that is an option as well off course :)
I'll see if i can spin patches like that if ever i find the time for it
> On Wed, May 11, 2011 at 02:07:53PM +0200, Rik Bobbaers wrote:
>> your patch won't work for a lot of instances...
>> one of the most important errors you made, is not changing the
>> atomic values, used by vserver to unchecked atomic values.
>> Bertl uses those counters so that they can (and are supposed
>> to) overflow. Therefore you have to change them to "unchecked"
>> atomics. You will get REFCOUNT overflow bugs in your kernel
>> (and corresponding panics).
>> So please... read my "vserver+grsec howto" from a couple of
>> months ago before taking this project on yourself ;)
> in general, I'd suggest to do patches _against_ a Linux-VServer
> patched kernel and not combination patches, as it simplifies
> the process of updating to newer Linux-VServer and kernel
> versions ...
>> read this thread first:
>> Rik Bobbaers
>> -- http://harry.enzoverder.be
>>> -- Sandino Araico Sánchez http://sandino.net
Received on Wed May 11 15:15:11 2011