merge: sync Rongsong's scheduler-framework into chatbot

Full merge of origin/scheduler-framework (scheduler-core restructure,
failover-adoption, WAL_JOB_DISPATCH, apiserver backend health/heartbeat,
build-time versioning, bug-1673/1676/1678/1679/1680).

Conflict resolution:
- scheduler: adopt Rongsong's scheduler-core architecture (driver loop,
  queue project/picker fields, resource bug-1569 fix)
- worker/controller_connector.c: keep Rongsong's reconnect/rejoin
  structure, PORT our bug-1509 resend_job_status onto the rejoin success
  path (his reconnect didn't re-report job state)
- apiserver backend: take Rongsong's health/heartbeat (supersedes our
  e505066 circuit breaker)
- group.c: keep our bug-1513 comment
- misc: take theirs for additive fixes (go_build versioning,
  test_group.c, persist.c RUSAGE_UPDATE, FilterRows bug-1680)

NOTE: our scheduler fixes (bug-1494 family dispatch-fail requeue,
PEND-reason classification) are NOT in Rongsong's branch — needs
verification/porting onto scheduler-core (separate follow-up).

Co-Authored-By: Claude <noreply@anthropic.com>

src/frontend/apiserver/backend/handler.go

@@ -4,6 +4,7 @@ import (
 	"apiserver/models"
 	"apiserver/store"
 	logger "apiserver/utils"
+	"sync"
 
 	"go.uber.org/zap"
 )
@@ -11,6 +12,66 @@ import (
 // DefaultHandler implements NotificationHandler by writing to stores.
 type DefaultHandler struct{}
 
+// defaultHostConfigName is the cluster-wide config sentinel in hosts.yaml
+// (never a real worker) whose fields worker.c's apply_join_state()/
+// find_matched_hostconfig() already falls back to on the C side for any
+// worker with no host-specific hosts.yaml entry of its own.
+const defaultHostConfigName = "default"
+
+// bug-1673: a dynamically-joined worker with no hosts.yaml entry under its
+// own literal hostname never receives its own OnHostConfigUpdated call, so
+// MaxSlot/HostModel/HostType/DisptWindow stayed at Go's zero-value forever
+// -- decoupled from the C-side scheduling path, which already resolves the
+// real value via the "default" wildcard entry. These package-level vars
+// cache that entry's fields plus which hosts have their OWN explicit
+// (non-default) config, so OnHostUpdated can apply the same fallback for
+// display purposes that worker.c already applies for real dispatch.
+var (
+	hostConfigMu            sync.Mutex
+	defaultHostConfig       *models.HostInfo
+	hostsWithExplicitConfig = map[string]bool{}
+)
+
+// resetHostConfigCache clears the cached "default" pseudo-host config and
+// explicit-config tracking. Test-only: called from this package's _test.go
+// store-reset helpers, so bug-1673 fallback state doesn't leak between
+// tests the way store.InitStores() already prevents for the stores
+// themselves.
+func resetHostConfigCache() {
+	hostConfigMu.Lock()
+	defaultHostConfig = nil
+	hostsWithExplicitConfig = map[string]bool{}
+	hostConfigMu.Unlock()
+}
+
+// applyDefaultHostConfigFallback copies the cached "default" pseudo-host's
+// config-sourced fields onto host, if one has been observed yet. No-op
+// otherwise (nothing to fall back to).
+//
+// A negative MaxSlot is HOSTCONFIG_MXJ_AUTO_DETECT (bug-1556): "no explicit
+// cap, use this host's real CPU count" -- the same resolution worker.c's
+// apply_join_state()/find_matched_hostconfig() already performs for actual
+// dispatch ("if (nslots<0) nslots = maxcpus"). Resolving it here too means
+// this display fallback shows the host's real usable capacity (matching
+// what dispatch will actually honor) instead of a bare "-1" that would
+// leave the reporter's original "is ncpus equal slots?" question (gitea
+// #40) just as unanswered as the "0" it replaces.
+func applyDefaultHostConfigFallback(host *models.HostInfo) {
+	hostConfigMu.Lock()
+	def := defaultHostConfig
+	hostConfigMu.Unlock()
+	if def == nil {
+		return
+	}
+	host.HostModel = def.HostModel
+	host.HostType = def.HostType
+	host.MaxSlot = def.MaxSlot
+	if host.MaxSlot < 0 && host.NumCPUs > 0 {
+		host.MaxSlot = host.NumCPUs
+	}
+	host.DisptWindow = def.DisptWindow
+}
+
 func (h *DefaultHandler) OnJobCreated(job *models.JobInfo) {
 	store.Jobs.Put(uint64(job.Unique_JobID), *job)
 	DefaultBroker.Publish(Event{Type: EventJobCreated, JobID: uint64(job.Unique_JobID), Status: models.JobStateStr(job.JobStatus), Job: job})
@@ -46,7 +107,10 @@ func (h *DefaultHandler) OnQueueUpdated(queue *models.QueueInfo) {
 
