[PATCH 1/4] sched_ext: Extract no-topology CID construction and table finalization

[Zhang Qiao <zhangqiao22@huawei.com>]

scx_cid_init() interleaves table allocation, the four-level topology
walk, the no-topo tail and the shard range finalization in a single
~230-line body. Start decomposing it by extracting the last two stages:

  - scx_cid_build_notopo() packs the possible-but-unassigned cpus into
    shards capped at min(shard_size, SCX_CID_SHARD_MAX_CPUS) and
    records the online cpus that ended up without topology info.
  - scx_cid_finalize_tables() derives shard_ranges[] from
    cid_to_shard[] and publishes nr_shards.

The counters are passed by pointer. Allocation, diagnostics and
cleanup stay in scx_cid_init(). No behavior change.

Signed-off-by: Zhang Qiao <zhangqiao22@huawei.com>
---
 kernel/sched/ext/cid.c | 120 +++++++++++++++++++++++------------------
 1 file changed, 69 insertions(+), 51 deletions(-)

diff --git a/kernel/sched/ext/cid.c b/kernel/sched/ext/cid.c
index a5330b760ecd..b6be8806a7d5 100644
--- a/kernel/sched/ext/cid.c
+++ b/kernel/sched/ext/cid.c
@@ -165,6 +165,71 @@ void scx_cid_retire_tables(void)
 	call_rcu(&tbls->rcu, scx_cid_tables_free_rcufn);
 }
 
+static void scx_cid_build_notopo(struct scx_cid_tables *tbls, u32 shard_size,
+				 u32 *next_cid, s32 *next_shard_idx,
+				 struct cpumask *online_no_topo)
+{
+	u32 notopo_in_shard;
+	s32 notopo_shard_cid, notopo_shard_idx;
+	s32 cpu, cid;
+
+	/*
+	 * No-topo section: any possible cpu without a cid - normally just the
+	 * not-online ones. Pack into shards of up to min(@shard_size,
+	 * SCX_CID_SHARD_MAX_CPUS) cids so that every cid has a valid shard
+	 * assignment and the hard cap holds even with a large @shard_size.
+	 * Collect any currently-online cpus that land here in @online_no_topo
+	 * so we can warn about them at the end.
+	 */
+	notopo_in_shard = min_t(u32, shard_size, SCX_CID_SHARD_MAX_CPUS);
+	notopo_shard_cid = -1;
+	notopo_shard_idx = -1;
+
+	for_each_cpu(cpu, cpu_possible_mask) {
+		if (tbls->cpu_to_cid[cpu] != -1)
+			continue;
+		if (cpu_online(cpu))
+			cpumask_set_cpu(cpu, online_no_topo);
+
+		cid = (*next_cid)++;
+		tbls->cid_to_cpu[cid] = cpu;
+		tbls->cpu_to_cid[cpu] = cid;
+
+		if (notopo_in_shard >= min_t(u32, shard_size, SCX_CID_SHARD_MAX_CPUS)) {
+			notopo_shard_cid = cid;
+			notopo_shard_idx = (*next_shard_idx)++;
+			notopo_in_shard = 0;
+		}
+		notopo_in_shard++;
+
+		tbls->cid_to_shard[cid] = notopo_shard_idx;
+		tbls->topo[cid] = SCX_CID_TOPO_NEG;
+		tbls->topo[cid].shard_cid = notopo_shard_cid;
+		tbls->topo[cid].shard_idx = notopo_shard_idx;
+	}
+}
+
+static void scx_cid_finalize_tables(struct scx_cid_tables *tbls, u32 next_cid,
+				    s32 next_shard_idx)
+{
+	s32 cid;
+
+	/*
+	 * Fill shard_ranges[] from cid_to_shard[]. Shards are contiguous
+	 * cid ranges by construction: base_cid is the first cid landing in a
+	 * shard, nr_cids is the count.
+	 */
+	for (cid = 0; cid < next_cid; cid++) {
+		s32 sidx = tbls->cid_to_shard[cid];
+
+		if (tbls->shard_ranges[sidx].nr_cids == 0)
+			tbls->shard_ranges[sidx].base_cid = cid;
+		tbls->shard_ranges[sidx].nr_cids++;
+	}
+
+	tbls->nr_shards = next_shard_idx;
+}
+
 /**
  * scx_cid_init - build the cid mapping
  * @sch: the scx_sched being initialized; used as the scx_error() target
@@ -193,9 +258,7 @@ s32 scx_cid_init(struct scx_sched *sch)
 	s32 next_cluster_idx = 0;
 	s32 next_shard_idx = 0;
 	u32 shard_size, max_cids;
-	u32 notopo_in_shard;
-	s32 notopo_shard_cid, notopo_shard_idx;
-	s32 cpu, cid, si;
+	s32 cpu, si;
 
