Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Fernando Oleo / Irvise]

[[Original post can be found in 
https://forum.ada-lang.io/t/ironclad-the-hard-real-time-capable-posix-like-kernel-written-in-spark-ada-received-an-nlnet-grant/1281]]

Dear community,

it fills me with joy to announce that Streaksu, the developer of 
Ironclad [1] has been given a grant from the nlnet foundation [2], 
announcement here [3].

If you would like to learn more about Ironclad and the Gloire 
distribution, Streaksu presented the project during AEiC 2024. You can 
find the slides and video here [4].

If you are interested in Ironclad/Gloire, you can check the git repos 
from the main webpage, join the Matrix/Element chatroom or maybe even 
sponsor Streaksu over at his liberapay page [5].

I am very happy to see that Streaksu applied for a grant and was 
accepted. This is what I want to see more of in the community. People 
creating very cool projects that they enjoy working on. Then promoting 
them and seeing if they can make the development of such projects a 
little bit more sustainable for themselves. This also disproves that 
“Ada projects” are not liked or as seen as old by foundations or other 
funding groups.

I am very happy for Streaksu and I hope to test Ironclad in a RISC-V 
board soon!

[1] https://ironclad.nongnu.org/
[2] https://nlnet.nl/
[3] https://nlnet.nl/news/2024/20241003-announcing-Core-call.html
[4] https://www.ada-europe.org/conference2024/adadev.html
[5] https://liberapay.com/streaksu/

Best regards and happy hacking!
Fer

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Thu, 3 Oct 2024 19:38:31 +0200, Fernando Oleo / Irvise wrote:

> [1] https://ironclad.nongnu.org/

It’s not microkernel-based, is it?

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 03/10/2024 23:12, Lawrence D'Oliveiro wrote:
> On Thu, 3 Oct 2024 19:38:31 +0200, Fernando Oleo / Irvise wrote:
> 
>> [1] https://ironclad.nongnu.org/
> 
> It’s not microkernel-based, is it?

That's something I intend to work on soon.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Fernando Oleo / Irvise]

On 10/4/24 00:12, Lawrence D'Oliveiro wrote:
> On Thu, 3 Oct 2024 19:38:31 +0200, Fernando Oleo / Irvise wrote:
> 
>> [1] https://ironclad.nongnu.org/
> 
> It’s not microkernel-based, is it?

AFAIK, no. It is POSIX-based/like. It is able to run quite a few *NIX 
applications. You can check the Gloire distribution [1] to test it out :)

If you are interested in microkernels written in Ada, see [2]

[1] https://github.com/Ironclad-Project/Gloire
[2] https://github.com/ohenley/awesome-ada?tab=readme-ov-file#os-and-kernels

Best regards,
Fer

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Kevin Chadwick]

>> [1] https://ironclad.nongnu.org/
>
>It’s not microkernel-based, is it?

Isn't it true that monolithic kernels become more attractive when Cs
 problens are removed with micro kernels swapping problems for new problems?

-- 
Regards, Kc

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Fri, 4 Oct 2024 18:56:58 +0200, Fernando Oleo / Irvise wrote:

> On 10/4/24 00:12, Lawrence D'Oliveiro wrote:
>>
>> On Thu, 3 Oct 2024 19:38:31 +0200, Fernando Oleo / Irvise wrote:
>> 
>>> [1] https://ironclad.nongnu.org/
>> 
>> It’s not microkernel-based, is it?
> 
> AFAIK, no. It is POSIX-based/like.

Pleased to hear that. ;)

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Fri, 4 Oct 2024 19:52:12 -0000 (UTC), Kevin Chadwick wrote:

> Isn't it true that monolithic kernels become more attractive when Cs
> problens are removed with micro kernels swapping problems for new
> problems?

The microkernel proponents still seem to think there is a point to their 
idea, even after decades of real-world experience to the contrary.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 04/10/2024 21:05, Lawrence D'Oliveiro wrote:
> On Fri, 4 Oct 2024 19:52:12 -0000 (UTC), Kevin Chadwick wrote:
> 
>> Isn't it true that monolithic kernels become more attractive when Cs
>> problens are removed with micro kernels swapping problems for new
>> problems?
> 
> The microkernel proponents still seem to think there is a point to their
> idea, even after decades of real-world experience to the contrary.

L4 have years of sticking a middle finger up at that.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Paul Rubin]

"Luke A. Guest" <laguest@archeia.com> writes:
> L4 have years of sticking a middle finger up at that.

I also have to wonder whether hypervisors count as microkernels.
Everyone is using them now.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Fri, 4 Oct 2024 23:19:09 +0100, Luke A. Guest wrote:

> On 04/10/2024 21:05, Lawrence D'Oliveiro wrote:
>
>> The microkernel proponents still seem to think there is a point to
>> their idea, even after decades of real-world experience to the
>> contrary.
> 
> L4 have years of sticking a middle finger up at that.

L4 is supposedly being used as the basis of the GNU Hurd kernel. 
Development of that started around the same time as Linux. There are grown 
adults walking the Earth right now who weren’t even born at that time, 
many of them now using Linux for production work, and Hurd still isn’t 
ready for prime time.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 05/10/2024 09:11, Lawrence D'Oliveiro wrote:
> On Fri, 4 Oct 2024 23:19:09 +0100, Luke A. Guest wrote:
> 
>> On 04/10/2024 21:05, Lawrence D'Oliveiro wrote:
>>
>>> The microkernel proponents still seem to think there is a point to
>>> their idea, even after decades of real-world experience to the
>>> contrary.
>>
>> L4 have years of sticking a middle finger up at that.
> 
> L4 is supposedly being used as the basis of the GNU Hurd kernel.
> Development of that started around the same time as Linux. There are grown
> adults walking the Earth right now who weren’t even born at that time,
> many of them now using Linux for production work, and Hurd still isn’t
> ready for prime time.

Hurd is never going to happen, because they keep starting, stopping, 
changing kernels, etc.

Anyway, I'm not really interested in yet another nix.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[DrPi]

Le 04/10/2024 à 22:05, Lawrence D'Oliveiro a écrit :
> On Fri, 4 Oct 2024 19:52:12 -0000 (UTC), Kevin Chadwick wrote:
> 
>> Isn't it true that monolithic kernels become more attractive when Cs
>> problens are removed with micro kernels swapping problems for new
>> problems?
> 
> The microkernel proponents still seem to think there is a point to their
> idea, even after decades of real-world experience to the contrary.
Any evidence of this assertion ?

You should try QNX.
My experience with QNX shows that it is far more stable than monolithic 
kernels since buggy drivers can't cause the kernel to panic.
Also, you don't have to recompile the kernel each time a driver needs to 
be recompiled.
I have many other arguments against monolithic kernels.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 05/10/2024 17:24, DrPi wrote:
> Le 04/10/2024 à 22:05, Lawrence D'Oliveiro a écrit :
>> On Fri, 4 Oct 2024 19:52:12 -0000 (UTC), Kevin Chadwick wrote:
>>
>>> Isn't it true that monolithic kernels become more attractive when Cs
>>> problens are removed with micro kernels swapping problems for new
>>> problems?
>>
>> The microkernel proponents still seem to think there is a point to their
>> idea, even after decades of real-world experience to the contrary.
> Any evidence of this assertion ?
> 
> You should try QNX.
> My experience with QNX shows that it is far more stable than monolithic 
> kernels since buggy drivers can't cause the kernel to panic.
> Also, you don't have to recompile the kernel each time a driver needs to 
> be recompiled.
> I have many other arguments against monolithic kernels.

Yeah, QNX was solid. Had the source at one point.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Sat, 5 Oct 2024 11:47:41 +0100, Luke A. Guest wrote:

> Hurd is never going to happen, because they keep starting, stopping,
> changing kernels, etc.

