5 files changed, 27 insertions, 69 deletions

diff --git a/src/backend/regex/regc_nfa.c b/src/backend/regex/regc_nfa.c
index ecc91a8ede9..fe8a4a12d17 100644
--- a/src/backend/regex/regc_nfa.c
+++ b/src/backend/regex/regc_nfa.c
@@ -63,7 +63,6 @@ newnfa(struct vars * v,
 	nfa->nstates = 0;
 	nfa->cm = cm;
 	nfa->v = v;
-	nfa->size = 0;
 	nfa->bos[0] = nfa->bos[1] = COLORLESS;
 	nfa->eos[0] = nfa->eos[1] = COLORLESS;
 	nfa->parent = parent;		/* Precedes newfstate so parent is valid. */
@@ -93,57 +92,6 @@ newnfa(struct vars * v,
 }
 
 /*
- * TooManyStates - checks if the max states exceeds the compile-time value
- */
-static int
-TooManyStates(struct nfa * nfa)
-{
-	struct nfa *parent = nfa->parent;
-	size_t		sz = nfa->size;
-
-	while (parent != NULL)
-	{
-		sz = parent->size;
-		parent = parent->parent;
-	}
-	if (sz > REG_MAX_STATES)
-		return 1;
-	return 0;
-}
-
-/*
- * IncrementSize - increases the tracked size of the NFA and its parents.
- */
-static void
-IncrementSize(struct nfa * nfa)
-{
-	struct nfa *parent = nfa->parent;
-
-	nfa->size++;
-	while (parent != NULL)
-	{
-		parent->size++;
-		parent = parent->parent;
-	}
-}
-
-/*
- * DecrementSize - decreases the tracked size of the NFA and its parents.
- */
-static void
-DecrementSize(struct nfa * nfa)
-{
-	struct nfa *parent = nfa->parent;
-
-	nfa->size--;
-	while (parent != NULL)
-	{
-		parent->size--;
-		parent = parent->parent;
-	}
-}
-
-/*
  * freenfa - free an entire NFA
  */
 static void
@@ -188,12 +136,6 @@ newstate(struct nfa * nfa)
 		return NULL;
 	}
 
-	if (TooManyStates(nfa))
-	{
-		NERR(REG_ETOOBIG);
-		return NULL;
-	}
-
 	if (nfa->free != NULL)
 	{
 		s = nfa->free;
@@ -201,12 +143,18 @@ newstate(struct nfa * nfa)
 	}
 	else
 	{
+		if (nfa->v->spaceused >= REG_MAX_COMPILE_SPACE)
+		{
+			NERR(REG_ETOOBIG);
+			return NULL;
+		}
 		s = (struct state *) MALLOC(sizeof(struct state));
 		if (s == NULL)
 		{
 			NERR(REG_ESPACE);
 			return NULL;
 		}
+		nfa->v->spaceused += sizeof(struct state);
 		s->oas.next = NULL;
 		s->free = NULL;
 		s->noas = 0;
@@ -230,8 +178,6 @@ newstate(struct nfa * nfa)
 	}
 	s->prev = nfa->slast;
 	nfa->slast = s;
-	/* track the current size and the parent size */
-	IncrementSize(nfa);
 	return s;
 }
 
@@ -294,7 +240,6 @@ freestate(struct nfa * nfa,
 	s->prev = NULL;
 	s->next = nfa->free;		/* don't delete it, put it on the free list */
 	nfa->free = s;
-	DecrementSize(nfa);
 }
 
 /*
@@ -312,11 +257,13 @@ destroystate(struct nfa * nfa,
 	{
 		abnext = ab->next;
 		FREE(ab);
+		nfa->v->spaceused -= sizeof(struct arcbatch);
 	}
 	s->ins = NULL;
 	s->outs = NULL;
 	s->next = NULL;
 	FREE(s);
+	nfa->v->spaceused -= sizeof(struct state);
 }
 
 /*
@@ -437,12 +384,18 @@ allocarc(struct nfa * nfa,
 		struct arcbatch *newAb;
 		int			i;
 
