[v8,3/3] mm: fix double page fault on arm64 if PTE_AF is cleared
diff mbox series

Message ID 20190921135054.142360-4-justin.he@arm.com
State New
Headers show
Series
  • fix double page fault on arm64
Related show

Commit Message

Jia He Sept. 21, 2019, 1:50 p.m. UTC
When we tested pmdk unit test [1] vmmalloc_fork TEST1 in arm64 guest, there
will be a double page fault in __copy_from_user_inatomic of cow_user_page.

Below call trace is from arm64 do_page_fault for debugging purpose
[  110.016195] Call trace:
[  110.016826]  do_page_fault+0x5a4/0x690
[  110.017812]  do_mem_abort+0x50/0xb0
[  110.018726]  el1_da+0x20/0xc4
[  110.019492]  __arch_copy_from_user+0x180/0x280
[  110.020646]  do_wp_page+0xb0/0x860
[  110.021517]  __handle_mm_fault+0x994/0x1338
[  110.022606]  handle_mm_fault+0xe8/0x180
[  110.023584]  do_page_fault+0x240/0x690
[  110.024535]  do_mem_abort+0x50/0xb0
[  110.025423]  el0_da+0x20/0x24

The pte info before __copy_from_user_inatomic is (PTE_AF is cleared):
[ffff9b007000] pgd=000000023d4f8003, pud=000000023da9b003, pmd=000000023d4b3003, pte=360000298607bd3

As told by Catalin: "On arm64 without hardware Access Flag, copying from
user will fail because the pte is old and cannot be marked young. So we
always end up with zeroed page after fork() + CoW for pfn mappings. we
don't always have a hardware-managed access flag on arm64."

This patch fix it by calling pte_mkyoung. Also, the parameter is
changed because vmf should be passed to cow_user_page()

Add a WARN_ON_ONCE when __copy_from_user_inatomic() returns error
in case there can be some obscure use-case.(by Kirill)

[1] https://github.com/pmem/pmdk/tree/master/src/test/vmmalloc_fork

Reported-by: Yibo Cai <Yibo.Cai@arm.com>
Signed-off-by: Jia He <justin.he@arm.com>
---
 mm/memory.c | 67 ++++++++++++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 61 insertions(+), 6 deletions(-)

Comments

Matthew Wilcox Sept. 21, 2019, 3:31 p.m. UTC | #1
On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> When we tested pmdk unit test [1] vmmalloc_fork TEST1 in arm64 guest, there
> will be a double page fault in __copy_from_user_inatomic of cow_user_page.
> 
> Below call trace is from arm64 do_page_fault for debugging purpose
> [  110.016195] Call trace:
> [  110.016826]  do_page_fault+0x5a4/0x690
> [  110.017812]  do_mem_abort+0x50/0xb0
> [  110.018726]  el1_da+0x20/0xc4
> [  110.019492]  __arch_copy_from_user+0x180/0x280
> [  110.020646]  do_wp_page+0xb0/0x860
> [  110.021517]  __handle_mm_fault+0x994/0x1338
> [  110.022606]  handle_mm_fault+0xe8/0x180
> [  110.023584]  do_page_fault+0x240/0x690
> [  110.024535]  do_mem_abort+0x50/0xb0
> [  110.025423]  el0_da+0x20/0x24
> 
> The pte info before __copy_from_user_inatomic is (PTE_AF is cleared):
> [ffff9b007000] pgd=000000023d4f8003, pud=000000023da9b003, pmd=000000023d4b3003, pte=360000298607bd3
> 
> As told by Catalin: "On arm64 without hardware Access Flag, copying from
> user will fail because the pte is old and cannot be marked young. So we
> always end up with zeroed page after fork() + CoW for pfn mappings. we
> don't always have a hardware-managed access flag on arm64."
> 
> This patch fix it by calling pte_mkyoung. Also, the parameter is
> changed because vmf should be passed to cow_user_page()
> 
> Add a WARN_ON_ONCE when __copy_from_user_inatomic() returns error
> in case there can be some obscure use-case.(by Kirill)
> 
> [1] https://github.com/pmem/pmdk/tree/master/src/test/vmmalloc_fork
> 
> Reported-by: Yibo Cai <Yibo.Cai@arm.com>
> Signed-off-by: Jia He <justin.he@arm.com>

Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Kirill A. Shutemov Sept. 23, 2019, 8:28 a.m. UTC | #2
On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> When we tested pmdk unit test [1] vmmalloc_fork TEST1 in arm64 guest, there
> will be a double page fault in __copy_from_user_inatomic of cow_user_page.
> 
> Below call trace is from arm64 do_page_fault for debugging purpose
> [  110.016195] Call trace:
> [  110.016826]  do_page_fault+0x5a4/0x690
> [  110.017812]  do_mem_abort+0x50/0xb0
> [  110.018726]  el1_da+0x20/0xc4
> [  110.019492]  __arch_copy_from_user+0x180/0x280
> [  110.020646]  do_wp_page+0xb0/0x860
> [  110.021517]  __handle_mm_fault+0x994/0x1338
> [  110.022606]  handle_mm_fault+0xe8/0x180
> [  110.023584]  do_page_fault+0x240/0x690
> [  110.024535]  do_mem_abort+0x50/0xb0
> [  110.025423]  el0_da+0x20/0x24
> 
> The pte info before __copy_from_user_inatomic is (PTE_AF is cleared):
> [ffff9b007000] pgd=000000023d4f8003, pud=000000023da9b003, pmd=000000023d4b3003, pte=360000298607bd3
> 
> As told by Catalin: "On arm64 without hardware Access Flag, copying from
> user will fail because the pte is old and cannot be marked young. So we
> always end up with zeroed page after fork() + CoW for pfn mappings. we
> don't always have a hardware-managed access flag on arm64."
> 
> This patch fix it by calling pte_mkyoung. Also, the parameter is
> changed because vmf should be passed to cow_user_page()
> 
> Add a WARN_ON_ONCE when __copy_from_user_inatomic() returns error
> in case there can be some obscure use-case.(by Kirill)
> 
> [1] https://github.com/pmem/pmdk/tree/master/src/test/vmmalloc_fork
> 
> Reported-by: Yibo Cai <Yibo.Cai@arm.com>
> Signed-off-by: Jia He <justin.he@arm.com>

Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Catalin Marinas Sept. 23, 2019, 5:04 p.m. UTC | #3
On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> @@ -2151,21 +2163,53 @@ static inline void cow_user_page(struct page *dst, struct page *src, unsigned lo
>  	 * fails, we just zero-fill it. Live with it.
>  	 */
>  	if (unlikely(!src)) {
> -		void *kaddr = kmap_atomic(dst);
> -		void __user *uaddr = (void __user *)(va & PAGE_MASK);
> +		void *kaddr;
> +		pte_t entry;
> +		void __user *uaddr = (void __user *)(addr & PAGE_MASK);
>  
> +		/* On architectures with software "accessed" bits, we would
> +		 * take a double page fault, so mark it accessed here.
> +		 */

Nitpick: please follow the kernel coding style for multi-line comments
(above and the for the rest of the patch):

		/*
		 * Your multi-line comment.
		 */

> +		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
> +			vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr,
> +						       &vmf->ptl);
> +			if (likely(pte_same(*vmf->pte, vmf->orig_pte))) {
> +				entry = pte_mkyoung(vmf->orig_pte);
> +				if (ptep_set_access_flags(vma, addr,
> +							  vmf->pte, entry, 0))
> +					update_mmu_cache(vma, addr, vmf->pte);
> +			} else {
> +				/* Other thread has already handled the fault
> +				 * and we don't need to do anything. If it's
> +				 * not the case, the fault will be triggered
> +				 * again on the same address.
> +				 */
> +				pte_unmap_unlock(vmf->pte, vmf->ptl);
> +				return false;
> +			}
> +			pte_unmap_unlock(vmf->pte, vmf->ptl);
> +		}

Another nit, you could rewrite this block slightly to avoid too much
indentation. Something like (untested):

		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
			vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr,
						       &vmf->ptl);
			if (unlikely(!pte_same(*vmf->pte, vmf->orig_pte))) {
				/*
				 * Other thread has already handled the fault
				 * and we don't need to do anything. If it's
				 * not the case, the fault will be triggered
				 * again on the same address.
				 */
				pte_unmap_unlock(vmf->pte, vmf->ptl);
				return false;
			}
			entry = pte_mkyoung(vmf->orig_pte);
			if (ptep_set_access_flags(vma, addr,
						  vmf->pte, entry, 0))
				update_mmu_cache(vma, addr, vmf->pte);
			pte_unmap_unlock(vmf->pte, vmf->ptl);
		}

> +
> +		kaddr = kmap_atomic(dst);

Since you moved the kmap_atomic() here, could the above
arch_faults_on_old_pte() run in a preemptible context? I suggested to
add a WARN_ON in patch 2 to be sure.

>  		/*
>  		 * This really shouldn't fail, because the page is there
>  		 * in the page tables. But it might just be unreadable,
>  		 * in which case we just give up and fill the result with
>  		 * zeroes.
>  		 */
> -		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
> +		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
> +			/* Give a warn in case there can be some obscure
> +			 * use-case
> +			 */
> +			WARN_ON_ONCE(1);

That's more of a question for the mm guys: at this point we do the
copying with the ptl released; is there anything else that could have
made the pte old in the meantime? I think unuse_pte() is only called on
anonymous vmas, so it shouldn't be the case here.

