rts: Define ticky macro stubs

These macros have long been undefined which has meant we were missing
reporting these allocations in ticky profiles.

The most critical missing definition was TICK_ALLOC_HEAP_NOCTR which was
missing all the RTS calls to allocate, this leads to a the overall
ALLOC_RTS_tot number to be severaly underreported.

Of particular interest though is the ALLOC_STACK_ctr and ALLOC_STACK_tot
counters which are useful to tracking stack allocations.

Fixes #23421

rts/RaiseAsync.c

@@ -877,7 +877,7 @@ raiseAsync(Capability *cap, StgTSO *tso, StgClosure *exception,
             write_barrier(); // XXX: Necessary?
             SET_HDR(ap,&stg_AP_STACK_info,
                     ((StgClosure *)frame)->header.prof.ccs /* ToDo */);
-            TICK_ALLOC_UP_THK(WDS(words+1),0);
+            TICK_ALLOC_UP_THK(AP_STACK_sizeW(words),0);
 
             //IF_DEBUG(scheduler,
             //       debugBelch("sched: Updating ");
@@ -927,7 +927,7 @@ raiseAsync(Capability *cap, StgTSO *tso, StgClosure *exception,
             }
 
             SET_HDR(ap,&stg_AP_STACK_NOUPD_info,stack->header.prof.ccs);
-            TICK_ALLOC_SE_THK(WDS(words+1),0);
+            TICK_ALLOC_SE_THK(AP_STACK_sizeW(words),0);
 
             stack->sp = sp;
             threadStackUnderflow(cap,tso);
@@ -963,7 +963,7 @@ raiseAsync(Capability *cap, StgTSO *tso, StgClosure *exception,
             // handler in this frame.
             //
             raise = (StgThunk *)allocate(cap,sizeofW(StgThunk)+1);
-            TICK_ALLOC_SE_THK(WDS(1),0);
+            TICK_ALLOC_SE_THK(sizeofW(StgThunk)+1,0);
             SET_HDR(raise,&stg_raise_info,cf->header.prof.ccs);
             raise->payload[0] = exception;
 

rts/RtsSymbols.c

@@ -402,9 +402,9 @@ extern char **environ;
       SymI_HasProto(ALLOC_PAP_gds)                      \
       SymI_HasProto(ALLOC_PAP_slp)                      \
       SymI_HasProto(ALLOC_TSO_ctr)                      \
-      SymI_HasProto(ALLOC_TSO_adm)                      \
-      SymI_HasProto(ALLOC_TSO_gds)                      \
-      SymI_HasProto(ALLOC_TSO_slp)                      \
+      SymI_HasProto(ALLOC_TSO_tot)                      \
+      SymI_HasProto(ALLOC_STACK_ctr)                    \
+      SymI_HasProto(ALLOC_STACK_tot)                    \
       SymI_HasProto(RET_NEW_ctr)                        \
       SymI_HasProto(RET_OLD_ctr)                        \
       SymI_HasProto(RET_UNBOXED_TUP_ctr)                \

rts/Threads.c

@@ -89,7 +89,7 @@ createThread(Capability *cap, W_ size)
     stack->marking      = 0;
 
     tso = (StgTSO *)allocate(cap, sizeofW(StgTSO));
-    TICK_ALLOC_TSO();
+    TICK_ALLOC_TSO(sizeofW(StgTSO));
     SET_HDR(tso, &stg_TSO_info, CCS_SYSTEM);
 
     // Always start with the compiled code evaluator
@@ -873,7 +873,7 @@ StgMutArrPtrs *listThreads(Capability *cap)
     StgMutArrPtrs *arr =
         (StgMutArrPtrs *)allocate(cap, sizeofW(StgMutArrPtrs) + size);
     SET_HDR(arr, &stg_MUT_ARR_PTRS_DIRTY_info, CCS_SYSTEM);
-    TICK_ALLOC_PRIM(sizeofW(StgMutArrPtrs), n, 0);
+    TICK_ALLOC_PRIM(sizeofW(StgMutArrPtrs), size, 0);
     arr->ptrs = n_threads;
     arr->size = size;
 

rts/Ticky.c

@@ -260,6 +260,12 @@ PrintTickyInfo(void)
   PR_CTR(ALLOC_PRIM_gds);
   PR_CTR(ALLOC_PRIM_slp);
 
