Thank you Calculex. I do have difficulties with this part of the Standard Model too, I am not working with it really. I really thought somebody would give correction to my post. I will try to fill the gaps by myself.
So indeed one starts in the standard model with massless particles. Well, the Higgs part would be :
{\cal L} _H=( {\cal D}_{\mu}H)^{\dagger}({\cal D}^{\mu}H)-V(H) with {\cal D}_{\mu}H = (\partial_\mu + \imath {\mathbf W}_\mu +\frac{\imath}{2}y_hB_\mu)H and V=-\mu^2H^{\dagger}H-\lambda(H^{\dagger}H)^2
So really \mu is the only "mass" term, but with the wrong sign (as usual, this is the mexican hat shape). y_h is the hypercharge of the Higgs, which so far is only a spinless boson, weak doublet (and color singlet), this being later broken into a single scalar. So anyway, one still has to identify all the physical fields.
My point to complete the previous post, was that I do not really understand the generation of mass for the fermions. Is it only :
one introduces a coupling to the Higgs and this in turn is analogous to a mass term because the Higgs is a scalar. Is that all ? I mean that in the case of the vector bosons mass generation, there is a mechanism to take one degree of freedom from the Goldstone (unwanted) to the longitudinal part of the (previously massless) vectors. On the contrary, one "only" has to break chiral invariance in the case of fermions. One "only" has to mix left and right. There is no need for any new degree of freedom (of course, only chirality gets broken, this is well known from the Dirac equation). This in turn is "only" adding a a term \sim h f\bar{f}=h(f^{\dagger}_Rf_L) and identify the coefficient to this term as the mass of the fermionic field
I am not saying "this is the easy part", especially for quarks where one has to
define the mass eigenstates for the up-like quarks.
Ref. : P. Ramond "Journeys Beyond the Standard Model" FIP n° 101
Well, I needed not the "beyond" part for this

This is an excellent one.
amazon