Hurd is never going to happen, because they are trying to build it as a 
microkernel.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Sat, 5 Oct 2024 18:24:39 +0200, DrPi wrote:

> Le 04/10/2024 à 22:05, Lawrence D'Oliveiro a écrit :
>
>> The microkernel proponents still seem to think there is a point to
>> their idea, even after decades of real-world experience to the
>> contrary.
>
> Any evidence of this assertion ?

Look around you, at what happened when people tried to use microkernels in 
real-world situations. I think Apple tried to use one in its “macOS” (née 
“OS X”), and performance suffered as a result.

> You should try QNX.

Was that used in any high-performance situation?

> Also, you don't have to recompile the kernel each time a driver needs to
> be recompiled.

Linux has supported loadable modules for maybe 30 years now.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[DrPi]

Le 06/10/2024 à 01:10, Lawrence D'Oliveiro a écrit :
> On Sat, 5 Oct 2024 18:24:39 +0200, DrPi wrote:
> 
>> Le 04/10/2024 à 22:05, Lawrence D'Oliveiro a écrit :
>>
>>> The microkernel proponents still seem to think there is a point to
>>> their idea, even after decades of real-world experience to the
>>> contrary.
>>
>> Any evidence of this assertion ?
> 
> Look around you, at what happened when people tried to use microkernels in
> real-world situations. I think Apple tried to use one in its “macOS” (née
> “OS X”), and performance suffered as a result.
MacOS-X is a hybrid kernel. Half way betwwen micro-kernel and monolithic 
kernel.

> 
>> You should try QNX.
> 
> Was that used in any high-performance situation?
Sure. QNX is designed for hard real time.

> 
>> Also, you don't have to recompile the kernel each time a driver needs to
>> be recompiled.
> 
> Linux has supported loadable modules for maybe 30 years now.
Yes, they exist but they are some sort of exception.
Look at linux kernel release logs like this one 
https://www.cnx-software.com/2024/09/16/linux-6-11-release-notable-changes-arm-risc-v-and-mips-architectures/
Most of the log content is about drivers.
And when time comes to debugging a kernel driver...

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[DrPi]

Le 06/10/2024 à 15:53, DrPi a écrit :
> Le 06/10/2024 à 01:10, Lawrence D'Oliveiro a écrit :
>> On Sat, 5 Oct 2024 18:24:39 +0200, DrPi wrote:
>>
>>> Le 04/10/2024 à 22:05, Lawrence D'Oliveiro a écrit :
>>>
>>>> The microkernel proponents still seem to think there is a point to
>>>> their idea, even after decades of real-world experience to the
>>>> contrary.
>>>
>>> Any evidence of this assertion ?
>>
>> Look around you, at what happened when people tried to use 
>> microkernels in
>> real-world situations. I think Apple tried to use one in its “macOS” (née
>> “OS X”), and performance suffered as a result.
Oh, I forgot, Minix (which is a micro-kernel OS) is embedded is every 
Intel x86 processor.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 06/10/2024 14:53, DrPi wrote:
> Le 06/10/2024 à 01:10, Lawrence D'Oliveiro a écrit :
>> On Sat, 5 Oct 2024 18:24:39 +0200, DrPi wrote:
>>
>>> Le 04/10/2024 à 22:05, Lawrence D'Oliveiro a écrit :
>>>
>>>> The microkernel proponents still seem to think there is a point to
>>>> their idea, even after decades of real-world experience to the
>>>> contrary.
>>>
>>> Any evidence of this assertion ?
>>
>> Look around you, at what happened when people tried to use 
>> microkernels in
>> real-world situations. I think Apple tried to use one in its “macOS” (née
>> “OS X”), and performance suffered as a result.
> MacOS-X is a hybrid kernel. Half way betwwen micro-kernel and monolithic 
> kernel.

MacOS uses Mach which is well known for being a terrible implementation 
of the microkernel, that's why it's a hybrid, same reason NT got changed 
into a hybrid too.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Sun, 6 Oct 2024 15:53:25 +0200, DrPi wrote:

> Le 06/10/2024 à 01:10, Lawrence D'Oliveiro a écrit :
>>
>> On Sat, 5 Oct 2024 18:24:39 +0200, DrPi wrote:
>> 
>>> You should try QNX.
>> 
>> Was that used in any high-performance situation?
>
> Sure. QNX is designed for hard real time.

