[PATCH][2.4] update ref counts on all allocated pages

[PATCH][2.4] update ref counts on all allocated pages

[Matt Porter]

The following patch sets the ref count on all pages of an
allocation.  This allows an allocation with order>0 to be freed
via individual __free_page() calls within vfree().
  
This is important on non cache coherent processors which
have a pci_alloc_consistent/consistent_alloc() implementation
that involves creating a vmalloc area cache inhibited mapping
of allocated pages.
  
Because of this implementation, the implementation
of pci_free_consistent/consistent_free is done via
a vfree().  Without this patch only the first page
of the allocated memory is freed since successive pages
all have ref counts of zero.

-Matt

===== mm/page_alloc.c 1.56 vs edited =====
--- 1.56/mm/page_alloc.c	Tue Aug 20 04:39:48 2002
+++ edited/mm/page_alloc.c	Tue Nov 26 12:03:38 2002
@@ -203,6 +203,7 @@
 	struct list_head *head, *curr;
 	unsigned long flags;
 	struct page *page;
+	int i;
 
 	spin_lock_irqsave(&zone->lock, flags);
 	do {
@@ -224,7 +225,10 @@
 			page = expand(zone, page, index, order, curr_order, area);
 			spin_unlock_irqrestore(&zone->lock, flags);
 
-			set_page_count(page, 1);
+			/* Set ref count on all pages */
+			for (i = 0; i < (1 << order); i++)
+				set_page_count(page+i, 1);
+
 			if (BAD_RANGE(zone,page))
 				BUG();
 			if (PageLRU(page))

-- 
Matt Porter
porter@cox.net
This is Linux Country. On a quiet night, you can hear Windows reboot.
-- 
Matt Porter
porter@cox.net
This is Linux Country. On a quiet night, you can hear Windows reboot.

Re: [PATCH][2.4] update ref counts on all allocated pages

[David S. Miller]

On Tue, 2002-11-26 at 16:07, Matt Porter wrote:
> The following patch sets the ref count on all pages of an
> allocation.  This allows an allocation with order>0 to be freed
> via individual __free_page() calls within vfree().

If your pci_alloc_consistent implementation is doing something
strang, _it_ should be doing the per-page reference counts.

Don't make every caller eat this overhead.

Re: [PATCH][2.4] update ref counts on all allocated pages

[Matt Porter]

On Wed, Nov 27, 2002 at 02:31:31AM -0800, David S. Miller wrote:
> On Tue, 2002-11-26 at 16:07, Matt Porter wrote:
> > The following patch sets the ref count on all pages of an
> > allocation.  This allows an allocation with order>0 to be freed
> > via individual __free_page() calls within vfree().
> 
> If your pci_alloc_consistent implementation is doing something
> strang, _it_ should be doing the per-page reference counts.
> 
> Don't make every caller eat this overhead.

I'm surprised that's the recommended solution.  It seems pretty
dangerous to have allocated pages in the system where the per-page
ref counts are bogus.

To clarify then, on an order>0 allocation, it is only valid/defined
to free the same order of pages.  Is that a true statement?  If so,
I'll submit a docs patch and adjust our our local implementation.

Thanks,
-- 
Matt Porter
porter@cox.net
This is Linux Country. On a quiet night, you can hear Windows reboot.

Re: [PATCH][2.4] update ref counts on all allocated pages

[David S. Miller]

On Wed, 2002-11-27 at 07:25, Matt Porter wrote:
> To clarify then, on an order>0 allocation, it is only valid/defined
> to free the same order of pages.  Is that a true statement?  If so,
> I'll submit a docs patch and adjust our our local implementation.

Yes, this is correct and it's what everyone does.  The exceptions
mess with the page counts themselves, look at the sparc64
pmd/pte dual-page allocations for example.