Re: [PATCH 1/2] ia64 kdump: add kdump_on_fatal_mca

[Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>]

>> I saw the proposed
>> patch that changes the current behavior without mentioning it
>> at the summary, so i brought it up. ;)

Sorry about that.  It's my fault.

But as the result it was lucky that we can have a good
discussion about "required action after a fatal MCA".

> So far SGI are voting for enabling dumping for MCA, and
> Fujitsu have implicitly voted to disable (as that is what
> Seto-san's patch sets).  Any other opinions?

Ah, my vote can be invalid.
AFAIK, Fujitsu have both situation where kdump_on_fatal_mca
should be enabled and where it should be disabled.
Therefore I don't mind which is the default.
Anyway we need to put a line to /etc/sysctl.conf of one of both.

So now "enable vs disable" is not 1:1 but 1:0 or 1.5:0.5.

Tony, (if there are no other votes until your poll closing)
would you apply my patch modifying the default to 1?
Or I need to resend you all just modified the one byte?

Thanks,
H.Seto