[libre-riscv-dev] daily kan-ban update 28may2020

[libre-riscv-dev] daily kan-ban update 28may2020

[Cole Poirier]

Yesterday:

Worked on proof_regfile.py, had to do a bit of a reorg at luke's
direction because I was mixing python and nmigen AST variables.

Today:

Continue working on proof_regfile.py with luke

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Luke Kenneth Casson Leighton]

On Thu, May 28, 2020 at 8:33 PM Cole Poirier <colepoirier@gmail.com> wrote:

> Continue working on proof_regfile.py with luke

(thank you to Michael for confirming about using "Initial()")

i'm currently doing a CompUnit version of ALU test_pipe_caller.py.
it's thrown a spanner in the works of MultiCompUnit interaction with
ready-valid signalling.  DummyALU's ready-valid signalling is not good
enough: it doesn't conform to the nmutil pipeline API, and neither
does MultiCompUnit.

l.

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Jacob Lifshay]

yesterday:
added Rust models of div and mod instructions -- tested to match POWER9,
also switched output format to JSON to allow easier ingestion

plans for today:
attend OpenPower virtual coffee meeting
work more on mul and div pipelines

Jacob
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Alain D D Williams]

I'm fighting a war of attrition ... geting the latex environment to build for documentation.

It turns out that the version on my main machine is too old - running Centos 6,
I'm planning upgrade to centos 8 in the summer/autumn.

A virtual centos 8 - packages that are on centos 6 are not available here :-(

A virtual Debian 10 - oh, good looks as if it will work ... just need a few new
packages .... damn ! RFS out of space; so move & symlink.

I'll get there.

-- 
Alain Williams
Linux/GNU Consultant - Mail systems, Web sites, Networking, Programmer, IT Lecturer.
+44 (0) 787 668 0256  https://www.phcomp.co.uk/
Parliament Hill Computers Ltd. Registration Information: https://www.phcomp.co.uk/Contact.html
#include <std_disclaimer.h>

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Luke Kenneth Casson Leighton]

On Fri, May 29, 2020 at 12:02 AM Alain D D Williams <addw@phcomp.co.uk> wrote:

> A virtual Debian 10 - oh, good looks as if it will work ... just need a few new
> packages .... damn ! RFS out of space; so move & symlink.

ah yes sorry, latex does historically come with a hell of a lot of
packages.  if you apt-get clean (or rm /var/lib/apt/cache/*.deb) you
should get 300mb back, at least.

> I'll get there.

:)

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Alain D D Williams]

On Fri, May 29, 2020 at 12:06:04AM +0100, Luke Kenneth Casson Leighton wrote:
> On Fri, May 29, 2020 at 12:02 AM Alain D D Williams <addw@phcomp.co.uk> wrote:
> 
> > A virtual Debian 10 - oh, good looks as if it will work ... just need a few new
> > packages .... damn ! RFS out of space; so move & symlink.
> 
> ah yes sorry, latex does historically come with a hell of a lot of
> packages.  if you apt-get clean (or rm /var/lib/apt/cache/*.deb) you
> should get 300mb back, at least.

The VM was more to play with the latest debian than do real work ... I just
installed not really worrying about how it partitioned the disk. It gave RFS
5.5 GB. Thanks to whoever working on BSD invented the symbolic link!

-- 
Alain Williams
Linux/GNU Consultant - Mail systems, Web sites, Networking, Programmer, IT Lecturer.
+44 (0) 787 668 0256  https://www.phcomp.co.uk/
Parliament Hill Computers Ltd. Registration Information: https://www.phcomp.co.uk/Contact.html
#include <std_disclaimer.h>

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Yehowshua]

Today:

Finished a website with strikingly similar objectives to Libre-SOC.
Could it be?
Nah…

https://systemeslibres.org <https://systemeslibres.org/>


_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Luke Kenneth Casson Leighton]

i like it.  we should be able to do 24 GFLOPs btw.  800mhz (0.8ghz) x
4 (quad-core) x 4-wide FP32 - and FMACs are counted as 2 FP "ops".

l.


On Fri, May 29, 2020 at 12:09 AM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
> Today:
>
> Finished a website with strikingly similar objectives to Libre-SOC.
> Could it be?
> Nah…
>
> https://systemeslibres.org <https://systemeslibres.org/>
>
>
> _______________________________________________
> libre-riscv-dev mailing list
> libre-riscv-dev@lists.libre-riscv.org
> http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Cole Poirier]

On May 28 2020, at 4:09 pm, Yehowshua <yimmanuel3@gatech.edu> wrote:
> Today:
>  
> Finished a website with strikingly similar objectives to Libre-SOC.
> Could it be?
> Nah…
>  
> https://systemeslibres.org <https://systemeslibres.org/>

Very, very, very cool!! Two quick notes, under the section "When can I
get a chip?" it says "Fall 2021 A single core Libre-SOC with a GPU and
open should be market ready by fall 2021." Which word has been
accidentally left out after open? Also on your updates the date is being
left out for some reason so it simply reads "Tuesday, May, 2020".  

Cole

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Luke Kenneth Casson Leighton]

https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fsystemeslibres.org

strange: this isn't completing, it sits at 99%.  does anyone else have
better luck?

l.

On Fri, May 29, 2020 at 12:32 AM Cole Poirier <colepoirier@gmail.com> wrote:
>
> On May 28 2020, at 4:09 pm, Yehowshua <yimmanuel3@gatech.edu> wrote:
> > Today:
> >
> > Finished a website with strikingly similar objectives to Libre-SOC.
> > Could it be?
> > Nah…
> >
> > https://systemeslibres.org <https://systemeslibres.org/>
>
> Very, very, very cool!! Two quick notes, under the section "When can I
> get a chip?" it says "Fall 2021 A single core Libre-SOC with a GPU and
> open should be market ready by fall 2021." Which word has been
> accidentally left out after open? Also on your updates the date is being
> left out for some reason so it simply reads "Tuesday, May, 2020".
>
> Cole
>
> _______________________________________________
> libre-riscv-dev mailing list
> libre-riscv-dev@lists.libre-riscv.org
> http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev

Re: [libre-riscv-dev] daily kan-ban update 28may2020

[Cole Poirier]

On May 28 2020, at 4:47 pm, Luke Kenneth Casson Leighton <lkcl@lkcl.net> wrote:
> https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fsystemeslibres.org
> 
> strange: this isn't completing, it sits at 99%.  does anyone else have
> better luck?
> 
> l.

Yes, just ran on my iphone 8 got 93/100. 2.7s largest contentful paint,
4.6s time to interactive.

Cole

_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev