[v5,1/5] mm: introduce debug_pagealloc_{map,unmap}_pages() helpers

From: Mike Rapoport <rppt@linux.ibm.com>

When CONFIG_DEBUG_PAGEALLOC is enabled, it unmaps pages from the kernel
direct mapping after free_pages(). The pages than need to be mapped back
before they could be used. Theese mapping operations use
__kernel_map_pages() guarded with with debug_pagealloc_enabled().

The only place that calls __kernel_map_pages() without checking whether
DEBUG_PAGEALLOC is enabled is the hibernation code that presumes
availability of this function when ARCH_HAS_SET_DIRECT_MAP is set.
Still, on arm64, __kernel_map_pages() will bail out when DEBUG_PAGEALLOC is
not enabled but set_direct_map_invalid_noflush() may render some pages not
present in the direct map and hibernation code won't be able to save such
pages.

To make page allocation debugging and hibernation interaction more robust,
the dependency on DEBUG_PAGEALLOC or ARCH_HAS_SET_DIRECT_MAP has to be made
more explicit.

Start with combining the guard condition and the call to
__kernel_map_pages() into debug_pagealloc_map_pages() and
debug_pagealloc_unmap_pages() functions to emphasize that
__kernel_map_pages() should not be called without DEBUG_PAGEALLOC and use
these new functions to map/unmap pages when page allocation debugging is
enabled.

Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
---
 include/linux/mm.h  | 15 +++++++++++++++
 mm/memory_hotplug.c |  3 +--
 mm/page_alloc.c     |  6 ++----
 mm/slab.c           | 16 +++++++---------
 4 files changed, 25 insertions(+), 15 deletions(-)

On 11/8/20 7:57 AM, Mike Rapoport wrote:
> --- a/mm/slab.c
> +++ b/mm/slab.c
> @@ -1428,21 +1428,19 @@ static bool is_debug_pagealloc_cache(struct kmem_cache *cachep)
>   	return false;
>   }
>   
> -#ifdef CONFIG_DEBUG_PAGEALLOC
>   static void slab_kernel_map(struct kmem_cache *cachep, void *objp, int map)
>   {
>   	if (!is_debug_pagealloc_cache(cachep))
>   		return;

Hmm, I didn't notice earlier, sorry.
The is_debug_pagealloc_cache() above includes a debug_pagealloc_enabled_static() 
check, so it should be fine to use
__kernel_map_pages() directly below. Otherwise we generate two static key checks 
for the same key needlessly.

>   
> -	kernel_map_pages(virt_to_page(objp), cachep->size / PAGE_SIZE, map);
> +	if (map)
> +		debug_pagealloc_map_pages(virt_to_page(objp),
> +					  cachep->size / PAGE_SIZE);
> +	else
> +		debug_pagealloc_unmap_pages(virt_to_page(objp),
> +					    cachep->size / PAGE_SIZE);
>   }
>   
> -#else
> -static inline void slab_kernel_map(struct kmem_cache *cachep, void *objp,
> -				int map) {}
> -
> -#endif
> -
>   static void poison_obj(struct kmem_cache *cachep, void *addr, unsigned char val)
>   {
>   	int size = cachep->object_size;
> @@ -2062,7 +2060,7 @@ int __kmem_cache_create(struct kmem_cache *cachep, slab_flags_t flags)
>   
>   #if DEBUG
>   	/*
> -	 * If we're going to use the generic kernel_map_pages()
> +	 * If we're going to use the generic debug_pagealloc_map_pages()
>   	 * poisoning, then it's going to smash the contents of
>   	 * the redzone and userword anyhow, so switch them off.
>   	 */
>

On Mon, Nov 09, 2020 at 12:33:46PM +0100, Vlastimil Babka wrote:
> On 11/8/20 7:57 AM, Mike Rapoport wrote:
> > --- a/mm/slab.c
> > +++ b/mm/slab.c
> > @@ -1428,21 +1428,19 @@ static bool is_debug_pagealloc_cache(struct kmem_cache *cachep)
> >   	return false;
> >   }
> > -#ifdef CONFIG_DEBUG_PAGEALLOC
> >   static void slab_kernel_map(struct kmem_cache *cachep, void *objp, int map)
> >   {
> >   	if (!is_debug_pagealloc_cache(cachep))
> >   		return;
> 
> Hmm, I didn't notice earlier, sorry.
> The is_debug_pagealloc_cache() above includes a
> debug_pagealloc_enabled_static() check, so it should be fine to use
> __kernel_map_pages() directly below. Otherwise we generate two static key
> checks for the same key needlessly.

Ok, I'll revert slab changes.

> > -	kernel_map_pages(virt_to_page(objp), cachep->size / PAGE_SIZE, map);
> > +	if (map)
> > +		debug_pagealloc_map_pages(virt_to_page(objp),
> > +					  cachep->size / PAGE_SIZE);
> > +	else
> > +		debug_pagealloc_unmap_pages(virt_to_page(objp),
> > +					    cachep->size / PAGE_SIZE);
> >   }
> > -#else
> > -static inline void slab_kernel_map(struct kmem_cache *cachep, void *objp,
> > -				int map) {}
> > -
> > -#endif
> > -
> >   static void poison_obj(struct kmem_cache *cachep, void *addr, unsigned char val)
> >   {
> >   	int size = cachep->object_size;
> > @@ -2062,7 +2060,7 @@ int __kmem_cache_create(struct kmem_cache *cachep, slab_flags_t flags)
> >   #if DEBUG
> >   	/*
> > -	 * If we're going to use the generic kernel_map_pages()
> > +	 * If we're going to use the generic debug_pagealloc_map_pages()
> >   	 * poisoning, then it's going to smash the contents of
> >   	 * the redzone and userword anyhow, so switch them off.
> >   	 */
> > 
>