[S390] tlb flush fix.

The current tlb flushing code for page table entries violates the
s390 architecture in a small detail. The relevant section from the
principles of operation (SA22-7832-02 page 3-47):

   "A valid table entry must not be changed while it is attached
   to any CPU and may be used for translation by that CPU except to
   (1) invalidate the entry by using INVALIDATE PAGE TABLE ENTRY or
   INVALIDATE DAT TABLE ENTRY, (2) alter bits 56-63 of a page-table
   entry, or (3) make a change by means of a COMPARE AND SWAP AND
   PURGE instruction that purges the TLB."

That means if one thread of a multithreaded applciation uses a vma
while another thread does an unmap on it, the page table entries of
that vma needs to get removed with IPTE, IDTE or CSP. In some strange
and rare situations a cpu could check-stop (die) because a entry has
been pushed out of the TLB that is still needed to complete a
(milli-coded) instruction. I've never seen it happen with the current
code on any of the supported machines, so right now this is a
theoretical problem. But I want to fix it nevertheless, to avoid
headaches in the futures.

To get this implemented correctly without changing common code the
primitives ptep_get_and_clear, ptep_get_and_clear_full and
ptep_set_wrprotect need to use the IPTE instruction to invalidate the
pte before the new pte value gets stored. If IPTE is always used for
the three primitives three important operations will have a performace
hit: fork, mprotect and exit_mmap. Time for some workarounds:

* 1: ptep_get_and_clear_full is used in unmap_vmas to remove page
tables entries in a batched tlb gather operation. If the mmu_gather
context passed to unmap_vmas has been started with full_mm_flush==1
or if only one cpu is online or if the only user of a mm_struct is the
current process then the fullmm indication in the mmu_gather context is
set to one. All TLBs for mm_struct are flushed by the tlb_gather_mmu
call. No new TLBs can be created while the unmap is in progress. In
this case ptep_get_and_clear_full clears the ptes with a simple store.

* 2: ptep_get_and_clear is used in change_protection to clear the
ptes from the page tables before they are reentered with the new
access flags. At the end of the update flush_tlb_range clears the
remaining TLBs. In general the ptep_get_and_clear has to issue IPTE
for each pte and flush_tlb_range is a nop. But if there is only one
user of the mm_struct then ptep_get_and_clear uses simple stores
to do the update and flush_tlb_range will flush the TLBs.

* 3: Similar to 2, ptep_set_wrprotect is used in copy_page_range
for a fork to make all ptes of a cow mapping read-only. At the end of
of copy_page_range dup_mmap will flush the TLBs with a call to
flush_tlb_mm.  Check for mm->mm_users and if there is only one user
avoid using IPTE in ptep_set_wrprotect and let flush_tlb_mm clear the
TLBs.

Overall for single threaded programs the tlb flush code now performs
better, for multi threaded programs it is slightly worse. In particular
exit_mmap() now does a single IDTE for the mm and then just frees every
page cache reference and every page table page directly without a delay
over the mmu_gather structure.

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

1 files changed, 71 insertions, 28 deletions

diff --git a/include/asm-s390/pgtable.h b/include/asm-s390/pgtable.h
index 39bb5192dc31..b424ab21f8bd 100644
--- a/include/asm-s390/pgtable.h
+++ b/include/asm-s390/pgtable.h
@@ -424,7 +424,8 @@ static inline pgd_t *get_shadow_pgd(pgd_t *pgdp)
  * within a page table are directly modified.  Thus, the following
  * hook is made available.
  */
-static inline void set_pte(pte_t *pteptr, pte_t pteval)
+static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
+			      pte_t *pteptr, pte_t pteval)
 {
 	pte_t *shadow_pte = get_shadow_pte(pteptr);
 
@@ -437,7 +438,6 @@ static inline void set_pte(pte_t *pteptr, pte_t pteval)
 			pte_val(*shadow_pte) = _PAGE_TYPE_EMPTY;
 	}
 }
-#define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval)
 
 /*
  * pgd/pmd/pte query functions
@@ -508,7 +508,8 @@ static inline int pte_file(pte_t pte)
 	return (pte_val(pte) & mask) == _PAGE_TYPE_FILE;
 }
 
-#define pte_same(a,b)	(pte_val(a) == pte_val(b))
+#define __HAVE_ARCH_PTE_SAME
+#define pte_same(a,b)  (pte_val(a) == pte_val(b))
 
 /*
  * query functions pte_write/pte_dirty/pte_young only work if
@@ -663,24 +664,19 @@ static inline pte_t pte_mkyoung(pte_t pte)
 	return pte;
 }
 
-static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep)
+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
+static inline int ptep_test_and_clear_young(struct vm_area_struct *vma,
+					    unsigned long addr, pte_t *ptep)
 {
 	return 0;
 }
 
