Coneprog: Exitflag -3, 1, -7

I want to convert my QP to SOCP and I followed the procedure described here.

I have

Asc = sqrtm(H);
Asc((end+1),(end+1)) = 1;
d = [zeros(size(f(:)));1];
gamma = -1;
b = zeros(size(d));
qp = secondordercone(Asc,b,d,gamma);
c=[f(:);1];
[u,fval,exitflag,output] = coneprog(c,qp)
exitflag

However, I immediately get this error

Problem is unbounded.

u =

[]


fval =

[]


exitflag =

-3


output = 

struct with fields:

iterations: 10
primalfeasibility: []
dualfeasibility: []
dualitygap: []
algorithm: 'interior-point'
linearsolver: 'augmented'
message: 'Problem is unbounded.'

I decided to add my second SOC constraint and then I get this output

Optimal solution found.

u_ =

1.0e+04 *

0.0068
-0.0212
2.4724


fval =

-2.4645e+04


exitflag =

1


output = 

struct with fields:

iterations: 9
primalfeasibility: 1.5796e-11
dualfeasibility: 3.0666e-10
dualitygap: 8.3902e-14
algorithm: 'interior-point'
linearsolver: 'augmented'
message: 'Optimal solution found.'


t =

0.7500

Optimal solution found.

u_ =

1.0e+04 *

0.0082
-0.0216
2.6814


fval =

-2.6732e+04


exitflag =

1


output = 

struct with fields:

iterations: 9
primalfeasibility: 1.4523e-11
dualfeasibility: 2.7386e-10
dualitygap: 7.2296e-14
algorithm: 'interior-point'
linearsolver: 'augmented'
message: 'Optimal solution found.'


t =

0.7600

Problem is unbounded.

u_ =

[]


fval =

[]


exitflag =

-3


output = 

struct with fields:

iterations: 9
primalfeasibility: []
dualfeasibility: []
dualitygap: []
algorithm: 'interior-point'
linearsolver: 'augmented'
message: 'Problem is unbounded.'

I also get -7 error. I would like to know why just this conversion (from link) doesn't work. What should I change to avoid having those errors?

Thank you!

[Matt J]

Isn't it obvious from the exit messages that the problem is unbounded? What do you get when you solve the original QP, using quadprog?

[Julius Dri]

Wanted result, no error, everything is fine