Re: 2.6.6-mm2

[Bill Davidsen <davidsen@tmr.com>]

Adrian Bunk wrote:
> On Thu, May 13, 2004 at 09:39:32AM -0400, Chris Mason wrote:
> 
>>On Thu, 2004-05-13 at 09:18, Adrian Bunk wrote:
>>
>>>On Thu, May 13, 2004 at 04:25:40AM -0700, Andrew Morton wrote:
>>>
>>>>...
>>>>Wim explained that any application changes now won't be widely deployed for
>>>>another year.  During that period the ability to run existing Oracle setups
>>>>requires that hugepage allocation be available to unprivileged
>>>>applications.
>>>>...
>>>>It means that if people install a kernel.org machine on their database
>>>>server, the database *just won't work*.  This is not good for those users,
>>>>for the kernel developers or for Linux's reputation in general.
>>>>...
>>>
>>>That sounds silly when talking about Oracle.
>>>
>>>Oracle says:
>>>  Which Kernels are supported?
>>>
>>>  Oracle does not support modified or recompiled kernels. Recompiled 
>>>  kernels are not supported with or without source modifications.
>>>
>>>
>>>I doubt there are many "existing Oracle setups" that will risk to lose 
>>>all Oracle support by installing a different kernel.
>>>
>>
>>No, I doubt so as well.  Then again, why force them into a vendor
>>kernel?  At the very least, it would be nice to be able to benchmark
>>vanilla against the vendors.
>>...
> 
> 
> I think I recall times when code contributions to the kernel were only 
> judged by their quality and not by the needs of some non-free apps or 
> what vendors did.
> 
> Either my memory is wrong, or these times are gone now...

I don't see that "quality" and "what vendors did" are mutually 
exclusive. What I don't see is why you think that having a capability 
control this is a bad thing. It would seem to be exactly the type of 
thing capabilities address, giving a selected bit of permission to a 
trusted application.

-- 
    -bill davidsen (davidsen@tmr.com)
"The secret to procrastination is to put things off until the
  last possible moment - but no longer"  -me