Linus is just telling you that trying to write to one
of dozens of possible hd driver interfaces from
a kernel that has just paniced is gambling with
"the money", and he isn't willing to merge (and
would not himself use) a patch that takes that risk.
Like he says, the situation is different on systems
that only have drivers for a tiny number of different
disk interfaces, and objecting to his attitude while
ignoring this issue is arguing in bad faith.
If you want to dump to the network or to a serial
port, that's different. How many different serial
port drivers are there?
The robust solution is to hook that serial port up
to a cheap little box sleeping on a serial port
interrupt whose only job is to react to that interrupt
and capture the crash dump to its own local disk,
which isn't being managed by a kernel that just
paniced. (If you want to give it a send-only network
interface to the outside world for rapid notification,
that's up to you. You could put ssh on it too, but the
more you add, the less robust a solution it is to its
real job, which is providing a stable interface to a
disk for crash dumps.)
Regards,
Clayton Weaver
<mailto: cgweav@email.com>
PS: This is my first message from email.com. Let me
know if it shows up in html or has any other
obnoxious features characteristic of clueless
email clients.
-- _______________________________________________ Sign-up for your own FREE Personalized E-mail at Mail.com http://www.mail.com/?sr=signupSingle & ready to mingle? lavalife.com: Where singles click. Free to Search! http://www.lavalife.com/mailcom.epl?a=2116
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/