From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00
	autolearn=unavailable autolearn_force=no version=3.4.4
Path: 
 eternal-september.org!reader01.eternal-september.org!feeder.eternal-september.org!news.swapon.de!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
From: Niklas Holsti <niklas.holsti@tidorum.invalid>
Newsgroups: comp.lang.ada
Subject: Re: Why couldn't an operating system be written in ada
Date: Thu, 28 Feb 2019 00:12:16 +0200
Organization: Tidorum Ltd
Message-ID: <gdoga0F4ia6U1@mid.individual.net>
References: <4s8rud$9j3@tribune>
 <792fba1b-7a54-4d00-ae85-e6bd0737f001@googlegroups.com>
 <gdkcmuF7b4gU1@mid.individual.net> <q54h7l$dmm$1@gioia.aioe.org>
 <1d9e7148-cc7d-423b-9c4a-2372d1af4ddd@googlegroups.com>
 <q56pnt$m6p$1@gioia.aioe.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
X-Trace: individual.net aGhgYgXE5+fyZyvnpbgO8gCTsr1f80JPVZhDudwlcRhrXwudZq
Cancel-Lock: sha1:y2B3ggHJC0Hif4ZkXCdekJ0ZhZA=
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:45.0)
 Gecko/20100101 Thunderbird/45.8.0
In-Reply-To: <q56pnt$m6p$1@gioia.aioe.org>
Xref: reader01.eternal-september.org comp.lang.ada:55713
Date: 2019-02-28T00:12:16+02:00
List-Id: <comp.lang.ada>

On 19-02-27 21:51 , russ lyttle wrote:
> On 2/27/19 2:10 PM, Shark8 wrote:
>>
>> Tasks would be ideal for a nice chunk OS design, especially
>> considering the move to multicore architectures.
>>
> I mean that the kernel should provide tasking, but tasks would not be
> required to write the kernel. It's a bit off topic for an Ada group, but
> just what would a kernel with native multi-core support look like?

The AdaCore run-time system (which implements tasks) that I have looked 
at in more detail -- the Small Foot-Print Ravenscar RTS for LEON2/LEON3 
SPARCs -- has optional multicore support; it didn't seem to add much 
code/logic to the basic code. Separate "ready queues" for each core is a 
straight-forward concept.

The implementation of protected objects may have to be somewhat 
different. In a single-core system, mutual exclusion can be implemented 
by the "ceiling priority" method, but that doesn't work in multi-core 
systems, so more conventional locks must likely be used.

-- 
Niklas Holsti
Tidorum Ltd
niklas holsti tidorum fi
       .      @       .