Merge 8.7

1 files changed, 29 insertions, 13 deletions

diff --git a/generic/tclStrToD.c b/generic/tclStrToD.c
index 9cd3811..a97ec04 100644
--- a/generic/tclStrToD.c
+++ b/generic/tclStrToD.c
@@ -1679,7 +1679,6 @@ MakeLowPrecisionDouble(
     int numSigDigs,		/* Number of digits in the significand */
     long exponent)		/* Power of ten */
 {
-    double retval;		/* Value of the number. */
     mp_int significandBig;	/* Significand expressed as a bignum. */
 
     /*
@@ -1687,18 +1686,25 @@ MakeLowPrecisionDouble(
      * This causes the result of double-precision calculations to be rounded
      * twice: once to the precision of double-extended and then again to the
      * precision of double. Double-rounding introduces gratuitous errors of 1
-     * ulp, so we need to change rounding mode to 53-bits.
+     * ulp, so we need to change rounding mode to 53-bits. We also make
+     * 'retval' volatile, so that it doesn't get promoted to a register.
      */
-
-    TCL_IEEE_DOUBLE_ROUNDING;
+    volatile double retval;		/* Value of the number. */
 
     /*
-     * Test for the easy cases.
+     * Test for zero significand, which requires explicit construction
+     * of -0.0. (Unary minus returns a positive zero.)
      */
-
     if (significand == 0) {
 	return copysign(0.0, -signum);
     }
+
+    /*
+     * Set the FP control word for 53 bits, WARNING: It must be reset
+     * before returning.
+     */
+    TCL_IEEE_DOUBLE_ROUNDING;
+
     if (numSigDigs <= QUICK_MAX) {
 	if (exponent >= 0) {
 	    if (exponent <= mmaxpow) {
@@ -1798,7 +1804,6 @@ MakeHighPrecisionDouble(
     int numSigDigs,		/* Number of significant digits */
     long exponent)		/* Power of 10 by which to multiply */
 {
-    double retval;
     int machexp = 0;		/* Machine exponent of a power of 10. */
 
     /*
@@ -1806,19 +1811,30 @@ MakeHighPrecisionDouble(
      * This causes the result of double-precision calculations to be rounded
      * twice: once to the precision of double-extended and then again to the
      * precision of double. Double-rounding introduces gratuitous errors of 1
-     * ulp, so we need to change rounding mode to 53-bits.
+     * ulp, so we need to change rounding mode to 53-bits. We also make
+     * 'retval' volatile to make sure that it doesn't get promoted to a
+     * register.
      */
-
-    TCL_IEEE_DOUBLE_ROUNDING;
+    volatile double retval;
 
     /*
-     * Quick checks for zero, and over/underflow. Be careful to avoid
-     * integer overflow when calculating with 'exponent'.
+     * A zero significand requires explicit construction of -0.0.
+     * (Unary minus returns positive zero.)
      */
-
     if (mp_iszero(significand)) {
 	return copysign(0.0, -signum);
     }
+
+    /*
+     * Set the 53-bit rounding mode. WARNING: It must be reset before
+     * returning.
+     */
+    TCL_IEEE_DOUBLE_ROUNDING;
+
+    /*
+     * Make quick checks for over/underflow. Be careful to avoid
+     * integer overflow when calculating with 'exponent'.
+     */
     if (exponent >= 0 && exponent-1 > maxDigits-numSigDigs) {
 	retval = HUGE_VAL;
 	goto returnValue;