[PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[Hillf Danton]

The flag, FAULT_FLAG_ALLOW_RETRY, was introduced by the patch,

	mm: retry page fault when blocking on disk transfer
	commit: d065bd810b6deb67d4897a14bfe21f8eb526ba99

for reducing mmap_sem hold times that are caused by waiting for disk
transfers when accessing file mapped VMAs.

To break COW, handle_mm_fault() is repeated with mmap_sem held, where
the introduced flag could be used again.

The straight way is to add changes in break_ksm(), but the function could be
under write-mode mmap_sem, so it has to be dupilcated.

Signed-off-by: Hillf Danton <dhillf@gmail.com>
---

--- a/mm/ksm.c	Sat Nov 19 16:08:10 2011
+++ b/mm/ksm.c	Sat Nov 19 19:33:49 2011
@@ -394,7 +394,31 @@ static void break_cow(struct rmap_item *
 		goto out;
 	if (!(vma->vm_flags & VM_MERGEABLE) || !vma->anon_vma)
 		goto out;
-	break_ksm(vma, addr);
+	for (;;) {
+		struct page *page;
+		int ret;
+
+		page = follow_page(vma, addr, FOLL_GET);
+		if (IS_ERR_OR_NULL(page))
+			break;
+
+		if (PageKsm(page))
+			ret = handle_mm_fault(mm, vma, addr,
+				FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_WRITE);
+		else
+			ret = VM_FAULT_WRITE;
+
+		put_page(page);
+
+		if (!(ret & (VM_FAULT_WRITE|VM_FAULT_SIGBUS|VM_FAULT_OOM))) {
+			if (ret & VM_FAULT_RETRY)
+				down_read(&mm->mmap_sem);
+		} else {
+			if (ret & VM_FAULT_RETRY)
+				return;
+			break;
+		}
+	}
 out:
 	up_read(&mm->mmap_sem);
 }

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[Hugh Dickins]

On Sat, 19 Nov 2011, Hillf Danton wrote:

> The flag, FAULT_FLAG_ALLOW_RETRY, was introduced by the patch,
> 
> 	mm: retry page fault when blocking on disk transfer
> 	commit: d065bd810b6deb67d4897a14bfe21f8eb526ba99
> 
> for reducing mmap_sem hold times that are caused by waiting for disk
> transfers when accessing file mapped VMAs.
> 
> To break COW, handle_mm_fault() is repeated with mmap_sem held, where
> the introduced flag could be used again.
> 
> The straight way is to add changes in break_ksm(), but the function could be
> under write-mode mmap_sem, so it has to be dupilcated.
> 
> Signed-off-by: Hillf Danton <dhillf@gmail.com>

Thank you for making the patch; but unless I'm mistaken - please
correct me if so - I think it's better to keep break_cow() simple
than add special FAULT_FLAG_ALLOW_RETRY handling there.  Do you
have any evidence that its down_read of mmap_sem is a problem in
some workload?  I sense that you're using it "because it's there".

I'm sceptical on several grounds.

One, break_cow() is itself only called on an "error" path: not
really an error, but when KSM's bet that it can merge pages turns
out to be wrong before it can complete the merge; not a rare case,
but not on the hot path.

Two, break_ksm()'s loop is required for correctness, but it
is a rare case that it actually needs to go round a second time.
The typical case it's needed (am I forgetting a more common one?)
is when userspace access flips a pte bit in between handle_pte_fault()
noting faulting pte, and the chosen fault handler checking pte_same()
before committing to its action.  With the page marked PageKsm, yet
not in the stable tree, even page reclaim is unable to interfere.

Three, FAULT_FLAG_ALLOW_RETRY is acted upon only in lock_page_or_retry(),
which is called only from filemap_fault() (not the case here since we
don't consider file pages for conversion to PageKsm) or do_swap_page();
yet the fault we're provoking would be handled by do_wp_page().

