On Tue, Aug 31, 2010 at 11:26 AM, Matthew McClintock [off-list ref] wr=
ote:

quoted

I'm not sure KERN_EMERG is warranted for this kind of error.

I'm not sure either - I left it as it was before.

My vote is to change it to KERN_ERR, but it's your call.

quoted

So if a node has an fsl,rstcr property, but the of_iomap() fails, we
jump to the next global-utilities node? =A0Perhaps you need a 'break'
after the printk()?

Or potentially a continue to be more robust? Or would two (or more) "has-=

rstcr" nodes be wrong?

My point is that if the of_iomap() call fails, looking for another
global-utilities node is not the right course of action.  of_iomap()
failing is a serious error that should result in an immediate exit.

quoted

Wouldn't it make more sense to assign fsl_rstcr_restart to
ppc_md.restart only if we find a valid fsl,has-rstcr property?

Again I'm not entirely sure, I left this as it was before. Is there anoth=

er way to reset the board if the rstcr node was not found correctly?

Not on our 85xx boards.  On 83xx, there's mpc83xx_restart().  I just
don't like "ppc_md.restart =3D=3D fsl_rstcr_restart", because it assumes
that the define_machine() entry is incorrect.

--=20
Timur Tabi
Linux kernel developer at Freescale