Fix zend_dval_to_lval outside 64bit integers range

cataphract · cataphract · commit 77566edbafb9 · 2013-02-23T17:23:49.000+01:00
PHP should preserve the least significant bits when casting from double
to long. Zend.m4 contains this:

AC_DEFINE([ZEND_DVAL_TO_LVAL_CAST_OK], 1, [Define if double cast to long preserves least significant bits])

If ZEND_DVAL_TO_LVAL_CAST_OK is not defined, zend_operators.h had an
inline implementation of zend_dval_to_lval() that would do a cast to an
int64_t (when sizeof(long) == 4), then a cast to unsigned long and
finally the cast to long.

While this works well for doubles inside the range of values of the type
used in the first cast (int64_t in the 32-bit version and unsigned long
in the 64-bit version), if outside the range, it is undefined behavior
that WILL give varying and not particularly useful results.

This commit uses fmod() to first put the double in a range that can
safely be cast to unsigned long and then casts this unsigned long to
long. This last cast is implementation defined, but it's very likely
that this gives the expected result (i.e. the internal 2's complement
representation is unchanged) on all platforms that PHP supports. In any
case, the previous implementationa already had this assumption.

This alternative code path is indeed significantly slower than simply
casting the double (almost an order of magnitude), but that should not
matter because casting doubles with a very high absolute value is a
rare event.
diff --git a/Zend/tests/bug39018.phpt b/Zend/tests/bug39018.phpt
@@ -64,6 +64,8 @@ print "\nDone\n";
 --EXPECTF--
 Notice: String offset cast occurred in %s on line %d
 
+Notice: Uninitialized string offset: %s in %s on line 6
+
 Notice: Uninitialized string offset: 0 in %s on line %d
 
 Notice: Uninitialized string offset: 0 in %s on line %d
diff --git a/Zend/tests/dval_to_lval_32.phpt b/Zend/tests/dval_to_lval_32.phpt
@@ -0,0 +1,29 @@
+--TEST--
+zend_dval_to_lval preserves low bits  (32 bit long)
+--SKIPIF--
+<?php
+if (PHP_INT_SIZE != 4)
+	 die("skip for machines with 32-bit longs");
+?>
+--FILE--
+<?php
+	/* test doubles around -4e21 */
+	$values = [
+		-4000000000000001048576.,
+		-4000000000000000524288.,
+		-4000000000000000000000.,
+		-3999999999999999475712.,
+		-3999999999999998951424.,
+	];
+
+	foreach ($values as $v) {
+		var_dump((int)$v);
+	}
+
+?>
+--EXPECT--
+int(-2056257536)
+int(-2055733248)
+int(-2055208960)
+int(-2054684672)
+int(-2054160384)
diff --git a/Zend/tests/dval_to_lval_64.phpt b/Zend/tests/dval_to_lval_64.phpt
@@ -0,0 +1,29 @@
+--TEST--
+zend_dval_to_lval preserves low bits  (64 bit long)
+--SKIPIF--
+<?php
+if (PHP_INT_SIZE != 8)
+	 die("skip for machines with 64-bit longs");
+?>
+--FILE--
+<?php
+	/* test doubles around -4e21 */
+	$values = [
+		-4000000000000001048576.,
+		-4000000000000000524288.,
+		-4000000000000000000000.,
+		-3999999999999999475712.,
+		-3999999999999998951424.,
+	];
+
+	foreach ($values as $v) {
+		var_dump((int)$v);
+	}
+
+?>
+--EXPECT--
+int(2943463994971652096)
+int(2943463994972176384)
+int(2943463994972700672)
+int(2943463994973224960)
+int(2943463994973749248)
diff --git a/Zend/zend_operators.h b/Zend/zend_operators.h
@@ -68,22 +68,36 @@ END_EXTERN_C()
 
 #if ZEND_DVAL_TO_LVAL_CAST_OK
 # define zend_dval_to_lval(d) ((long) (d))
-#elif SIZEOF_LONG == 4 && defined(HAVE_ZEND_LONG64)
+#elif SIZEOF_LONG == 4
 static zend_always_inline long zend_dval_to_lval(double d)
 {
 	if (d > LONG_MAX || d < LONG_MIN) {
-		return (long)(unsigned long)(zend_long64) d;
+		double	two_pow_32 = pow(2., 32.),
+				dmod;
+
+		dmod = fmod(d, two_pow_32);
+		if (dmod < 0) {
+			dmod += two_pow_32;
+		}
+		return (long)(unsigned long)dmod;
 	}
-	return (long) d;
+	return (long)d;
 }
 #else
 static zend_always_inline long zend_dval_to_lval(double d)
 {
 	/* >= as (double)LONG_MAX is outside signed range */
-	if (d >= LONG_MAX) {
-		return (long)(unsigned long) d;
+	if (d >= LONG_MAX || d < LONG_MIN) {
+		double	two_pow_64 = pow(2., 64.),
+				dmod;
+
+		dmod = fmod(d, two_pow_64);
+		if (dmod < 0) {
+			dmod += two_pow_64;
+		}
+		return (long)(unsigned long)dmod;
 	}
-	return (long) d;
+	return (long)d;
 }
 #endif
 /* }}} */
