* Re: [libre-riscv-dev] 2pm UK time, 17 Jun 2020: Marketnext hackathon preliminary setup and introduction call
From: Yehowshua @ 2020-06-15 18:27 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <D4D469BD-1020-4A85-BEE2-E31A1384BB51@getmailspring.com>
> On Jun 15, 2020, at 2:26 PM, Cole Poirier <colepoirier@gmail.com> wrote:
>
>>>
>>> That’s 9PM EST for me.
>>> I’d be glad to hop on.
>>
>> fantastic. anyone else?
>
> Me. 6pm PDT. I'll create a bug report under libre-soc-org
Hold up - meant to say AM. 9AM for me, 6AM for you.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] 2pm UK time, 17 Jun 2020: Marketnext hackathon preliminary setup and introduction call
From: Cole Poirier @ 2020-06-15 18:38 UTC (permalink / raw)
To: Libre-RISCV General Development; +Cc: Libre-RISCV General Development
In-Reply-To: <D2AEACB6-30D7-4C6C-9420-55A3A9637D08@gatech.edu>
On Jun 15 2020, at 11:27 am, Yehowshua <yimmanuel3@gatech.edu> wrote:
> Hold up - meant to say AM. 9AM for me, 6AM for you.
Guess I'll be up early :0
Cole
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Luke Kenneth Casson Leighton @ 2020-06-15 19:32 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <FC91F7BD-FFC3-4C4F-87AD-0B0CAAAC0B9F@gatech.edu>
On Mon, Jun 15, 2020 at 5:47 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
> >
> > maybe. we will have to evaluate it. as i explained in the crossover
> > message there are other higher priority design requirements that are unique
> > and specially required for the design.
>
> I watched the video and read through the pages.
> I understood high level what is going on - the details are
> not entirely clear to me at this stage.
this is understandable: it's literally taken me 2 years to learn.
> At some point, I can imagine that you’ll want to connect
> up the memory systems.
yes. *right now*!
i just got LD/ST CompUnit working in test_core.py and it is relying on
a hard-link to a Test Memory limited to 64 8-byte words.
i am holding off putting in Wishbone interfaces onto the
(prototype/test) L0CacheBuffer because it is yet another layer of
complexity and it needs unit tests of a wishbone link in isolation,
first, before becoming comfortable with it.
https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/experiment/l0_cache.py;h=5213ee986da7524b41ba8fb6842a15dc2c861d49;hb=HEAD#l318
that in turn requires that we have an actual SRAM *behind* that
Wishbone interface in order to be able to actually *do* those tests.
this is what i've been able to find. harry points out that he did not
write any unit tests (yet).
https://github.com/HarryMakes/nmigen-soc/blob/wishbone_interconnect/nmigen_soc/wishbone/sram.py
at that point it is a matter of connect-connect-connect and finally we
are in a position to flip out the test infrastructure at the Wishbone
level and drop in Litex instead.
if Minerva's Load/Store interface is used / adapted then there is no
need to write-code-from-scratch, and it also means that enabling L1
caches is a brain-dead runtime configuration switch.
https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/minerva/units/loadstore.py;hb=HEAD#l75
however there are *zero* unit tests for this code and so that will be
important to write.
> How much labor do you think is needed for this?
if we have the right components and the right people, probably only
about 2-3 weeks.
> Assuming I’m free starting at the end of July,
> It would probably take me a couple weeks to catch up to speed
> on the intricacies of the scoreboard.
with the current resources the scoreboard is not an immediate
incremental priority: getting a functional core that executes
instructions is a top priority... *and then* the scoreboard can be
incrementally added to it if there is time to do so.
this all changes if the people from Marketnext are good quality
engineers, who can get up to speed rapidly. subtasks can be allocated
to them and at that point yes there will be plenty of time to do do
heavy-duty sustained focus on complex tasks.
in the meantime i know you really want to get to understand the
precise-augmented scoreboard system. the only thing i can suggest is:
read and re-read Mitch's book chapters again and again and again, run
the unit tests again and again, go over the Tomasulo "morph" page
again and again, and look at the graphviz diagrams and match them to
Mitch's diagrams.
it littteerrallly takes mmmooonnnths of sustained dedicated focus on
this and *nothing* else, for the design to properly sink in. remember
it took me *four* months to find a bug in the later sections of
Mitch's second book chapter, which not even he had noticed.
l.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Yehowshua @ 2020-06-15 19:54 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDztS0TrODPL7B53SK8XcF-NegpBYLW8TiKfhvqVq8yrWA@mail.gmail.com>
>
> with the current resources the scoreboard is not an immediate
> incremental priority: getting a functional core that executes
> instructions is a top priority... *and then* the scoreboard can be
> incrementally added to it if there is time to do so.
