Give Python a debug-mode pymalloc, much as sketched on Python-Dev.

When WITH_PYMALLOC is defined, define PYMALLOC_DEBUG to enable the debug
allocator.  This can be done independent of build type (release or debug).
A debug build automatically defines PYMALLOC_DEBUG when pymalloc is
enabled.  It's a detected error to define PYMALLOC_DEBUG when pymalloc
isn't enabled.

Two debugging entry points defined only under PYMALLOC_DEBUG:

+ _PyMalloc_DebugCheckAddress(const void *p) can be used (e.g., from gdb)
  to sanity-check a memory block obtained from pymalloc.  It sprays
  info to stderr (see next) and dies via Py_FatalError if the block is
  detectably damaged.

+ _PyMalloc_DebugDumpAddress(const void *p) can be used to spray info
  about a debug memory block to stderr.

A tiny start at implementing "API family" checks isn't good for
anything yet.

_PyMalloc_DebugRealloc() has been optimized to do little when the new
size is <= old size.  However, if the new size is larger, it really
can't call the underlying realloc() routine without either violating its
contract, or knowing something non-trivial about how the underlying
realloc() works.  A memcpy is always done in this case.

This was a disaster for (and only) one of the std tests:  test_bufio
creates single text file lines up to a million characters long.  On
Windows, fileobject.c's get_line() uses the horridly funky
getline_via_fgets(), which keeps growing and growing a string object
hoping to find a newline.  It grew the string object 1000 bytes each
time, so for a million-character string it took approximately forever
(I gave up after a few minutes).

So, also:

fileobject.c, getline_via_fgets():  When a single line is outrageously
long, grow the string object at a mildly exponential rate, instead of
just 1000 bytes at a time.

That's enough so that a debug-build test_bufio finishes in about 5 seconds
on my Win98SE box.  I'm curious to try this on Win2K, because it has very
different memory behavior than Win9X, and test_bufio always took a factor
of 10 longer to complete on Win2K.  It *could* be that the endless
reallocs were simply killing it on Win2K even in the release build.

4 files changed, 346 insertions, 28 deletions

diff --git a/Include/Python.h b/Include/Python.h
index 9724fb7..d1ffc73 100644
--- a/Include/Python.h
+++ b/Include/Python.h
@@ -61,6 +61,15 @@
 
 #include "pyport.h"
 
+/* Debug-mode build with pymalloc implies PYMALLOC_DEBUG.
+ *  PYMALLOC_DEBUG is in error if pymalloc is not in use.
+ */
+#if defined(Py_DEBUG) && defined(WITH_PYMALLOC) && !defined(PYMALLOC_DEBUG)
+#define PYMALLOC_DEBUG
+#endif
+#if defined(PYMALLOC_DEBUG) && !defined(WITH_PYMALLOC)
+#error "PYMALLOC_DEBUG requires WITH_PYMALLOC"
+#endif
 #include "pymem.h"
 
 #include "object.h"
diff --git a/Include/pymem.h b/Include/pymem.h
index 10e915f..a321f5b 100644
--- a/Include/pymem.h
+++ b/Include/pymem.h
@@ -89,20 +89,34 @@ extern DL_IMPORT(void) PyMem_Free(void *);
    it is recommended to write the test explicitly in the code.
    Note that according to ANSI C, free(NULL) has no effect. */
 
