On Fri, Sep 23, 2016 at 11:27:25AM +0900, Masahiro Yamada wrote:
> Hi Tom,
>
>
> 2016-09-23 10:11 GMT+09:00 Tom Rini <tr...@konsulko.com>:
> > We cannot call on the CONFIG_CLK based clk_get_rate function unless
> > CONFIG_CLK is set.
> >
> > Signed-off-by: Tom Rini <tr...@konsulko.com>
> > ---
> > drivers/serial/ns16550.c | 7 +++++--
> > include/clk.h | 25 ++++++++++++-------------
> > 2 files changed, 17 insertions(+), 15 deletions(-)
> >
> > diff --git a/drivers/serial/ns16550.c b/drivers/serial/ns16550.c
> > index 3f6ea4d27591..765499dab646 100644
> > --- a/drivers/serial/ns16550.c
> > +++ b/drivers/serial/ns16550.c
> > @@ -13,6 +13,7 @@
> > #include <serial.h>
> > #include <watchdog.h>
> > #include <linux/types.h>
> > +#include <linux/compiler.h>
> > #include <asm/io.h>
> >
> > DECLARE_GLOBAL_DATA_PTR;
> > @@ -353,8 +354,8 @@ int ns16550_serial_ofdata_to_platdata(struct udevice
> > *dev)
> > {
> > struct ns16550_platdata *plat = dev->platdata;
> > fdt_addr_t addr;
> > - struct clk clk;
> > - int err;
> > + __maybe_unused struct clk clk;
> > + __maybe_unused int err;
> >
> > /* try Processor Local Bus device first */
> > addr = dev_get_addr(dev);
> > @@ -401,6 +402,7 @@ int ns16550_serial_ofdata_to_platdata(struct udevice
> > *dev)
> > plat->reg_shift = fdtdec_get_int(gd->fdt_blob, dev->of_offset,
> > "reg-shift", 0);
> >
> > +#ifdef CONFIG_CLK
> > err = clk_get_by_index(dev, 0, &clk);
> > if (!err) {
> > err = clk_get_rate(&clk);
> > @@ -410,6 +412,7 @@ int ns16550_serial_ofdata_to_platdata(struct udevice
> > *dev)
> > debug("ns16550 failed to get clock\n");
> > return err;
> > }
> > +#endif
> >
> > if (!plat->clock)
> > plat->clock = fdtdec_get_int(gd->fdt_blob, dev->of_offset,
>
>
>
> Why did you need this patch?
Because we have non-CONFIG_CLK clk_get_rate() functions, so we get
function prototype mismatches.
> > diff --git a/include/clk.h b/include/clk.h
> > index 94c003714700..7273127bb4fb 100644
> > --- a/include/clk.h
> > +++ b/include/clk.h
> > @@ -98,19 +98,6 @@ int clk_get_by_index(struct udevice *dev, int index,
> > struct clk *clk);
> > * @return 0 if OK, or a negative error code.
> > */
> > int clk_get_by_name(struct udevice *dev, const char *name, struct clk
> > *clk);
> > -#else
> > -static inline int clk_get_by_index(struct udevice *dev, int index,
> > - struct clk *clk)
> > -{
> > - return -ENOSYS;
> > -}
> > -
> > -static inline int clk_get_by_name(struct udevice *dev, const char *name,
> > - struct clk *clk)
> > -{
> > - return -ENOSYS;
> > -}
> > -#endif
> >
> > /**
> > * clk_request - Request a clock by provider-specific ID.
> > @@ -175,5 +162,17 @@ int clk_enable(struct clk *clk);
> > int clk_disable(struct clk *clk);
> >
> > int soc_clk_dump(void);
> > +#else
> > +static inline int clk_get_by_index(struct udevice *dev, int index,
> > + struct clk *clk)
> > +{
> > + return -ENOSYS;
> > +}
> >
> > +static inline int clk_get_by_name(struct udevice *dev, const char *name,
> > + struct clk *clk)
> > +{
> > + return -ENOSYS;
> > +}
> > +#endif
> > #endif
>
>
> This just made the situation worse; it will cause
> implicit declaration of clk_enable(), clk_set_rate(), etc
> unless we patch clk consumers around with #ifdef CONFIG_CLK.
There's none today, and I think it makes it more obvious when we have a
clock "framework" that's not using CONFIG_CLK and needs to be moved
over.
Without this k2*evm fails to build.
--
Tom
signature.asc
Description: Digital signature
_______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
