On Thu, Jun 06, 2019 at 12:27:40PM +0100, Catalin Marinas wrote:
> On Thu, Jun 06, 2019 at 10:24:01AM +0530, Anshuman Khandual wrote:
> > On 06/04/2019 08:26 PM, Catalin Marinas wrote:
> > > On Mon, Jun 03, 2019 at 12:11:25PM +0530, Anshuman Khandual wrote:
> > >> diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
> > >> index 4bb65f3..41fa905 100644
> > >> --- a/arch/arm64/mm/fault.c
> > >> +++ b/arch/arm64/mm/fault.c
> > >> @@ -397,37 +397,29 @@ static void do_bad_area(unsigned long addr,
> > >> unsigned int esr, struct pt_regs *re
> > >> static vm_fault_t __do_page_fault(struct mm_struct *mm, unsigned long
> > >> addr,
> > >> unsigned int mm_flags, unsigned long
> > >> vm_flags)
> > >> {
> > >> - struct vm_area_struct *vma;
> > >> - vm_fault_t fault;
> > >> + struct vm_area_struct *vma = find_vma(mm, addr);
> > >>
> > >> - vma = find_vma(mm, addr);
> > >> - fault = VM_FAULT_BADMAP;
> > >> if (unlikely(!vma))
> > >> - goto out;
> > >> - if (unlikely(vma->vm_start > addr))
> > >> - goto check_stack;
> > >> + return VM_FAULT_BADMAP;
> > >>
> > >> /*
> > >> * Ok, we have a good vm_area for this memory access, so we can
> > >> handle
> > >> * it.
> > >> */
> > >> -good_area:
> > >> + if (unlikely(vma->vm_start > addr)) {
> > >> + if (!(vma->vm_flags & VM_GROWSDOWN))
> > >> + return VM_FAULT_BADMAP;
> > >> + if (expand_stack(vma, addr))
> > >> + return VM_FAULT_BADMAP;
> > >> + }
> > >
> > > You could have a single return here:
> > >
> > > if (unlikely(vma->vm_start > addr) &&
> > > (!(vma->vm_flags & VM_GROWSDOWN) || expand_stack(vma, addr)))
> > > return VM_FAULT_BADMAP;
> > >
> > > Not sure it's any clearer though.
> >
> > TBH the proposed one seems clearer as it separates effect (vma->vm_start >
> > addr)
> > from required permission check (vma->vm_flags & VM_GROWSDOWN) and required
> > action
> > (expand_stack(vma, addr)). But I am happy to change as you have mentioned
> > if that
> > is preferred.
>
> Not bothered really. You can leave them as in your proposal (I was just
> seeing the VM_GROWSDOWN check tightly coupled with the expand_stack(),
> it's fine either way).
Personally, I find it clearer as separate statements, so I'd suggest
keeping it as per Anshuman's proposal.
Thanks,
Mark.