OK. So right now, it sounds like instructions execute in order with
scoreboard?
Are we simply stalling with instructional dependencies?
I’m assuming that right now there is no multi-issue or anything of that
nature at the moment?
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] daily kan-ban update 15jun2020
From: Luke Kenneth Casson Leighton @ 2020-06-15 19:58 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDy+jJ6ifvjMSOrn51v-Bs6qdXa6ZELT_r8TDRH7qx8uyQ@mail.gmail.com>
On Mon, Jun 15, 2020 at 4:11 PM Luke Kenneth Casson Leighton
<lkcl@lkcl.net> wrote:
> * look at why the 2nd LD/ST instruction remains "busy" in test_core.py
> where in the test_ldst_compunit.py it is fine
found it quickly. turns out that you must not link signals together
*after* you create a nmigen Simulator instance. the nmigen
Simulator(m) instance must be the last thing done.
now the LD/ST unit tests also work in test_core.py as well. this is a
significant step because with LD/ST, branch, and add (etc.) all
confirmed working, simple loops can be created once we have
instruction read/feed in place.
> * if that's found quickly, continue investigating the next step which
> is to first add in an instruction "reader" (I-Cache path) and second
> to point the PC at it and actually get instructions from an actual
> I-Cache / SRAM.
this is now the next high priority task: getting an SRAM that's on the
other side of a wishbone bus operational (and properly unit tested)
then dropping it into place on the L0CacheBuffer.
on a separate path an instruction queue can also be added... however
agaaain, that needs to be reading from a Memory, preferably via a
wishbone interface.
what i might do for now is just drop in a small SRAM just to get
something working.
l.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] daily kan-ban update 15jun2020
From: Yehowshua @ 2020-06-15 20:03 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDx1hevEgNrMvEdORv3cLmFyBqgrA4eAK5Z4Vuo0f=SqcA@mail.gmail.com>
>
> what i might do for now is just drop in a small SRAM just to get
> something working.
You might already know, but you can do an SRAM like so:
**should also map to ECP5 BRAM blocks in Yosys
```
from nmigen import Memory
gprf = Memory(width=32, depth=32)
gprf_rp1 = gprf.read_port()
gprf_rp2 = gprf.read_port()
gprf_wp = gprf.write_port()
```
https://github.com/lambdaconcept/minerva/blob/master/minerva/core.py#L227
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Luke Kenneth Casson Leighton @ 2020-06-15 20:10 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <077FCA11-14A9-4607-9833-BF7D50A93747@gatech.edu>
On Mon, Jun 15, 2020 at 8:54 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
> >
> > with the current resources the scoreboard is not an immediate
> > incremental priority: getting a functional core that executes
> > instructions is a top priority... *and then* the scoreboard can be
> > incrementally added to it if there is time to do so.
>
> OK. So right now, it sounds like instructions execute in order with
> scoreboard?
no scoreboard at all. no dependency checking at all.
> Are we simply stalling with instructional dependencies?
correct. it's why it's nicknamed the brain-dead core
https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/simple/core.py;hb=HEAD
https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/simple/test/test_core.py;hb=HEAD
at the moment it has three major top-level components:
* Function Unit block
* Register Files
* TSTL0CacheBuffer which contains a single memory port to a single
nmigen Memory with 64 8-words.
the unit test feeds instructions directly into one (and only one) of
the pipelines, and waits for the busy flag to go LOW.
> I’m assuming that right now there is no multi-issue or anything of that
> nature at the moment?
correct. incremental development. each step - each new piece of
functionality - confirmed by a unit test.
* when the core's executing instructions, we can start writing more
comprehensive unit tests.
* when the scoreboard is added, we re-run those unit tests. if they
work, chances are scoreboard is functional.
