[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20080704041758.GH12945@secretlab.ca>
Date: Thu, 3 Jul 2008 22:17:58 -0600
From: Grant Likely <grant.likely@...retlab.ca>
To: Chen Gong <G.Chen@...escale.com>
Cc: david-b@...bell.net, fabrizio.garetto@...il.com,
linuxppc-dev@...abs.org, spi-devel-general@...ts.sourceforge.net,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 4/5] spi: Add OF binding support for SPI busses
On Fri, Jul 04, 2008 at 11:54:57AM +0800, Chen Gong wrote:
> Grant Likely wrote:
> > + /* Mode (clock phase/polarity/etc.) */
> > + if (of_find_property(nc, "spi,cpha", NULL))
> > + spi->mode |= SPI_CPHA;
> > + if (of_find_property(nc, "spi,cpol", NULL))
> > + spi->mode |= SPI_CPOL;
>
> so becuase in function spi_alloc_deive, spi is allocated by kzalloc,
> how about writing as follows:
> /* Mode (clock phase/polarity/etc.) */
> prop = of_get_property(nc, "spi,cpha", NULL))
> if (prop)
> spi->mode |= *prop;
> prop = of_get_property(nc, "spi,cpol", NULL))
> if (prop)
> spi->mode |= *prop;
spi,cpha and spi,cpol are defined as empty properties. The presence of
the property in the node means I need to set the flag in the spi_device.
> > + /* Select device driver */
> > + sprop = of_get_property(nc, "linux,modalias", &len);
> > + if (sprop && len > 0)
> > + strncpy(spi->modalias, sprop, KOBJ_NAME_LEN);
> > + else
> > + strncpy(spi->modalias, "spidev", KOBJ_NAME_LEN);
> > +
> how about writing as follows:
>
> if (sprop && len > 0)
> strncpy(spi->modalias, sprop, KOBJ_NAME_LEN -
> 1);
> else
> strncpy(spi->modalias, "spidev", KOBJ_NAME_LEN -
> 1);
Actually, neither are very good. What it should really be is:
if (sprop && len > 0)
strlcpy(spi->modalias, sprop, sizeof (spi->modalias));
else
strlcpy(spi->modalias, "spidev", sizeof (spi->modalias));
This ensures that the string is always null terminated and always the
right size.
... But the whole argument is a bit moot. The fact that I even defined
a "linux,modalias" property is a great big hairy hack that I would
never want my mother to see. Instead, I need a method to bind to an SPI
driver based on the compatible list. Jon Smirl has done some work
in this regard for i2c, but I haven't looked at it very deeply, and so
do not at all understand it (yet).
Thanks for the comments.
g.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists