' : 'A');
>> -
>> -            domain_crash(current->domain);
>> -        }
>> -        if ( pf_type != real_fault )
>> -            return;
>> -    }
>
> You may want to consider reducing pf_type's scope at the same
> time.

This isn't too much of a change, so I've included it.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel