Libre RISC-V Mailing list Archives (accessible via public-inbox)
 help / color / Atom feed
* [libre-riscv-dev] [Bug 215] evaluate minerva for base in libre-soc
From: bugzilla-daemon @ 2020-03-11 15:37 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-215-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=215

--- Comment #4 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jacob Lifshay from comment #3)
> do note that compressed texture decoding needs to be able to load 128-bit
> wide values (a single compressed texture block), 

okaaay.

> so our scheduling circuitry
> should be designed to support that. They should always be aligned, so we
> won't need to worry about that in the realignment network.

whew.

so that's 128-bit-wide for _textures_... that's on the *load* side.  are there
any simultaneous (overlapping) "store" requirements? are the code-loops tight
enough to require simultaneous 128-bit LD *and* 128-bit ST?

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 178] first coriolis2 tutorial, workflow and "test project" page
From: bugzilla-daemon @ 2020-03-11 15:35 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-178-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=178

--- Comment #204 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jock Tanner from comment #203)

> Can the solution looks like this:
> 
> > ('experiment5/alu_heir.py')
> >
> > def create_pads(top_element, pins, file_name):
> >     for pin in pins:
> >         # some introspective magic
> >         # as well as a bunch of guessing
> >         pass
> > 
> >     with open(file_name, "w") as ioring:
> >         # a bit of templating magic with jinja2

ermmm ermermerm i like the idea of keeping things in an external file

*thinks*... actually, CSV format would be better, because it's 3 lines of code:

    import csv # standard python library
    with open(file_path, 'r') as csvfile:
        return list(csv.DictReader(csvfile))

remember, we need to minimise the dependencies (see HDL_workflow, and see
discussion we just had!)

in CSV format, it could be:

SIDE,pinname,offset,comment
NORTH,clk,,,

the defaults would be, if there is no offset, to make one.  the order of
the pins in the CSV file would define the order in which they appear on
that side.

i think that would work really well.  later we can consider actually embedding
the CSV file contents *into* the docstring (or a function) of the actual nmigen
source code.

or something.

if it turns out to be more sophisticated than that, we can revisit this (or use
an alternative format).

basic guiding principle: keep it *real* simple.  i won't go into full detail of
the nightmare story of someone doing a less-functional "Model-View-Controller"
replacement of 800 lines of python i did, once.  he wrote 16 *THOUSAND* lines
of code - in 5 weeks.  for a simple "real-time CPU usage" web app.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 215] evaluate minerva for base in libre-soc
From: bugzilla-daemon @ 2020-03-11 15:06 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-215-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=215

--- Comment #3 from Jacob Lifshay <programmerjake@gmail.com> ---
do note that compressed texture decoding needs to be able to load 128-bit wide
values (a single compressed texture block), so our scheduling circuitry should
be designed to support that. They should always be aligned, so we won't need to
worry about that in the realignment network.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 181] test and install public-inbox
From: bugzilla-daemon @ 2020-03-11 14:55 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-181-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=181

--- Comment #29 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jacob Lifshay from comment #28)

> I think this is because the git repo contains everything sent by mailman as
> well as everything received, where it shouldn't include email generated by
> mailman.

it's configured as a "subscribed user", just like all other mailman users, and
that "user" happens to be an account that happens to be local on the host that
happens to be named "libre-riscv.org".  therefore, *all* mail sent by mailman
goes *to* that user - publicinbox@libre-riscv.org - it ends up in
/home/publicinbox/Maildir - and public-inbox-watch picks it up from there.

however only things that have the list-id tag are supposed to show up: i
configured public-inbox to "filter" messages only having that tag (otherwise
messages from other lists end up in the git repo).

it looks like the tag may have been included even when mailman sent "admin" and
"subscribe" messages.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 178] first coriolis2 tutorial, workflow and "test project" page
From: bugzilla-daemon @ 2020-03-11 14:51 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-178-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=178

--- Comment #203 from Jock Tanner <tanner.of.kha@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #201)
> let me know how you get on with the analogy of creating something similar to
> ioring.py, to create "Pads" that you can see in the experiments5 and
> alu_hier nmigen benchs.  you might even be able to use the code *from*
> coriolis2 that imports ioring.py as a base.

