mm: delay rmap removal until after TLB flushmmu_gather-race-fix

When we remove a page table entry, we are very careful to only free the
page after we have flushed the TLB, because other CPUs could still be
using the page through stale TLB entries until after the flush.

However, we have removed the rmap entry for that page early, which means
that functions like folio_mkclean() would end up not serializing with
the page table lock because the page had already been made invisible to
rmap.

And that is a problem, because while the TLB entry exists, we could end
up with the following situation:

 (a) one CPU could come in and clean it, never seeing our mapping of
     the page

 (b) another CPU could continue to use the stale and dirty TLB entry
     and continue to write to said page

resulting in a page that has been dirtied, but then marked clean again,
all while another CPU might have dirtied it some more.

End result: possibly lost dirty data.

This commit uses the same old TLB gather array that we use to delay the
freeing of the page to also say 'remove from rmap after flush', so that
we can keep the rmap entries alive until all TLB entries have been
flushed.

It might be worth noting that this means that the page_zap_pte_rmap() is
now called outside the page table lock.  That was never mutual exclusion
(since the same page could be mapped under multiple different page
tables), but it does mean that it needs to use the more careful version
of dec_lruvec_page_state() that doesn't depend on being called in a
non-preemptable context.

NOTE! While the "possibly lost dirty data" sounds catastrophic, for this
all to happen you need to have a user thread doing either madvise() with
MADV_DONTNEED or a full re-mmap() of the area concurrently with another
thread continuing to use said mapping.

So arguably this is about user space doing crazy things, but from a VM
consistency standpoint it's better if we track the dirty bit properly
even when user space goes off the rails.

Reported-by: Nadav Amit <nadav.amit@gmail.com>
Link: https://lore.kernel.org/all/B88D3073-440A-41C7-95F4-895D3F657EF2@gmail.com/
Cc: Will Deacon <will@kernel.org>
Cc: Aneesh Kumar <aneesh.kumar@linux.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Nick Piggin <npiggin@gmail.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: Sven Schnelle <svens@linux.ibm.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com> # s390
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

1 files changed, 33 insertions, 6 deletions

diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index 492dce43236e..b6045149d0b9 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -73,7 +73,8 @@
  *    __tlb_remove_page_size() is the basic primitive that queues a page for
  *    freeing. __tlb_remove_page() assumes PAGE_SIZE. Both will return a
  *    boolean indicating if the queue is (now) full and a call to
- *    tlb_flush_mmu() is required.
+ *    tlb_flush_mmu() is required. They take a 'flags' parameter that
+ *    states whether the rmap of the page should be removed after TLB flush.
  *
  *    tlb_remove_page() and tlb_remove_page_size() imply the call to
  *    tlb_flush_mmu() when required and has no return value.
@@ -187,6 +188,7 @@
  *  This is useful if your architecture already flushes TLB entries in the
  *  various ptep_get_and_clear() functions.
  */
+#define TLB_ZAP_RMAP 1ul
 
 #ifdef CONFIG_MMU_GATHER_TABLE_FREE
 
@@ -238,11 +240,36 @@ extern void tlb_remove_table(struct mmu_gather *tlb, void *table);
  */
 #define MMU_GATHER_BUNDLE	8
 
+/*
+ * Fake type for an encoded page with flag bits in the low bits.
+ *
+ * Right now just one bit, but we could have more depending on the
+ * alignment of 'struct page'.
+ */
+struct encoded_page;
+#define ENCODE_PAGE_BITS (TLB_ZAP_RMAP)
+
+static inline struct encoded_page *encode_page(struct page *page, unsigned long flags)
+{
+	flags &= ENCODE_PAGE_BITS;
+	return (struct encoded_page *)(flags | (unsigned long)page);
+}
+
+static inline bool encoded_page_flags(struct encoded_page *page)
+{
+	return ENCODE_PAGE_BITS & (unsigned long)page;
+}
+
+static inline struct page *encoded_page_ptr(struct encoded_page *page)
+{
+	return (struct page *)(~ENCODE_PAGE_BITS & (unsigned long)page);
+}
+
 struct mmu_gather_batch {
 	struct mmu_gather_batch	*next;
 	unsigned int		nr;
 	unsigned int		max;
-	struct page		*pages[];
+	struct encoded_page	*encoded_pages[];
 };
 
 #define MAX_GATHER_BATCH	\
@@ -257,7 +284,7 @@ struct mmu_gather_batch {
 #define MAX_GATHER_BATCH_COUNT	(10000UL/MAX_GATHER_BATCH)
 
 extern bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page,
-				   int page_size);
+				   int page_size, unsigned int flags);
 #endif
 
 /*
@@ -431,13 +458,13 @@ static inline void tlb_flush_mmu_tlbonly(struct mmu_gather *tlb)
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
-	if (__tlb_remove_page_size(tlb, page, page_size))
+	if (__tlb_remove_page_size(tlb, page, page_size, 0))
 		tlb_flush_mmu(tlb);
 }
 
-static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
+static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page, unsigned int flags)
 {
-	return __tlb_remove_page_size(tlb, page, PAGE_SIZE);
+	return __tlb_remove_page_size(tlb, page, PAGE_SIZE, flags);
 }
 
 /* tlb_remove_page