diff mbox series

[2/2] mm: Indicate pfmemalloc pages in compound_head

Message ID 20210416230724.2519198-3-willy@infradead.org (mailing list archive)
State Not Applicable
Delegated to: Netdev Maintainers
Headers show
Series Change struct page layout for page_pool | expand

Checks

Context Check Description
netdev/cover_letter success Link
netdev/fixes_present success Link
netdev/patch_count success Link
netdev/tree_selection success Guessed tree name to be net-next
netdev/subject_prefix success Link
netdev/cc_maintainers warning 12 maintainers not CCed: jgg@ziepe.ca akpm@linux-foundation.org walken@google.com hughd@google.com guoqing.jiang@cloud.ionos.com guro@fb.com peterz@infradead.org will@kernel.org yuzhao@google.com fenghua.yu@intel.com peterx@redhat.com vbabka@suse.cz
netdev/source_inline success Was 0 now: 0
netdev/verify_signedoff success Link
netdev/module_param success Was 0 now: 0
netdev/build_32bit success Errors and warnings before: 23287 this patch: 23287
netdev/kdoc success Errors and warnings before: 4 this patch: 4
netdev/verify_fixes success Link
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 42 lines checked
netdev/build_allmodconfig_warn success Errors and warnings before: 22777 this patch: 22777
netdev/header_inline success Link

Commit Message

Matthew Wilcox (Oracle) April 16, 2021, 11:07 p.m. UTC
The net page_pool wants to use a magic value to identify page pool pages.
The best place to put it is in the first word where it can be clearly a
non-pointer value.  That means shifting dma_addr up to alias with ->index,
which means we need to find another way to indicate page_is_pfmemalloc().
Since page_pool doesn't want to set its magic value on pages which are
pfmemalloc, we can use bit 1 of compound_head to indicate that the page
came from the memory reserves.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 include/linux/mm.h       | 12 +++++++-----
 include/linux/mm_types.h |  7 +++----
 2 files changed, 10 insertions(+), 9 deletions(-)

Comments

David Laight April 17, 2021, 9:13 p.m. UTC | #1
From: Matthew Wilcox (Oracle) <willy@infradead.org>
> Sent: 17 April 2021 00:07
> 
> The net page_pool wants to use a magic value to identify page pool pages.
> The best place to put it is in the first word where it can be clearly a
> non-pointer value.  That means shifting dma_addr up to alias with ->index,
> which means we need to find another way to indicate page_is_pfmemalloc().
> Since page_pool doesn't want to set its magic value on pages which are
> pfmemalloc, we can use bit 1 of compound_head to indicate that the page
> came from the memory reserves.
> 
...
>  		struct {	/* page_pool used by netstack */
> -			/**
> -			 * @dma_addr: might require a 64-bit value on
> -			 * 32-bit architectures.
> -			 */
> +			unsigned long pp_magic;
> +			unsigned long xmi;
> +			unsigned long _pp_mapping_pad;
>  			unsigned long dma_addr[2];
>  		};

You've deleted the comment.

I also think there should be a comment that dma_addr[0]
must be aliased to ->index.
(Or whatever all the exact requirements are.)

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
Matthew Wilcox (Oracle) April 17, 2021, 9:28 p.m. UTC | #2
On Sat, Apr 17, 2021 at 09:13:45PM +0000, David Laight wrote:
> >  		struct {	/* page_pool used by netstack */
> > -			/**
> > -			 * @dma_addr: might require a 64-bit value on
> > -			 * 32-bit architectures.
> > -			 */
> > +			unsigned long pp_magic;
> > +			unsigned long xmi;
> > +			unsigned long _pp_mapping_pad;
> >  			unsigned long dma_addr[2];
> >  		};
> 
> You've deleted the comment.

Yes.  It no longer added any value.  You can see dma_addr now occupies
two words.

> I also think there should be a comment that dma_addr[0]
> must be aliased to ->index.

That's not a requirement.  Moving the pfmemalloc indicator is a
requirement so that we _can_ use index, but there's no requirement about
how index is used.
diff mbox series

Patch

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 8ba434287387..44eab3f6d5ae 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1629,10 +1629,12 @@  struct address_space *page_mapping_file(struct page *page);
 static inline bool page_is_pfmemalloc(const struct page *page)
 {
 	/*
-	 * Page index cannot be this large so this must be
-	 * a pfmemalloc page.
+	 * This is not a tail page; compound_head of a head page is unused
+	 * at return from the page allocator, and will be overwritten
+	 * by callers who do not care whether the page came from the
+	 * reserves.
 	 */
-	return page->index == -1UL;
+	return page->compound_head & 2;
 }
 
 /*
@@ -1641,12 +1643,12 @@  static inline bool page_is_pfmemalloc(const struct page *page)
  */
 static inline void set_page_pfmemalloc(struct page *page)
 {
-	page->index = -1UL;
+	page->compound_head = 2;
 }
 
 static inline void clear_page_pfmemalloc(struct page *page)
 {
-	page->index = 0;
+	page->compound_head = 0;
 }
 
 /*
diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index 5aacc1c10a45..39f7163dcace 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -96,10 +96,9 @@  struct page {
 			unsigned long private;
 		};
 		struct {	/* page_pool used by netstack */
-			/**
-			 * @dma_addr: might require a 64-bit value on
-			 * 32-bit architectures.
-			 */
+			unsigned long pp_magic;
+			unsigned long xmi;
+			unsigned long _pp_mapping_pad;
 			unsigned long dma_addr[2];
 		};
 		struct {	/* slab, slob and slub */