>  			clear_page(kaddr);
> +		}
>  		kunmap_atomic(kaddr);
>  		flush_dcache_page(dst);
>  	} else
> -		copy_user_highpage(dst, src, va, vma);
> +		copy_user_highpage(dst, src, addr, vma);
> +
> +	return true;
>  }
Jia He Sept. 24, 2019, 6:43 a.m. UTC | #4
Hi Catalin
Please see an important comment inline, thanks

> -----Original Message-----
> From: Catalin Marinas <catalin.marinas@arm.com>
> Sent: 2019年9月24日 1:05
> To: Justin He (Arm Technology China) <Justin.He@arm.com>
> Cc: Will Deacon <will@kernel.org>; Mark Rutland
> <Mark.Rutland@arm.com>; James Morse <James.Morse@arm.com>; Marc
> Zyngier <maz@kernel.org>; Matthew Wilcox <willy@infradead.org>; Kirill A.
> Shutemov <kirill.shutemov@linux.intel.com>; linux-arm-
> kernel@lists.infradead.org; linux-kernel@vger.kernel.org; linux-
> mm@kvack.org; Suzuki Poulose <Suzuki.Poulose@arm.com>; Punit
> Agrawal <punitagrawal@gmail.com>; Anshuman Khandual
> <Anshuman.Khandual@arm.com>; Alex Van Brunt
> <avanbrunt@nvidia.com>; Robin Murphy <Robin.Murphy@arm.com>;
> Thomas Gleixner <tglx@linutronix.de>; Andrew Morton <akpm@linux-
> foundation.org>; Jérôme Glisse <jglisse@redhat.com>; Ralph Campbell
> <rcampbell@nvidia.com>; hejianet@gmail.com; Kaly Xin (Arm Technology
> China) <Kaly.Xin@arm.com>; nd <nd@arm.com>
> Subject: Re: [PATCH v8 3/3] mm: fix double page fault on arm64 if PTE_AF
> is cleared
> 
> On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> > @@ -2151,21 +2163,53 @@ static inline void cow_user_page(struct page
> *dst, struct page *src, unsigned lo
> >  	 * fails, we just zero-fill it. Live with it.
> >  	 */
> >  	if (unlikely(!src)) {
> > -		void *kaddr = kmap_atomic(dst);
> > -		void __user *uaddr = (void __user *)(va & PAGE_MASK);
> > +		void *kaddr;
> > +		pte_t entry;
> > +		void __user *uaddr = (void __user *)(addr & PAGE_MASK);
> >
> > +		/* On architectures with software "accessed" bits, we would
> > +		 * take a double page fault, so mark it accessed here.
> > +		 */
> 
> Nitpick: please follow the kernel coding style for multi-line comments
> (above and the for the rest of the patch):
> 
> 		/*
> 		 * Your multi-line comment.
> 		 */
> 
> > +		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte))
> {
> > +			vmf->pte = pte_offset_map_lock(mm, vmf->pmd,
> addr,
> > +						       &vmf->ptl);
> > +			if (likely(pte_same(*vmf->pte, vmf->orig_pte))) {
> > +				entry = pte_mkyoung(vmf->orig_pte);
> > +				if (ptep_set_access_flags(vma, addr,
> > +							  vmf->pte, entry, 0))
> > +					update_mmu_cache(vma, addr, vmf-
> >pte);
> > +			} else {
> > +				/* Other thread has already handled the
> fault
> > +				 * and we don't need to do anything. If it's
> > +				 * not the case, the fault will be triggered
> > +				 * again on the same address.
> > +				 */
> > +				pte_unmap_unlock(vmf->pte, vmf->ptl);
> > +				return false;
> > +			}
> > +			pte_unmap_unlock(vmf->pte, vmf->ptl);
> > +		}
> 
> Another nit, you could rewrite this block slightly to avoid too much
> indentation. Something like (untested):
> 
> 		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte))
> {
> 			vmf->pte = pte_offset_map_lock(mm, vmf->pmd,
> addr,
> 						       &vmf->ptl);
> 			if (unlikely(!pte_same(*vmf->pte, vmf->orig_pte))) {
> 				/*
> 				 * Other thread has already handled the fault
> 				 * and we don't need to do anything. If it's
> 				 * not the case, the fault will be triggered
> 				 * again on the same address.
> 				 */
> 				pte_unmap_unlock(vmf->pte, vmf->ptl);
> 				return false;
> 			}
> 			entry = pte_mkyoung(vmf->orig_pte);
> 			if (ptep_set_access_flags(vma, addr,
> 						  vmf->pte, entry, 0))
> 				update_mmu_cache(vma, addr, vmf->pte);
> 			pte_unmap_unlock(vmf->pte, vmf->ptl);
> 		}
> 
> > +
> > +		kaddr = kmap_atomic(dst);
> 
> Since you moved the kmap_atomic() here, could the above
> arch_faults_on_old_pte() run in a preemptible context? I suggested to
> add a WARN_ON in patch 2 to be sure.

Should I move kmap_atomic back to the original line? Thus, we can make sure
that arch_faults_on_old_pte() is in the context of preempt_disabled?
Otherwise, arch_faults_on_old_pte() may cause plenty of warning if I add
a WARN_ON in arch_faults_on_old_pte.  I tested it when I enable the PREEMPT=y
on a ThunderX2 qemu guest.


