[PATCH] resource: harden resource_contains

[PATCH] resource: harden resource_contains

[alucerop]

From: Alejandro Lucero <alucerop@amd.com>

While resource_contains checks for IORESOURCE_UNSET flag for the
resources given, if r1 was initialized with 0 size, the function
returns a false positive. This is so because resource start and
end fields are unsigned with end initialised to size - 1 by current
resource macros.

Make the function to check for the resource size for both resources
since r2 with size 0 should not be considered as valid for the function
purpose.

Signed-off-by: Alejandro Lucero <alucerop@amd.com>
Suggested-by: Alison Schofield <alison.schofield@intel.com>
---
 include/linux/ioport.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/include/linux/ioport.h b/include/linux/ioport.h
index 6e9fb667a1c5..6cb8a8494508 100644
--- a/include/linux/ioport.h
+++ b/include/linux/ioport.h
@@ -264,6 +264,8 @@ static inline unsigned long resource_ext_type(const struct resource *res)
 /* True iff r1 completely contains r2 */
 static inline bool resource_contains(const struct resource *r1, const struct resource *r2)
 {
+	if (!resource_size(r1) || !resource_size(r2))
+		return false;
 	if (resource_type(r1) != resource_type(r2))
 		return false;
 	if (r1->flags & IORESOURCE_UNSET || r2->flags & IORESOURCE_UNSET)
-- 
2.17.1

Re: [PATCH] resource: harden resource_contains

[Alison Schofield]

+ linux-cxl

On Fri, Nov 29, 2024 at 09:15:12AM +0000, alucerop@amd.com wrote:
> From: Alejandro Lucero <alucerop@amd.com>
> 
> While resource_contains checks for IORESOURCE_UNSET flag for the
> resources given, if r1 was initialized with 0 size, the function
> returns a false positive. This is so because resource start and
> end fields are unsigned with end initialised to size - 1 by current
> resource macros.
> 
> Make the function to check for the resource size for both resources
> since r2 with size 0 should not be considered as valid for the function
> purpose.
> 

Hi Alejandro,

Can this patch be included in the CXL Type-2 patchset, as a replacement for:

[PATCH v6 10/28] cxl: harden resource_contains checks to handle zero size resources
https://lore.kernel.org/20241202171222.62595-11-alejandro.lucero-palau@amd.com/T/#u

Keeping it in that set also keeps the discussion history.

DaveJ may be able to take it through the CXL tree with Bjorn's ACK.

--Alison


Reviewed-by: Alison Schofield <alison.schofield@intel.com>


> Signed-off-by: Alejandro Lucero <alucerop@amd.com>
> Suggested-by: Alison Schofield <alison.schofield@intel.com>
> ---
>  include/linux/ioport.h | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/include/linux/ioport.h b/include/linux/ioport.h
> index 6e9fb667a1c5..6cb8a8494508 100644
> --- a/include/linux/ioport.h
> +++ b/include/linux/ioport.h
> @@ -264,6 +264,8 @@ static inline unsigned long resource_ext_type(const struct resource *res)
>  /* True iff r1 completely contains r2 */
>  static inline bool resource_contains(const struct resource *r1, const struct resource *r2)
>  {
> +	if (!resource_size(r1) || !resource_size(r2))
> +		return false;
>  	if (resource_type(r1) != resource_type(r2))
>  		return false;
>  	if (r1->flags & IORESOURCE_UNSET || r2->flags & IORESOURCE_UNSET)
> -- 
> 2.17.1
> 

Re: [PATCH] resource: harden resource_contains

[Bjorn Helgaas]

On Fri, Nov 29, 2024 at 09:15:12AM +0000, alucerop@amd.com wrote:
> From: Alejandro Lucero <alucerop@amd.com>
> 
> While resource_contains checks for IORESOURCE_UNSET flag for the
> resources given, if r1 was initialized with 0 size, the function
> returns a false positive. This is so because resource start and
> end fields are unsigned with end initialised to size - 1 by current
> resource macros.
> 
> Make the function to check for the resource size for both resources
> since r2 with size 0 should not be considered as valid for the function
> purpose.
> 
> Signed-off-by: Alejandro Lucero <alucerop@amd.com>
> Suggested-by: Alison Schofield <alison.schofield@intel.com>

Seems reasonable to me.  FWIW,

Acked-by: Bjorn Helgaas <bhelgaas@google.com>

> ---
>  include/linux/ioport.h | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/include/linux/ioport.h b/include/linux/ioport.h
> index 6e9fb667a1c5..6cb8a8494508 100644
> --- a/include/linux/ioport.h
> +++ b/include/linux/ioport.h
> @@ -264,6 +264,8 @@ static inline unsigned long resource_ext_type(const struct resource *res)
>  /* True iff r1 completely contains r2 */
>  static inline bool resource_contains(const struct resource *r1, const struct resource *r2)
>  {
> +	if (!resource_size(r1) || !resource_size(r2))
> +		return false;
>  	if (resource_type(r1) != resource_type(r2))
>  		return false;
>  	if (r1->flags & IORESOURCE_UNSET || r2->flags & IORESOURCE_UNSET)
> -- 
> 2.17.1
> 

