[Tarantool-patches] [PATCH v2 1/2] core: introduce various platform metrics

[Sergey Kaplun <skaplun@tarantool.org>]

This patch introduces the following counters:
- overall amount of allocated tables, cdata and udata objects
- number of incremental GC steps grouped by GC state
- number of string hashes hits and misses
- amount of allocated and freed memory
- number of trace aborts and restored snapshots

Interfaces to obtain these metrics via both Lua and C API are
introduced in the next patch.

Part of tarantool/tarantool#5187
---
 src/lj_cdata.c |  2 ++
 src/lj_cdata.h |  2 ++
 src/lj_gc.c    |  4 ++++
 src/lj_gc.h    |  6 +-----
 src/lj_jit.h   |  3 +++
 src/lj_obj.h   | 22 ++++++++++++++++++++++
 src/lj_snap.c  |  1 +
 src/lj_state.c |  2 +-
 src/lj_str.c   |  5 +++++
 src/lj_tab.c   |  2 ++
 src/lj_trace.c |  5 ++++-
 src/lj_udata.c |  2 ++
 12 files changed, 49 insertions(+), 7 deletions(-)

diff --git a/src/lj_cdata.c b/src/lj_cdata.c
index 68e16d7..0dd8553 100644
--- a/src/lj_cdata.c
+++ b/src/lj_cdata.c
@@ -46,6 +46,7 @@ GCcdata *lj_cdata_newv(lua_State *L, CTypeID id, CTSize sz, CTSize align)
   cd->marked |= 0x80;
   cd->gct = ~LJ_TCDATA;
   cd->ctypeid = id;
+  g->gc.cdatanum++;
   return cd;
 }
 
@@ -82,6 +83,7 @@ void LJ_FASTCALL lj_cdata_free(global_State *g, GCcdata *cd)
   } else {
     lj_mem_free(g, memcdatav(cd), sizecdatav(cd));
   }
+  g->gc.cdatanum--;
 }
 
 void lj_cdata_setfin(lua_State *L, GCcdata *cd, GCobj *obj, uint32_t it)
diff --git a/src/lj_cdata.h b/src/lj_cdata.h
index 5bb0f5d..66b023b 100644
--- a/src/lj_cdata.h
+++ b/src/lj_cdata.h
@@ -45,6 +45,7 @@ static LJ_AINLINE GCcdata *lj_cdata_new(CTState *cts, CTypeID id, CTSize sz)
   cd = (GCcdata *)lj_mem_newgco(cts->L, sizeof(GCcdata) + sz);
   cd->gct = ~LJ_TCDATA;
   cd->ctypeid = ctype_check(cts, id);
+  G(cts->L)->gc.cdatanum++;
   return cd;
 }
 
@@ -54,6 +55,7 @@ static LJ_AINLINE GCcdata *lj_cdata_new_(lua_State *L, CTypeID id, CTSize sz)
   GCcdata *cd = (GCcdata *)lj_mem_newgco(L, sizeof(GCcdata) + sz);
   cd->gct = ~LJ_TCDATA;
   cd->ctypeid = id;
+  G(L)->gc.cdatanum++;
   return cd;
 }
 
diff --git a/src/lj_gc.c b/src/lj_gc.c
index 1c8e6ce..44c8aa1 100644
--- a/src/lj_gc.c
+++ b/src/lj_gc.c
@@ -634,6 +634,7 @@ static void atomic(global_State *g, lua_State *L)
 static size_t gc_onestep(lua_State *L)
 {
   global_State *g = G(L);
+  g->gc.state_count[g->gc.state]++;
   switch (g->gc.state) {
   case GCSpause:
     gc_mark_start(g);  /* Start a new GC cycle by marking all GC roots. */
@@ -857,6 +858,8 @@ void *lj_mem_realloc(lua_State *L, void *p, GCSize osz, GCSize nsz)
   lua_assert((nsz == 0) == (p == NULL));
   lua_assert(checkptrGC(p));
   g->gc.total = (g->gc.total - osz) + nsz;
+  g->gc.allocated += nsz;
+  g->gc.freed += osz;
   return p;
 }
 
@@ -869,6 +872,7 @@ void * LJ_FASTCALL lj_mem_newgco(lua_State *L, GCSize size)
     lj_err_mem(L);
   lua_assert(checkptrGC(o));
   g->gc.total += size;
+  g->gc.allocated += size;
   setgcrefr(o->gch.nextgc, g->gc.root);
   setgcref(g->gc.root, o);
   newwhite(g, o);
diff --git a/src/lj_gc.h b/src/lj_gc.h
index 669bbe9..2051220 100644
--- a/src/lj_gc.h
+++ b/src/lj_gc.h
@@ -8,11 +8,6 @@
 
 #include "lj_obj.h"
 
