fmincon code generation for single-precision floating point hardware

I am using fmincon solver to solve an optimization problem with linear equality and inequality constraints and non-linear objective function. I have set up a Simulink model to generate C code. However, I just realized that fmincon supports code generation for double-precision floating point targets only. I was wondering if there is a way to generate code for single-precision processors? I am also open to using other solvers. I'll appreciate your inputs.

[Walter Roberson]

Someone asked the same question about a month ago, and the answer from someone at Mathworks was No, fmincon requires double precision and cannot be generated for single precision.

[Maziar Mobarrez]

Thanks Walter, I am still interested to know what limits the code gen to generate fmincon code for single-precision hardware and find the potential solution.

[Walter Roberson]

See https://www.mathworks.com/matlabcentral/answers/1594944-need-help-is-there-any-single-precision-optimization-code#answer_848915