Verification is the fundamental step that any turbulence simulation code ă has to be submitted in order to assess the proper implementation of the ă underlying equations. We have carried out a cross comparison of three ă flux tube gyrokinetic codes, GENE [F. Jenko et al., Phys. Plasmas 7, ă 1904 (2000)], GKW [A. G. Peeters et al., Comput. Phys. Commun. 180, ă 2650 (2009)], and GS2 [W. Dorland et al., Phys. Rev. Lett. 85, 5579 ă (2000)], focusing our attention on the effect of realistic geometries ă described by a series of MHD equilibria with increasing shaping ă complexity. To simplify the effort, the benchmark has been limited to ă the electrostatic collisionless linear behaviour of the system. A fully ă gyrokinetic model has been used to describe the dynamics of both ions ă and electrons. Several tests have been carried out looking at linear ă stability at ion and electron scales, where for the assumed profiles Ion ă Temperature Gradient (ITG)/Trapped Electron Modes and Electron ă Temperature Gradient modes are unstable. The capability of the codes to ă handle a non-zero ballooning angle has been successfully benchmarked in ă the ITG regime. Finally, the standard Rosenbluth-Hinton test has been ă successfully carried out looking at the effect of shaping on Zonal Flows ă (ZFs) and Geodesic Acoustic Modes (GAMs). Inter-code comparison as well ă as validation of simulation results against analytical estimates has ă been accomplished. All the performed tests confirm that plasma ă elongation strongly stabilizes plasma instabilities as well as leads to ă a strong increase in ZF residual and GAM damping.