--
Cheers,
Justin (Jia He)


> 
> >  		/*
> >  		 * This really shouldn't fail, because the page is there
> >  		 * in the page tables. But it might just be unreadable,
> >  		 * in which case we just give up and fill the result with
> >  		 * zeroes.
> >  		 */
> > -		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
> > +		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
> > +			/* Give a warn in case there can be some obscure
> > +			 * use-case
> > +			 */
> > +			WARN_ON_ONCE(1);
> 
> That's more of a question for the mm guys: at this point we do the
> copying with the ptl released; is there anything else that could have
> made the pte old in the meantime? I think unuse_pte() is only called on
> anonymous vmas, so it shouldn't be the case here.
> 
> >  			clear_page(kaddr);
> > +		}
> >  		kunmap_atomic(kaddr);
> >  		flush_dcache_page(dst);
> >  	} else
> > -		copy_user_highpage(dst, src, va, vma);
> > +		copy_user_highpage(dst, src, addr, vma);
> > +
> > +	return true;
> >  }
> 
> --
> Catalin
Catalin Marinas Sept. 24, 2019, 10:33 a.m. UTC | #5
On Tue, Sep 24, 2019 at 06:43:06AM +0000, Justin He (Arm Technology China) wrote:
> Catalin Marinas wrote:
> > On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> > > @@ -2151,21 +2163,53 @@ static inline void cow_user_page(struct page *dst, struct page *src, unsigned lo
> > >  	 * fails, we just zero-fill it. Live with it.
> > >  	 */
> > >  	if (unlikely(!src)) {
> > > -		void *kaddr = kmap_atomic(dst);
> > > -		void __user *uaddr = (void __user *)(va & PAGE_MASK);
> > > +		void *kaddr;
> > > +		pte_t entry;
> > > +		void __user *uaddr = (void __user *)(addr & PAGE_MASK);
> > >
> > > +		/* On architectures with software "accessed" bits, we would
> > > +		 * take a double page fault, so mark it accessed here.
> > > +		 */
[...]
> > > +		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
> > > +			vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr,
> > > +						       &vmf->ptl);
> > > +			if (likely(pte_same(*vmf->pte, vmf->orig_pte))) {
> > > +				entry = pte_mkyoung(vmf->orig_pte);
> > > +				if (ptep_set_access_flags(vma, addr,
> > > +							  vmf->pte, entry, 0))
> > > +					update_mmu_cache(vma, addr, vmf->pte);
> > > +			} else {
> > > +				/* Other thread has already handled the fault
> > > +				 * and we don't need to do anything. If it's
> > > +				 * not the case, the fault will be triggered
> > > +				 * again on the same address.
> > > +				 */
> > > +				pte_unmap_unlock(vmf->pte, vmf->ptl);
> > > +				return false;
> > > +			}
> > > +			pte_unmap_unlock(vmf->pte, vmf->ptl);
> > > +		}
[...]
> > > +
> > > +		kaddr = kmap_atomic(dst);
> > 
> > Since you moved the kmap_atomic() here, could the above
> > arch_faults_on_old_pte() run in a preemptible context? I suggested to
> > add a WARN_ON in patch 2 to be sure.
> 
> Should I move kmap_atomic back to the original line? Thus, we can make sure
> that arch_faults_on_old_pte() is in the context of preempt_disabled?
> Otherwise, arch_faults_on_old_pte() may cause plenty of warning if I add
> a WARN_ON in arch_faults_on_old_pte.  I tested it when I enable the PREEMPT=y
> on a ThunderX2 qemu guest.

So we have two options here:

1. Change arch_faults_on_old_pte() scope to the whole system rather than
   just the current CPU. You'd have to wire up a new arm64 capability
   for the access flag but this way we don't care whether it's
   preemptible or not.

2. Keep the arch_faults_on_old_pte() per-CPU but make sure we are not
   preempted here. The kmap_atomic() move would do but you'd have to
   kunmap_atomic() before the return.

I think the answer to my question below also has some implication on
which option to pick:

> > >  		/*
> > >  		 * This really shouldn't fail, because the page is there
> > >  		 * in the page tables. But it might just be unreadable,
> > >  		 * in which case we just give up and fill the result with
> > >  		 * zeroes.
> > >  		 */
> > > -		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
> > > +		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
> > > +			/* Give a warn in case there can be some obscure
> > > +			 * use-case
> > > +			 */
> > > +			WARN_ON_ONCE(1);
> > 
> > That's more of a question for the mm guys: at this point we do the
> > copying with the ptl released; is there anything else that could have
> > made the pte old in the meantime? I think unuse_pte() is only called on
> > anonymous vmas, so it shouldn't be the case here.