* when that's added, we add multi-issue. exactly the same same thing
again: run the unit tests.
i am absolutely shocked to find that there is not a single other
"open" processor that takes this approach. they all - all of them -
rely heavily on "high level" integration tests. test suites written
by other people. or just a series of binaries which test the *whole*
system.
one error and the only way to find it is to compile up one massive
core. if that takes several minutes to do (minerva takes a *minute*
to compile to verilog), tough luck.
we can run the *local* unit tests for the *local* module that is
believed to be the problem and rapidly track the problem down.
btw my brother dan pointed out that this type of approach - which is
not carried out *AT ALL* by *ANY* commercial team (especially Intel
right now) let alone an "open source" one, is a critically important
"Unique Selling Point" to investors.
by having a combination of low-level unit tests (and formal
correctness proofs) that test components individually prior to use in
a higher level that then *ALSO* has a unit test, this type of approach
is *EXACTLY* the kind of thing that will give investors a high degree
of confidence that we will be a safe bet.
l.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* [libre-riscv-dev] Introduction
From: Anais Poirier @ 2020-06-15 20:06 UTC (permalink / raw)
To: libre-riscv-dev
Hello All,
My name is Anais, I'm a third year electrical and computer engineering
student at the University of Toronto, and I am Cole Poirier's sister! I'm
most interested in gaining knowledge and experience working on a hardware
project. I have limited experience with Assembly and Verilog as well as
extensive experience with c/c++ but I am excited to learn Nmigen and hope
to contribute to the Power Core. I am also the one who created the visual
slides for the pitch last week and would like to help with the Elm
Kanban/bug tracker visual front end for Bugzilla.
I have read and agreed to the charter.
Anais
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Yehowshua @ 2020-06-15 20:16 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAKdzrAz4XnsYNXrrc2Au5ci9XaqHNWO3iEwPtPxcQ4x0jcG=Cw@mail.gmail.com>
> I have limited experience with Assembly and Verilog as well as
> extensive experience with c/c++ but I am excited to learn Nmigen and hope
> to contribute to the Power Core
Hello.
I’ll let Luke follow up a bit, but also do subscribe to the LibreSOC misc mailing list below[1].
Reason being is that I and others can help newcomers with more specific FPGA, RTL,
and nMigen concepts allowing us to keep the main dev mailing list lower traffic.
And as always, Welcome!
Yehowshua
[1]: http://lists.libre-soc.org/mailman/listinfo/libre-soc-misc
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Yehowshua @ 2020-06-15 20:23 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDwVL4Ac+soRavWw5u8a21-+Tj055bBBoruW4-0oCP7v2Q@mail.gmail.com>
> * when the core's executing instructions, we can start writing more
> comprehensive unit tests.
> * when the scoreboard is added, we re-run those unit tests. if they
> work, chances are scoreboard is functional.
> * when that's added, we add multi-issue. exactly the same same thing
> again: run the unit tests.
I completely support this approach.
> if that takes several minutes to do (minerva takes a *minute*
> to compile to verilog), tough luck.
Minerva took 8 seconds to compile on my Mac…
> a higher level that then *ALSO* has a unit test, this type of approach
> is *EXACTLY* the kind of thing that will give investors a high degree
> of confidence that we will be a safe bet.
Exactly! I plan to mention this in my pitch deck.
My mentor has been helpful in getting the meat of the ideas out of my head.
Also, I’m fiddling around with the Minerva cache.
Not quite sure about a few things:
1. Why are there s1 and s2 inputs? What does 1 vs 2 mean?
https://github.com/lambdaconcept/minerva/blob/master/minerva/cache.py#L43 <https://github.com/lambdaconcept/minerva/blob/master/minerva/cache.py#L43>
2. Why does s1_stall switch the memory read ports addresses from s1_addr to s2_addr?
https://github.com/lambdaconcept/minerva/blob/master/minerva/cache.py#L141 <https://github.com/lambdaconcept/minerva/blob/master/minerva/cache.py#L141>
I reached out to Jean Paul from lambda concepts about this.
Once I figure out exactly how Minerva core works - I’ll write a
tiny unit bench for it. Hopefully this would be helpful in modifying
it for our needs.