-	
+
 /* pymalloc (private to the interpreter) */
 #ifdef WITH_PYMALLOC
 DL_IMPORT(void *) _PyMalloc_Malloc(size_t nbytes);
 DL_IMPORT(void *) _PyMalloc_Realloc(void *p, size_t nbytes);
 DL_IMPORT(void) _PyMalloc_Free(void *p);
+
+#ifdef PYMALLOC_DEBUG
+DL_IMPORT(void *) _PyMalloc_DebugMalloc(size_t nbytes, int family);
+DL_IMPORT(void *) _PyMalloc_DebugRealloc(void *p, size_t nbytes, int family);
+DL_IMPORT(void) _PyMalloc_DebugFree(void *p, int family);
+DL_IMPORT(void) _PyMalloc_DebugDumpAddress(const void *p);
+DL_IMPORT(void) _PyMalloc_DebugCheckAddress(const void *p);
+#define _PyMalloc_MALLOC(N) _PyMalloc_DebugMalloc(N, 0)
+#define _PyMalloc_REALLOC(P, N) _PyMalloc_DebugRealloc(P, N, 0)
+#define _PyMalloc_FREE(P) _PyMalloc_DebugFree(P, 0)
+
+#else	/* WITH_PYMALLOC && ! PYMALLOC_DEBUG */
 #define _PyMalloc_MALLOC _PyMalloc_Malloc
 #define _PyMalloc_REALLOC _PyMalloc_Realloc
 #define _PyMalloc_FREE _PyMalloc_Free
-#else
+#endif
+
+#else	/* ! WITH_PYMALLOC */
 #define _PyMalloc_MALLOC PyMem_MALLOC
 #define _PyMalloc_REALLOC PyMem_REALLOC
 #define _PyMalloc_FREE PyMem_FREE
-#endif
+#endif	/* WITH_PYMALLOC */
 
 
 #ifdef __cplusplus
diff --git a/Objects/fileobject.c b/Objects/fileobject.c
index 47e6b17..6a82cce 100644
--- a/Objects/fileobject.c
+++ b/Objects/fileobject.c
@@ -772,13 +772,9 @@ getline_via_fgets(FILE *fp)
  * cautions about boosting that.  300 was chosen because the worst real-life
  * text-crunching job reported on Python-Dev was a mail-log crawler where over
  * half the lines were 254 chars.
- * INCBUFSIZE is the amount by which we grow the buffer, if MAXBUFSIZE isn't
- * enough.  It doesn't much matter what this is set to: we only get here for
- * absurdly long lines anyway.
  */
 #define INITBUFSIZE 100
 #define MAXBUFSIZE 300
-#define INCBUFSIZE 1000
 	char* p;	/* temp */
 	char buf[MAXBUFSIZE];
 	PyObject* v;	/* the string object result */
@@ -786,6 +782,7 @@ getline_via_fgets(FILE *fp)
 	char* pvend;    /* address one beyond last free slot */
 	size_t nfree;	/* # of free buffer slots; pvend-pvfree */
 	size_t total_v_size;  /* total # of slots in buffer */
+	size_t increment;	/* amount to increment the buffer */
 
 	/* Optimize for normal case:  avoid _PyString_Resize if at all
 	 * possible via first reading into stack buffer "buf".
@@ -853,7 +850,7 @@ getline_via_fgets(FILE *fp)
 	/* The stack buffer isn't big enough; malloc a string object and read
 	 * into its buffer.
 	 */
-	total_v_size = MAXBUFSIZE + INCBUFSIZE;
+	total_v_size = MAXBUFSIZE << 1;
 	v = PyString_FromStringAndSize((char*)NULL, (int)total_v_size);
 	if (v == NULL)
 		return v;
@@ -897,7 +894,8 @@ getline_via_fgets(FILE *fp)
 		}
 		/* expand buffer and try again */
 		assert(*(pvend-1) == '\0');
-		total_v_size += INCBUFSIZE;
+		increment = total_v_size >> 2;	/* mild exponential growth */
+		total_v_size += increment;
 		if (total_v_size > INT_MAX) {
 			PyErr_SetString(PyExc_OverflowError,
 			    "line is longer than a Python string can hold");
@@ -907,14 +905,13 @@ getline_via_fgets(FILE *fp)
 		if (_PyString_Resize(&v, (int)total_v_size) < 0)
 			return NULL;
 		/* overwrite the trailing null byte */
-		pvfree = BUF(v) + (total_v_size - INCBUFSIZE - 1);
+		pvfree = BUF(v) + (total_v_size - increment - 1);
 	}
 	if (BUF(v) + total_v_size != p)
 		_PyString_Resize(&v, p - BUF(v));
 	return v;
 #undef INITBUFSIZE
 #undef MAXBUFSIZE
-#undef INCBUFSIZE
 }
 #endif	/* ifdef USE_FGETS_IN_GETLINE */
 
diff --git a/Objects/obmalloc.c b/Objects/obmalloc.c
index 4e141c7..3e9465f 100644
--- a/Objects/obmalloc.c
+++ b/Objects/obmalloc.c
@@ -623,24 +623,10 @@ _PyMalloc_Realloc(void *p, size_t nbytes)
 	return (void *)bp;
 }
 