Can the solution looks like this:

> ('experiment5/alu_heir.py')
>
> def create_pads(top_element, pins, file_name):
>     for pin in pins:
>         # some introspective magic
>         # as well as a bunch of guessing
>         pass
> 
>     with open(file_name, "w") as ioring:
>         # a bit of templating magic with jinja2
>
>
>         ...
>
>
> if __name__ == "__main__":
>     alu = ALU(width=16)
>     pins = [alu.op, alu.a, alu.b, alu.o]
>     create_pads(alu, pins, "coriolis2/ioring.py")
>     create_ilang(alu, pins, "alu_hier")
>

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 215] evaluate minerva for base in libre-soc
From: bugzilla-daemon @ 2020-03-11 14:50 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-215-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=215

--- Comment #2 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jacob Lifshay from comment #1)
> will need adjusting to make the datapath between the core and L1 wider --
> 64-bit at the very least, 128-bit or wider preferred.

yes.  four LD/STs @ 32-bit is the minimum viable data width to the L1 cache,
realistically.  preferably four LD/STs @ 64 bit.

this is a monster we're designing!

address widths also need to be updated: i'm going to suggest parameterising
them because we might not have time to do an MMU (compliant with the
POWER ISA), just have to see how it goes.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 181] test and install public-inbox
From: bugzilla-daemon @ 2020-03-11 14:40 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-181-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=181

--- Comment #28 from Jacob Lifshay <programmerjake@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #27)
> (In reply to Jacob Lifshay from comment #26)
> > note that the list password for publicinbox@libre-riscv.org is exposed:
> 
> ah well done me! :)  well, the plan is to move to libre-soc with new mailing
> lists anyway.

I think this is because the git repo contains everything sent by mailman as
well as everything received, where it shouldn't include email generated by
mailman.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 181] test and install public-inbox
From: bugzilla-daemon @ 2020-03-11 14:37 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-181-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=181

--- Comment #27 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jacob Lifshay from comment #26)
> note that the list password for publicinbox@libre-riscv.org is exposed:

ah well done me! :)  well, the plan is to move to libre-soc with new mailing
lists anyway.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 215] evaluate minerva for base in libre-soc
From: bugzilla-daemon @ 2020-03-11 14:36 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-215-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=215

Jacob Lifshay <programmerjake@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |programmerjake@gmail.com

--- Comment #1 from Jacob Lifshay <programmerjake@gmail.com> ---
will need adjusting to make the datapath between the core and L1 wider --
64-bit at the very least, 128-bit or wider preferred.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 215] New: evaluate minerva for base in libre-soc
From: bugzilla-daemon @ 2020-03-11 14:31 UTC (permalink / raw)
  To: libre-riscv-dev

http://bugs.libre-riscv.org/show_bug.cgi?id=215

            Bug ID: 215
           Summary: evaluate minerva for base in libre-soc
           Product: Libre Shakti M-Class
           Version: unspecified
          Hardware: PC
                OS: Linux
            Status: CONFIRMED
          Severity: enhancement
          Priority: ---
         Component: Source Code
          Assignee: lkcl@lkcl.net
          Reporter: lkcl@lkcl.net
                CC: libre-riscv-dev@lists.libre-riscv.org
   NLnet milestone: ---

https://github.com/lambdaconcept/minerva

looks really good: clean design, uses wishbone to the L1 caches and
to the main core.  this will help when it comes to adding SMP down
the line.  the decoder is also very clean.

the only non-obvious bit is how the core works, with source/sink
being created in class "Stage" (which is fine), and with the
inter-transfer layouts being the same (source on previous stage
equals sink on the next stage), that's obvious: the bit that's
not obvious is what-gets-connected-to-what.

i think this is because "sinks" are set up at the start of core.py
whilst "sources" are set up much further down.  in the libre-soc
pipeline code, the layouts are done via objects, and the modules
"take care" of placing data into the "output" inherently.  here,
it's messy, and the separation makes understanding difficult.

other than that, though, it's pretty damn good.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 181] test and install public-inbox
From: bugzilla-daemon @ 2020-03-11 14:19 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-181-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=181

Jacob Lifshay <programmerjake@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |programmerjake@gmail.com

--- Comment #26 from Jacob Lifshay <programmerjake@gmail.com> ---
note that the list password for publicinbox@libre-riscv.org is exposed:

base64 decode the body of
https://git.libre-riscv.org/?p=mailman.git;a=blob_plain;f=44/a87dceac4d54a9ec7d712d8c2d57701d164680;h=5658427eaea7de80a828649c8b62f2603b9885c7;hb=14bd89885655679bec6707132377d37478fb17e9

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 178] first coriolis2 tutorial, workflow and "test project" page
From: bugzilla-daemon @ 2020-03-11 12:59 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-178-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=178

--- Comment #202 from Jock Tanner <tanner.of.kha@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #201)
> this does not surprise me at all.  and you won't find anything anyway:
> there are zero docstrings 
> 