Re: [PATCH] resource: harden resource_contains

[Alejandro Lucero Palau]

On 12/4/24 23:06, Alison Schofield wrote:
> + linux-cxl
>
> On Fri, Nov 29, 2024 at 09:15:12AM +0000, alucerop@amd.com wrote:
>> From: Alejandro Lucero <alucerop@amd.com>
>>
>> While resource_contains checks for IORESOURCE_UNSET flag for the
>> resources given, if r1 was initialized with 0 size, the function
>> returns a false positive. This is so because resource start and
>> end fields are unsigned with end initialised to size - 1 by current
>> resource macros.
>>
>> Make the function to check for the resource size for both resources
>> since r2 with size 0 should not be considered as valid for the function
>> purpose.
>>
> Hi Alejandro,
>
> Can this patch be included in the CXL Type-2 patchset, as a replacement for:
>
> [PATCH v6 10/28] cxl: harden resource_contains checks to handle zero size resources
> https://lore.kernel.org/20241202171222.62595-11-alejandro.lucero-palau@amd.com/T/#u
>
> Keeping it in that set also keeps the discussion history.
>
> DaveJ may be able to take it through the CXL tree with Bjorn's ACK.


Sure. I'll do so in the almost ready v7.

>
> --Alison
>
>
> Reviewed-by: Alison Schofield <alison.schofield@intel.com>


Thank you


>
>> Signed-off-by: Alejandro Lucero <alucerop@amd.com>
>> Suggested-by: Alison Schofield <alison.schofield@intel.com>
>> ---
>>   include/linux/ioport.h | 2 ++
>>   1 file changed, 2 insertions(+)
>>
>> diff --git a/include/linux/ioport.h b/include/linux/ioport.h
>> index 6e9fb667a1c5..6cb8a8494508 100644
>> --- a/include/linux/ioport.h
>> +++ b/include/linux/ioport.h
>> @@ -264,6 +264,8 @@ static inline unsigned long resource_ext_type(const struct resource *res)
>>   /* True iff r1 completely contains r2 */
>>   static inline bool resource_contains(const struct resource *r1, const struct resource *r2)
>>   {
>> +	if (!resource_size(r1) || !resource_size(r2))
>> +		return false;
>>   	if (resource_type(r1) != resource_type(r2))
>>   		return false;
>>   	if (r1->flags & IORESOURCE_UNSET || r2->flags & IORESOURCE_UNSET)
>> -- 
>> 2.17.1
>>

Re: [PATCH] resource: harden resource_contains

[Alejandro Lucero Palau]

On 12/4/24 23:24, Bjorn Helgaas wrote:
> On Fri, Nov 29, 2024 at 09:15:12AM +0000, alucerop@amd.com wrote:
>> From: Alejandro Lucero <alucerop@amd.com>
>>
>> While resource_contains checks for IORESOURCE_UNSET flag for the
>> resources given, if r1 was initialized with 0 size, the function
>> returns a false positive. This is so because resource start and
>> end fields are unsigned with end initialised to size - 1 by current
>> resource macros.
>>
>> Make the function to check for the resource size for both resources
>> since r2 with size 0 should not be considered as valid for the function
>> purpose.
>>
>> Signed-off-by: Alejandro Lucero <alucerop@amd.com>
>> Suggested-by: Alison Schofield <alison.schofield@intel.com>
> Seems reasonable to me.  FWIW,
>
> Acked-by: Bjorn Helgaas <bhelgaas@google.com>


Thanks!


>> ---
>>   include/linux/ioport.h | 2 ++
>>   1 file changed, 2 insertions(+)
>>
>> diff --git a/include/linux/ioport.h b/include/linux/ioport.h
>> index 6e9fb667a1c5..6cb8a8494508 100644
>> --- a/include/linux/ioport.h
>> +++ b/include/linux/ioport.h
>> @@ -264,6 +264,8 @@ static inline unsigned long resource_ext_type(const struct resource *res)
>>   /* True iff r1 completely contains r2 */
>>   static inline bool resource_contains(const struct resource *r1, const struct resource *r2)
>>   {
>> +	if (!resource_size(r1) || !resource_size(r2))
>> +		return false;
>>   	if (resource_type(r1) != resource_type(r2))
>>   		return false;
>>   	if (r1->flags & IORESOURCE_UNSET || r2->flags & IORESOURCE_UNSET)
>> -- 
>> 2.17.1
>>