perf(wave1): reduce Config caps, reorder Client/Monitor for cache

Wave 1 changes:
- Reduce CFG_MAX_KEYBINDS 256→128, CFG_MAX_STRLEN 256→64, CFG_MAX_ARGS 16→8
  → saves ~2.5 MB BSS from Config struct
- Add truncation warnings to config_load when keybind/button/scroll limit reached
- Reorder Client struct: HOT fields (type, mon, tags, floating flags, geom) in first
  cache line, COLD fields (wl_listeners) at end — saves cache misses in arrange()
- Reorder LayerSurface struct similarly
- Split Monitor into Monitor/MonitorCold: hot fields (~120B) stay inline, cold layout
  state lazily allocated on first dwindle/master use
- All cold-field accesses through m->cold-> indirection
- cleanupmon frees m->cold
- Zero-warning build with clang -Werror -Wpedantic

include/client.h

@@ -20,23 +20,36 @@ struct wlr_scene_shadow;
 enum { XDGShell, LayerShell, X11 };
 
 typedef struct Client {
-	/* Must keep this field first */
-	unsigned int type; /* XDGShell or X11* */
+	/* Must keep this field first — union-cast discriminant */
+	unsigned int type; /* XDGShell or X11 */
 
+	/* HOT — accessed every frame / arrange / focus / resize */
 	Monitor *mon;
-	struct wlr_scene_tree *scene;
-	struct wlr_scene_rect *border_bg; /* single border rect with clipped content hole */
-	struct wlr_scene_tree *scene_surface;
-	struct wl_list link;
-	struct wl_list flink;
+	uint32_t tags;
+	int isfloating;
+	int isurgent;
+	int isfullscreen;
+	int isscratchpad;
 	struct wlr_box geom;   /* layout-relative, includes border */
 	struct wlr_box prev;   /* layout-relative, includes border */
 	struct wlr_box bounds; /* only width and height are used */
+	unsigned int bw;
+	int corner_radius;
+	struct wlr_scene_shadow *shadow;
+
+	/* WARM — accessed less frequently */
+	struct wl_list link;
+	struct wl_list flink;
 	union {
 		struct wlr_xdg_surface *xdg;
 		struct wlr_xwayland_surface *xwayland;
 	} surface;
+	struct wlr_scene_tree *scene;
+	struct wlr_scene_rect *border_bg; /* single border rect with clipped content hole */
+	struct wlr_scene_tree *scene_surface;
 	struct wlr_xdg_toplevel_decoration_v1 *decoration;
+
+	/* COLD — wl_listeners (set up once, rarely touched after) */
 	struct wl_listener commit;
 	struct wl_listener map;
 	struct wl_listener maximize;
@@ -53,26 +66,25 @@ typedef struct Client {
 	struct wl_listener configure;
 	struct wl_listener set_hints;
 #endif
-	unsigned int bw;
-	int corner_radius;
-	struct wlr_scene_shadow *shadow;
-	uint32_t tags;
-	int isfloating, isurgent, isfullscreen, isscratchpad;
 	uint32_t resize; /* configure serial of a pending resize */
 } Client;
 
 typedef struct {
-	/* Must keep this field first */
+	/* Must keep this field first — union-cast discriminant */
 	unsigned int type; /* LayerShell */
 
+	/* HOT */
 	Monitor *mon;
 	struct wlr_scene_tree *scene;
 	struct wlr_scene_tree *popups;
 	struct wlr_scene_layer_surface_v1 *scene_layer;
-	struct wl_list link;
 	int mapped;
 	struct wlr_layer_surface_v1 *layer_surface;
 
+	/* WARM */
+	struct wl_list link;
+
+	/* COLD — wl_listeners */
 	struct wl_listener destroy;
 	struct wl_listener unmap;
 	struct wl_listener surface_commit;

include/monitor.h

@@ -21,6 +21,20 @@ typedef struct {
 	void (*arrange)(Monitor *);
 } Layout;
 
+/* Cold state — accessed only on layout changes, not every frame.
+ * Allocated lazily on first dwindle/master layout use. */
+typedef struct MonitorCold {
+	const Layout *lt[TAGCOUNT][2];
+	unsigned int sellt[TAGCOUNT];
+	float mfact;
+	int nmaster;
+	char ltsymbol[TAGCOUNT][16];
+	DwindleNode *dwindle_root[TAGCOUNT];
+	struct Client *dwindle_focus[TAGCOUNT];
+	struct Client *master_master[TAGCOUNT];
+	int master_side[TAGCOUNT];
+} MonitorCold;
+
 struct Monitor {
 	struct wl_list link;
 	struct wlr_output *wlr_output;
@@ -34,21 +48,12 @@ struct Monitor {
 	struct wlr_box m;         /* monitor area, layout-relative */
 	struct wlr_box w;         /* window area, layout-relative */
 	struct wl_list layers[4]; /* LayerSurface.link */
-	const Layout *lt[TAGCOUNT][2];  /* per-tag layout slots */
+	struct MonitorCold *cold; /* lazy-allocated layout state */
 	int gaps;
 	unsigned int seltags;
 	struct wlr_ext_workspace_group_handle_v1 *ext_group;
-	unsigned int sellt[TAGCOUNT];   /* per-tag layout toggle index */
 	uint32_t tagset[2];
-	float mfact;
-	int nmaster;
-	char ltsymbol[TAGCOUNT][16];    /* per-tag layout symbol */
 	int asleep;
-	DwindleNode *dwindle_root[TAGCOUNT]; /* one tree per tag */
-	struct Client *dwindle_focus[TAGCOUNT];     /* insertion anchor, one per tag */
-	/* master/stack layout state */
-	struct Client *master_master[TAGCOUNT]; /* the master client (NULL if none) */
-	int master_side[TAGCOUNT];       /* 0 = master on left, 1 = master on right */
 	int scratchpad_visible;          /* whether scratchpad is shown */
 	struct Client *scratchpad_prev_focus;   /* client focused before scratchpad opened */
 	struct Client *scratchpad_current;      /* currently visible scratchpad client */