-static inline int
-ptep_clear_flush_young(struct vm_area_struct *vma,
-			unsigned long address, pte_t *ptep)
+#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
+static inline int ptep_clear_flush_young(struct vm_area_struct *vma,
+					 unsigned long address, pte_t *ptep)
 {
 	/* No need to flush TLB; bits are in storage key */
-	return ptep_test_and_clear_young(vma, address, ptep);
-}
-
-static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
-{
-	pte_t pte = *ptep;
-	pte_clear(mm, addr, ptep);
-	return pte;
+	return 0;
 }
 
 static inline void __ptep_ipte(unsigned long address, pte_t *ptep)
@@ -709,6 +705,32 @@ static inline void ptep_invalidate(unsigned long address, pte_t *ptep)
 		__ptep_ipte(address, ptep);
 }
 
+/*
+ * This is hard to understand. ptep_get_and_clear and ptep_clear_flush
+ * both clear the TLB for the unmapped pte. The reason is that
+ * ptep_get_and_clear is used in common code (e.g. change_pte_range)
+ * to modify an active pte. The sequence is
+ *   1) ptep_get_and_clear
+ *   2) set_pte_at
+ *   3) flush_tlb_range
+ * On s390 the tlb needs to get flushed with the modification of the pte
+ * if the pte is active. The only way how this can be implemented is to
+ * have ptep_get_and_clear do the tlb flush. In exchange flush_tlb_range
+ * is a nop.
+ */
+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
+#define ptep_get_and_clear(__mm, __address, __ptep)			\
+({									\
+	pte_t __pte = *(__ptep);					\
+	if (atomic_read(&(__mm)->mm_users) > 1 ||			\
+	    (__mm) != current->active_mm)				\
+		ptep_invalidate(__address, __ptep);			\
+	else								\
+		pte_clear((__mm), (__address), (__ptep));		\
+	__pte;								\
+})
+
+#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
 static inline pte_t ptep_clear_flush(struct vm_area_struct *vma,
 				     unsigned long address, pte_t *ptep)
 {
@@ -717,12 +739,40 @@ static inline pte_t ptep_clear_flush(struct vm_area_struct *vma,
 	return pte;
 }
 
-static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+/*
+ * The batched pte unmap code uses ptep_get_and_clear_full to clear the
+ * ptes. Here an optimization is possible. tlb_gather_mmu flushes all
+ * tlbs of an mm if it can guarantee that the ptes of the mm_struct
+ * cannot be accessed while the batched unmap is running. In this case
+ * full==1 and a simple pte_clear is enough. See tlb.h.
+ */
+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
+static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
+					    unsigned long addr,
+					    pte_t *ptep, int full)
 {
-	pte_t old_pte = *ptep;
-	set_pte_at(mm, addr, ptep, pte_wrprotect(old_pte));
+	pte_t pte = *ptep;
+
+	if (full)
+		pte_clear(mm, addr, ptep);
+	else
+		ptep_invalidate(addr, ptep);
+	return pte;
 }
 
+#define __HAVE_ARCH_PTEP_SET_WRPROTECT
+#define ptep_set_wrprotect(__mm, __addr, __ptep)			\
+({									\
+	pte_t __pte = *(__ptep);					\
+	if (pte_write(__pte)) {						\
+		if (atomic_read(&(__mm)->mm_users) > 1 ||		\
+		    (__mm) != current->active_mm)			\
+			ptep_invalidate(__addr, __ptep);		\
+		set_pte_at(__mm, __addr, __ptep, pte_wrprotect(__pte));	\
+	}								\
+})
+
+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
 #define ptep_set_access_flags(__vma, __addr, __ptep, __entry, __dirty)	\
 ({									\
 	int __changed = !pte_same(*(__ptep), __entry);			\
@@ -740,11 +790,13 @@ static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr,
  * should therefore only be called if it is not mapped in any
  * address space.
  */
+#define __HAVE_ARCH_PAGE_TEST_DIRTY
 static inline int page_test_dirty(struct page *page)
 {
 	return (page_get_storage_key(page_to_phys(page)) & _PAGE_CHANGED) != 0;
 }
 
+#define __HAVE_ARCH_PAGE_CLEAR_DIRTY
 static inline void page_clear_dirty(struct page *page)
 {
 	page_set_storage_key(page_to_phys(page), PAGE_DEFAULT_KEY);
@@ -753,6 +805,7 @@ static inline void page_clear_dirty(struct page *page)
 /*
  * Test and clear referenced bit in storage key.
  */
+#define __HAVE_ARCH_PAGE_TEST_AND_CLEAR_YOUNG
 static inline int page_test_and_clear_young(struct page *page)
 {
 	unsigned long physpage = page_to_phys(page);
@@ -930,16 +983,6 @@ extern int remove_shared_memory(unsigned long start, unsigned long size);
 #define __HAVE_ARCH_MEMMAP_INIT
 extern void memmap_init(unsigned long, int, unsigned long, unsigned long);
 
-#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
-#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
-#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
-#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
-#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
-#define __HAVE_ARCH_PTEP_SET_WRPROTECT
-#define __HAVE_ARCH_PTE_SAME
-#define __HAVE_ARCH_PAGE_TEST_DIRTY
-#define __HAVE_ARCH_PAGE_CLEAR_DIRTY
-#define __HAVE_ARCH_PAGE_TEST_AND_CLEAR_YOUNG
 #include <asm-generic/pgtable.h>
 
 #endif /* _S390_PAGE_H */