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Yehowshua @ 2020-06-15 20:24 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <F1BA77AC-E94C-4025-B476-F6E19D1ED1F2@gatech.edu>
>
> Minerva core works
Meant to say Minerva cache.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Yehowshua @ 2020-06-15 20:27 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAKdzrAz4XnsYNXrrc2Au5ci9XaqHNWO3iEwPtPxcQ4x0jcG=Cw@mail.gmail.com>
Also, I believe Cole had mentioned that you’re using a Mac?
If this is the case, use home-brew - it makes things notably easier.
The first thing I would do to get started is try
section 5.1 here: https://libre-soc.org
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Luke Kenneth Casson Leighton @ 2020-06-15 20:27 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAKdzrAz4XnsYNXrrc2Au5ci9XaqHNWO3iEwPtPxcQ4x0jcG=Cw@mail.gmail.com>
On Mon, Jun 15, 2020 at 9:13 PM Anais Poirier <anais.shiraz@gmail.com> wrote:
>
> Hello All,
>
> My name is Anais, I'm a third year electrical and computer engineering
> student at the University of Toronto, and I am Cole Poirier's sister!
hi, lovely to hear from you. Cole mentioned you a couple days ago
when i had a monster 3 hour call with him and he took 10 pages of
notes, bless 'im. that his head didn't explode or steam come out his
ears is a good sign.
> I'm most interested in gaining knowledge and experience working on a hardware
> project.
ya picked a Monster, you do realise that, right? :)
> I have limited experience with Assembly and Verilog as well as
> extensive experience with c/c++
oo, good. this will come in real handy, there are 3D drivers to do,
we need a c/c++ simulator for SimpleV (on top of an existing POWERPC
simulator), there's linux kernel drivers and blah blah.
> but I am excited to learn Nmigen and hope
> to contribute to the Power Core.
excellent. well Cole's been through the process already, we threw him
in the deep end and he hasn't sunk yet :) i find it hilarious that he
knew nothing about gate-level design so we found (or wrote) some
tutorials and he went through them, came back and said, "ok now what?"
:)
i recommend reading this - it will give you some background and insights:
https://libre-soc.org/3d_gpu/tutorial/
you've done verilog, that's a really good start. try this one, it's
very good - an introduction to nmigen:
https://github.com/RobertBaruch/nmigen-tutorial
basically jump to the end of this page and go through them one by one
https://libre-soc.org/HDL_workflow/
once you've done *that*, then there are two *relatively* simple
pipelines that need work:
https://bugs.libre-soc.org/show_bug.cgi?id=348
https://bugs.libre-soc.org/show_bug.cgi?id=325
and - jacob's working on DIV, so hypothetically MUL is free:
https://bugs.libre-soc.org/show_bug.cgi?id=323
> I am also the one who created the visual
> slides for the pitch last week
niiice :)
> and would like to help with the Elm
> Kanban/bug tracker visual front end for Bugzilla.
this was just an idea which Cole and i sort-of went "hmm that would be
useful" and i'd honestly forgotten about - if you can raise a
bugreport about it on bugs.libre-soc.org under the "website" product
that would be really useful. ask Cole for help, there.
> I have read and agreed to the charter.
excellent.
ok so you've probably patiently read this https://libre-soc.org/HDL_workflow/
and Cole either already has the dev-environment set up on his dev
machine: if you have your own machine _great_, do ask him to help you
get set up. also if you dare be awake at 6am on Wednesday it would be
good to have you also on the Marketnext call, we will do some
introduction then and go over things that you should also be aware of.
great to have you on board.
warmest,
l.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Yehowshua @ 2020-06-15 20:30 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <F1BA77AC-E94C-4025-B476-F6E19D1ED1F2@gatech.edu>
>
> I reached out to Jean Paul from lambda concepts about this.
> Once I figure out exactly how Minerva core works - I’ll write a
> tiny unit bench for it. Hopefully this would be helpful in modifying
> it for our needs.
Actually - I figured it out.
Stage1 and stage2.
Stage1 is address pre-fetch.
Stage 2 is actual fetch.
