[PATCH] powerpc/512x: add function for CS parameter configuration

[Anatolij Gustschin]

On Fri, 1 Feb 2013 22:31:51 -0600
Timur Tabi <timur at tabi.org> wrote:

> On Fri, Feb 1, 2013 at 7:28 AM, Anatolij Gustschin <agust at denx.de> wrote:
> > Add ability to configure CS parameters for devices that need
> > different CS parameters setup after their configuration. I.e.
> > an FPGA device on LP bus can require different CS parameters
> > for its bus interface after loading firmware into it. A driver
> > can easily reconfigure the LPC CS parameters using this function.
> 
> Could you define "CS" somewhere?

I'll define it in revised commit log in v2.

> > +struct mpc512x_lpc {
> > +       u32     cs_cfg[8];      /* CS config */
> > +       u32     cs_ctrl;        /* CS Control Register */
> > +       u32     cs_status;      /* CS Status Register */
> > +       u32     burst_ctrl;     /* CS Burst Control Register */
> > +       u32     deadcycle_ctrl; /* CS Deadcycle Control Register */
> > +       u32     holdcycle_ctrl; /* CS Holdcycle Control Register */
> > +       u32     alt;            /* Address Latch Timing Register */
> > +};
> 
> These should be __be32.

Why? To add a new bunch of sparse warnings?

...
> > +       if (cs < 0 || cs > 7)
> > +               return -EINVAL;
> 
> If you make cs an unsigned int, you won't need to test for < 0.

can do that, yes.

> > +
> > +       if (!lpc) {
> > +               np = of_find_compatible_node(NULL, NULL, "fsl,mpc5121-lpc");
> > +               lpc = of_iomap(np, 0);
> > +               of_node_put(np);
> > +               if (!lpc)
> > +                       return -ENOMEM;
> > +       }
> > +
> > +       out_be32(&lpc->cs_cfg[cs], val);
> 
> You forgot the iounmap() if lpc == NULL.

No, it is intentional, no need to map/unmap again and again for all
subsequent calls.


> > +       return 0;
> > +}
> > +EXPORT_SYMBOL_GPL(mpc512x_cs_config);
> 
> EXPORT_SYMBOL, please, to match the rest of the Freescale platforms.

I'll change it in v2. Thanks!

Anatolij