[S390] pgtables: Fix race in enable_sie vs. page table ops

The current enable_sie code sets the mm->context.pgstes bit to tell
dup_mm that the new mm should have extended page tables. This bit is also
used by the s390 specific page table primitives to decide about the page
table layout - which means context.pgstes has two meanings. This can cause
any kind of bugs. For example  - e.g. shrink_zone can call
ptep_clear_flush_young while enable_sie is running. ptep_clear_flush_young
will test for context.pgstes. Since enable_sie changed that value of the old
struct mm without changing the page table layout ptep_clear_flush_young will
do the wrong thing.
The solution is to split pgstes into two bits
- one for the allocation
- one for the current state

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

1 files changed, 16 insertions, 3 deletions

diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h
index 4c2fbf48c9c4..28ec870655af 100644
--- a/arch/s390/include/asm/mmu_context.h
+++ b/arch/s390/include/asm/mmu_context.h
@@ -20,12 +20,25 @@ static inline int init_new_context(struct task_struct *tsk,
 #ifdef CONFIG_64BIT
         mm->context.asce_bits |= _ASCE_TYPE_REGION3;
 #endif
-        if (current->mm->context.pgstes) {
+        if (current->mm->context.alloc_pgste) {
+                /*
+                 * alloc_pgste indicates, that any NEW context will be created
+                 * with extended page tables. The old context is unchanged. The
+                 * page table allocation and the page table operations will
+                 * look at has_pgste to distinguish normal and extended page
+                 * tables. The only way to create extended page tables is to
+                 * set alloc_pgste and then create a new context (e.g. dup_mm).
+                 * The page table allocation is called after init_new_context
+                 * and if has_pgste is set, it will create extended page
+                 * tables.
+                 */
                 mm->context.noexec = 0;
-                mm->context.pgstes = 1;
+                mm->context.has_pgste = 1;
+                mm->context.alloc_pgste = 1;
         } else {
                 mm->context.noexec = s390_noexec;
-                mm->context.pgstes = 0;
+                mm->context.has_pgste = 0;
+                mm->context.alloc_pgste = 0;
         }
         mm->context.asce_limit = STACK_TOP_MAX;
         crst_table_init((unsigned long *) mm->pgd, pgd_entry_type(mm));