[PATCH 6/6] mm/mremap: hold the rmap lock in write mode when moving page table entries.

["Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>]

To avoid a race between rmap walk and mremap, mremap does take_rmap_locks().
The lock was taken to ensure that rmap walk don't miss a page table entry due to
PTE moves via move_pagetables(). The kernel does further optimization of
this lock such that if we are going to find the newly added vma after the
old vma, the rmap lock is not taken. This is because rmap walk would find the
vmas in the same order and if we don't find the page table attached to
older vma we would find it with the new vma which we would iterate later.

As explained in commit eb66ae030829 ("mremap: properly flush TLB before releasing the page")
mremap is special in that it doesn't take ownership of the page. The
optimized version for PUD/PMD aligned mremap also doesn't hold the ptl lock.
This can result in stale TLB entries as show below.

This patch updates the rmap locking requirement in mremap to handle the race condition
explained below with optimized mremap::

Optmized PMD move

    CPU 1                           CPU 2                                   CPU 3

    mremap(old_addr, new_addr)      page_shrinker/try_to_unmap_one

    mmap_write_lock_killable()

                                    addr = old_addr
                                    lock(pte_ptl)
    lock(pmd_ptl)
    pmd = *old_pmd
    pmd_clear(old_pmd)
    flush_tlb_range(old_addr)

    *new_pmd = pmd
                                                                            *new_addr = 10; and fills
                                                                            TLB with new addr
                                                                            and old pfn

    unlock(pmd_ptl)
                                    ptep_clear_flush()
                                    old pfn is free.
                                                                            Stale TLB entry

Optimized PUD move also suffers from a similar race.
Both the above race condition can be fixed if we force mremap path to take rmap lock.

Fixes: 2c91bd4a4e2e ("mm: speed up mremap by 20x on large regions")
Fixes: c49dd3401802 ("mm: speedup mremap on 1GB or larger regions")
Link: https://lore.kernel.org/linux-mm/CAHk-=wgXVR04eBNtxQfevontWnP6FDm+oj5vauQXP3S-huwbPw@mail.gmail.com
Acked-by: Hugh Dickins <hughd@google.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 mm/mremap.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/mremap.c b/mm/mremap.c
index dacfa9111ab1..b8eed7645cea 100644
--- a/mm/mremap.c
+++ b/mm/mremap.c
@@ -504,7 +504,7 @@ unsigned long move_page_tables(struct vm_area_struct *vma,
 		} else if (IS_ENABLED(CONFIG_HAVE_MOVE_PUD) && extent == PUD_SIZE) {
 
 			if (move_pgt_entry(NORMAL_PUD, vma, old_addr, new_addr,
-					   old_pud, new_pud, need_rmap_locks))
+					   old_pud, new_pud, true))
 				continue;
 		}
 
@@ -531,7 +531,7 @@ unsigned long move_page_tables(struct vm_area_struct *vma,
 			 * moving at the PMD level if possible.
 			 */
 			if (move_pgt_entry(NORMAL_PMD, vma, old_addr, new_addr,
-					   old_pmd, new_pmd, need_rmap_locks))
+					   old_pmd, new_pmd, true))
 				continue;
 		}
 
-- 
2.31.1