Not the same thing. “Hard real time” does not imply “high performance”. 
Ask the folks working on the very subject of this thread.

Is the QNX “microkernel” similar in style to the old Amiga “microkernel”? 
That is, instead of copying message buffers around, it just passed around 
pointers to those message buffers? That would give you speed, but 
sacrifice much of the robustness that a microkernel is supposed to offer.

>> Linux has supported loadable modules for maybe 30 years now.
>
> Yes, they exist but they are some sort of exception.

No they are not. They typically make up the majority of the kernel. 
Looking at my current kernel, the base kernel file is less than 10MB in 
size, whereas the directory containing loadable modules for that kernel is 
about 10× that.

Of course, this is your choice; some prefer to build all the options they 
need into a single monolithic image, for faster loading. Linux is all 
about having the choice.

> And when time comes to debugging a kernel driver...

Hurd should have found that easy to do, if microkernels really make things 
easy to do. But it didn’t.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Sun, 6 Oct 2024 16:19:29 +0200, DrPi wrote:

> Oh, I forgot, Minix (which is a micro-kernel OS) is embedded is every
> Intel x86 processor.

Bad example. <https://www.theregister.com/2018/08/29/intel_jtag_flaw/>

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Sun, 6 Oct 2024 20:46:16 +0100, Luke A. Guest wrote:

> MacOS uses Mach which is well known for being a terrible implementation
> of the microkernel, that's why it's a hybrid, same reason NT got changed
> into a hybrid too.

And yet they are both still outperformed by Linux on the same hardware.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 06/10/2024 22:30, Lawrence D'Oliveiro wrote:
> On Sun, 6 Oct 2024 20:46:16 +0100, Luke A. Guest wrote:
> 
>> MacOS uses Mach which is well known for being a terrible implementation
>> of the microkernel, that's why it's a hybrid, same reason NT got changed
>> into a hybrid too.
> 
> And yet they are both still outperformed by Linux on the same hardware.

What's your point? I said Mach is the worst example of a microkernel, 
it's been proven, decades ago.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Mon, 7 Oct 2024 00:02:34 +0100, Luke A. Guest wrote:

> On 06/10/2024 22:30, Lawrence D'Oliveiro wrote:
>>
>> On Sun, 6 Oct 2024 20:46:16 +0100, Luke A. Guest wrote:
>> 
>>> MacOS uses Mach which is well known for being a terrible
>>> implementation of the microkernel, that's why it's a hybrid, same
>>> reason NT got changed into a hybrid too.
>> 
>> And yet they are both still outperformed by Linux on the same hardware.
> 
> What's your point? I said Mach is the worst example of a microkernel,
> it's been proven, decades ago.

So where is there a better one? It’s long been established that 
microkernel performance is terrible, and the theoretical reliability 
advantages have failed to materialize. What reason is there left to use 
them? None.

After 40 or more years trying to tout the idea, it’s time to give up.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:
> On Mon, 7 Oct 2024 00:02:34 +0100, Luke A. Guest wrote:
> 
>> On 06/10/2024 22:30, Lawrence D'Oliveiro wrote:
>>>
>>> On Sun, 6 Oct 2024 20:46:16 +0100, Luke A. Guest wrote:
>>>
>>>> MacOS uses Mach which is well known for being a terrible
>>>> implementation of the microkernel, that's why it's a hybrid, same
>>>> reason NT got changed into a hybrid too.
>>>
>>> And yet they are both still outperformed by Linux on the same hardware.
>>
>> What's your point? I said Mach is the worst example of a microkernel,
>> it's been proven, decades ago.
> 
> So where is there a better one? It’s long been established that
> microkernel performance is terrible, and the theoretical reliability
> advantages have failed to materialize. What reason is there left to use
> them? None.
> 
> After 40 or more years trying to tout the idea, it’s time to give up.

