linux-kernel - Re: [PATCH v2] ACPI: NUMA: debug invalid unused PXM value for CFMWs

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Z9MHvp6GA_iGwfg0@aschofie-mobl2.lan>
Date: Thu, 13 Mar 2025 09:28:46 -0700
From: Alison Schofield <alison.schofield@...el.com>
To: Yuquan Wang <wangyuquan1236@...tium.com.cn>
Cc: rafael@...nel.org, lenb@...nel.org, dave@...olabs.net,
	jonathan.cameron@...wei.com, dave.jiang@...el.com,
	vishal.l.verma@...el.com, ira.weiny@...el.com,
	dan.j.williams@...el.com, rrichter@....com, bfaccini@...dia.com,
	rppt@...nel.org, haibo1.xu@...el.com, chenbaozi@...tium.com.cn,
	linux-acpi@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-cxl@...r.kernel.org
Subject: Re: [PATCH v2] ACPI: NUMA: debug invalid unused PXM value for CFMWs

On Thu, Mar 13, 2025 at 02:09:07PM +0800, Yuquan Wang wrote:
> The absence of SRAT would cause the fake_pxm to be -1 and increment
> to 0, then send to acpi_parse_cfmws(). If there exists CXL memory
> ranges that are defined in the CFMWS and not already defined in the
> SRAT, the new node (node0) for the CXL memory would be invalid, as
> node0 is already in "used".
> 
> This utilizes disable_srat() & srat_disabled() to fail CXL init.

Seems like this fixup has drifted from adjusting the fake_pxm to 
shutting down CXL parsing. More below -

> 
> Signed-off-by: Yuquan Wang <wangyuquan1236@...tium.com.cn>
> ---
> 
> Changes in v2:
> - Add disable_srat() when fake_pxm is invalid
> - Add srat_disabled() check in cxl_acpi_probe() and acpi_parse_cfmws()
> 
> 
>  drivers/acpi/numa/srat.c | 10 ++++++++++
>  drivers/cxl/acpi.c       |  4 ++++
>  2 files changed, 14 insertions(+)
> 
> diff --git a/drivers/acpi/numa/srat.c b/drivers/acpi/numa/srat.c
> index 00ac0d7bb8c9..2dac25c9258a 100644
> --- a/drivers/acpi/numa/srat.c
> +++ b/drivers/acpi/numa/srat.c
> @@ -441,6 +441,11 @@ static int __init acpi_parse_cfmws(union acpi_subtable_headers *header,
>  	start = cfmws->base_hpa;
>  	end = cfmws->base_hpa + cfmws->window_size;
>  
> +	if (srat_disabled()) {
> +		pr_err("SRAT is missing or bad while processing CFMWS.\n");
> +		return -EINVAL;
> +	}
> +

This goes too far by shutting down cfmws parsing for lack of SRAT.

>  	/*
>  	 * The SRAT may have already described NUMA details for all,
>  	 * or a portion of, this CFMWS HPA range. Extend the memblks
> @@ -646,6 +651,11 @@ int __init acpi_numa_init(void)
>  		if (node_to_pxm_map[i] > fake_pxm)
>  			fake_pxm = node_to_pxm_map[i];
>  	}
> +
> +	/* Make sure CFMWs fake nodes start at node[1] */
> +	if (fake_pxm < 0)
> +		disable_srat();
> +

How does the code above make sure fake node starts at node[1]?
Would an explicit adjustment like this work?

-       last_real_pxm = fake_pxm;
-       fake_pxm++;
+       fake_pxm = max(fake_pxm, 1);
+       last_real_pxm = fake_pxm--;

>  	last_real_pxm = fake_pxm;
>  	fake_pxm++;
>  	acpi_table_parse_cedt(ACPI_CEDT_TYPE_CFMWS, acpi_parse_cfmws,
> diff --git a/drivers/cxl/acpi.c b/drivers/cxl/acpi.c
> index cb14829bb9be..e75a8ead99f6 100644
> --- a/drivers/cxl/acpi.c
> +++ b/drivers/cxl/acpi.c
> @@ -829,6 +829,10 @@ static int cxl_acpi_probe(struct platform_device *pdev)
>  	if (rc)
>  		return rc;
>  
> +	/* CXL must be in a NUMA system */
> +	if (srat_disabled())
> +		return -EINVAL;
> +
>  	cxl_res = devm_kzalloc(host, sizeof(*cxl_res), GFP_KERNEL);
>  	if (!cxl_res)
>  		return -ENOMEM;
> -- 
> 2.34.1
>