-/* Garbage collector states. Order matters. */
-enum {
-  GCSpause, GCSpropagate, GCSatomic, GCSsweepstring, GCSsweep, GCSfinalize
-};
-
 /* Bitmasks for marked field of GCobj. */
 #define LJ_GC_WHITE0	0x01
 #define LJ_GC_WHITE1	0x02
@@ -117,6 +112,7 @@ LJ_FUNC void *lj_mem_grow(lua_State *L, void *p,
 static LJ_AINLINE void lj_mem_free(global_State *g, void *p, size_t osize)
 {
   g->gc.total -= (GCSize)osize;
+  g->gc.freed += osize;
   g->allocf(g->allocd, p, osize, 0);
 }
 
diff --git a/src/lj_jit.h b/src/lj_jit.h
index 7eb3d2a..90c1914 100644
--- a/src/lj_jit.h
+++ b/src/lj_jit.h
@@ -475,6 +475,9 @@ typedef struct jit_State {
   size_t szmcarea;	/* Size of current mcode area. */
   size_t szallmcarea;	/* Total size of all allocated mcode areas. */
 
+  size_t nsnaprestore;	/* Overall number of snap restores for this jit_State. */
+  size_t ntraceabort;	/* Overall number of abort traces for this jit_State. */
+
   TValue errinfo;	/* Additional info element for trace errors. */
 
 #if LJ_HASPROFILE
diff --git a/src/lj_obj.h b/src/lj_obj.h
index f368578..18df173 100644
--- a/src/lj_obj.h
+++ b/src/lj_obj.h
@@ -571,6 +571,17 @@ typedef enum {
 #define basemt_obj(g, o)	((g)->gcroot[GCROOT_BASEMT+itypemap(o)])
 #define mmname_str(g, mm)	(strref((g)->gcroot[GCROOT_MMNAME+(mm)]))
 
+/* Garbage collector states. Order matters. */
+enum {
+  GCSpause,
+  GCSpropagate,
+  GCSatomic,
+  GCSsweepstring,
+  GCSsweep,
+  GCSfinalize,
+  GCSmax
+};
+
 typedef struct GCState {
   GCSize total;		/* Memory currently allocated. */
   GCSize threshold;	/* Memory threshold. */
@@ -578,6 +589,9 @@ typedef struct GCState {
   uint8_t state;	/* GC state. */
   uint8_t nocdatafin;	/* No cdata finalizer called. */
   uint8_t unused2;
+  size_t freed;		/* Total amount of freed memory. */
+  size_t allocated;	/* Total amount of allocated memory. */
+  size_t state_count[GCSmax]; /* Count of incremental GC steps per state. */
   MSize sweepstr;	/* Sweep position in string table. */
   GCRef root;		/* List of all collectable objects. */
   MRef sweep;		/* Sweep position in root list. */
@@ -589,6 +603,12 @@ typedef struct GCState {
   GCSize estimate;	/* Estimate of memory actually in use. */
   MSize stepmul;	/* Incremental GC step granularity. */
   MSize pause;		/* Pause between successive GC cycles. */
+
+  size_t tabnum;	/* Amount of allocated table objects. */
+  size_t udatanum;	/* Amount of allocated udata objects. */
+#ifdef LJ_HASFFI
+  size_t cdatanum;	/* Amount of allocated cdata objects. */
+#endif
 } GCState;
 
 /* Global state, shared by all threads of a Lua universe. */
@@ -602,6 +622,8 @@ typedef struct global_State {
     BloomFilter next[2];
   } strbloom;
 #endif
+  size_t strhash_hit;	/* Strings amount founded in string hash. */
+  size_t strhash_miss;	/* Strings amount allocated and put into string hash. */
   lua_Alloc allocf;	/* Memory allocator. */
   void *allocd;		/* Memory allocator data. */
   GCState gc;		/* Garbage collector. */
diff --git a/src/lj_snap.c b/src/lj_snap.c
index 7554caf..4cf27e7 100644
--- a/src/lj_snap.c
+++ b/src/lj_snap.c
@@ -904,6 +904,7 @@ const BCIns *lj_snap_restore(jit_State *J, void *exptr)
     L->top = frame + snap->nslots;
     break;
   }
+  J->nsnaprestore++;
   return pc;
 }
 
