Hi all!
commit in mainline 10a0a8d4e3f6bf2d077f94344441909abe670f5a is go in
the satble 2.6.22
the grund for this question is http://hup.hu/node/49773 .
--
Thanks,
Oliver
Oliver Pinter (Pint?r Oliv?r) wrote:
> Hi all!
>
> commit in mainline 10a0a8d4e3f6bf2d077f94344441909abe670f5a is go in
> the satble 2.6.22
>
> the grund for this question is http://hup.hu/node/49773 .
>
Is there a compelling reason? I added it to remove some duplicate code,
but it isn't a bugfix or anything.
J
On 1/24/08, Jeremy Fitzhardinge <[email protected]> wrote:
> Oliver Pinter (Pint?r Oliv?r) wrote:
> > Hi all!
> >
> > commit in mainline 10a0a8d4e3f6bf2d077f94344441909abe670f5a is go in
> > the satble 2.6.22
> >
> > the grund for this question is http://hup.hu/node/49773 .
> >
>
> Is there a compelling reason?
no, but in forum some person have the question, why the kernel call
userspace program and it is secure or not...
I added it to remove some duplicate code,
> but it isn't a bugfix or anything.
>
> J
>
>
--
Thanks,
Oliver
Oliver Pinter (Pint?r Oliv?r) wrote:
> On 1/24/08, Jeremy Fitzhardinge <[email protected]> wrote:
>
>> Oliver Pinter (Pint?r Oliv?r) wrote:
>>
>>> Hi all!
>>>
>>> commit in mainline 10a0a8d4e3f6bf2d077f94344441909abe670f5a is go in
>>> the satble 2.6.22
>>>
>>> the grund for this question is http://hup.hu/node/49773 .
>>>
>>>
>> Is there a compelling reason?
>>
>
> no, but in forum some person have the question, why the kernel call
> userspace program and it is secure or not...
>
Ah, my non-existent knowledge of Hungarian wasn't up to interpreting
that page ;)
It should be secure if done properly. Of course, if you set the
"poweroff" executable to something untrustworthy then you'll have a
problem, but there's no inherent security problem there. And if you
want to do an orderly shutdown, you need to have some amount of help
from the userspace side of things, and invoking a command is the normal
way to do this.
J
On 1/24/08, Jeremy Fitzhardinge <[email protected]> wrote:
> Oliver Pinter (Pint?r Oliv?r) wrote:
> > On 1/24/08, Jeremy Fitzhardinge <[email protected]> wrote:
> >
> >> Oliver Pinter (Pint?r Oliv?r) wrote:
> >>
> >>> Hi all!
> >>>
> >>> commit in mainline 10a0a8d4e3f6bf2d077f94344441909abe670f5a is go in
> >>> the satble 2.6.22
> >>>
> >>> the grund for this question is http://hup.hu/node/49773 .
> >>>
> >>>
> >> Is there a compelling reason?
> >>
> >
> > no, but in forum some person have the question, why the kernel call
> > userspace program and it is secure or not...
> >
>
> Ah, my non-existent knowledge of Hungarian wasn't up to interpreting
> that page ;)
>
> It should be secure if done properly. Of course, if you set the
> "poweroff" executable to something untrustworthy then you'll have a
> problem, but there's no inherent security problem there. And if you
> want to do an orderly shutdown, you need to have some amount of help
> from the userspace side of things, and invoking a command is the normal
> way to do this.
yes, i know this
>
> J
>
the original topic name is : why linux kernel crap ...
the substance is, when from kernel call users-pace program, then come
up more problems with security...
when the fs crash out or the userspace code is breaked, then the
system not shuting down ..
and when it is only code duplication cleanup, then is unneaded for
2.6.22, thanks and, sorry for bad spelling.
--
Thanks,
Oliver (pinyo_villany)
Oliver Pinter (Pint?r Oliv?r) wrote:
> the original topic name is : why linux kernel crap ...
>
> the substance is, when from kernel call users-pace program, then come
> up more problems with security...
> when the fs crash out or the userspace code is breaked, then the
> system not shuting down ..
>
orderly_poweroff() has the option to force a hard shutdown if the exec
of the usermode command fails.
J
thanks
On 1/24/08, Jeremy Fitzhardinge <[email protected]> wrote:
> Oliver Pinter (Pint?r Oliv?r) wrote:
> > the original topic name is : why linux kernel crap ...
> >
> > the substance is, when from kernel call users-pace program, then come
> > up more problems with security...
> > when the fs crash out or the userspace code is breaked, then the
> > system not shuting down ..
> >
>
> orderly_poweroff() has the option to force a hard shutdown if the exec
> of the usermode command fails.
>
> J
>
--
Thanks,
Oliver