If we need to hold the ptl here, you could as well have an enclosing
kmap/kunmap_atomic (option 2) with some goto instead of "return false".
Kirill A. Shutemov Sept. 24, 2019, 11:59 a.m. UTC | #6
On Tue, Sep 24, 2019 at 11:33:25AM +0100, Catalin Marinas wrote:
> On Tue, Sep 24, 2019 at 06:43:06AM +0000, Justin He (Arm Technology China) wrote:
> > Catalin Marinas wrote:
> > > On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> > > > @@ -2151,21 +2163,53 @@ static inline void cow_user_page(struct page *dst, struct page *src, unsigned lo
> > > >  	 * fails, we just zero-fill it. Live with it.
> > > >  	 */
> > > >  	if (unlikely(!src)) {
> > > > -		void *kaddr = kmap_atomic(dst);
> > > > -		void __user *uaddr = (void __user *)(va & PAGE_MASK);
> > > > +		void *kaddr;
> > > > +		pte_t entry;
> > > > +		void __user *uaddr = (void __user *)(addr & PAGE_MASK);
> > > >
> > > > +		/* On architectures with software "accessed" bits, we would
> > > > +		 * take a double page fault, so mark it accessed here.
> > > > +		 */
> [...]
> > > > +		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
> > > > +			vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr,
> > > > +						       &vmf->ptl);
> > > > +			if (likely(pte_same(*vmf->pte, vmf->orig_pte))) {
> > > > +				entry = pte_mkyoung(vmf->orig_pte);
> > > > +				if (ptep_set_access_flags(vma, addr,
> > > > +							  vmf->pte, entry, 0))
> > > > +					update_mmu_cache(vma, addr, vmf->pte);
> > > > +			} else {
> > > > +				/* Other thread has already handled the fault
> > > > +				 * and we don't need to do anything. If it's
> > > > +				 * not the case, the fault will be triggered
> > > > +				 * again on the same address.
> > > > +				 */
> > > > +				pte_unmap_unlock(vmf->pte, vmf->ptl);
> > > > +				return false;
> > > > +			}
> > > > +			pte_unmap_unlock(vmf->pte, vmf->ptl);
> > > > +		}
> [...]
> > > > +
> > > > +		kaddr = kmap_atomic(dst);
> > > 
> > > Since you moved the kmap_atomic() here, could the above
> > > arch_faults_on_old_pte() run in a preemptible context? I suggested to
> > > add a WARN_ON in patch 2 to be sure.
> > 
> > Should I move kmap_atomic back to the original line? Thus, we can make sure
> > that arch_faults_on_old_pte() is in the context of preempt_disabled?
> > Otherwise, arch_faults_on_old_pte() may cause plenty of warning if I add
> > a WARN_ON in arch_faults_on_old_pte.  I tested it when I enable the PREEMPT=y
> > on a ThunderX2 qemu guest.
> 
> So we have two options here:
> 
> 1. Change arch_faults_on_old_pte() scope to the whole system rather than
>    just the current CPU. You'd have to wire up a new arm64 capability
>    for the access flag but this way we don't care whether it's
>    preemptible or not.
> 
> 2. Keep the arch_faults_on_old_pte() per-CPU but make sure we are not
>    preempted here. The kmap_atomic() move would do but you'd have to
>    kunmap_atomic() before the return.
> 
> I think the answer to my question below also has some implication on
> which option to pick:
> 
> > > >  		/*
> > > >  		 * This really shouldn't fail, because the page is there
> > > >  		 * in the page tables. But it might just be unreadable,
> > > >  		 * in which case we just give up and fill the result with
> > > >  		 * zeroes.
> > > >  		 */
> > > > -		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
> > > > +		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
> > > > +			/* Give a warn in case there can be some obscure
> > > > +			 * use-case
> > > > +			 */
> > > > +			WARN_ON_ONCE(1);
> > > 
> > > That's more of a question for the mm guys: at this point we do the
> > > copying with the ptl released; is there anything else that could have
> > > made the pte old in the meantime? I think unuse_pte() is only called on
> > > anonymous vmas, so it shouldn't be the case here.
> 
> If we need to hold the ptl here, you could as well have an enclosing
> kmap/kunmap_atomic (option 2) with some goto instead of "return false".

Yeah, look like we need to hold ptl for longer. There is nothing I see
that would prevent clearing young bit under us otherwise.
Jia He Sept. 24, 2019, 3:29 p.m. UTC | #7
Hi Catalin

