Question about Gaussian elimination about converting a matrix to a upper-triangle matrix

function x=rowGauss(A,b)
[N,N]=size(A);
x=zeros(1,N+1);
Ag=[A,b];
for p=1:N-1
  [y,j]=max(abs(Ag(p:N,p)))
  C=Ag(p,:);
  Ag(p,:)=Ag(j+p-1,:);
  Ag(j+p-1,:)=c;
  if Ag(p,p)==0
      disp('Error')
      break;
  end
  for k=p+1:N
      m=Ag(k,p)/Ag(p,p);
      Ag(k,p:N+1)=Ag(k,p:N+1)-m*Ag(p,p:N+1);
  end
end
end

I have a question about the codes below

Ag(p,:)=Ag(j+p-1,:);
Ag(j+p-1,:)=c;
if Ag(p,p)==0

Find the maximum element then exchange with the pivot. But I think it should be like this:

 C=Ag(p,:);
 Ag(p,:)=Ag(j,:);
 Ag(j,:)=c; 

Could anybody tell me the difference? Thanks