On Wed, Sep 12, 2018 at 09:03:55AM -0400, Jerome Glisse wrote: > On Wed, Sep 12, 2018 at 02:49:21PM +0800, Peter Xu wrote: > > Add an extra check on page dirty bit in change_pte_range() since there > > might be case where PTE dirty bit is unset but it's actually dirtied. > > One example is when a huge PMD is splitted after written: the dirty bit > > will be set on the compound page however we won't have the dirty bit set > > on each of the small page PTEs. > > > > I noticed this when debugging with a customized kernel that implemented > > userfaultfd write-protect. In that case, the dirty bit will be critical > > since that's required for userspace to handle the write protect page > > fault (otherwise it'll get a SIGBUS with a loop of page faults). > > However it should still be good even for upstream Linux to cover more > > scenarios where we shouldn't need to do extra page faults on the small > > pages if the previous huge page is already written, so the dirty bit > > optimization path underneath can cover more. > > > > So as said by Kirill NAK you are not looking at the right place for > your bug please first apply the below patch and read my analysis in > my last reply. Just to be clear you are trying to fix a userspace bug that is hidden for non THP pages by a kernel space bug inside userfaultfd by making the kernel space bug of userfaultfd buggy for THP too. > > Below patch fix userfaultfd bug. I am not posting it as it is on a > branch and i am not sure when Andrea plan to post. Andrea feel free > to squash that fix. > > > From 35cdb30afa86424c2b9f23c0982afa6731be961c Mon Sep 17 00:00:00 2001 > From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Glisse?= <jglisse@xxxxxxxxxx> > Date: Wed, 12 Sep 2018 08:58:33 -0400 > Subject: [PATCH] userfaultfd: do not set dirty accountable when changing > protection > MIME-Version: 1.0 > Content-Type: text/plain; charset=UTF-8 > Content-Transfer-Encoding: 8bit > > mwriteprotect_range() has nothing to do with the dirty accountable > optimization so do not set it as it opens a door for userspace to > unwrite protect pages in a range that is write protected ie the vma > !(vm_flags & VM_WRITE). > > Signed-off-by: Jérôme Glisse <jglisse@xxxxxxxxxx> > --- > mm/userfaultfd.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c > index a0379c5ffa7c..59db1ce48fa0 100644 > --- a/mm/userfaultfd.c > +++ b/mm/userfaultfd.c > @@ -632,7 +632,7 @@ int mwriteprotect_range(struct mm_struct *dst_mm, unsigned long start, > newprot = vm_get_page_prot(dst_vma->vm_flags); > > change_protection(dst_vma, start, start + len, newprot, > - !enable_wp, 0); > + false, 0); > > err = 0; > out_unlock: > -- > 2.17.1 >
