<p>Mmm... as long as the platforms call pxa2xx_mfp_config(), I was thinking the pxa2xx part can remain intact, saving some bits.</p>
<p>Soon will be on the flight, will reply two days later. Thanks.</p>
<p><blockquote type="cite">On Oct 18, 2009 2:43 PM, "Haojian Zhuang" <<a href="mailto:haojian.zhuang@gmail.com">haojian.zhuang@gmail.com</a>> wrote:<br><br><p><font color="#500050">On Sun, Oct 18, 2009 at 2:03 AM, Eric Miao <<a href="mailto:eric.y.miao@gmail.com">eric.y.miao@gmail.com</a>> wrote:
> On Sun, Oct 18, 2009 at ...</font></p>I think that mfp_to_gpio() macro is also a part of mfp driver. This<br>
macro is used widely<br>
in platform driver in order to allocate platform irq resource.<br>
NR_BUILTIN_GPIO is similar<br>
to this macro.<br>
<br>
Because they're used in structure initialization, we can't use<br>
detecting CPU method to<br>
distinguish how much mfp/gpio pins can be applied on current CPU.<br>
Since we couldn't<br>
import variable or function into structure intialization code. Current<br>
patch is the easiest<br>
modification.<br>
<br>
If there's a better solution, I'm glad to adopt it. :)<br>
<br>
Thanks<br>
<font color="#888888">Haojian<br>
</font></blockquote></p>