[PATCH v2 0/2] Add bounds check for Hotplugged memory

[PATCH v2 0/2] Add bounds check for Hotplugged memory

[Alastair D'Silva]

From: Alastair D'Silva <alastair@d-silva.org>

This series adds bounds checks for hotplugged memory, ensuring that
it is within the physically addressable range (for platforms that
define MAX_(POSSIBLE_)PHYSMEM_BITS.

This allows for early failure, rather than attempting to access
bogus section numbers.

Changelog:
 V2:
   - Don't use MAX_POSSIBLE_PHYSMEM_BITS as it's wider that what
     may be available

Alastair D'Silva (2):
  memory_hotplug: Add a bounds check to check_hotplug_memory_range()
  mm: Add a bounds check in devm_memremap_pages()

 include/linux/memory_hotplug.h |  1 +
 mm/memory_hotplug.c            | 13 ++++++++++++-
 mm/memremap.c                  |  8 ++++++++
 3 files changed, 21 insertions(+), 1 deletion(-)

-- 
2.21.0

[PATCH v2 1/2] memory_hotplug: Add a bounds check to check_hotplug_memory_range()

[Alastair D'Silva]

From: Alastair D'Silva <alastair@d-silva.org>

On PowerPC, the address ranges allocated to OpenCAPI LPC memory
are allocated from firmware. These address ranges may be higher
than what older kernels permit, as we increased the maximum
permissable address in commit 4ffe713b7587
("powerpc/mm: Increase the max addressable memory to 2PB"). It is
possible that the addressable range may change again in the
future.

In this scenario, we end up with a bogus section returned from
__section_nr (see the discussion on the thread "mm: Trigger bug on
if a section is not found in __section_nr").

Adding a check here means that we fail early and have an
opportunity to handle the error gracefully, rather than rumbling
on and potentially accessing an incorrect section.

Further discussion is also on the thread ("powerpc: Perform a bounds
check in arch_add_memory").

Signed-off-by: Alastair D'Silva <alastair@d-silva.org>
---
 include/linux/memory_hotplug.h |  1 +
 mm/memory_hotplug.c            | 13 ++++++++++++-
 2 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h
index f46ea71b4ffd..bc477e98a310 100644
--- a/include/linux/memory_hotplug.h
+++ b/include/linux/memory_hotplug.h
@@ -110,6 +110,7 @@ extern void __online_page_increment_counters(struct page *page);
 extern void __online_page_free(struct page *page);
 
 extern int try_online_node(int nid);
+int check_hotplug_memory_addressable(u64 start, u64 size);
 
 extern int arch_add_memory(int nid, u64 start, u64 size,
 			struct mhp_restrictions *restrictions);
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index c73f09913165..02cb9a74f561 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -1030,6 +1030,17 @@ int try_online_node(int nid)
 	return ret;
 }
 
+int check_hotplug_memory_addressable(u64 start, u64 size)
+{
+#ifdef MAX_PHYSMEM_BITS
+	if ((start + size - 1) >> MAX_PHYSMEM_BITS)
+		return -E2BIG;
+#endif
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(check_hotplug_memory_addressable);
+
 static int check_hotplug_memory_range(u64 start, u64 size)
 {
 	/* memory range must be block size aligned */
@@ -1040,7 +1051,7 @@ static int check_hotplug_memory_range(u64 start, u64 size)
 		return -EINVAL;
 	}
 
-	return 0;
+	return check_hotplug_memory_addressable(start, size);
 }
 
 static int online_memory_block(struct memory_block *mem, void *arg)
-- 
2.21.0

[PATCH v2 2/2] mm: Add a bounds check in devm_memremap_pages()

[Alastair D'Silva]

From: Alastair D'Silva <alastair@d-silva.org>

The call to check_hotplug_memory_addressable() validates that the memory
is fully addressable.

Without this call, it is possible that we may remap pages that is
not physically addressable, resulting in bogus section numbers
being returned from __section_nr().

Signed-off-by: Alastair D'Silva <alastair@d-silva.org>
---
 mm/memremap.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/mm/memremap.c b/mm/memremap.c
index 86432650f829..fd00993caa3e 100644
--- a/mm/memremap.c
+++ b/mm/memremap.c
@@ -269,6 +269,13 @@ void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
 
 	mem_hotplug_begin();
 
+	error = check_hotplug_memory_addressable(res->start,
+						 resource_size(res));
+	if (error) {
+		mem_hotplug_done();
+		goto err_checkrange;
+	}
+
 	/*
 	 * For device private memory we call add_pages() as we only need to
 	 * allocate and initialize struct page for the device memory. More-
@@ -324,6 +331,7 @@ void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
 
  err_add_memory:
 	kasan_remove_zero_shadow(__va(res->start), resource_size(res));
+ err_checkrange:
  err_kasan:
 	untrack_pfn(NULL, PHYS_PFN(res->start), resource_size(res));
  err_pfn_remap:
-- 
2.21.0

Re: [PATCH v2 2/2] mm: Add a bounds check in devm_memremap_pages()

[David Hildenbrand]

On 16.09.19 08:05, Alastair D'Silva wrote:
> From: Alastair D'Silva <alastair@d-silva.org>
> 
> The call to check_hotplug_memory_addressable() validates that the memory
> is fully addressable.
> 
> Without this call, it is possible that we may remap pages that is
> not physically addressable, resulting in bogus section numbers
> being returned from __section_nr().
> 
> Signed-off-by: Alastair D'Silva <alastair@d-silva.org>
> ---
>  mm/memremap.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/mm/memremap.c b/mm/memremap.c
> index 86432650f829..fd00993caa3e 100644
> --- a/mm/memremap.c
> +++ b/mm/memremap.c
> @@ -269,6 +269,13 @@ void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
>  
>  	mem_hotplug_begin();
>  
> +	error = check_hotplug_memory_addressable(res->start,
> +						 resource_size(res));
> +	if (error) {
> +		mem_hotplug_done();
> +		goto err_checkrange;
> +	}

As I said in reply to v1, please move this out of the memory hotplug
lock. These are static checks.

> +
>  	/*
>  	 * For device private memory we call add_pages() as we only need to
>  	 * allocate and initialize struct page for the device memory. More-
> @@ -324,6 +331,7 @@ void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
>  
>   err_add_memory:
>  	kasan_remove_zero_shadow(__va(res->start), resource_size(res));
> + err_checkrange:
>   err_kasan:
>  	untrack_pfn(NULL, PHYS_PFN(res->start), resource_size(res));
>   err_pfn_remap:
> 


-- 

Thanks,

David / dhildenb