-/* calloc */
-
-/* -- unused --
-void *
-_PyMalloc_Calloc(size_t nbel, size_t elsz)
-{
-        void *p;
-	size_t nbytes;
-
-	nbytes = nbel * elsz;
-	p = _PyMalloc_Malloc(nbytes);
-	if (p != NULL)
-		memset(p, 0, nbytes);
-	return p;
-}
-*/
-
 #else	/* ! WITH_PYMALLOC */
+
+/*==========================================================================*/
+/* pymalloc not enabled:  Redirect the entry points to the PyMem family. */
 void *
 _PyMalloc_Malloc(size_t n)
 {
@@ -686,3 +672,315 @@ _PyMalloc_Del(PyObject *op)
 {
 	_PyMalloc_FREE(op);
 }
+
+#ifdef PYMALLOC_DEBUG
+/*==========================================================================*/
+/* A x-platform debugging allocator. */
+
+#define PYMALLOC_CLEANBYTE      0xCB    /* uninitialized memory */
+#define PYMALLOC_DEADBYTE       0xDB    /* free()ed memory */
+#define PYMALLOC_FORBIDDENBYTE  0xFB    /* unusable memory */
+
+static ulong serialno = 0;	/* incremented on each debug {m,re}alloc */
+
+/* Read 4 bytes at p as a big-endian ulong. */
+static ulong
+read4(const void *p)
+{
+	const unsigned char *q = (unsigned char *)p;
+	return ((ulong)q[0] << 24) |
+	       ((ulong)q[1] << 16) |
+	       ((ulong)q[2] <<  8) |
+	        (ulong)q[3];
+}
+
+/* Write the 4 least-significant bytes of n as a big-endian unsigned int,
+   MSB at address p, LSB at p+3. */
+static void
+write4(void *p, ulong n)
+{
+	unsigned char *q = (unsigned char *)p;
+	q[0] = (unsigned char)((n >> 24) & 0xff);
+	q[1] = (unsigned char)((n >> 16) & 0xff);
+	q[2] = (unsigned char)((n >>  8) & 0xff);
+	q[3] = (unsigned char)( n        & 0xff);
+}
+
+static void
+check_family(const void *p, int family)
+{
+	const uchar *q = (const uchar *)p;
+	int original_family;
+	char buf[200];
+
+	assert(p != NULL);
+	original_family = (int)*(q-4);
+	if (family != original_family) {
+		/* XXX better msg */
+		PyOS_snprintf(buf, sizeof(buf),
+			"free or realloc from family #%d called, "
+			"but block was allocated by family #%d",
+			family, original_family);
+		_PyMalloc_DebugDumpAddress(p);
+		Py_FatalError(buf);
+	}
+}
+
+/* The debug malloc asks for 16 extra bytes and fills them with useful stuff,
+   here calling the underlying malloc's result p:
+
+p[0:4]
+    Number of bytes originally asked for.  4-byte unsigned integer,
+    big-endian (easier to read in a memory dump).
+p[4]
+    The API "family" this malloc call belongs to.  XXX todo XXX
+p[5:8]
+    Copies of PYMALLOC_FORBIDDENBYTE.  Used to catch under- writes
+    and reads.
+p[8:8+n]
+    The requested memory, filled with copies of PYMALLOC_CLEANBYTE.
+    Used to catch reference to uninitialized memory.
+    &p[8] is returned.  Note that this is 8-byte aligned if PyMalloc
+    handled the request itself.
+p[8+n:8+n+4]
+    Copies of PYMALLOC_FORBIDDENBYTE.  Used to catch over- writes
+    and reads.
+p[8+n+4:8+n+8]
+    A serial number, incremented by 1 on each call to _PyMalloc_DebugMalloc
+    and _PyMalloc_DebugRealloc.
+    4-byte unsigned integer, big-endian.
+    If "bad memory" is detected later, the serial number gives an