 	/* CMASK_MAX_WORDS in cid.bpf.h covers NR_CPUS up to 8192 */
 	BUILD_BUG_ON(NR_CPUS > 8192);
@@ -351,40 +414,8 @@ s32 scx_cid_init(struct scx_sched *sch)
 		}
 	}
 
-	/*
-	 * No-topo section: any possible cpu without a cid - normally just the
-	 * not-online ones. Pack into shards of up to min(@shard_size,
-	 * SCX_CID_SHARD_MAX_CPUS) cids so that every cid has a valid shard
-	 * assignment and the hard cap holds even with a large @shard_size.
-	 * Collect any currently-online cpus that land here in @online_no_topo
-	 * so we can warn about them at the end.
-	 */
-	notopo_in_shard = min_t(u32, shard_size, SCX_CID_SHARD_MAX_CPUS);
-	notopo_shard_cid = -1;
-	notopo_shard_idx = -1;
-
-	for_each_cpu(cpu, cpu_possible_mask) {
-		if (tbls->cpu_to_cid[cpu] != -1)
-			continue;
-		if (cpu_online(cpu))
-			cpumask_set_cpu(cpu, online_no_topo);
-
-		cid = next_cid++;
-		tbls->cid_to_cpu[cid] = cpu;
-		tbls->cpu_to_cid[cpu] = cid;
-
-		if (notopo_in_shard >= min_t(u32, shard_size, SCX_CID_SHARD_MAX_CPUS)) {
-			notopo_shard_cid = cid;
-			notopo_shard_idx = next_shard_idx++;
-			notopo_in_shard = 0;
-		}
-		notopo_in_shard++;
-
-		tbls->cid_to_shard[cid] = notopo_shard_idx;
-		tbls->topo[cid] = SCX_CID_TOPO_NEG;
-		tbls->topo[cid].shard_cid = notopo_shard_cid;
-		tbls->topo[cid].shard_idx = notopo_shard_idx;
-	}
+	scx_cid_build_notopo(tbls, shard_size, &next_cid, &next_shard_idx,
+			     online_no_topo);
 
 	if (!cpumask_empty(llc_fallback))
 		pr_warn("scx_cid: cpus without cacheinfo, using node mask as llc: %*pbl\n",
@@ -393,20 +424,7 @@ s32 scx_cid_init(struct scx_sched *sch)
 		pr_warn("scx_cid: online cpus with no usable topology: %*pbl\n",
 			cpumask_pr_args(online_no_topo));
 
-	/*
-	 * Fill cid_shard_ranges[] from cid_to_shard[]. Shards are contiguous
-	 * cid ranges by construction: base_cid is the first cid landing in a
-	 * shard, nr_cids is the count.
-	 */
-	for (cid = 0; cid < next_cid; cid++) {
-		s32 sidx = tbls->cid_to_shard[cid];
-
-		if (tbls->shard_ranges[sidx].nr_cids == 0)
-			tbls->shard_ranges[sidx].base_cid = cid;
-		tbls->shard_ranges[sidx].nr_cids++;
-	}
-
-	tbls->nr_shards = next_shard_idx;
+	scx_cid_finalize_tables(tbls, next_cid, next_shard_idx);
 	return 0;
 }
 
-- 
2.43.0