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] 2pm UK time, 17 Jun 2020: Marketnext hackathon preliminary setup and introduction call
From: Hendrik Boom @ 2020-06-15 20:48 UTC (permalink / raw)
To: libre-riscv-dev
In-Reply-To: <CAPweEDw3atuUZ7=fXffyF5u8XXxVUqB+bixm_zg-uaVrys_C1g@mail.gmail.com>
On Mon, Jun 15, 2020 at 07:09:48PM +0100, Luke Kenneth Casson Leighton wrote:
> On Mon, Jun 15, 2020 at 6:56 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
> >
> >
> > > if anyone else has time to participate and help pass on knowledge and
> > > answer questions that would be really helpful. i will follow up with
> > > the call dial-in details when i have them.
> >
> > That’s 9PM EST for me.
> > I’d be glad to hop on.
>
> fantastic. anyone else?
I might be interested if I get enough sleep the night before (I'm having
problems with that).
How long is the session likely to last? I have an early afternoon
appointment.
And yes, I'm aware that it's really 9AM EST. But that probably means
9AM EDT, because of daylight savings time? People keep writing EST for
EDT.
-- hendrik
>
> l.
>
> _______________________________________________
> 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
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Luke Kenneth Casson Leighton @ 2020-06-15 20:49 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <C34E9E0A-96F0-4F07-B038-B3A2195CAE2E@gatech.edu>
On Mon, Jun 15, 2020 at 9:28 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
> Also, I believe Cole had mentioned that you’re using a Mac?
ah. right. okay. we're not supporting macs, and we're not going to
do development on macs. there are two aspects to this:
1). when it comes to the ASIC layout, it is absolutely critical - and
is Industry Standard Practice - to have full reproducible builds. not
even the username is different. everyone doing the builds sets up
carbon-copy machines, where you could literally do a file-system
byte-level comparison and find absolutely everything is the same.
this is abbbsolutely essential when you are about to drop USD $5m on a
chip tape-out... oh but the last person who checked the build used
some totally different library and now "it doesn't work", finding the
discrepancy results in time lost, you miss the tape-out window
entirely, and the company loses a USD $100 million dollar deal because
of it.
so it's actually *really serious* not to be using the exact same
development workflow as everyone else.
2) applying this same underlying logic to the actual HDL development
is not an unreasonable step.
also we are a small team and do not need the hassle of "i have a Mac,
i can't get it to work, what do i do, what do i do" and because it's
not something that everyone has, and the problem turns out to be
something related to homebrew, or the native mac tools, we literally
*can't* help you... and at a crucial moment for a deadline *may not
have time*.
at that point you become a liability to the team, not an asset...
because you arbitrarily chose not to respect and follow Industry
Standard practices.
therefore - really - please do not use MacOSX for development. the
lowest common denominator OS which everyone has access to is debian,
and we chose (specifically) debian 10 because its default version of
python is reasonably close enough to what nmigen requires.
you do not have to install debian 10 native: you can install homebrew,
followed by installing qemu, followed by installing debian 10 under
qemu and then using remote terminal login (using the Mac OS as a
"conduit"), or network loop-back filesystem so that you can use quotes
favourite quotes editors under the lovely nice pretty MacOS but
actually run the unit tests entirely under debian 10... *not* native
on the mac and *not* under homebrew.
however you will quickly find that this is really quite painful, slow
and inconvenient, particularly when it comes to setting up yet
*another* virtualisation - chroot within qemu - and that installing
ReFIT (or ReFIND) and actually dual-booting the mac will be much
easier.
OR
as you're in the same house, Cole has a development machine, which is
running debian native: ask him nicely and you can set up a 2nd login
on it. whilst doing that, you can look at getting set up,
independently, and yet still be productive.
i actually use an 8-year-old Mac downstairs. i ssh in to my laptop
upstairs, run 3 "screen" commands, one in each window, and thus if the
WIFI goes down i can get back to each of the three terminals, exactly
where they were disconnected.
hilariously i also installed an openssh client on my Samsung N9005 and
do the same trick: run "screen" and then if i am particularly tired or
if it is late, i do not have a bright LCD screen to stare at, just a
small hand-held one.
l.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Luke Kenneth Casson Leighton @ 2020-06-15 20:53 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <7BA190FA-6551-4A23-BAAE-5CFC2B3838AE@gatech.edu>
On Mon, Jun 15, 2020 at 9:31 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
> >
> > I reached out to Jean Paul from lambda concepts about this.
