[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Xen-devel] [PATCH 7/9] x86: skip check for spurious faults for non-present faults
- To: "H. Peter Anvin" <hpa@xxxxxxxxx>
- From: Keir Fraser <keir.xen@xxxxxxxxx>
- Date: Thu, 15 May 2014 20:22:43 +0100
- Cc: x86@xxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, Dave Hansen <dave.hansen@xxxxxxxxx>, Ingo Molnar <mingo@xxxxxxxxxx>, David Vrabel <david.vrabel@xxxxxxxxxx>, xen-devel@xxxxxxxxxxxxxxxxxxxx, Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>, Thomas Gleixner <tglx@xxxxxxxxxxxxx>, Mel Gorman <mgorman@xxxxxxx>
- Delivery-date: Thu, 15 May 2014 19:23:14 +0000
- List-id: Xen developer discussion <xen-devel.lists.xen.org>
H. Peter Anvin wrote:
On 04/15/2014 07:15 AM, David Vrabel wrote:
If a fault on a kernel address is due to a non-present page, then it
cannot be the result of stale TLB entry from a protection change (RO
to RW or NX to X). Thus the pagetable walk in spurious_fault() can be
skipped.
Erk... this code is screaming WTF to me. The x86 architecture is such
that the CPU is responsible for avoiding these faults.
Not in this case...
<dig> <dig> <dig>
5b727a3b0158a129827c21ce3bfb0ba997e8ddd0
x86: ignore spurious faults
When changing a kernel page from RO->RW, it's OK to leave stale TLB
entries around, since doing a global flush is expensive and they
pose no security problem. They can, however, generate a spurious
fault, which we should catch and simply return from (which will
have the side-effect of reloading the TLB to the current PTE).
This can occur when running under Xen, because it frequently changes
kernel pages from RW->RO->RW to implement Xen's pagetable semantics.
It could also occur when using CONFIG_DEBUG_PAGEALLOC, since it
avoids doing a global TLB flush after changing page permissions.
Signed-off-by: Jeremy Fitzhardinge<jeremy@xxxxxxxxxxxxx>
Cc: Harvey Harrison<harvey.harrison@xxxxxxxxx>
Signed-off-by: Ingo Molnar<mingo@xxxxxxx>
Signed-off-by: Thomas Gleixner<tglx@xxxxxxxxxxxxx>
Again WTF?
Are we chasing hardware errata here? Or did someone go off and *assume*
that the x86 hardware architecture work a certain way? Or is there
something way more subtle going on?
See Intel Developer's Manual Vol 3 Section 4.10.4.3, 3rd bullet... This
is expected behaviour, probably to make copy-on-write faults faster.
-- Keir
I guess next step is mailing list archaeology...
Does anyone still have contacts with Jeremy, and if so, could they poke
him perhaps?
-hpa
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel
|
