Re: INT 18h BIOS call

[Jun OKAJIMA]

>Greetings developers; I have sort of a question and feature request. My 
>question is why GRUB stage1 does not use the BIOS INT 18h call on failure 
>so that the BIOS knows that the bootstrap has failed and can either try 
>the next bootable device or ask the user if they want to retry the boot 
>sequence (whatever the BIOS is configured to do) instead of just sitting 
>in a loop? It would be much more robust to use this feature, especially in 
>high-availability environments. Any response is greatly appreciated.
>Thank you.
>- David Barksdale
>address@hidden
>


I am not sure what INT 18h really does, but sounds
very good to my purpose.

What we need is, safe way to upgrade kernel on a dedicated server.
It is easy if you were a guy who never make any mistake, but actually not.
# vi /boot/grub/menu.lst
# reboot ...  wait, wait, wait a few minutes.
then, you can get the server coming up? If not, how to fix it?

What we want is, in the case it does not boot, you can boot the alternative
( it would be an old kernel) in the next booting automatically.
If this way works, what we have to do in a boot problem is just pushing a reset
button by e-mailing to an iDC company.

Your INT 18h way helps this way?

          --- Okajima, Jun. Tokyo, Japan.