I already said, L4. But you don't want to listen you just want to slag 
off something you don't or won't understand and bang on about Linux, 
which was always meant to be a server OS, not a desktop one.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:

> After 40 or more years trying to tout the idea, it’s time to give up.

Why are you wanging on about this on an Ada group? Try this on 
alt.os.development.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 07/10/2024 09:21, Luke A. Guest wrote:
>> After 40 or more years trying to tout the idea, it’s time to give up.
> 
> I already said, L4. But you don't want to listen you just want to slag 
> off something you don't or won't understand and bang on about Linux, 
> which was always meant to be a server OS, not a desktop one.

There's also MorphOS which is based on L3.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Chris Townley]

On 07/10/2024 09:22, Luke A. Guest wrote:
> On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:
> 
>> After 40 or more years trying to tout the idea, it’s time to give up.
> 
> Why are you wanging on about this on an Ada group? Try this on 
> alt.os.development.

He is just trolling!

-- 
Chris

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 07/10/2024 11:25, Chris Townley wrote:
> On 07/10/2024 09:22, Luke A. Guest wrote:
>> On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:
>>
>>> After 40 or more years trying to tout the idea, it’s time to give up.
>>
>> Why are you wanging on about this on an Ada group? Try this on 
>> alt.os.development.
> 
> He is just trolling!
> 

I know. I just I'd point him in the right direction and see if he dare 
go there.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Nioclásán Caileán de Ghlostéir]

On Sun, 6 Oct 2024, Lawrence D'Oliveiro wrote:
"On Sun, 6 Oct 2024 16:19:29 +0200, DrPi wrote:

> Oh, I forgot, Minix (which is a micro-kernel OS) is embedded is every
> Intel x86 processor.

Bad example. <https://www.theregister.com/2018/08/29/intel_jtag_flaw/>
"


Hello.

I wrote in "Four gigabytes of RAM but "Not enough core" during 
installation" in news:g77u2n$mfs$1@aioe.org in news:comp.os.minix  16 
years ago in 2008:
"Dear all,
On a computer with four gigabytes of RAM, I booted
Minix_IDE-3.1.2a.iso which reported
"[..]
Physical memory: total 6196 KB, system 5700 KB, free 496 KB.
[..]
/etc/rc: /bin/service: Not enough core
[..]"
for the standard >= 16 MB RAM installation option,
and it reported a total of less than 2000 KB for
the 8192 KB RAM installation option, again moaning
"Not enough core".

Do you have any suggestions?

With kind regards,
Nick Paul Colin Gloster"

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Mon, 7 Oct 2024 09:21:26 +0100, Luke A. Guest wrote:

> On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:
>
>> After 40 or more years trying to tout the idea, it’s time to give up.
> 
> I already said, L4.

The one that Hurd has been trying to use, without success.

Has anybody else made production use of it?

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Luke A. Guest]

On 07/10/2024 21:30, Lawrence D'Oliveiro wrote:
> On Mon, 7 Oct 2024 09:21:26 +0100, Luke A. Guest wrote:
> 
>> On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:
>>
>>> After 40 or more years trying to tout the idea, it’s time to give up.
>>
>> I already said, L4.
> 
> The one that Hurd has been trying to use, without success.
> 
> Has anybody else made production use of it?

I already said in another message, about L3.

Re: Ironclad, the hard-Real Time capable POSIX-like kernel written in SPARK/Ada, received an nlnet grant

[Lawrence D'Oliveiro]

On Mon, 7 Oct 2024 21:47:18 +0100, Luke A. Guest wrote:

> On 07/10/2024 21:30, Lawrence D'Oliveiro wrote:
>
>> On Mon, 7 Oct 2024 09:21:26 +0100, Luke A. Guest wrote:
>> 
>>> On 07/10/2024 00:48, Lawrence D'Oliveiro wrote:
>>>
>>>> After 40 or more years trying to tout the idea, it’s time to give up.
>>>
>>> I already said, L4.
>> 
>> The one that Hurd has been trying to use, without success.
>> 
>> Has anybody else made production use of it?
> 
> I already said in another message, about L3.

This is L4 we’re talking about, not L3.