// This file is part of crlibm and is distributed under the GNU Public Licence
// See file COPYING for details
// The following lines are either comments (beginning with a #)
-// or give
+// or give
// 1/ a rounding mode : RN|RU|RD|RZ (crlibm syntax) or N|P|M|Z (libmcr syntax)
// 2/ The high and low hexadecimal halves of an input
// 3/ The high and low hexadecimal halves of the expected corresponding output
-// Some special cases. Much to do here
+// Some special cases. Much to do here
T(RN, 0x0p+0, 0x0p+0, 0x1p+0, 0x0p+0, 0)
-T(RN, 0x1p+1, -0x1.0ccp+10, 0x0p+0, -0x1p-1, INEXACT|UNDERFLOW) // 2^-1075
-T(RN, 0x1p+1, -0x1.0c8p+10, 0x1p-1074, 0x0p+0, 0) // 2^-1074
+T(RN, 0x1p+1, -0x1.0ccp+10, 0x0p+0, -0x1p-1, INEXACT|UNDERFLOW) // 2^-1075
+T(RN, 0x1p+1, -0x1.0c8p+10, 0x1p-1074, 0x0p+0, 0) // 2^-1074
// Some exact mid-point cases. Subnormal cases are missing.
T(RN, 0x1.cfdd8p+17, 0x1.8p+1, 0x1.7cbf0571d6e62p+53, 0x1p-1, INEXACT)
T(RN, 0x1.927a0f5b2p+35, 0x1.8p+0, 0x1.64d79ee5e5b8ep+53, -0x1p-1, INEXACT)