> > Once I figure out exactly how Minerva core works - I’ll write a
> > tiny unit bench for it. Hopefully this would be helpful in modifying
> > it for our needs.
yes please. please drop it into soc/src/soc/minerva/test/... ah!
excellent! there *is* a test_cache.py already in that directory.
(i'm currently making all imports explicit)
> Actually - I figured it out.
ah excellent
> Stage1 and stage2.
>
> Stage1 is address pre-fetch.
> Stage 2 is actual fetch.
okaay. which isn't documented. no code-comments (at all). sigh.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Cole Poirier @ 2020-06-15 20:55 UTC (permalink / raw)
To: Libre-RISCV General Development; +Cc: Libre-RISCV General Development
In-Reply-To: <CAPweEDwvfxDLw1LYkOfYQTXhTJO7BFaH1DcmZxDQs-xUz2cfjg@mail.gmail.com>
On Jun 15 2020, at 1:49 pm, Luke Kenneth Casson Leighton <lkcl@lkcl.net> wrote:
> On Mon, Jun 15, 2020 at 9:28 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>>
>> Also, I believe Cole had mentioned that you’re using a Mac?
>
> ah. right. okay. we're not supporting macs, and we're not going to
> do development on macs. there are two aspects to this:
>
> 1). when it comes to the ASIC layout, it is absolutely critical - and
> is Industry Standard Practice - to have full reproducible builds. not
> even the username is different. everyone doing the builds sets up
> carbon-copy machines, where you could literally do a file-system
> byte-level comparison and find absolutely everything is the same.
Don't worry Luke, I set up her development machine, running native
debian 10.3 on bare-metal, like mine, when I set up my own. She has a
mac that she will most likely do part of the Elm work on, but she's not
planning on using her mac for HDL work as I have already set up the
proper development on a dedicated desktop.
Cole
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Yehowshua @ 2020-06-15 20:58 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDwvfxDLw1LYkOfYQTXhTJO7BFaH1DcmZxDQs-xUz2cfjg@mail.gmail.com>
> ah. right. okay. we're not supporting macs, and we're not going to
> do development on macs. there are two aspects to this:
Dang. I’ve been using a Mac this whole time.
Not really keen to switch unless I have to do Coriolis stuff.
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Luke Kenneth Casson Leighton @ 2020-06-15 21:04 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <B53F08B2-212D-43DF-9717-D230D39B4EF5@getmailspring.com>
On Mon, Jun 15, 2020 at 9:55 PM Cole Poirier <colepoirier@gmail.com> wrote:
> Don't worry Luke, I set up her development machine, running native
> debian 10.3 on bare-metal, like mine, when I set up my own.
staaaar. whew.
On Mon, Jun 15, 2020 at 9:58 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
>
> > ah. right. okay. we're not supporting macs, and we're not going to
> > do development on macs. there are two aspects to this:
>
> Dang. I’ve been using a Mac this whole time.
it _is_ in the HDL workflow.
> Not really keen to switch unless I have to do Coriolis stuff.
i know - however think it through. it's not the same software. there
could be compilation-level bugs introduced by different compile-time
switches actually in /usr/bin/python, due to a different gcc compiler.
this could result in something incredibly important being missed, and
we have to become OS debugging experts to find out why.
jacob and i experimented with pypy3 about 18 months. it didn't go
well :) although the tests that did run, they did run faster, it's
just... it's far too risky.
if we were doing something that was a GUI app, for some "open"
project, there would be no problem - not least because you have to get
the GUI app working for Macs anyway.
but this? this is *literally* USD 100 million territory, and using a
"personal preferred OS" is just... you just don't do that. sorry!
l.
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Minerva L1 Cache
From: Yehowshua @ 2020-06-15 21:08 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDw5OVhVo=Z-OyDu-w=ON=URcSzBcs7wersJdpeBe_oBFw@mail.gmail.com>
>
> there *is* a test_cache.py already in that directory.
> (i'm currently making all imports explicit)
Right this is the formal.
The formal doesn’t make it clear how to actually
use the thing.
Another question: what is the interface you want to the cache?
I think you said you wanted to go with Minerva’s loadstore frontend?
