Fix the wrong implementation of the xblat1.f implementations. (Fixes: #963)

Since this is a regression after upgrading from v3.11.0 to v3.12.0, we can
narrow down the range of the bug into the newly added SB1NRM2 subroutine.
According to the buildlog and the documentation in the code, the VALUES(9),
calculated as SXVALS(XX,2) should be infty. But the current code is returning
a zero (or randomly) initialized variable YY, which does not make sense.

In fact, if you go back to the reference implementation, namely the
supplementary material of this paper
  https://dl.acm.org/doi/abs/10.1145/3061665
You can find a similar implementation of the SXVALS function in the
`la_xxvals.F90` file. This patch corrests the test following the reference
code.

Author: cdluminate

BLAS/TESTING/cblat1.f

@@ -995,12 +995,13 @@ REAL FUNCTION SXVALS(XX,K)
       REAL              XX
       INTEGER           K
 *     .. Local Scalars ..
-      REAL              X, Y, YY, Z
+      REAL              X, Y, Z
 *     .. Intrinsic Functions ..
       INTRINSIC         HUGE
 *     .. Executable Statements ..
+      X = ZERO
       Y = HUGE(XX)
-      Z = YY
+      Z = Y*Y
       IF (K.EQ.1) THEN
          X = -Z
       ELSE IF (K.EQ.2) THEN

BLAS/TESTING/dblat1.f

@@ -1327,12 +1327,13 @@ DOUBLE PRECISION FUNCTION DXVALS(XX,K)
       DOUBLE PRECISION  XX
       INTEGER           K
 *     .. Local Scalars ..
-      DOUBLE PRECISION  X, Y, YY, Z
+      DOUBLE PRECISION  X, Y, Z
 *     .. Intrinsic Functions ..
       INTRINSIC         HUGE
 *     .. Executable Statements ..
+      X = ZERO
       Y = HUGE(XX)
-      Z = YY
+      Z = Y*Y
       IF (K.EQ.1) THEN
          X = -Z
       ELSE IF (K.EQ.2) THEN

BLAS/TESTING/sblat1.f

@@ -1279,12 +1279,13 @@ REAL FUNCTION SXVALS(XX,K)
       REAL              XX
       INTEGER           K
 *     .. Local Scalars ..
-      REAL              X, Y, YY, Z
+      REAL              X, Y, Z
 *     .. Intrinsic Functions ..
       INTRINSIC         HUGE
 *     .. Executable Statements ..
+      X = ZERO
       Y = HUGE(XX)
-      Z = YY
+      Z = Y*Y
       IF (K.EQ.1) THEN
          X = -Z
       ELSE IF (K.EQ.2) THEN

BLAS/TESTING/zblat1.f

@@ -995,12 +995,13 @@ DOUBLE PRECISION FUNCTION DXVALS(XX,K)
       DOUBLE PRECISION  XX
       INTEGER           K
 *     .. Local Scalars ..
-      DOUBLE PRECISION  X, Y, YY, Z
+      DOUBLE PRECISION  X, Y, Z
 *     .. Intrinsic Functions ..
       INTRINSIC         HUGE
 *     .. Executable Statements ..
+      X = ZERO
       Y = HUGE(XX)
-      Z = YY
+      Z = Y*Y
       IF (K.EQ.1) THEN
          X = -Z
       ELSE IF (K.EQ.2) THEN