[Wien] different eigenvalues at the same k point

[yxl@email.jlu.edu.cn]

Dear users,
    I got the different eigenvalues at (0,0,0) from case.klist and case.klist_band. I think this
 is unreasonable, maybe I have something wrong with my process.
I am doing GGA+U calculations for pure ZnO, U(Znd)=7.5eV
runsp_lapw -p -orb -cc 0.001 -ec 0.0001,
After that I got the eigenvalues at (0,0,0) for up spin are
K=   0.00000   0.00000   0.00000            1
:RKM  : MATRIX SIZE  371LOs:  34  RKM= 6.98  WEIGHT= 1.00  PGR:
       EIGENVALUES ARE:
        -5.3728632   -5.3726646   -5.3720193   -5.3720193   -5.3720013
        -5.3720013   -1.0604865   -1.0117592   -0.3038482   -0.3038479
        -0.3021278   -0.2924885   -0.2924884   -0.2485623   -0.2485622
        -0.2355277   -0.2355277   -0.2295515   -0.1254418    0.1735083
         0.1735085    0.2235002    0.2259273    0.2259274    0.3317865
         0.6572532    0.8743296    1.2174851    1.2174853    1.3120212
         1.3126602    1.3126614    1.4813955    1.6015259    1.6015262
         1.7272975    1.7272977    1.7333897    1.8434873    1.9786480
       ********************************************************
:FER  : F E R M I - ENERGY(TETRAH.M.)=   0.22593

Then I calculated the band structure
'x lapw1 -band -up -p' 
I found the eigenvalues in case.output1up are
K=   0.00000   0.00000   0.00000   GAMMA
      MATRIX SIZE  371  WEIGHT= 2.00  PGR:
     EIGENVALUES ARE:
     -5.3728631   -5.3726644   -5.3720192   -5.3720192   -5.3720012
     -5.3720012   -1.0604835   -1.0064367   -0.1393203   -0.1148119
     -0.1148116   -0.1086308   -0.1007132   -0.1007131   -0.0109564
     -0.0109563    0.0066803    0.0066803    0.0235110    0.2286371
      0.2286373    0.2787604    0.2875916    0.2875918    0.3318913
      0.6578704    0.8795310    1.2226986    1.2226988    1.3120498
      1.3149472    1.3149484    1.4930243    1.6279692    1.6279695
      1.7430521    1.7449112    1.7449114    1.8470804    1.9802465
            0 EIGENVALUES BELOW THE ENERGY  -10.00000
       ********************************************************
So if I enter the fermi energy 0.22593 in ZnO.insp, I will get the metalic band structure.
But according to  the ZnO.scf and DOS(do not need recalculate lapw1) it is semiconductor, 
and in fact it should be semiconduter.
What's wrong with my calculation? Have you found some strange? Any suggestions will
 be appreciated.
yours sincerely,
hongxia