Well, actually 'help(Hurricane)' gives a lot of info, that Pycharm should use
for syntax check, code completion, et c. I wonder why it has failed this time.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 178] first coriolis2 tutorial, workflow and "test project" page
From: bugzilla-daemon @ 2020-03-11 12:45 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-178-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=178

--- Comment #201 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jock Tanner from comment #200)

> Well, if “no extra software” is the main concern, then I guess my initial
> idea of installing Pycharm locally in chroot will do just fine. It's
> completely self-contained and requires no modification to the system, with
> the exception of:
> 
> - setup folder (can be installed in home folder, no root required),
> - global settings folder (also in home folder),
> - project settings folder (can be put in '.git/info/exclude' or
> '.gitignore', which I usually do).
> 
> Delete that 3 objects, and you'll never remember that you used it.
> 
> Actually, I've tried it already, and all went well 

excellent

> with one exception: I can't get binary extensions indexed. 

this does not surprise me at all.  and you won't find anything anyway:
there are zero docstrings 

> Yes, I'd like to walk through Hurricane
> attributes in editor. =) I think it's worth a little investigation.

you will likely need the coriolisEnv.py settings. i believe pycharm
actually *executes* the module (imports it, which can have side-effects).

clearly, to execute the Hurricane.so that involves being able to load
the dynamic libraries it's linked to.

that in turn involves having coriolisEnv.py environment variables set up

it also implies having access to the entire coriolis binaries and libraries.

anyway - focus.

let me know how you get on with the analogy of creating something similar to
ioring.py, to create "Pads" that you can see in the experiments5 and alu_hier
nmigen benchs.  you might even be able to use the code *from* coriolis2 that
imports ioring.py as a base.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 178] first coriolis2 tutorial, workflow and "test project" page
From: bugzilla-daemon @ 2020-03-11 12:35 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-178-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=178

--- Comment #200 from Jock Tanner <tanner.of.kha@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #197)
>
> yesyes i know i said "no extra software" then suggested installing pyro.  if
> i'm installing vim and editing inside the chroot, you can sneak pyro in :)

Well, if “no extra software” is the main concern, then I guess my initial idea
of installing Pycharm locally in chroot will do just fine. It's completely
self-contained and requires no modification to the system, with the exception
of:

- setup folder (can be installed in home folder, no root required),
- global settings folder (also in home folder),
- project settings folder (can be put in '.git/info/exclude' or '.gitignore',
which I usually do).

Delete that 3 objects, and you'll never remember that you used it.

Actually, I've tried it already, and all went well with one exception: I can't
get binary extensions indexed. Yes, I'd like to walk through Hurricane
attributes in editor. =) I think it's worth a little investigation.

(In reply to Luke Kenneth Casson Leighton from comment #198)
> absolutely and critically dependent on the website - pypi.org - for its
> security.  do you GPG sign packages before uploading them? what happens if
> pypi - the website - is hacked?

Yes, it depends on pypi.org to store and transfer packages. And on SSL chain of
trust to know if pypi.org is authentic. If pypi.org or chain of trust are
hacked, we're all hacked.

On the other hand, the processing of signatures is significantly increasing an
attack surface by inclusion of the corresponding tools in the process, on both
client and server sides. And, if a process is more complex, people are tend to
make more mistakes.

The model of trust is also important here. Should I trust the owner of a
certain site? The issuer of a certain PGP signature? The ISP?

Thankfully, all components, involved in the process: twine, cheeseshop, pip and
setuptools − are quite versatile, and the process can be easily tailored to
your model of trust. You can use the VCS of your choice on a trusted host to
store packages, or set up your own PyPI.

You can also place your signature somewhere in your PyPI package, and then
modify setuptools to verify the signature along with the package hash before
running setup.py. I think it should be fairly easy to implement.

> a solution is to *pre-install* debian python dependency packages *before*
> running "python setup.py develop".  this *stops* pip going out and
> arbitrarily and blithely downloading random crap that's completely unsigned
> and you have absolutely no idea if it's been vetted.

Package pinning in pip is out there since I don't remember when. I think there
were some difficulties in pinning pip itself, but I suppose they are gone now.
pip should be ready for reproducible builds.

The problem of malicious packages in PyPI can be also mitigated by package
pinning, although this is more a social problem, than a technical one.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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] [Bug 181] test and install public-inbox
From: bugzilla-daemon @ 2020-03-11 12:34 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-181-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=181

--- Comment #25 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
https://git.libre-riscv.org/?p=mailman.git;a=tree

success!  a test list, which i'm going to use an excuse to do another
test...

-- 
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
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:       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