Re: [tip: x86/urgent] x86/resctrl: Fix SNC detection

[Babu Moger <babu.moger@amd.com>]

Hi Tony,

On 3/4/26 09:46, tip-bot2 for Tony Luck wrote:
> The following commit has been merged into the x86/urgent branch of tip:
> 
> Commit-ID:     59674fc9d0bfd96ce8a776680ee1cf22c28c9ac7
> Gitweb:        https://git.kernel.org/tip/59674fc9d0bfd96ce8a776680ee1cf22c28c9ac7
> Author:        Tony Luck <tony.luck@intel.com>
> AuthorDate:    Tue, 03 Mar 2026 11:55:44 +01:00
> Committer:     Peter Zijlstra <peterz@infradead.org>
> CommitterDate: Wed, 04 Mar 2026 16:35:09 +01:00
> 
> x86/resctrl: Fix SNC detection
> 
> Now that the x86 topology code has a sensible nodes-per-package
> measure, that does not depend on the online status of CPUs, use this
> to divinate the SNC mode.
> 
> Note that when Cluster on Die (CoD) is configured on older systems this
> will also show multiple NUMA nodes per package. Intel Resource Director
> Technology is incomaptible with CoD. Print a warning and do not use the
> fixup MSR_RMID_SNC_CONFIG.
> 
> Signed-off-by: Tony Luck <tony.luck@intel.com>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> Reviewed-by: Ingo Molnar <mingo@kernel.org>
> Tested-by: Zhang Rui <rui.zhang@intel.com>
> Tested-by: Chen Yu <yu.c.chen@intel.com>
> Link: https://patch.msgid.link/aaCxbbgjL6OZ6VMd@agluck-desk3
> Link: https://patch.msgid.link/20260303110100.367976706@infradead.org
> ---
>   arch/x86/kernel/cpu/resctrl/monitor.c | 36 +++-----------------------
>   1 file changed, 5 insertions(+), 31 deletions(-)
> 
> diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c b/arch/x86/kernel/cpu/resctrl/monitor.c
> index e6a1542..9bd87ba 100644
> --- a/arch/x86/kernel/cpu/resctrl/monitor.c
> +++ b/arch/x86/kernel/cpu/resctrl/monitor.c
> @@ -364,7 +364,7 @@ void arch_mon_domain_online(struct rdt_resource *r, struct rdt_l3_mon_domain *d)
>   		msr_clear_bit(MSR_RMID_SNC_CONFIG, 0);
>   }
>   
> -/* CPU models that support MSR_RMID_SNC_CONFIG */
> +/* CPU models that support SNC and MSR_RMID_SNC_CONFIG */
>   static const struct x86_cpu_id snc_cpu_ids[] __initconst = {
>   	X86_MATCH_VFM(INTEL_ICELAKE_X, 0),
>   	X86_MATCH_VFM(INTEL_SAPPHIRERAPIDS_X, 0),
> @@ -375,40 +375,14 @@ static const struct x86_cpu_id snc_cpu_ids[] __initconst = {
>   	{}
>   };
>   
> -/*
> - * There isn't a simple hardware bit that indicates whether a CPU is running
> - * in Sub-NUMA Cluster (SNC) mode. Infer the state by comparing the
> - * number of CPUs sharing the L3 cache with CPU0 to the number of CPUs in
> - * the same NUMA node as CPU0.
> - * It is not possible to accurately determine SNC state if the system is
> - * booted with a maxcpus=N parameter. That distorts the ratio of SNC nodes
> - * to L3 caches. It will be OK if system is booted with hyperthreading
> - * disabled (since this doesn't affect the ratio).
> - */
>   static __init int snc_get_config(void)
>   {
> -	struct cacheinfo *ci = get_cpu_cacheinfo_level(0, RESCTRL_L3_CACHE);
> -	const cpumask_t *node0_cpumask;
> -	int cpus_per_node, cpus_per_l3;
> -	int ret;
> -
> -	if (!x86_match_cpu(snc_cpu_ids) || !ci)
> -		return 1;
> +	int ret = topology_num_nodes_per_package();
>   
> -	cpus_read_lock();
> -	if (num_online_cpus() != num_present_cpus())
> -		pr_warn("Some CPUs offline, SNC detection may be incorrect\n");
> -	cpus_read_unlock();

Saw these messages on AMD system(7.1.0-rc1).

sudo dmesg |grep CoD
[    6.614667] resctrl: CoD enabled system? Resctrl not supported


Might confuse some users. Not a big deal. But we should fix it in next 
version.

Thanks
Babu