On Mon, Sep 06, 2010 at 12:49:17PM +0800, Zang Roy-R61911 wrote: > > On Fri, Aug 06, 2010 at 10:51:35AM +0800, Roy Zang wrote: > > [...] > > > > > > +static struct fsl_elbc_fcm_ctrl *elbc_fcm_ctrl; > > > + > > > > Are you sure that you want it as a global var? A bit scary change. > > > > Oh, you probably don't need it, as you can get it from > > fsl_lbc_ctrl_dev->nand? > Get it form fsl_lbc_ctrl_dev->nand or assign it to > fsl_lbc_ctrl_dev->nand in probe?
I meant to get it from fsl_lbc_ctrl_dev->nand. I.e. in probe() you do: fsl_lbc_ctrl_dev->nand = elbc_fcm_ctrl, so you probably don't need the global var. (fsl_lbc_ctrl_dev seems to be global as well, duh. Well, one variable less in the global name space. But I'd probably use lbc_np->data to store the LBC private struct). Scott seem to be fine with it as there are probably no plans to to add several localbus controllers into the SoCs. But, I saw a custom board with two MPC82xx SoCs connected together, one as a master (core + peripheral devs), and other as a slave (its core was halted, and only slave's CPM peripheral devices were used by the master CPU). I think it is possible to connect two (or more) SoCs in a such way so that two or more LBC controllers would be visible for the Linux. Thanks, -- Anton Vorontsov email: cbouatmai...@gmail.com irc://irc.freenode.net/bd2 _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