On 2019/9/24 18:33, Catalin Marinas wrote:
> On Tue, Sep 24, 2019 at 06:43:06AM +0000, Justin He (Arm Technology China) wrote:
>> Catalin Marinas wrote:
>>> On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
>>>> @@ -2151,21 +2163,53 @@ static inline void cow_user_page(struct page *dst, struct page *src, unsigned lo
>>>>   	 * fails, we just zero-fill it. Live with it.
>>>>   	 */
>>>>   	if (unlikely(!src)) {
>>>> -		void *kaddr = kmap_atomic(dst);
>>>> -		void __user *uaddr = (void __user *)(va & PAGE_MASK);
>>>> +		void *kaddr;
>>>> +		pte_t entry;
>>>> +		void __user *uaddr = (void __user *)(addr & PAGE_MASK);
>>>>
>>>> +		/* On architectures with software "accessed" bits, we would
>>>> +		 * take a double page fault, so mark it accessed here.
>>>> +		 */
> [...]
>>>> +		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
>>>> +			vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr,
>>>> +						       &vmf->ptl);
>>>> +			if (likely(pte_same(*vmf->pte, vmf->orig_pte))) {
>>>> +				entry = pte_mkyoung(vmf->orig_pte);
>>>> +				if (ptep_set_access_flags(vma, addr,
>>>> +							  vmf->pte, entry, 0))
>>>> +					update_mmu_cache(vma, addr, vmf->pte);
>>>> +			} else {
>>>> +				/* Other thread has already handled the fault
>>>> +				 * and we don't need to do anything. If it's
>>>> +				 * not the case, the fault will be triggered
>>>> +				 * again on the same address.
>>>> +				 */
>>>> +				pte_unmap_unlock(vmf->pte, vmf->ptl);
>>>> +				return false;
>>>> +			}
>>>> +			pte_unmap_unlock(vmf->pte, vmf->ptl);
>>>> +		}
> [...]
>>>> +
>>>> +		kaddr = kmap_atomic(dst);
>>> Since you moved the kmap_atomic() here, could the above
>>> arch_faults_on_old_pte() run in a preemptible context? I suggested to
>>> add a WARN_ON in patch 2 to be sure.
>> Should I move kmap_atomic back to the original line? Thus, we can make sure
>> that arch_faults_on_old_pte() is in the context of preempt_disabled?
>> Otherwise, arch_faults_on_old_pte() may cause plenty of warning if I add
>> a WARN_ON in arch_faults_on_old_pte.  I tested it when I enable the PREEMPT=y
>> on a ThunderX2 qemu guest.
> So we have two options here:
>
> 1. Change arch_faults_on_old_pte() scope to the whole system rather than
>     just the current CPU. You'd have to wire up a new arm64 capability
>     for the access flag but this way we don't care whether it's
>     preemptible or not.
>
> 2. Keep the arch_faults_on_old_pte() per-CPU but make sure we are not
>     preempted here. The kmap_atomic() move would do but you'd have to
>     kunmap_atomic() before the return.
>
> I think the answer to my question below also has some implication on
> which option to pick:
>
>>>>   		/*
>>>>   		 * This really shouldn't fail, because the page is there
>>>>   		 * in the page tables. But it might just be unreadable,
>>>>   		 * in which case we just give up and fill the result with
>>>>   		 * zeroes.
>>>>   		 */
>>>> -		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
>>>> +		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
>>>> +			/* Give a warn in case there can be some obscure
>>>> +			 * use-case
>>>> +			 */
>>>> +			WARN_ON_ONCE(1);
>>> That's more of a question for the mm guys: at this point we do the
>>> copying with the ptl released; is there anything else that could have
>>> made the pte old in the meantime? I think unuse_pte() is only called on
>>> anonymous vmas, so it shouldn't be the case here.
> If we need to hold the ptl here, you could as well have an enclosing
> kmap/kunmap_atomic (option 2) with some goto instead of "return false".

I am not 100% sure that I understand your suggestion well, so I drafted the patch

here:

Changes: optimize the indentions

      hold the ptl longer


-static inline void cow_user_page(struct page *dst, struct page *src, unsigned 
long va, struct vm_area_struct *vma)
+static inline bool cow_user_page(struct page *dst, struct page *src,
+                 struct vm_fault *vmf)
  {
+    struct vm_area_struct *vma = vmf->vma;
+    struct mm_struct *mm = vma->vm_mm;
+    unsigned long addr = vmf->address;
+    bool ret;
+    pte_t entry;
+    void *kaddr;
+    void __user *uaddr;
+
      debug_dma_assert_idle(src);

+    if (likely(src)) {
+        copy_user_highpage(dst, src, addr, vma);
+        return true;
+    }
+
      /*
       * If the source page was a PFN mapping, we don't have
       * a "struct page" for it. We do a best-effort copy by
       * just copying from the original user address. If that
       * fails, we just zero-fill it. Live with it.
       */
-    if (unlikely(!src)) {
-        void *kaddr = kmap_atomic(dst);
-        void __user *uaddr = (void __user *)(va & PAGE_MASK);
+    kaddr = kmap_atomic(dst);
+    uaddr = (void __user *)(addr & PAGE_MASK);
+
+    /*
+     * On architectures with software "accessed" bits, we would
+     * take a double page fault, so mark it accessed here.
+     */
+    vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr, &vmf->ptl);
+    if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
+        if (!likely(pte_same(*vmf->pte, vmf->orig_pte))) {
+            /*
+             * Other thread has already handled the fault
+             * and we don't need to do anything. If it's
+             * not the case, the fault will be triggered
+             * again on the same address.
+             */
+            ret = false;
+            goto pte_unlock;
+        }
+
+        entry = pte_mkyoung(vmf->orig_pte);
+        if (ptep_set_access_flags(vma, addr, vmf->pte, entry, 0))
+            update_mmu_cache(vma, addr, vmf->pte);
+    }

