fix(nvidia): surface Xid 79/154 GPU bus-fall-off as a physical-reboot-required signal

nvidia-bug-report.sh appends .gz to --output-file when gzip is available,
which silently produced empty nvidia-bug-report.txt in support bundles
(cat looked for the uncompressed name that never existed).

Also: a GPU that falls off the PCIe/NVLink bus (Xid 79) or gets flagged
for Node Reboot Required (Xid 154) mid-SAT-run left every downstream test
failing with generic, unrelated-looking errors (CUDA "unknown error",
"unable to determine device handle") with no indication the GPU needed a
physical power-cycle to recover. Detect these codes from SAT run logs and
surface a plain-English "physical reboot required" message in the task's
failure detail, the persisted component-status DB, a dashboard banner on
the Hardware Summary card, and topology diagram GPU-node severity.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

audit/internal/webui/page_topo.go

@@ -587,6 +587,19 @@ type topoEdge struct {
 }
 
 func renderTopoMainDiagram(hw schema.HardwareSnapshot, exportDir string) string {
+	// A GPU hardware fault that needs a physical reboot (Xid 79 "fallen off
+	// the bus", Xid 154 "Node Reboot Required" — see gpuNeedsPhysicalReboot)
+	// isn't visible in dev.Status: the collector only sets that from PCIe
+	// link-speed checks, not from SAT/kmsg results. Without this, a GPU that
+	// dropped off the bus mid-test still renders green here even though the
+	// Hardware Summary card is showing a critical banner for it.
+	gpuHardwareFault := false
+	if db, err := app.OpenComponentStatusDB(filepath.Join(exportDir, "component-status.json")); err == nil {
+		if _, needsReboot := gpuNeedsPhysicalReboot(matchedRecords(db.All(), nil, []string{"pcie:gpu:"})); needsReboot {
+			gpuHardwareFault = true
+		}
+	}
+
 	socketIdx := buildSocketIndex(hw.CPUs)
 	numCols := len(hw.CPUs)
 	if numCols == 0 {
@@ -810,6 +823,9 @@ func renderTopoMainDiagram(hw schema.HardwareSnapshot, exportDir string) string
 				if kind == "gpu" && crossNUMAWarnBDF[p.bdf] && sev < 2 {
 					sev = 2
 				}
+				if kind == "gpu" && gpuHardwareFault && sev < 3 {
+					sev = 3
+				}
 				tally.add(sev)
 				if i == 0 && p.dev.Model != nil {
 					model = *p.dev.Model