Max sharpe ratio portfolio with constrains, using FMINCON + NONLCON

Hi all, i'm trying to find the weights of the max sharpe portfolio (wh) with an upperbound on risk contribution (URC). The code works well, but the NONLCON input function does not, i get the problem "too many output arguments".

%% MAX SHARPE PORTFOLIO CODE %%

[A,B]                  =      xlsread('US_stocks') ;                    
prices                 =      A;       
logret                 =      diff(log(prices));                 
sigma                  =      cov(logret);
mu                     =      mean(logret);
std                    =      sqrt(diag(sigma));

f = @(w) -((mu*w)/sqrt(w'*sigma*w));

opts2=optimset('Display','iter','algorithm','sqp');
lb=zeros(1,N);      
ub=ones(1,N);       
Amat=ones(1,N);     
Bmat=1;             
w0=1/N*ones(N,1);

wh = fmincon(f,w0,[],[],Amat,Bmat,lb,ub,@GRCcon,opts2);

% total risk
R=sqrt((wh')*sigma*wh);

% un-standardized Marginal Risks
mR = sigma*wh;

% standardized Risk Contributions
RC= wh.*mR/R;

trn=mu*wh;

%% GRCcon FUNCTION %%

    function [cineq] = GRCcon(w)
    global sigma;

% total risk
  R=sqrt((w')*sigma*w);

% un-standardized Marginal Risks
  mR = sigma*w;

% standardized Risk Contributions
  RC  = w.*mR/R;

% upper bound
  URC = 0.5*ones(size(mR));

% Compute inequalities
  cineq = (RC-URC);   

end