Btw, just wante dto say I really love microkernels.  I like how clean they
keep things by having packages of code neatly contained and isolated.

Good work, guys.

On Fri, Aug 14, 2015 at 10:50 AM, Raymond Jennings <[email protected]>
wrote:

> Greatly appreciated.
>
> I'll definitely be taking a look at it.
>
> As far as funding, I'm actually just a hobbyist interested in microkernels
> and I am not here in any corporate or business capacity.
>
> I suppose someone could always add A/D bit access themselves by submitting
> a pull request on github?
>
> On Fri, Aug 14, 2015 at 6:56 AM, Gernot Heiser <[email protected]>
> wrote:
>
>> Hi Raymond,
>>
>> I think that allowing the hardware bits to be queried (and probably
>> changed atomically while being read) would be useful.
>>
>> Would such a feature be added in the future?  I have a crazy hunch (but
>> nowhere near the experience to prove it) that avoiding such "canary" style
>> page faults could optimize things.
>>
>>
>> As Kevin said: it’s an optimisation that was until recently only
>> available on x86 processors. Because its an optimisation, it adds no
>> functionality, only performance.
>>
>> We take performance very seriously. In fact, while seL4’s design is
>> driven primarily by security, our philosophy is also that security is no
>> excuse for bad performance.
>>
>> But, as Kevin also said, we’ve yet to see an important use case that will
>> significantly benefit from this optimisation. There is no reason why we
>> couldn’t support this optimisation, but, until someone convinces us that
>> it’ll make a real difference, we have far more important work to do. A
>> “would be nice” is not sufficient, I’m afraid.
>>
>> So, if you have an important and convincing use case that depends on
>> this, then let us know. If you can come up with funding for the work, and
>> we’re convinced it really helps, we’ll increase the priority.
>>
>> Soon (probably later today) I’ll post our development roadmap on
>> [email protected]. You’ll see that we’ve got plenty on our plate,
>> all of which is important functionality that indisputably adds to seL4’s
>> real-world usefulness, and therefore has priority. In fact, if we had more
>> people, we would put them on accelerating the roadmap before we work on
>> nice-to-haves.
>>
>> Gernot
>>
>> ------------------------------
>>
>> The information in this e-mail may be confidential and subject to legal
>> professional privilege and/or copyright. National ICT Australia Limited
>> accepts no liability for any damage caused by this email or its attachments.
>>
>
>
_______________________________________________
Devel mailing list
[email protected]
https://sel4.systems/lists/listinfo/devel

Reply via email to