+    excellent way to set a breakpoint on the next run, to capture the
+    instant at which this block was passed out.
+*/
+
+void *
+_PyMalloc_DebugMalloc(size_t nbytes, int family)
+{
+	uchar *p;	/* base address of malloc'ed block */
+	uchar *q;	/* p + 8 + nbytes +  */
+	size_t total;	/* nbytes + 16 */
+
+	assert(family == 0);
+
+	++serialno;
+	total = nbytes + 16;
+	if (total < nbytes || (total >> 31) > 1) {
+		/* overflow, or we can't represent it in 4 bytes */
+		/* Obscure:  can't do (total >> 32) != 0 instead, because
+		   C doesn't define what happens for a right-shift of 32
+		   when size_t is a 32-bit type.  At least C guarantees
+		   size_t is an unsigned type. */
+		return NULL;
+	}
+
+	p = _PyMalloc_Malloc(total);	/* XXX derive from family */
+	if (p == NULL)
+		return NULL;
+
+	write4(p, nbytes);
+	p[4] = (uchar)family;
+	p[5] = p[6] = p[7] = PYMALLOC_FORBIDDENBYTE;
+
+	if (nbytes > 0)
+		memset(p+8, PYMALLOC_CLEANBYTE, nbytes);
+
+	q = p + 8 + nbytes;
+	q[0] = q[1] = q[2] = q[3] = PYMALLOC_FORBIDDENBYTE;
+	write4(q+4, serialno);
+
+	return p+8;
+}
+
+/* The debug free first uses the address to find the number of bytes
+   originally asked for, then checks the 8 bytes on each end for
+   sanity (in particular, that the PYMALLOC_FORBIDDENBYTEs are still
+   intact).
+   Then fills the original bytes with PYMALLOC_DEADBYTE.
+   Then calls the underlying free.
+*/
+void
+_PyMalloc_DebugFree(void *p, int family)
+{
+	uchar *q = (uchar*)p;
+	size_t nbytes;
+
+	assert(family == 0);
+
+	if (p == NULL)
+		return;
+	check_family(p, family);
+	_PyMalloc_DebugCheckAddress(p);
+	nbytes = read4(q-8);
+	if (nbytes > 0)
+		memset(q, PYMALLOC_DEADBYTE, nbytes);
+	_PyMalloc_Free(q-8);	/* XXX derive from family */
+}
+
+void *
+_PyMalloc_DebugRealloc(void *p, size_t nbytes, int family)
+{
+	uchar *q = (uchar *)p;
+	size_t original_nbytes;
+	uchar *fresh;	/* new memory block, if needed */
+
+	assert(family == 0);
+
+	if (p == NULL)
+		return _PyMalloc_DebugMalloc(nbytes, family);
+
+	check_family(p, family);
+	_PyMalloc_DebugCheckAddress(p);
+
+	original_nbytes = read4(q-8);
+	if (nbytes == original_nbytes) {
+		/* note that this case is likely to be common due to the
+		   way Python appends to lists */
+		++serialno;
+		write4(q + nbytes + 4, serialno);
+		return p;
+	}
+
+	if (nbytes < original_nbytes) {
+		/* shrinking -- leave the guts alone, except to
+		   fill the excess with DEADBYTE */
+		const size_t excess = original_nbytes - nbytes;
+		++serialno;
+		write4(q-8, nbytes);
+		/* kill the excess bytes plus the trailing 8 pad bytes */
+		memset(q + nbytes, PYMALLOC_DEADBYTE, excess + 8);
+		q += nbytes;
+		q[0] = q[1] = q[2] = q[3] = PYMALLOC_FORBIDDENBYTE;
+		write4(q+4, serialno);
+		return p;
+	}
+
+	/* More memory is needed:  get it, copy over the first original_nbytes
+	   of the original data, and free the original memory. */
+	fresh = (uchar *)_PyMalloc_DebugMalloc(nbytes, family);
+	if (fresh != NULL && original_nbytes > 0)
+		memcpy(fresh, p, original_nbytes);
+	_PyMalloc_DebugFree(p, family);
