Hi all, Before gcc-4.7.0, the PowerPC cpu type "8548" was set as a dummy entry, pointing to "8540" settings. Thus, it was bound to the ppc8540_cost structure, in which L2 cache is set to 256 (see gcc/config/rs6000/rs6000.c). Freescale MPC8548E PowerQUICC III processors are based on e500v2 cores, whose L2 cache size is 512 KB. Would it mean the PowerPC "8548" (e500v2 family core) target is not optimally processed in gcc, at least regarding the L2 cache ? (the "8548" cpu type has even been fired, starting from gcc-4.7.x series). Thanks, Yon