Re: [PATCH 4/5] cxl: Add HDM-DB region creation

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Type-2 (CXL_DEVTYPE_DEVMEM) endpoints lacking Back-Invalidation (BI) support are incorrectly allowed to attach to HDM-DB regions, resulting in an invalid HDM-D configuration within a BI-restricted window.
--

commit 17274d5d8a1db5fc909bbe15ef4736f49f5f5bb1
Author: Davidlohr Bueso <dave@stgolabs.net>

cxl: Add HDM-DB region creation

A region inherits its coherency from the chosen root decoder: HDM-DB if the
root has CXL_DECODER_F_BI, otherwise HDM-H. This commit surfaces the
topology through sysfs and enables HDM-DB capability checking.

> diff --git a/drivers/cxl/core/region.c b/drivers/cxl/core/region.c
> index e50dc716d4e82..3f9ddacfae4ea 100644
> --- a/drivers/cxl/core/region.c
> +++ b/drivers/cxl/core/region.c
[ ... ]
> @@ -2083,10 +2073,31 @@ static int cxl_region_attach(struct cxl_region *cxlr,
>  		return -ENXIO;
>  	}
>  
> -	if (cxled->cxld.target_type != cxlr->type) {
> -		dev_dbg(&cxlr->dev, "%s:%s type mismatch: %d vs %d\n",
> -			dev_name(&cxlmd->dev), dev_name(&cxled->cxld.dev),
> -			cxled->cxld.target_type, cxlr->type);
> +	/*
> +	 * Verify the device and HDM are capable of the region's flavor before
> +	 * proceeding. The endpoint decoder's target_type is then inherited
> +	 * from cxlr->type later in cxl_rr_assign_decoder().
> +	 */
> +	if (cxlr->type == CXL_DECODER_DEVMEM &&
> +	    cxlds->type == CXL_DEVTYPE_CLASSMEM && !cxlds->bi) {

[Severity: High]
Does this capability check incorrectly allow Type-2 devices without back
invalidation support to bypass the validation?

Since this check requires cxlds->type == CXL_DEVTYPE_CLASSMEM (Type-3
devices), Type-2 devices bypass it. If they attach to an HDM-DB region,
cxld_set_type() in drivers/cxl/core/hdm.c will configure the decoder as
HDM-D (BI=0) because of this logic:

cxld_set_type()
    bi = bi && cxlds->bi;

Will this violate the host bridge's coherency expectations by allowing an
HDM-D configuration within a BI-restricted window?

> +		dev_err(&cxlr->dev, "%s:%s BI not enabled on device\n",
> +			dev_name(&cxlmd->dev), dev_name(&cxled->cxld.dev));
> +		return -ENXIO;
> +	}
> +
> +	cxlhdm = dev_get_drvdata(&ep_port->dev);
> +	if (!cxlhdm)
> +		return -ENXIO;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260605003329.2584012-1-dave@stgolabs.net?part=4