+		if (nfa->v->spaceused >= REG_MAX_COMPILE_SPACE)
+		{
+			NERR(REG_ETOOBIG);
+			return NULL;
+		}
 		newAb = (struct arcbatch *) MALLOC(sizeof(struct arcbatch));
 		if (newAb == NULL)
 		{
 			NERR(REG_ESPACE);
 			return NULL;
 		}
+		nfa->v->spaceused += sizeof(struct arcbatch);
 		newAb->next = s->oas.next;
 		s->oas.next = newAb;
 
diff --git a/src/backend/regex/regcomp.c b/src/backend/regex/regcomp.c
index 6b031075e75..324fea5ffb0 100644
--- a/src/backend/regex/regcomp.c
+++ b/src/backend/regex/regcomp.c
@@ -248,6 +248,7 @@ struct vars
 	struct cvec *cv2;			/* utility cvec */
 	struct subre *lacons;		/* lookahead-constraint vector */
 	int			nlacons;		/* size of lacons */
+	size_t		spaceused;		/* approx. space used for compilation */
 };
 
 /* parsing macros; most know that `v' is the struct vars pointer */
@@ -363,6 +364,7 @@ pg_regcomp(regex_t *re,
 	v->cv2 = NULL;
 	v->lacons = NULL;
 	v->nlacons = 0;
+	v->spaceused = 0;
 	re->re_magic = REMAGIC;
 	re->re_info = 0;			/* bits get set during parse */
 	re->re_csize = sizeof(chr);
diff --git a/src/include/regex/regerrs.h b/src/include/regex/regerrs.h
index 809b5112660..41e25f7ff00 100644
--- a/src/include/regex/regerrs.h
+++ b/src/include/regex/regerrs.h
@@ -75,7 +75,7 @@
 },
 
 {
-	REG_ETOOBIG, "REG_ETOOBIG", "nfa has too many states"
+	REG_ETOOBIG, "REG_ETOOBIG", "regular expression is too complex"
 },
 
 {
diff --git a/src/include/regex/regex.h b/src/include/regex/regex.h
index 3020b0ff0f7..5e1b692d26c 100644
--- a/src/include/regex/regex.h
+++ b/src/include/regex/regex.h
@@ -152,7 +152,7 @@ typedef struct
 #define REG_INVARG	16			/* invalid argument to regex function */
 #define REG_MIXED	17			/* character widths of regex and string differ */
 #define REG_BADOPT	18			/* invalid embedded option */
-#define REG_ETOOBIG 19			/* nfa has too many states */
+#define REG_ETOOBIG 19			/* regular expression is too complex */
 #define REG_ECOLORS 20			/* too many colors */
 #define REG_CANCEL	21			/* operation cancelled */
 /* two specials for debugging and testing */
diff --git a/src/include/regex/regguts.h b/src/include/regex/regguts.h
index 357891aa254..19fe991c74f 100644
--- a/src/include/regex/regguts.h
+++ b/src/include/regex/regguts.h
@@ -334,9 +334,6 @@ struct nfa
 	struct colormap *cm;		/* the color map */
 	color		bos[2];			/* colors, if any, assigned to BOS and BOL */
 	color		eos[2];			/* colors, if any, assigned to EOS and EOL */
-	size_t		size;			/* Current NFA size; differs from nstates as
-								 * it also counts the number of states in
-								 * children of this NFA. */
 	struct vars *v;				/* simplifies compile error reporting */
 	struct nfa *parent;			/* parent NFA, if any */
 };
@@ -384,10 +381,16 @@ struct cnfa
 #define NULLCNFA(cnfa)	((cnfa).nstates == 0)
 
 /*
- * Used to limit the maximum NFA size to something sane. [Tcl Bug 1810264]
+ * This symbol limits the transient heap space used by the regex compiler,
+ * and thereby also the maximum complexity of NFAs that we'll deal with.
+ * Currently we only count NFA states and arcs against this; the other
+ * transient data is generally not large enough to notice compared to those.
+ * Note that we do not charge anything for the final output data structures
+ * (the compacted NFA and the colormap).
  */
-#ifndef REG_MAX_STATES
-#define REG_MAX_STATES	100000
+#ifndef REG_MAX_COMPILE_SPACE
+#define REG_MAX_COMPILE_SPACE  \
+	(100000 * sizeof(struct state) + 100000 * sizeof(struct arcbatch))
 #endif
 
 /*