Four, lock_page_or_retry() is called in those places when there's a
possibility that the page is being read in from disk, to drop the
mmap_sem across the slow I/O.  There is no precedent for dropping
mmap_sem here while allocating a new page, nor when pte_same() fails:
in the former case it could only be a win when the system is already
slowed by memory pressure, in the latter case there's little point,
since mmap_sem would be reacquired in a moment.

I think that amounts to a genial Nack!

Hugh

> ---
> 
> --- a/mm/ksm.c	Sat Nov 19 16:08:10 2011
> +++ b/mm/ksm.c	Sat Nov 19 19:33:49 2011
> @@ -394,7 +394,31 @@ static void break_cow(struct rmap_item *
>  		goto out;
>  	if (!(vma->vm_flags & VM_MERGEABLE) || !vma->anon_vma)
>  		goto out;
> -	break_ksm(vma, addr);
> +	for (;;) {
> +		struct page *page;
> +		int ret;
> +
> +		page = follow_page(vma, addr, FOLL_GET);
> +		if (IS_ERR_OR_NULL(page))
> +			break;
> +
> +		if (PageKsm(page))
> +			ret = handle_mm_fault(mm, vma, addr,
> +				FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_WRITE);
> +		else
> +			ret = VM_FAULT_WRITE;
> +
> +		put_page(page);
> +
> +		if (!(ret & (VM_FAULT_WRITE|VM_FAULT_SIGBUS|VM_FAULT_OOM))) {
> +			if (ret & VM_FAULT_RETRY)
> +				down_read(&mm->mmap_sem);
> +		} else {
> +			if (ret & VM_FAULT_RETRY)
> +				return;
> +			break;
> +		}
> +	}
>  out:
>  	up_read(&mm->mmap_sem);
>  }

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[Hillf Danton]

On Mon, Nov 21, 2011 at 12:16 PM, Hugh Dickins <hughd@google.com> wrote:
> On Sat, 19 Nov 2011, Hillf Danton wrote:
>
>> The flag, FAULT_FLAG_ALLOW_RETRY, was introduced by the patch,
>>
>>       mm: retry page fault when blocking on disk transfer
>>       commit: d065bd810b6deb67d4897a14bfe21f8eb526ba99
>>
>> for reducing mmap_sem hold times that are caused by waiting for disk
>> transfers when accessing file mapped VMAs.
>>
>> To break COW, handle_mm_fault() is repeated with mmap_sem held, where
>> the introduced flag could be used again.
>>
>> The straight way is to add changes in break_ksm(), but the function could be
>> under write-mode mmap_sem, so it has to be dupilcated.
>>
>> Signed-off-by: Hillf Danton <dhillf@gmail.com>
>
> Thank you for making the patch; but unless I'm mistaken - please
> correct me if so - I think it's better to keep break_cow() simple
> than add special FAULT_FLAG_ALLOW_RETRY handling there.  Do you
> have any evidence that its down_read of mmap_sem is a problem in
> some workload?  I sense that you're using it "because it's there".
>
> I'm sceptical on several grounds.
>
> One, break_cow() is itself only called on an "error" path: not
> really an error, but when KSM's bet that it can merge pages turns
> out to be wrong before it can complete the merge; not a rare case,
> but not on the hot path.
>
> Two, break_ksm()'s loop is required for correctness, but it
> is a rare case that it actually needs to go round a second time.
> The typical case it's needed (am I forgetting a more common one?)
> is when userspace access flips a pte bit in between handle_pte_fault()
> noting faulting pte, and the chosen fault handler checking pte_same()
> before committing to its action.  With the page marked PageKsm, yet
> not in the stable tree, even page reclaim is unable to interfere.
>
> Three, FAULT_FLAG_ALLOW_RETRY is acted upon only in lock_page_or_retry(),
> which is called only from filemap_fault() (not the case here since we
> don't consider file pages for conversion to PageKsm) or do_swap_page();
> yet the fault we're provoking would be handled by do_wp_page().
>
> Four, lock_page_or_retry() is called in those places when there's a
> possibility that the page is being read in from disk, to drop the
> mmap_sem across the slow I/O.  There is no precedent for dropping
> mmap_sem here while allocating a new page, nor when pte_same() fails:
> in the former case it could only be a win when the system is already
> slowed by memory pressure, in the latter case there's little point,
> since mmap_sem would be reacquired in a moment.
>
> I think that amounts to a genial Nack!
>
Hello Hugh,