+  PR_CTR(ALLOC_TSO_ctr);
+  PR_CTR(ALLOC_TSO_tot);
+
+  PR_CTR(ALLOC_STACK_ctr);
+  PR_CTR(ALLOC_STACK_tot);
+
   PR_CTR(ENT_VIA_NODE_ctr);
   PR_CTR(ENT_STATIC_CON_ctr);
   PR_CTR(ENT_DYN_CON_ctr);

rts/include/stg/Ticky.h

@@ -170,9 +170,10 @@ EXTERN StgInt ALLOC_PAP_gds INIT(0);
 EXTERN StgInt ALLOC_PAP_slp INIT(0);
 
 EXTERN StgInt ALLOC_TSO_ctr INIT(0);
-EXTERN StgInt ALLOC_TSO_adm INIT(0);
-EXTERN StgInt ALLOC_TSO_gds INIT(0);
-EXTERN StgInt ALLOC_TSO_slp INIT(0);
+EXTERN StgInt ALLOC_TSO_tot INIT(0);
+
+EXTERN StgInt ALLOC_STACK_ctr INIT(0);
+EXTERN StgInt ALLOC_STACK_tot INIT(0);
 
 EXTERN StgInt RET_NEW_ctr INIT(0);
 EXTERN StgInt RET_OLD_ctr INIT(0);
@@ -215,15 +216,37 @@ EXTERN StgInt RET_UNBOXED_TUP_hst[TICKY_BIN_COUNT] INIT({0});
 
 #define TICK_BUMP(ctr)      TICK_BUMP_BY(ctr,1)
 
-#define TICK_ALLOC_PRIM(x,y,z)        // FIXME: update counter
 #define TICK_UPD_OLD_IND()            TICK_BUMP(UPD_OLD_IND_ctr)
 #define TICK_UPD_NEW_IND()            TICK_BUMP(UPD_NEW_IND_ctr)
 #define TICK_UPD_SQUEEZED()           TICK_BUMP(UPD_SQUEEZED_ctr)
-#define TICK_ALLOC_HEAP_NOCTR(bytes)  // FIXME: update counter
-#define TICK_GC_FAILED_PROMOTION()    // FIXME: update counter
-#define TICK_ALLOC_TSO()              // FIXME: update counter
-#define TICK_ALLOC_STACK(g)           // FIXME: update counter
-#define TICK_ALLOC_UP_THK(g,s)        // FIXME: update counter
-#define TICK_ALLOC_SE_THK(g,s)        // FIXME: update counter
 
+#define TICK_ALLOC_PRIM(hdr,goods,slop)\
+  TICK_BUMP(ALLOC_PRIM_ctr);\
+  TICK_BUMP_BY(ALLOC_PRIM_adm,hdr);\
+  TICK_BUMP_BY(ALLOC_PRIM_gds,goods);\
+  TICK_BUMP_BY(ALLOC_PRIM_slp,slop);
+
+#define TICK_GC_FAILED_PROMOTION() TICK_BUMP(GC_FAILED_PROMOTION_ctr)
+
+#define TICK_ALLOC_TSO(n)\
+  TICK_BUMP(ALLOC_TSO_ctr);\
+  TICK_BUMP_BY(ALLOC_TSO_tot,n);
+
+#define TICK_ALLOC_STACK(n)\
+  TICK_BUMP(ALLOC_STACK_ctr);\
+  TICK_BUMP_BY(ALLOC_STACK_tot,n);
+
+#define TICK_ALLOC_UP_THK(g,s)\
+  TICK_BUMP(ALLOC_UP_THK_ctr);\
+  TICK_BUMP_BY(ALLOC_THK_gds,g);\
+  TICK_BUMP_BY(ALLOC_THK_slp,s);\
+
+#define TICK_ALLOC_SE_THK(g,s)\
+  TICK_BUMP(ALLOC_SE_THK_ctr);\
+  TICK_BUMP_BY(ALLOC_THK_gds,g);\
+  TICK_BUMP_BY(ALLOC_THK_slp,s);\
+
+#define TICK_ALLOC_HEAP_NOCTR(bytes)\
+  TICK_BUMP(ALLOC_RTS_ctr);\
+  TICK_BUMP_BY(ALLOC_RTS_tot,n);
 #endif