xscvdphp
VSX Scalar Convert Double to Half-Precision
Converts a double-precision floating-point value to a half-precision floating-point value with rounding.
Details
The xscvdphp instruction converts the double-precision floating-point value in doubleword element 0 of VSR[XB] to a half-precision floating-point value, rounded according to the FPSCR.RN setting. The result is zero-extended and placed into doubleword element 0 of VSR[XT]. Doubleword element 1 of VSR[XT] is set to 0.
Pseudocode Operation
if MSR.VSX=0 then VSX_Unavailable()
reset_flags()
src ← bfp_CONVERT_FROM_BFP64(VSR[BX×32+B].dword[0])
rnd ← bfp_ROUND_TO_BFP16(FPSCR.RN, src)
result ← bfp16_CONVERT_FROM_BFP(rnd)
if vxsnan_flag=1 then SetFX(FPSCR.VXSNAN)
if ox_flag=1 then SetFX(FPSCR.OX)
if ux_flag=1 then SetFX(FPSCR.UX)
if xx_flag=1 then SetFX(FPSCR.XX)
vex_flag ← FPSCR.VE & vxsnan_flag
if vex_flag=0 then do
VSR[TX×32+T].hword[0:2] ← 0x0000_0000_0000
VSR[TX×32+T].dword[1] ← 0x0000_0000_0000_0000
FPSCR.FPRF ← fprf_CLASS_BFP16(result)
end
FPSCR.FR ← (vex_flag=0) & inc_flag
FPSCR.FI ← (vex_flag=0) & xx_flag
Programming Note
['This instruction can be used to operate on a single-precision source operand.', 'Previous versions of the architecture allowed the contents of doubleword 1 of the result register to be undefined. However, all processors that support this instruction write 0s into doubleword 1 of the result register, as is required by this version of the architecture.']
Example
Encoding
Operands
-
XT
Target -
XB
Source