Is there an existing CVA6 bug for this?
Bug Description
At least with my version of Questa and with CVA6 configured with active H-ext, the expression
lu_gpaddr_o[12+2*CVA6Cfg.VpnLen/CVA6Cfg.PtLevels-1:12] = lu_vaddr_i[12+2*(CVA6Cfg.VpnLen/CVA6Cfg.PtLevels)-1:12];
assigns the wrong number of bits to the element on the left. VpnLen is 29, PteLevels is 3: 2*CVA6Cfg.VpnLen/CVA6Cfg.PtLevels (on the left) is 19, while 2*(CVA6Cfg.VpnLen/CVA6Cfg.PtLevels) (on the right) is 18.
Is there an existing CVA6 bug for this?
Bug Description
At least with my version of Questa and with CVA6 configured with active H-ext, the expression
assigns the wrong number of bits to the element on the left.
VpnLenis 29,PteLevelsis 3:2*CVA6Cfg.VpnLen/CVA6Cfg.PtLevels(on the left) is 19, while2*(CVA6Cfg.VpnLen/CVA6Cfg.PtLevels)(on the right) is 18.