+	return fresh;
+}
+
+void
+_PyMalloc_DebugCheckAddress(const void *p)
+{
+	const uchar *q = (const uchar *)p;
+	char *msg = NULL;
+
+	if (p == NULL)
+		msg = "didn't expect a NULL pointer";
+
+	else if (*(q-3) != PYMALLOC_FORBIDDENBYTE ||
+	    	 *(q-2) != PYMALLOC_FORBIDDENBYTE ||
+	    	 *(q-1) != PYMALLOC_FORBIDDENBYTE)
+	    	msg = "bad leading pad byte";
+
+	else {
+		const ulong nbytes = read4(q-8);
+		const uchar *tail = q + nbytes;
+		int i;
+		for (i = 0; i < 4; ++i) {
+			if (tail[i] != PYMALLOC_FORBIDDENBYTE) {
+				msg = "bad trailing pad byte";
+				break;
+			}
+		}
+	}
+
+	if (msg != NULL) {
+		_PyMalloc_DebugDumpAddress(p);
+		Py_FatalError(msg);
+	}
+}
+
+void
+_PyMalloc_DebugDumpAddress(const void *p)
+{
+	const uchar *q = (const uchar *)p;
+	const uchar *tail;
+	ulong nbytes, serial;
+
+	fprintf(stderr, "Debug memory block at address p=%p:\n", p);
+	if (p == NULL)
+		return;
+
+	nbytes = read4(q-8);
+	fprintf(stderr, "    %lu bytes originally allocated\n", nbytes);
+	fprintf(stderr, "    from API family #%d\n", *(q-4));
+
+	/* In case this is nuts, check the pad bytes before trying to read up
+	   the serial number (the address deref could blow up). */
+
+	fprintf(stderr, "    the 3 pad bytes at p-3 are ");
+	if (*(q-3) == PYMALLOC_FORBIDDENBYTE &&
+	    *(q-2) == PYMALLOC_FORBIDDENBYTE &&
+	    *(q-1) == PYMALLOC_FORBIDDENBYTE) {
+		fprintf(stderr, "PYMALLOC_FORBIDDENBYTE, as expected\n");
+	}
+	else {
+		int i;
+		fprintf(stderr, "not all PYMALLOC_FORBIDDENBYTE (0x%02x):\n",
+			PYMALLOC_FORBIDDENBYTE);
+		for (i = 3; i >= 1; --i) {
+			const uchar byte = *(q-i);
+			fprintf(stderr, "        at p-%d: 0x%02x", i, byte);
+			if (byte != PYMALLOC_FORBIDDENBYTE)
+				fputs(" *** OUCH", stderr);
+			fputc('\n', stderr);
+		}
+	}
+
+	tail = q + nbytes;
+	fprintf(stderr, "    the 4 pad bytes at tail=%p are ", tail);
+	if (tail[0] == PYMALLOC_FORBIDDENBYTE &&
+	    tail[1] == PYMALLOC_FORBIDDENBYTE &&
+	    tail[2] == PYMALLOC_FORBIDDENBYTE &&
+	    tail[3] == PYMALLOC_FORBIDDENBYTE) {
+		fprintf(stderr, "PYMALLOC_FORBIDDENBYTE, as expected\n");
+	}
+	else {
+		int i;
+		fprintf(stderr, "not all PYMALLOC_FORBIDDENBYTE (0x%02x):\n",
+			PYMALLOC_FORBIDDENBYTE);
+		for (i = 0; i < 4; ++i) {
+			const uchar byte = tail[i];
+			fprintf(stderr, "        at tail+%d: 0x%02x",
+				i, byte);
+			if (byte != PYMALLOC_FORBIDDENBYTE)
+				fputs(" *** OUCH", stderr);
+			fputc('\n', stderr);
+		}
+	}
+
+	serial = read4(tail+4);
+	fprintf(stderr, "    the block was made by call #%lu to "
+	                "debug malloc/realloc\n", serial);
+
+	if (nbytes > 0) {
+		int i = 0;
+		fprintf(stderr, "    data at p:");
+		/* print up to 8 bytes at the start */
+		while (q < tail && i < 8) {
+			fprintf(stderr, " %02x", *q);
+			++i;
+			++q;
+		}
+		/* and up to 8 at the end */
+		if (q < tail) {
+			if (tail - q > 8) {
+				fprintf(stderr, " ...");
+				q = tail - 8;
+			}
+			while (q < tail) {
+				fprintf(stderr, " %02x", *q);
+				++q;
+			}
+		}
+		fprintf(stderr, "\n");
+	}
+}
+
+#endif	/* PYMALLOC_DEBUG */