diff --git a/src/lj_state.c b/src/lj_state.c
index 632dd07..1d9c628 100644
--- a/src/lj_state.c
+++ b/src/lj_state.c
@@ -214,7 +214,7 @@ LUA_API lua_State *lua_newstate(lua_Alloc f, void *ud)
   g->gc.state = GCSpause;
   setgcref(g->gc.root, obj2gco(L));
   setmref(g->gc.sweep, &g->gc.root);
-  g->gc.total = sizeof(GG_State);
+  g->gc.allocated = g->gc.total = sizeof(GG_State);
   g->gc.pause = LUAI_GCPAUSE;
   g->gc.stepmul = LUAI_GCMUL;
   lj_dispatch_init((GG_State *)L);
diff --git a/src/lj_str.c b/src/lj_str.c
index 24e90ca..8ff955e 100644
--- a/src/lj_str.c
+++ b/src/lj_str.c
@@ -222,6 +222,7 @@ GCstr *lj_str_new(lua_State *L, const char *str, size_t lenx)
                       str_fastcmp(str, strdata(sx), len) == 0) {
 	/* Resurrect if dead. Can only happen with fixstring() (keywords). */
 	if (isdead(g, o)) flipwhite(o);
+	g->strhash_hit++;
 	return sx;  /* Return existing string. */
       }
       o = gcnext(o);
@@ -234,6 +235,7 @@ GCstr *lj_str_new(lua_State *L, const char *str, size_t lenx)
                       memcmp(str, strdata(sx), len) == 0) {
 	/* Resurrect if dead. Can only happen with fixstring() (keywords). */
 	if (isdead(g, o)) flipwhite(o);
+	g->strhash_hit++;
 	return sx;  /* Return existing string. */
       }
       o = gcnext(o);
@@ -266,6 +268,7 @@ GCstr *lj_str_new(lua_State *L, const char *str, size_t lenx)
 	    if (sx->hash == fh && sx->len == len && str_fastcmp(str, strdata(sx), len) == 0) {
 	      /* Resurrect if dead. Can only happen with fixstring() (keywords). */
 	      if (isdead(g, o)) flipwhite(o);
+	      g->strhash_hit++;
 	      return sx;  /* Return existing string. */
 	    }
 	    o = gcnext(o);
@@ -276,6 +279,7 @@ GCstr *lj_str_new(lua_State *L, const char *str, size_t lenx)
 	    if (sx->hash == fh && sx->len == len && memcmp(str, strdata(sx), len) == 0) {
 	      /* Resurrect if dead. Can only happen with fixstring() (keywords). */
 	      if (isdead(g, o)) flipwhite(o);
+	      g->strhash_hit++;
 	      return sx;  /* Return existing string. */
 	    }
 	    o = gcnext(o);
@@ -293,6 +297,7 @@ GCstr *lj_str_new(lua_State *L, const char *str, size_t lenx)
     }
   }
 #endif
+  g->strhash_miss++;
   /* Nope, create a new string. */
   s = lj_mem_newt(L, sizeof(GCstr)+len+1, GCstr);
   newwhite(g, s);
diff --git a/src/lj_tab.c b/src/lj_tab.c
index ff216f3..c5f358e 100644
--- a/src/lj_tab.c
+++ b/src/lj_tab.c
@@ -141,6 +141,7 @@ static GCtab *newtab(lua_State *L, uint32_t asize, uint32_t hbits)
   }
   if (hbits)
     newhpart(L, t, hbits);
+  G(L)->gc.tabnum++;
   return t;
 }
 
@@ -240,6 +241,7 @@ void LJ_FASTCALL lj_tab_free(global_State *g, GCtab *t)
     lj_mem_free(g, t, sizetabcolo((uint32_t)t->colo & 0x7f));
   else
     lj_mem_freet(g, t);
+  g->gc.tabnum--;
 }
 
 /* -- Table resizing ------------------------------------------------------ */
diff --git a/src/lj_trace.c b/src/lj_trace.c
index d85b47f..b17bde3 100644
--- a/src/lj_trace.c
+++ b/src/lj_trace.c
@@ -538,8 +538,10 @@ static int trace_downrec(jit_State *J)
   /* Restart recording at the return instruction. */
   lua_assert(J->pt != NULL);
   lua_assert(bc_isret(bc_op(*J->pc)));
-  if (bc_op(*J->pc) == BC_RETM)
+  if (bc_op(*J->pc) == BC_RETM) {
+    J->ntraceabort++;
     return 0;  /* NYI: down-recursion with RETM. */
+  }
   J->parent = 0;
   J->exitno = 0;
   J->state = LJ_TRACE_RECORD;
@@ -616,6 +618,7 @@ static int trace_abort(jit_State *J)
     return trace_downrec(J);
   else if (e == LJ_TRERR_MCODEAL)
     lj_trace_flushall(L);
+  J->ntraceabort++;
   return 0;
 }
 
diff --git a/src/lj_udata.c b/src/lj_udata.c
index bd0321b..70c722a 100644
--- a/src/lj_udata.c
+++ b/src/lj_udata.c
@@ -24,11 +24,13 @@ GCudata *lj_udata_new(lua_State *L, MSize sz, GCtab *env)
   /* Chain to userdata list (after main thread). */
   setgcrefr(ud->nextgc, mainthread(g)->nextgc);
   setgcref(mainthread(g)->nextgc, obj2gco(ud));
+  g->gc.udatanum++;
   return ud;
 }
 
 void LJ_FASTCALL lj_udata_free(global_State *g, GCudata *ud)
 {
+  g->gc.udatanum--;
   lj_mem_free(g, ud, sizeudata(ud));
 }
 
-- 
2.24.1