8 Aug
2008
8 Aug
'08
3:14 p.m.
On Aug 7, 2008, at 9:37 PM, Jerry Van Baren wrote:
Kumar Gala wrote:
One of the things that wasn't clear to me is if we are ok with maintaining state between 'bootm' subcommand inside u-boot or if we really require passing all state via arguments and env. While I know it would be nice if the subcommands were stateless I dont think this is practical. state we'd have to keep track of:
- arguments to the "top level" bootm command
- type of arguments (fit vs plain addresses)
- Image information, for FIT we get something like:
[snip of killer state information]
- entry point of OS image
- region tracking of memory regions used by previous subcommands
(OS image, bd_t, fdt, initrd, etc.) This seems like a lot of state to pass around in the env and via arguments to commands. My vote is for stateful sub_commands.
- k
Useful info and analysis. I agree, it looks like we will need to be stateful.
Wolfgang do you agree?
- k