+    /*
+     * This really shouldn't fail, because the page is there
+     * in the page tables. But it might just be unreadable,
+     * in which case we just give up and fill the result with
+     * zeroes.
+     */
+    if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
          /*
-         * This really shouldn't fail, because the page is there
-         * in the page tables. But it might just be unreadable,
-         * in which case we just give up and fill the result with
-         * zeroes.
+         * Give a warn in case there can be some obscure
+         * use-case
           */
-        if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
-            clear_page(kaddr);
-        kunmap_atomic(kaddr);
-        flush_dcache_page(dst);
-    } else
-        copy_user_highpage(dst, src, va, vma);
+        WARN_ON_ONCE(1);
+        clear_page(kaddr);
+    }
+
+    ret = true;
+
+pte_unlock:
+    pte_unmap_unlock(vmf->pte, vmf->ptl);
+    kunmap_atomic(kaddr);
+    flush_dcache_page(dst);
+
+    return ret;
  }


---
Cheers,
Justin (Jia He)
Catalin Marinas Sept. 24, 2019, 4:35 p.m. UTC | #8
On Tue, Sep 24, 2019 at 11:29:07PM +0800, Jia He wrote:
> On 2019/9/24 18:33, Catalin Marinas wrote:
> > On Tue, Sep 24, 2019 at 06:43:06AM +0000, Justin He (Arm Technology China) wrote:
> > > Catalin Marinas wrote:
> > > > On Sat, Sep 21, 2019 at 09:50:54PM +0800, Jia He wrote:
> > > > >   		/*
> > > > >   		 * This really shouldn't fail, because the page is there
> > > > >   		 * in the page tables. But it might just be unreadable,
> > > > >   		 * in which case we just give up and fill the result with
> > > > >   		 * zeroes.
> > > > >   		 */
> > > > > -		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
> > > > > +		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
> > > > > +			/* Give a warn in case there can be some obscure
> > > > > +			 * use-case
> > > > > +			 */
> > > > > +			WARN_ON_ONCE(1);
> > > > That's more of a question for the mm guys: at this point we do the
> > > > copying with the ptl released; is there anything else that could have
> > > > made the pte old in the meantime? I think unuse_pte() is only called on
> > > > anonymous vmas, so it shouldn't be the case here.
> >
> > If we need to hold the ptl here, you could as well have an enclosing
> > kmap/kunmap_atomic (option 2) with some goto instead of "return false".
> 
> I am not 100% sure that I understand your suggestion well, so I
> drafted the patch

Well, however you think the code is cleaner really.

The copy/paste didn't work well, tabs disappeared (or rather the
Exchange server corrupting outgoing emails) but I'll try to comment
below:

> -static inline void cow_user_page(struct page *dst, struct page *src,
>   unsigned long va, struct vm_area_struct *vma)
> +static inline bool cow_user_page(struct page *dst, struct page *src,
> +                 struct vm_fault *vmf)
>  {
> +    struct vm_area_struct *vma = vmf->vma;
> +    struct mm_struct *mm = vma->vm_mm;
> +    unsigned long addr = vmf->address;
> +    bool ret;
> +    pte_t entry;
> +    void *kaddr;
> +    void __user *uaddr;
> +
>      debug_dma_assert_idle(src);
> 
> +    if (likely(src)) {
> +        copy_user_highpage(dst, src, addr, vma);
> +        return true;
> +    }
> +
>      /*
>       * If the source page was a PFN mapping, we don't have
>       * a "struct page" for it. We do a best-effort copy by
>       * just copying from the original user address. If that
>       * fails, we just zero-fill it. Live with it.
>       */
> -    if (unlikely(!src)) {
> -        void *kaddr = kmap_atomic(dst);
> -        void __user *uaddr = (void __user *)(va & PAGE_MASK);
> +    kaddr = kmap_atomic(dst);
> +    uaddr = (void __user *)(addr & PAGE_MASK);
> +
> +    /*
> +     * On architectures with software "accessed" bits, we would
> +     * take a double page fault, so mark it accessed here.
> +     */
> +    vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr, &vmf->ptl);
> +    if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {

I'd move the pte_offset_map_lock() inside the 'if' block as we don't
want to affect architectures that handle old ptes automatically.

> +        if (!likely(pte_same(*vmf->pte, vmf->orig_pte))) {
> +            /*
> +             * Other thread has already handled the fault
> +             * and we don't need to do anything. If it's
> +             * not the case, the fault will be triggered
> +             * again on the same address.
> +             */
> +            ret = false;
> +            goto pte_unlock;
> +        }
> +
> +        entry = pte_mkyoung(vmf->orig_pte);
> +        if (ptep_set_access_flags(vma, addr, vmf->pte, entry, 0))
> +            update_mmu_cache(vma, addr, vmf->pte);
> +    }
> 
> +    /*
> +     * This really shouldn't fail, because the page is there
> +     * in the page tables. But it might just be unreadable,
> +     * in which case we just give up and fill the result with
> +     * zeroes.
> +     */
> +    if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
>          /*
> -         * This really shouldn't fail, because the page is there
> -         * in the page tables. But it might just be unreadable,
> -         * in which case we just give up and fill the result with
> -         * zeroes.
> +         * Give a warn in case there can be some obscure
> +         * use-case
>           */
> -        if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
> -            clear_page(kaddr);
> -        kunmap_atomic(kaddr);
> -        flush_dcache_page(dst);
> -    } else
> -        copy_user_highpage(dst, src, va, vma);
> +        WARN_ON_ONCE(1);
> +        clear_page(kaddr);
> +    }
> +
> +    ret = true;
> +
> +pte_unlock:
> +    pte_unmap_unlock(vmf->pte, vmf->ptl);