After reading your reply and the comments in break_ksm(), if the patch does
not mess up
	"The important thing is to not let VM_MERGEABLE be cleared while any
	 such pages might remain in the area",
and
	"because handle_mm_fault() may back out if there's
	 any difficulty e.g. if pte accessed bit gets updated concurrently",

then if the path in which lock_page_or_retry() is called is not involved,
mmap_sem is not upped, so the patch has nearly same behavior with break_ksm.

And the overhead of the patch, I think, could match break_ksm.

With dozen cases of writers of mmap_sem in the mm directory, the patch looks
more flexible in rare and rare corners.

Best regards
Hillf

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[Hugh Dickins]

On Mon, 21 Nov 2011, Hillf Danton wrote:
> On Mon, Nov 21, 2011 at 12:16 PM, Hugh Dickins <hughd@google.com> wrote:
> 
> After reading your reply and the comments in break_ksm(), if the patch does
> not mess up
> 	"The important thing is to not let VM_MERGEABLE be cleared while any
> 	 such pages might remain in the area",
> and
> 	"because handle_mm_fault() may back out if there's
> 	 any difficulty e.g. if pte accessed bit gets updated concurrently",
> 
> then if the path in which lock_page_or_retry() is called is not involved,
> mmap_sem is not upped, so the patch has nearly same behavior with break_ksm.
> 
> And the overhead of the patch, I think, could match break_ksm.
> 
> With dozen cases of writers of mmap_sem in the mm directory, the patch looks
> more flexible in rare and rare corners.

But what's the point in enlarging the kernel, adding code to make
break_cow() look more complicated, when there's no way in which the
addition can make an improvement?

Adding in a FAULT_FLAG_ALLOW_RETRY flag is not enough for mmap_sem
to be dropped for retry: you'd need a lock_page_or_retry() on the
faulting path and I do not see that here - please point it out to
me if you can see it.

(And I'll be somewhat sceptical if you respond with patches adding
lock_page_or_retry() all over, in order to meet this objection!)

Hugh

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[Hillf Danton]

On Tue, Nov 22, 2011 at 6:23 AM, Hugh Dickins <hughd@google.com> wrote:
> But what's the point in enlarging the kernel, adding code to make
> break_cow() look more complicated, when there's no way in which the
> addition can make an improvement?
>
Hello Hugh

Thanks for correcting me, there is no real point to complicate the function.

Hillf

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] ksm: use FAULT_FLAG_ALLOW_RETRY in breaking COW

[Michel Lespinasse]

On Sat, Nov 19, 2011 at 3:50 AM, Hillf Danton <dhillf@gmail.com> wrote:
> The flag, FAULT_FLAG_ALLOW_RETRY, was introduced by the patch,
>
>        mm: retry page fault when blocking on disk transfer
>        commit: d065bd810b6deb67d4897a14bfe21f8eb526ba99
>
> for reducing mmap_sem hold times that are caused by waiting for disk
> transfers when accessing file mapped VMAs.
>
> To break COW, handle_mm_fault() is repeated with mmap_sem held, where
> the introduced flag could be used again.
>
> The straight way is to add changes in break_ksm(), but the function could be
> under write-mode mmap_sem, so it has to be dupilcated.
>
> Signed-off-by: Hillf Danton <dhillf@gmail.com>

I have to concur with Hugh here - FAULT_FLAG_ALLOW_RETRY was
introduced to avoid holding mmap_sem while we block on a disk read,
but you shouldn't hit this case in the break COW case, so there seems
to be little point in adding the flag there.

-- 
Michel "Walken" Lespinasse
A program is never fully debugged until the last user dies.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>