Hi York, On 25 April 2014 16:29, York Sun <york...@freescale.com> wrote: > On 03/22/2014 04:14 PM, Simon Glass wrote: >> We should move forward to remove the old board init code. Add a >> prominent message to encourage maintainers to get started on this >> work. >> >> Signed-off-by: Simon Glass <s...@chromium.org> >> --- >> > Simon, > > How much test has been done on powerpc boards? I randomly pick one MPC8536DS > and > it doesn't work. I have spent hours to track the difference and I am still > struggling. I understand the first one could be more work, but if you have > done > any, especially 85xx family, please let me know.
I don't have any of those boards. I believe Wolfgang tested three of them some time ago. See my other message - board_init_f() doesn't zero the global_data structure. It needs to be zeroed so make sure it is done elsewhere. Originally the code uses the cache which may have been zeroed, but now global_data is on the stack (still cache but maybe a different part?). Sorry I don't understand PowerPC very well, so I'm not sure of the pre-conditions. Also, can you please be more specific about the failure? Regards, Simon _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
