[PATCH] cxl/core: Check existence of cxl_memdev_state in poison test

[PATCH] cxl/core: Check existence of cxl_memdev_state in poison test

[Alison Schofield]

Before now, all CXL memdevs were assumed to have a mailbox-backed
cxl_memdev_state, so poison command checks could safely dereference
the @mds.

With the introduction of Type 2 devices, a memdev may not implement
a mailbox interface, and so there is no associated cxl_memdev_state.
Guard against this case by returning false when @mds is absent.

Signed-off-by: Alison Schofield <alison.schofield@intel.com>
---
 drivers/cxl/core/memdev.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/cxl/core/memdev.c b/drivers/cxl/core/memdev.c
index 273c22118d3d..591425866045 100644
--- a/drivers/cxl/core/memdev.c
+++ b/drivers/cxl/core/memdev.c
@@ -204,6 +204,9 @@ bool cxl_memdev_has_poison_cmd(struct cxl_memdev *cxlmd,
 {
 	struct cxl_memdev_state *mds = to_cxl_memdev_state(cxlmd->cxlds);
 
+	if (!mds)
+		return 0;
+
 	return test_bit(cmd, mds->poison.enabled_cmds);
 }
 
-- 
2.37.3

Re: [PATCH] cxl/core: Check existence of cxl_memdev_state in poison test

[Alejandro Lucero Palau]

Hi Allison,


I think the introduction of type2 is weakening some assumptions of the 
core design, and this check is another example after similar ones added 
for type2. Not sure how this could be addressed properly, so I will look 
at how to make this integration seamless.


By now, as this check is necessary, at least as a sanity check:


Reviewed-by: Alejandro Lucero <alucerop@amd.com>


On 3/31/26 01:50, Alison Schofield wrote:
> Before now, all CXL memdevs were assumed to have a mailbox-backed
> cxl_memdev_state, so poison command checks could safely dereference
> the @mds.
>
> With the introduction of Type 2 devices, a memdev may not implement
> a mailbox interface, and so there is no associated cxl_memdev_state.
> Guard against this case by returning false when @mds is absent.
>
> Signed-off-by: Alison Schofield <alison.schofield@intel.com>
> ---
>   drivers/cxl/core/memdev.c | 3 +++
>   1 file changed, 3 insertions(+)
>
> diff --git a/drivers/cxl/core/memdev.c b/drivers/cxl/core/memdev.c
> index 273c22118d3d..591425866045 100644
> --- a/drivers/cxl/core/memdev.c
> +++ b/drivers/cxl/core/memdev.c
> @@ -204,6 +204,9 @@ bool cxl_memdev_has_poison_cmd(struct cxl_memdev *cxlmd,
>   {
>   	struct cxl_memdev_state *mds = to_cxl_memdev_state(cxlmd->cxlds);
>   
> +	if (!mds)
> +		return 0;
> +
>   	return test_bit(cmd, mds->poison.enabled_cmds);
>   }
>   

Re: [PATCH] cxl/core: Check existence of cxl_memdev_state in poison test

[Dave Jiang]

On 3/30/26 5:50 PM, Alison Schofield wrote:
> Before now, all CXL memdevs were assumed to have a mailbox-backed
> cxl_memdev_state, so poison command checks could safely dereference
> the @mds.
> 
> With the introduction of Type 2 devices, a memdev may not implement
> a mailbox interface, and so there is no associated cxl_memdev_state.
> Guard against this case by returning false when @mds is absent.
> 
> Signed-off-by: Alison Schofield <alison.schofield@intel.com>

Applied to cxl/next
261a02b93d9b

> ---
>  drivers/cxl/core/memdev.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/cxl/core/memdev.c b/drivers/cxl/core/memdev.c
> index 273c22118d3d..591425866045 100644
> --- a/drivers/cxl/core/memdev.c
> +++ b/drivers/cxl/core/memdev.c
> @@ -204,6 +204,9 @@ bool cxl_memdev_has_poison_cmd(struct cxl_memdev *cxlmd,
>  {
>  	struct cxl_memdev_state *mds = to_cxl_memdev_state(cxlmd->cxlds);
>  
> +	if (!mds)
> +		return 0;
> +
>  	return test_bit(cmd, mds->poison.enabled_cmds);
>  }
>