On 02/17/2020 03:57 PM, Kirill A. Shutemov wrote: > On Mon, Feb 17, 2020 at 10:33:52AM +0530, Anshuman Khandual wrote: >> This replaces all remaining open encodings with vma_set_anonymous(). >> >> Cc: Sudeep Dutt <sudeep.dutt@xxxxxxxxx> >> Cc: Ashutosh Dixit <ashutosh.dixit@xxxxxxxxx> >> Cc: Arnd Bergmann <arnd@xxxxxxxx> >> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> >> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx> >> Cc: Kate Stewart <kstewart@xxxxxxxxxxxxxxxxxxx> >> Cc: Allison Randal <allison@xxxxxxxxxxx> >> Cc: Richard Fontana <rfontana@xxxxxxxxxx> >> Cc: linux-kernel@xxxxxxxxxxxxxxx >> Signed-off-by: Anshuman Khandual <anshuman.khandual@xxxxxxx> >> --- >> drivers/misc/mic/scif/scif_mmap.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/drivers/misc/mic/scif/scif_mmap.c b/drivers/misc/mic/scif/scif_mmap.c >> index a151d416f39c..1f0dec5df994 100644 >> --- a/drivers/misc/mic/scif/scif_mmap.c >> +++ b/drivers/misc/mic/scif/scif_mmap.c >> @@ -580,7 +580,7 @@ static void scif_munmap(struct vm_area_struct *vma) >> * The kernel probably zeroes these out but we still want >> * to clean up our own mess just in case. >> */ >> - vma->vm_ops = NULL; >> + vma_set_anonymous(vma); >> vma->vm_private_data = NULL; >> kref_put(&vmapvt->ref, vma_pvt_release); >> scif_delete_vma(ep, vma); > > This is misleading. The VMA doesn't become anonymous here. This is undo of > the previously overwritten vm_ops. I think we should leave it opencodded. I was bit unsure about it as well, thanks for the clarification. Will drop it.
