great exciting band

As it has been established, Co2MnSi, the most famous full-Heusler alloy, has semi-metallic gap in its spin minority, here, spin-down channel. But in my case, when calculation converges, the gap occurs in the spin-up DOS and the whole of dos is exactly reversed for spin-up and down. In order to clarifying what exactly i did, the input file is attached. Please advice me if there is a mistake.

Regards

############################################
<?xml version='1.0' encoding='UTF-8'?>
<input>
<title> Co2MnSi</title>
<structure speciespath="/home/soft/exciting/species/">
<crystal scale="10.6399956208">
<basevect> 0.500000 0.500000 0.000000</basevect>
<basevect> 0.500000 0.000000 0.500000</basevect>
<basevect> 0.000000 0.500000 0.500000</basevect>
</crystal>
<species speciesfile="Co.xml">
<atom coord=" 0.25000000 0.25000000 0.25000000 " bfcmt="0.0 0.0 1.0"/>
<atom coord=" 0.75000000 0.75000000 0.75000000 " bfcmt="0.0 0.0 1.5"/>
</species>
<species speciesfile="Mn.xml">
<atom coord=" 0.0000000 0.0000000 0.0000000 " bfcmt="0.0 0.0 2.0"/>
</species>
<species speciesfile="Si.xml">
<atom coord=" 0.50000000 0.50000000 0.50000000 " bfcmt="0.0 0.0 -0.1"/>
</species>
</structure>
<groundstate
do="fromscratch"
xctype="GGA_PBE"
rgkmax="7.0"
ngridk="8 8 8"
nempty="20">

<spin/>

</groundstate>
<properties>
<dos
nsmdos="2"
winddos="-0.3 0.3"/>
</properties>

</input>