lkml.org 
[lkml]   [2018]   [Feb]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
SubjectRe: [tip:x86/pti] x86/speculation: Use Indirect Branch Prediction Barrier in context switch
From
Date
On 02/05/2018 06:18 AM, David Woodhouse wrote:
> On Tue, 2018-01-30 at 14:39 -0800, tip-bot for Tim Chen wrote:
>> Thanks to the reviewers and Andy Lutomirski for the suggestion of
>> using ctx_id which got rid of the problem of mm pointer recycling.
>
> That one doesn't backport well to 4.9. Suggestions welcome.
>

Will something like the following work for 4.9 using active_mm?
This patch is not really tested, but just
want to put it out here to see if this is a reasonable backport.

Tim

Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
---
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index a7655f6..4994db2 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -9,6 +9,7 @@

#include <asm/tlbflush.h>
#include <asm/mmu_context.h>
+#include <asm/nospec-branch.h>
#include <asm/cache.h>
#include <asm/apic.h>
#include <asm/uv/uv.h>
@@ -75,6 +76,9 @@ void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct *next,
struct task_struct *tsk)
{
unsigned cpu = smp_processor_id();
+#ifdef CONFIG_SMP
+ struct mm_struct *active_mm = this_cpu_read(cpu_tlbstate.active_mm);
+#endif

if (likely(prev != next)) {
if (IS_ENABLED(CONFIG_VMAP_STACK)) {
@@ -91,6 +95,28 @@ void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct *next,
set_pgd(pgd, init_mm.pgd[stack_pgd_index]);
}

+ /*
+ * Avoid user/user BTB poisoning by flushing the branch
+ * predictor when switching between processes. This stops
+ * one process from doing Spectre-v2 attacks on another.
+ *
+ * As an optimization, flush indirect branches only when
+ * switching into processes that disable dumping. This
+ * protects high value processes like gpg, without having
+ * too high performance overhead. IBPB is *expensive*!
+ *
+ * This will not flush branches when switching into kernel
+ * threads. It will also not flush if we switch to idle
+ * thread and back to the same process. It will flush if we
+ * switch to a different non-dumpable process.
+ */
+ if (tsk && tsk->mm &&
+#ifdef CONFIG_SMP
+ next != active_mm &&
+#endif
+ get_dumpable(tsk->mm) != SUID_DUMP_USER)
+ indirect_branch_prediction_barrier();
+
#ifdef CONFIG_SMP
this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK);
this_cpu_write(cpu_tlbstate.active_mm, next);
\
 
 \ /
  Last update: 2018-02-05 20:35    [W:0.089 / U:1.196 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site