29 Apr
2011
29 Apr
'11
10:24 p.m.
On Fri, 29 Apr 2011 15:16:59 -0500 Kumar Gala galak@kernel.crashing.org wrote:
On Apr 29, 2011, at 12:36 PM, Tabi Timur-B04825 wrote:
What about this patch:
http://patchwork.ozlabs.org/patch/71614/
It was ACK'd by gvb:
http://lists.denx.de/pipermail/u-boot/2011-April/089791.html
I'd rather we just fix the addresses up rather than verifying them.
As I pointed out before, you can't fix up every possible error. If someone passes a 32-bit device tree to a 36-bit U-Boot, we can't fix that up sanely. But we can detect it, and give a useful message rather than a hang.
Perhaps some subset of the things that are checked (e.g. PCI addresses) would me more appropriately updated by U-Boot, though.
-Scott