On 1/27/07, [email protected] <[email protected]> wrote:
> diff -puN drivers/sbus/char/openprom.c~dynamic-kernel-command-line-sparc64-fix drivers/sbus/char/openprom.c
> --- a/drivers/sbus/char/openprom.c~dynamic-kernel-command-line-sparc64-fix
> +++ a/drivers/sbus/char/openprom.c
> @@ -283,7 +283,7 @@ static int oprompath2node(void __user *a
>
> static int opromgetbootargs(void __user *argp, struct openpromio *op, int bufsize)
> {
> - char *buf = saved_command_line;
> + char *buf = boot_command_line;
> int len = strlen(buf);
>
> if (len > bufsize)
Hello Andrew,
I don't think this is correct.
The boot_command_line should only be used by init code.
Drivers should continue to use saved_command_line.
I don't know why saved_command_line cannot be resolved and
boot_command_line can be.
They are both located in linux/init.h
extern char __initdata boot_command_line[];
extern char *saved_command_line;
And the drivers/sbus/char/openprom.c do include it and I assume
__ASSEMBLY__ is not defined.
Best Regards,
Alon Bar-Lev.
On Sat, 27 Jan 2007 13:14:57 +0200
"Alon Bar-Lev" <[email protected]> wrote:
> I don't know why saved_command_line cannot be resolved and
> boot_command_line can be.
> They are both located in linux/init.h
>
> extern char __initdata boot_command_line[];
> extern char *saved_command_line;
>
> And the drivers/sbus/char/openprom.c do include it and I assume
> __ASSEMBLY__ is not defined.
hm. It went away. Maybe I dreamed it.
y'know, I've been plugging away at this crappile for over twelve hours,
have applied 55 distinct compile fixes so far and it's still not there.
Tomorrow I get to try to make it all run. That should be amusing.