 // OnHostUpdated merges worker-sourced fields into an existing host entry.
 // Config-sourced fields (HostModel, HostType, MaxSlot, DisptWindow) are
-// preserved from the prior entry if one exists.
+// preserved from the prior entry if one exists. If this is the host's
+// first-ever registration (no prior entry), fall back to the cached
+// "default" pseudo-host's config (bug-1673) rather than leaving those
+// fields at Go's zero-value -- mirrors the C-side wildcard fallback.
 func (h *DefaultHandler) OnHostUpdated(incoming *models.HostInfo) {
 	merged := *incoming
 	if existing, ok := store.Hosts.Get(merged.HostName); ok {
@@ -54,6 +118,8 @@ func (h *DefaultHandler) OnHostUpdated(incoming *models.HostInfo) {
 		merged.HostType = existing.HostType
 		merged.MaxSlot = existing.MaxSlot
 		merged.DisptWindow = existing.DisptWindow
+	} else {
+		applyDefaultHostConfigFallback(&merged)
 	}
 	store.Hosts.Put(merged.HostName, merged)
 }
@@ -61,7 +127,35 @@ func (h *DefaultHandler) OnHostUpdated(incoming *models.HostInfo) {
 // OnHostConfigUpdated applies config-sourced fields onto an existing worker-registered host.
 // Hosts not yet registered by a worker are skipped — config entries like "default" (which sets
 // cluster-wide defaults) or hosts whose daemons haven't started must not appear in the host list.
+//
+// The "default" pseudo-host entry itself is cached (bug-1673) rather than
+// simply discarded, then retroactively backfilled onto any already-
+// registered host with no explicit config of its own -- "default" can
+// arrive either before or after a given worker's own OnHostUpdated call.
 func (h *DefaultHandler) OnHostConfigUpdated(incoming *models.HostInfo) {
+	if incoming.HostName == defaultHostConfigName {
+		hostConfigMu.Lock()
+		cp := *incoming
+		defaultHostConfig = &cp
+		hostConfigMu.Unlock()
+
+		store.Hosts.Range(func(name string, host models.HostInfo) bool {
+			hostConfigMu.Lock()
+			explicit := hostsWithExplicitConfig[name]
+			hostConfigMu.Unlock()
+			if !explicit {
+				applyDefaultHostConfigFallback(&host)
+				store.Hosts.Put(name, host)
+			}
+			return true
+		})
+		return
+	}
+
+	hostConfigMu.Lock()
+	hostsWithExplicitConfig[incoming.HostName] = true
+	hostConfigMu.Unlock()
+
 	existing, ok := store.Hosts.Get(incoming.HostName)
 	if !ok {
 		return
@@ -96,6 +190,36 @@ func (h *DefaultHandler) OnResourceUpdated(res *models.ResourceDefinitionInfo) {
 func (h *DefaultHandler) OnResourceLocationUpdated(res *models.ResourceDistributionInfo) {
 	key := res.ResourceName + ":" + res.Location
 	store.ResourceLocations.Put(key, *res)
+
+	// resource-value-verification task 4.2 (Requirements 5.1, 5.2, 5.4): trace
+	// the stored value at its point of application, mirroring the scheduler's
+	// sibling trace log added in task 4.1 (core_events.c). One typed "value"
+	// field carrying whichever of StringValue/NumericValue/BooleanValue is
+	// live for res.ValueType, so a downstream grep sees exactly what was
+	// stored, not every possibly-zero field on the struct.
+	//
+	// DO NOT change the "resource-location stored" message string, or the
+	// resource/host/value field names below, without also updating
+	// query_apiserver_resvals() in
+	// .claude/skills/xce-test-resources/run-resource-verify.sh (design.md's
+	// own Revalidation Triggers list this exact pair) -- that script greps
+	// this exact message and parses zap's console-encoder trailing JSON
+	// block for these field names; an unrelated logging cleanup here
+	// silently turns every future verification run into a false
+	// UNREACHABLE/timeout, indistinguishable from a real regression.
+	var valueField zap.Field
+	switch res.ValueType {
+	case int(models.STRING):
+		valueField = zap.String("value", res.StringValue)
+	case int(models.BOOLEAN):
+		valueField = zap.Bool("value", res.BooleanValue)
+	default:
+		valueField = zap.Float64("value", res.NumericValue)
+	}
+	logger.Debug("resource-location stored",
+		zap.String("resource", res.ResourceName),
+		zap.String("host", res.Location),
+		valueField)
 }
 
 // dispatchJobArray delivers a locally-expanded compact array push to the