Since the locking would be moved in the 'if' block above, we need
another check here before unlocking:

	if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte))
		pte_unmap_unlock(vmf->pte, vmf->ptl);

You could probably replace the two calls to arch_faults_on_old_pte()
with a single bool variable initialisation, something like:

	force_mkyoung = arch_faults_on_old_pte() &&
		!pte_young(vmf->orig_pte)

and only check for "if (force_mkyoung)" in both cases.

> +    kunmap_atomic(kaddr);
> +    flush_dcache_page(dst);
> +
> +    return ret;
>  }

Patch
diff mbox series

diff --git a/mm/memory.c b/mm/memory.c
index e2bb51b6242e..ae09b070b04d 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -118,6 +118,13 @@  int randomize_va_space __read_mostly =
 					2;
 #endif
 
+#ifndef arch_faults_on_old_pte
+static inline bool arch_faults_on_old_pte(void)
+{
+	return false;
+}
+#endif
+
 static int __init disable_randmaps(char *s)
 {
 	randomize_va_space = 0;
@@ -2140,8 +2147,13 @@  static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd,
 	return same;
 }
 
-static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va, struct vm_area_struct *vma)
+static inline bool cow_user_page(struct page *dst, struct page *src,
+				 struct vm_fault *vmf)
 {
+	struct vm_area_struct *vma = vmf->vma;
+	struct mm_struct *mm = vma->vm_mm;
+	unsigned long addr = vmf->address;
+
 	debug_dma_assert_idle(src);
 
 	/*
@@ -2151,21 +2163,53 @@  static inline void cow_user_page(struct page *dst, struct page *src, unsigned lo
 	 * fails, we just zero-fill it. Live with it.
 	 */
 	if (unlikely(!src)) {
-		void *kaddr = kmap_atomic(dst);
-		void __user *uaddr = (void __user *)(va & PAGE_MASK);
+		void *kaddr;
+		pte_t entry;
+		void __user *uaddr = (void __user *)(addr & PAGE_MASK);
 
+		/* On architectures with software "accessed" bits, we would
+		 * take a double page fault, so mark it accessed here.
+		 */
+		if (arch_faults_on_old_pte() && !pte_young(vmf->orig_pte)) {
+			vmf->pte = pte_offset_map_lock(mm, vmf->pmd, addr,
+						       &vmf->ptl);
+			if (likely(pte_same(*vmf->pte, vmf->orig_pte))) {
+				entry = pte_mkyoung(vmf->orig_pte);
+				if (ptep_set_access_flags(vma, addr,
+							  vmf->pte, entry, 0))
+					update_mmu_cache(vma, addr, vmf->pte);
+			} else {
+				/* Other thread has already handled the fault
+				 * and we don't need to do anything. If it's
+				 * not the case, the fault will be triggered
+				 * again on the same address.
+				 */
+				pte_unmap_unlock(vmf->pte, vmf->ptl);
+				return false;
+			}
+			pte_unmap_unlock(vmf->pte, vmf->ptl);
+		}
+
+		kaddr = kmap_atomic(dst);
 		/*
 		 * This really shouldn't fail, because the page is there
 		 * in the page tables. But it might just be unreadable,
 		 * in which case we just give up and fill the result with
 		 * zeroes.
 		 */
-		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
+		if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) {
+			/* Give a warn in case there can be some obscure
+			 * use-case
+			 */
+			WARN_ON_ONCE(1);
 			clear_page(kaddr);
+		}
 		kunmap_atomic(kaddr);
 		flush_dcache_page(dst);
 	} else
-		copy_user_highpage(dst, src, va, vma);
+		copy_user_highpage(dst, src, addr, vma);
+
+	return true;
 }
 
 static gfp_t __get_fault_gfp_mask(struct vm_area_struct *vma)
@@ -2318,7 +2362,18 @@  static vm_fault_t wp_page_copy(struct vm_fault *vmf)
 				vmf->address);
 		if (!new_page)
 			goto oom;
-		cow_user_page(new_page, old_page, vmf->address, vma);
+
+		if (!cow_user_page(new_page, old_page, vmf)) {
+			/* COW failed, if the fault was solved by other,
+			 * it's fine. If not, userspace would re-fault on
+			 * the same address and we will handle the fault
+			 * from the second attempt.
+			 */
+			put_page(new_page);
+			if (old_page)
+				put_page(old_page);
+			return 0;
+		}
 	}
 
 	if (mem_cgroup_try_charge_delay(new_page, mm, GFP_KERNEL, &memcg, false))