Vetting/agent/tests/psu.go

package tests

import (
	"context"
	"fmt"
	"os"
	"path/filepath"
	"strconv"
	"strings"
)

// PSU walks /sys/class/hwmon for in*_input (mV) and in*_label to find
// PSU rails. In home-lab hosts the kernel surfaces a handful of named
// rails (12V, 5V, 3V3). No rails → auto-skip. Any rail outside a ±10%
// window of its nominal value → fail.
func PSU(ctx context.Context, d Deps) Outcome {
	rails := scanPSURails()
	if len(rails) == 0 {
		d.Info("PSU: no voltage rails found under /sys/class/hwmon — skipping stage")
		return Outcome{
			Passed:  true,
			Summary: "skipped (no PSU sensors)",
			Extras:  map[string]any{"skipped": true, "reason": "no_hwmon_voltages"},
		}
	}

	var samples []Sample
	problems := []string{}
	for _, rail := range rails {
		samples = append(samples, Sample{Kind: "psu_volt", Key: rail.Label, Value: rail.Volts, Unit: "V"})
		if ok, why := voltageInRange(rail); !ok {
			problems = append(problems, fmt.Sprintf("%s=%.2fV (%s)", rail.Label, rail.Volts, why))
		}
	}
	if d.Sensor != nil {
		_ = d.Sensor(ctx, samples)
	}

	extras := map[string]any{
		"rails":    rails,
		"problems": problems,
	}
	if len(problems) > 0 {
		d.Error("PSU: out-of-range rails: " + strings.Join(problems, ", "))
		return Outcome{
			Passed:  false,
			Message: "PSU rails out of range: " + strings.Join(problems, ", "),
			Summary: fmt.Sprintf("%d rails, %d failing", len(rails), len(problems)),
			Extras:  extras,
		}
	}
	d.Info(fmt.Sprintf("PSU: %d rails within ±10%% nominal", len(rails)))
	return Outcome{
		Passed:  true,
		Summary: fmt.Sprintf("%d rails nominal", len(rails)),
		Extras:  extras,
	}
}

type psuRail struct {
	Label string  `json:"label"`
	Volts float64 `json:"volts"`
}

// scanPSURails walks every hwmon chip looking for in*_input files with
// an accompanying in*_label that mentions a known rail name. Unknown
// labels are skipped rather than flagged — motherboard VRMs report many
// rails that aren't PSU outputs.
func scanPSURails() []psuRail {
	root := "/sys/class/hwmon"
	chips, err := os.ReadDir(root)
	if err != nil {
		return nil
	}
	var out []psuRail
	for _, c := range chips {
		base := filepath.Join(root, c.Name())
		files, err := os.ReadDir(base)
		if err != nil {
			continue
		}
		for _, f := range files {
			name := f.Name()
			if !strings.HasPrefix(name, "in") || !strings.HasSuffix(name, "_input") {
				continue
			}
			n := strings.TrimSuffix(strings.TrimPrefix(name, "in"), "_input")
			labelPath := filepath.Join(base, "in"+n+"_label")
			label := strings.TrimSpace(readFileStr(labelPath))
			if !isPSULabel(label) {
				continue
			}
			raw := strings.TrimSpace(readFileStr(filepath.Join(base, name)))
			mv, err := strconv.Atoi(raw)
			if err != nil {
				continue
			}
			out = append(out, psuRail{Label: label, Volts: float64(mv) / 1000})
		}
	}
	return out
}

// isPSULabel filters labels that look like PSU rails. Keeps a small
// allowlist to avoid flagging CPU VRM rails as PSU failures.
func isPSULabel(label string) bool {
	l := strings.ToLower(label)
	switch {
	case strings.Contains(l, "12v"), strings.Contains(l, "5v"),
		strings.Contains(l, "3.3v"), strings.Contains(l, "3v3"),
		strings.Contains(l, "vccin"):
		return true
	}
	return false
}

// voltageInRange returns (ok, reason). A label like "12V" has a 12.0V
// nominal; we accept ±10%. Unknown labels pass.
func voltageInRange(r psuRail) (bool, string) {
	nom := nominalFor(r.Label)
	if nom == 0 {
		return true, ""
	}
	delta := r.Volts - nom
	if delta < 0 {
		delta = -delta
	}
	if delta/nom > 0.10 {
		return false, fmt.Sprintf("expected ~%.1fV", nom)
	}
	return true, ""
}

func nominalFor(label string) float64 {
	l := strings.ToLower(label)
	switch {
	case strings.Contains(l, "12v"):
		return 12.0
	case strings.Contains(l, "5v"):
		return 5.0
	case strings.Contains(l, "3.3v"), strings.Contains(l, "3v3"):
		return 3.3
	}
	return 0
}

func readFileStr(p string) string {
	b, err := os.ReadFile(p)
	if err != nil {
		return ""
	}
	return string(b)
}