Re: [PATCH v3 3/3] powerpc/numa: Support coregroup on PowerNV

* Shrikanth Hegde [off-list ref] [2026-09-04 15:34:50]:

On 9/2/26 6:05 PM, Srikar Dronamraju wrote:

quoted

Coregroup support on powerpc has so far been limited to PowerVM LPARs.
However, PowerNV can also support coregroups when firmware exposes the
required coregroup information through the associativity hierarchy.

Existing firmware does expose this info already?

quoted

Detect coregroup support by checking whether primary_domain_index is the
penultimate domain in the CPU node's ibm,associativity property. On
PowerNV, a non-penultimate primary_domain_index indicates that firmware
provides an additional level for coregroup information.

This keeps the logic compatible with PowerVM systems, where
primary_domain_index is likewise not the penultimate associativity
domain.

Could you please put the ibm,associativity on this powernv? as well
PowerVM's so that one understands it better?

quoted

Signed-off-by: Srikar Dronamraju <redacted>
---
Changelog from v2:
- Handle comments from Ritesh (one hunk needed to be moved from patch 2 to
   patch 3 to build correctly)

Changelog from v1: https://lkml.kernel.org/r/20260524010017.140408-1-srikar@linux.ibm.com
- Handle comments from Christophe Leroy; make code more flat

  arch/powerpc/mm/numa.c | 58 ++++++++++++++++++++++++++++++++++--------
  1 file changed, 48 insertions(+), 10 deletions(-)

diff --git a/arch/powerpc/mm/numa.c b/arch/powerpc/mm/numa.c
index 5f326b005a2a..e97b624203ea 100644
--- a/arch/powerpc/mm/numa.c
+++ b/arch/powerpc/mm/numa.c

@@ -889,12 +889,32 @@ static int __init numa_setup_drmem_lmb(struct drmem_lmb *lmb,
  	return 0;
  }
+/*
+ * If hierarchy extends beyond primary_domain_index + 1, then next
+ * level corresponds to coregroup.
+ */
+static int detect_and_enable_coregroup(const __be32 *associativity, int index)
+{
+	if (!associativity || index == -1)
+		goto out;
+
+	index = of_read_number(associativity, 1);
+
+	if (index > primary_domain_index + 1) {
+		coregroup_enabled = 1;
+		return index;
+	}
+out:
+	coregroup_enabled = 0;
+	return -1;
+}
+
  static int __init parse_numa_properties(void)
  {
  	struct device_node *memory, *pci;
-	int default_nid = 0;
-	unsigned long i;
+	int default_nid = 0, index = 0;
  	const __be32 *associativity;
+	unsigned long i;
  	if (numa_enabled == 0) {
  		pr_warn("disabled by user\n");

@@ -927,7 +947,6 @@ static int __init parse_numa_properties(void)
  	 */
  	for_each_present_cpu(i) {
  		__be32 vphn_assoc[VPHN_ASSOC_BUFSIZE];
-		struct device_node *cpu;
  		int nid = NUMA_NO_NODE;
  		memset(vphn_assoc, 0, VPHN_ASSOC_BUFSIZE * sizeof(__be32));

@@ -935,7 +954,9 @@ static int __init parse_numa_properties(void)
  		if (__vphn_get_associativity(i, vphn_assoc) == 0) {
  			nid = associativity_to_nid(vphn_assoc);
  			initialize_form1_numa_distance(vphn_assoc);
+			index = detect_and_enable_coregroup(vphn_assoc, index);

nit: I don't see return value being used.

index is getting used within the loop. Do note, we are passing index as a
parameter to detect_and_enable_coregroup().
For example, index is set to -1, in the first iteration, then
in the next iterations,  detect_and_enable_coregroup will not retry.
So the return value gets used.

Other than that, rest looks good to me.
Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com>

quoted

  		} else {

-- 
Thanks and Regards
Srikar Dronamraju