Could be a bit tricky as the loadstore frontend uses stage specific
bit masks. Might be easier to morph the L1 cache to your needs directly.
From your video, it seems that there is no instruction cache vs. data cache.
Can you specify what interface you want to see from the cache’s perspective?
I suspect the only change we need to make is set s1_stall to Const(0), and disable
s2_addr in Minerva’s L1Cache.
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Yehowshua @ 2020-06-15 21:08 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <CAPweEDxnQSfDz92PDPepq63mVWG06kTqeJ7ZLe8_0XAKePsAYg@mail.gmail.com>
>
> this could result in something incredibly important being missed, and
> we have to become OS debugging experts to find out why.
When I can afford a desktop, I’ll use Debian.
Mac doesn’t have much storage for now so don’t want to do QEMU or dual boot.
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] 2pm UK time, 17 Jun 2020: Marketnext hackathon preliminary setup and introduction call
From: Luke Kenneth Casson Leighton @ 2020-06-15 21:09 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <20200615204854.cwgsburg5i6rmiy4@topoi.pooq.com>
On Mon, Jun 15, 2020 at 9:48 PM Hendrik Boom <hendrik@topoi.pooq.com> wrote:
> I might be interested if I get enough sleep the night before (I'm having
> problems with that).
you and me both :)
> How long is the session likely to last? I have an early afternoon
> appointment.
no idea: feel free to drop in when you have the time
> And yes, I'm aware that it's really 9AM EST. But that probably means
> 9AM EDT, because of daylight savings time? People keep writing EST for
> EDT.
the time is set for 2pm UK local time. everything is relative to
that. here's a timezone converter.
https://www.timeanddate.com/worldclock/meetingdetails.html?year=2020&month=6&day=17&hour=13&min=0&sec=0&p1=136&p2=224&p3=179&p4=83&p5=176&p6=256
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] daily kan-ban update 15jun2020
From: Cole Poirier @ 2020-06-15 21:12 UTC (permalink / raw)
To: Libre-RISCV General Development; +Cc: Libre-RISCV General Development
In-Reply-To: <A3CC338E-A1BB-46A6-8347-760CA3881DBD@gatech.edu>
Yesterday: Mostly processing the huge list of todos I generated over my
3 hour convo with Luke on Saturday, as well as running some errands.
Today:
On boarding two new project members (my two sisters hehe)
Talking with Dan Leighton (Luke's brother) about Elm
Start organizing my todo list for myself, which includes figuring out
how to organize the project 'todo list' and bug tracking,
budget/timeline tracking, roadmap etc. My personal to do list will be
done on a private online Kanban-like service, which I will use to
organize myself enough that I can start putting all of the project
related todos into my wiki page or elsewhere. Will be discussing what
I'm doing on Libre-SOC Org bug reports, but keeping my private 'to do'
list duplicated in the private kanban app so I can organize myself
visually. I will end up cloning something like this, with significant
improvements for the Elm application that I will be deveopling with Dan
and Anais over the next couple months. Then I will no longer maintain
this private to do list as I will prefer using our augmented to our
needs kanban-like application.
Going through every diagram on the wiki and picking out the most basic,
legible ones for Elle to start turning into SVGs.
Doing Robert Baruch's nmigen tutorial with anais
Tomorrow:
Continuing much of the above
Starting to write the bpermd HDL tutorial as I work through it with Anais
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
* Re: [libre-riscv-dev] Introduction
From: Yehowshua @ 2020-06-15 21:13 UTC (permalink / raw)
To: Libre-RISCV General Development
In-Reply-To: <8E6E70EE-DE5E-44E3-AC63-DDC020530DE6@gatech.edu>
Actually - I can do testing in Debian docker.
Share the filesystem with docker mount.
Fair enough!
Yehowshua
_______________________________________________
libre-riscv-dev mailing list
libre-riscv-dev@lists.libre-riscv.org
http://lists.libre-riscv.org/mailman/listinfo/libre-riscv-dev
^ permalink raw reply
page: next prev latest
Libre RISC-V Mailing list Archives (accessible via public-inbox)
Archives are clonable: git clone --mirror https://inbox.libre-soc.org/libre-riscv-dev
Example config snippet for mirrors
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git