@ne_one,

Maybe 1%. At least they have a way to get around it, at least from what I read. The new exec will be able to keep on track.

Anyone still remembers AmiWest 2013?

_________________
This is just like television, only you can see much further.

As far as I remember few time ago it was told that the scheduler was working and what was holding 4.2 was gallium development implicitally stating that exec was completed. This last anuncement seems to contradict that.

Last edited by ilbarbax on 26-Feb-2015 at 06:57 AM.

@NutsAboutAmiga

Ok, so If I understand it correctly:

- No existing programs have to be adapted to run on the SMP AOS, the schedular arranges all CPU load and tells programs where to run.
- Almost all existing programs will benefit as CPU load per core should be les then when everyting runs on 1 core.
- Programs that use one thread can only run on one core.
- Programs that use more then one thread will run on both cores.
- But programs specially written for use on more then one core will benefit more as they are better suited to do so.

In short, this is a perfect solution to OS4. Right ?

_________________
Theres a time to live and a time to die
When its time to meet the maker
Theres a time to live but isnt it strange
That as soon as you're born you're dying

@BSzili

Yeah

I remember the key "bullets" but now there seems to be some extra text between them =)

@ilbarbax

Quote:

5. AmigaOS 4.2 Depends on Gallium3D release May or may not depend on multicore support

http://blog.hyperion-entertainment.biz/?p=1010

_________________
Theres a time to live and a time to die
When its time to meet the maker
Theres a time to live but isnt it strange
That as soon as you're born you're dying

That is a very interesting status update and explained well.
It would be great if the scheduler goes back to assembly at some point, but I'm pleased with the progress either way.

Quote:

The first step was to rewrite the scheduler in C for easier accessibility. In the very end, this step might be reversed again, rewriting the then SMP capable scheduler back into assembly language.

_________________
http://www.hostcove.com
http://www.youtube.com/hostcove
Sam460 : X1000 : X5000

@Develin

Which bears a striking resemblance to the AmiWest 2013 Skype conversation with Jason McMullan.

_________________
This is just like television, only you can see much further.

@BSzili

i knew ive heard that stuff before, good find;)

@BSzili

I suppose that there isn't a million way of doing SMP on Amiga and still maintain compatibility.
The difference is that Jason stopped working on it while Hyperion continue even at a slow pace.

Kamelito

Last edited by kamelit0 on 26-Feb-2015 at 12:07 PM.
Last edited by kamelit0 on 26-Feb-2015 at 12:06 PM.
Last edited by kamelit0 on 26-Feb-2015 at 12:06 PM.

@kamelit0

Jason stopped but where do you know that work on it stopped? I do not know but the last picture posted on facebook indicates something different. Have you inside informations from Arix project you want to share with us?

@kamelit0

Quote:

Jason stopped working on it

Who says Jason was the only one working on it? There were at least 3 devs working with him at the time. Jason did stop working on it at the time however others continued. ;)

Last edited by terminills on 26-Feb-2015 at 12:26 PM.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@kamelit0
I'm not talking about about who came up with how to do the SMP first. I was questioning the news value of the blogpost, which seems to merely reiterate which has been already said 2 years ago. The state of the SMP is the same as it was in 2013: new scheduler written in C, decoupled from ExecBase, doesn't run on the additional cores yet, etc.:
http://blog.hyperion-entertainment.biz/?p=863
http://blog.hyperion-entertainment.biz/?p=1010

I have no dog in this fight, but I don't want people to be lead on, and think that there have been progress without any evidence.

_________________
This is just like television, only you can see much further.

@BSzili

Quote:

BSzili wrote: @kamelit0 ...I don't want people to be lead on, and think that there have been progress without any evidence.

http://www.stat.amigaworld.net/modules/newbb/viewtopic.php?mode=viewtopic&topic_id=38284&forum=2&start=380&viewmode=flat&order=0#721070

Good luck, I gave up on that concept over a year ago.

_________________
Troll - n., A disenfranchised former potential customer who remains interested enough to stay informed and express critical opinions.
opp., the vast majority who voted silently with their feet.

@BSzili

Yeah, it felt very familiar =)

@Metalheart

4 X-Kernel Update

Done - Task scheduler rewritten in C

Done - Removed reliance on data structures (e.g. ExecBase task lists and ThisTask pointer)

Dropped - Moving scheduler to run on auxiliary cores

Dropped - All cores schedule tasks independently.
(That won't work to resource hugging)

New/Done - One core will take care of scheduling.

Not done yet - Load balancing between cores

Not listed from 2013:

Done - Testing, they have now tested forbid()/permit() under SMP and its working, as concept.

Basically a report of the end of tunnel regarding SMP support and AmigaOS.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@BSzili

Well the plan is more or less the same, but some things are dropped, and some things are done, so it progress nonetheless.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@NutsAboutAmiga

looks like o1i's branch of sillysmp.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@NutsAboutAmiga
You've said "Moving scheduler to run on auxiliary cores" is dropped, yet the new blog post states it as a future goal:
Quote:

The next step is to have each core in the development system (currently, the X1000) to run the scheduler.

What are the things that are done now, but they weren't in 2013? It took 2 years to replace Forbid/Permit with locks running on a single core?

_________________
This is just like television, only you can see much further.

@BSzili

I suspect testing might be a big part of it and that takes time. With something as integrated as this you have to make sure there are no bugs before moving to the next step otherwise it becomes extremely difficult to debug.

_________________

José