Re: 2.6.9-mm1

From: Andrew Morton
Date: Sun Oct 24 2004 - 03:52:30 EST


Alexander Nyberg <alexn@xxxxxxxxx> wrote:
>
> > - kexec and crashdump: this all allegedly works, but I want to *see* it
> > work first.
>
> - sys_reboot() calls device_shutdown() which naturally makes my disks go
> to sleep and immediatly after spin up when the disk initialization code
> comes in. Is there some specific reason why this is needed? Appears to
> work for me just removing the function call.

There's special-case code in the IDE driver to prevent this. The kexec
code manages to defeat it. This should fix:

--- 25/kernel/sys.c~kexec-ide-spindown-fix 2004-10-24 01:46:52.028277048 -0700
+++ 25-akpm/kernel/sys.c 2004-10-24 01:47:20.911886072 -0700
@@ -531,8 +531,9 @@ asmlinkage long sys_reboot(int magic1, i
return -EINVAL;
}
notifier_call_chain(&reboot_notifier_list, SYS_RESTART, NULL);
- system_state = SYSTEM_BOOTING;
+ system_state = SYSTEM_RESTART;
device_shutdown();
+ system_state = SYSTEM_BOOTING;
printk(KERN_EMERG "Starting new kernel\n");
machine_shutdown();
machine_kexec(image);
_

> - 3c59x driver together with my 3c509C-TX card hits:
> "ff:ff:ff:ff:ff:ff<3>*** EEPROM MAC address invalid"
> after doing a kexec-reboot. I tried reseting it at bootup but I couldn't
> get it kicking. I couldn't find any specs nor maintainer for this one...
>

3c59x power management is a bit flakey. I'll have a poke at this when I
get onto playing with kexec.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/