Libre RISC-V Mailing list Archives (accessible via public-inbox)
 help / color / Atom feed
* [libre-riscv-dev] [Bug 306] Formal Correctness Proof for ALU pipeline
From: bugzilla-daemon @ 2020-05-20 15:20 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-306-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=306

Luke Kenneth Casson Leighton <lkcl@lkcl.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           See Also|                            |https://bugs.libre-soc.org/
                   |                            |show_bug.cgi?id=331

-- 
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 331] Formal Correctness Proof for LOGICAL pipeline
From: bugzilla-daemon @ 2020-05-20 15:20 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-331-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=331

Luke Kenneth Casson Leighton <lkcl@lkcl.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
    parent task for|                            |198
  budget allocation|                            |
    NLnet milestone|---                         |NLNet.2019.10.Formal

-- 
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 331] Formal Correctness Proof for LOGICAL pipeline
From: bugzilla-daemon @ 2020-05-20 15:21 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-331-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=331

Luke Kenneth Casson Leighton <lkcl@lkcl.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 total budget (EUR)|0                           |400
  for completion of|                            |
       task and all|                            |
           subtasks|                            |

-- 
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 198] Formal correctness proofs are needed for low-level libraries in LibreSOC
From: bugzilla-daemon @ 2020-05-20 15:21 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-198-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=198

Luke Kenneth Casson Leighton <lkcl@lkcl.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   budget (EUR) for|8250                        |7850
 completion of task|                            |
   (excludes budget|                            |
       allocated to|                            |
          subtasks)|                            |

-- 
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 331] Formal Correctness Proof for LOGICAL pipeline
From: bugzilla-daemon @ 2020-05-20 15:22 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-331-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=331

--- Comment #1 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
* OP_POPCOUNT proof
https://git.libre-soc.org/?p=soc.git;a=commitdiff;h=0bf0e9b743aac05f25669a0891c638958f01360c
* OP_PARITY proof
https://git.libre-soc.org/?p=soc.git;a=commitdiff;h=8dd3c64cbd6b6112894f990a2cdd6661e79dd9bf

-- 
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

* Re: [libre-riscv-dev] daily kan-ban update 20may2020
From: Michael Nolan @ 2020-05-20 15:50 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <CAPweEDxxHh-1kjXtP-ZB-u5AoGE8b1+zt6XmqVVe1fHTtw5bEA@mail.gmail.com>

It looks like I'm working formal proofs today! I started out by making 
sure the alu's 32 bit carry was working correctly and kinda branched out 
from there... After lunch I think I'll work on the overflow and 
overflow32 ports for alu and checking the outputs of the CR pipeline.

--Michael



_______________________________________________
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 331] Formal Correctness Proof for LOGICAL pipeline
From: bugzilla-daemon @ 2020-05-20 15:59 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-331-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=331

Michael Nolan <mtnolan2640@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|lkcl@lkcl.net               |mtnolan2640@gmail.com
                 CC|                            |mtnolan2640@gmail.com

-- 
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 330] create POWER9 Logic Pipeline
From: bugzilla-daemon @ 2020-05-20 16:00 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-330-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=330

Michael Nolan <mtnolan2640@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|lkcl@lkcl.net               |mtnolan2640@gmail.com
                 CC|                            |mtnolan2640@gmail.com

-- 
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

* Re: [libre-riscv-dev] daily kan-ban update 20may2020
From: Luke Kenneth Casson Leighton @ 2020-05-20 16:23 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <d98dc7c3-6795-7603-dbaa-93cfb891f011@gmail.com>

On Wed, May 20, 2020 at 4:50 PM Michael Nolan <mtnolan2640@gmail.com> wrote:
>
> It looks like I'm working formal proofs today! I started out by making
> sure the alu's 32 bit carry was working correctly and kinda branched out
> from there... After lunch I think I'll work on the overflow and
> overflow32 ports for alu and checking the outputs of the CR pipeline.

nice.  can you raise the associated bugreports (like i just did here
https://bugs.libre-soc.org/show_bug.cgi?id=331)

i'll help fill them in.

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] [Bug 331] Formal Correctness Proof for LOGICAL pipeline
From: bugzilla-daemon @ 2020-05-20 16:28 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-331-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=331

--- Comment #2 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
* OP_CNTLZ proof
https://git.libre-soc.org/?p=soc.git;a=commitdiff;h=fcc9bc0960283f91288a17eb0eb672ac5fdcc6f0

-- 
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] RISCV-V Extension
From: Yehowshua @ 2020-05-20 16:31 UTC (permalink / raw)
  To: Libre-RISCV General Development

It seems pretty similar to SimpleV.
Register Length agnostic vector instructions that then get unrolled onto hardware as necessary.

https://gms.tf/riscv-vector.html <https://gms.tf/riscv-vector.html>

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

* [libre-riscv-dev] [Bug 332] New: Formal correctness proof needed for CR pipeline
From: bugzilla-daemon @ 2020-05-20 16:34 UTC (permalink / raw)
  To: libre-riscv-dev

https://bugs.libre-soc.org/show_bug.cgi?id=332

            Bug ID: 332
           Summary: Formal correctness proof needed for CR pipeline
           Product: Libre-SOC's first SoC
           Version: unspecified
          Hardware: PC
                OS: Windows
            Status: CONFIRMED
          Severity: enhancement
          Priority: ---
         Component: Formal Verification
          Assignee: lkcl@lkcl.net
          Reporter: mtnolan2640@gmail.com
                CC: libre-riscv-dev@lists.libre-riscv.org
   NLnet milestone: ---

-- 
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 332] Formal correctness proof needed for CR pipeline
From: bugzilla-daemon @ 2020-05-20 16:34 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-332-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=332

--- Comment #1 from Michael Nolan <mtnolan2640@gmail.com> ---
Needed: 
 - OP_MCRF
 - OP_CROP - this is crand, cror, crxor, etc.
 - OP_MTCRF
 - OP_MFCR

-- 
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 332] Formal correctness proof needed for CR pipeline
From: bugzilla-daemon @ 2020-05-20 16:35 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-332-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=332

Michael Nolan <mtnolan2640@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|lkcl@lkcl.net               |mtnolan2640@gmail.com

-- 
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

* Re: [libre-riscv-dev] Debug port (was Re: minimum viable ASIC)
From: Staf Verhaegen @ 2020-05-20 16:44 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <CAPweEDx5zV3K3y=gjpArhpZ5VrKYL01=938EYA4xd0PjROqppg@mail.gmail.com>

[-- Attachment #1.1: Type: text/plain, Size: 2338 bytes --]

Luke Kenneth Casson Leighton schreef op vr 08-05-2020 om 23:19 [+0100]:
> ---crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
> 
> On Fri, May 8, 2020 at 6:23 PM Staf Verhaegen <staf@fibraservi.eu> wrote:
> 
> ah.  as a software engineer, the practice of using wildcard imports isan extremely bad one.  i strongly advocate *not* getting into thehabit of doing "from nmigen import *" everywhere - it will make yourlife - and other users lives - absolute hell when it comes to tryingto track and debug code.
> minerva has a JTAG interface as well.  i corrected the practice ofusing "import *" in this one

Would 'import nmigen as nm' work for you?I did not touch code for a few months and in the mean time I do use a more PEP compliant code style. But I am annoyed by the individual list of things to import. For nmigen or Hurricane there is typically a few lines of these includes. When I change for example a Pad to a Rectilinear I may have to update the import list. I don't see value add of this and it annoys me a lot.So I am thinking of switching to 'import Hurricane as hur' and 'import nmigen as nm'To me the only valid reason for not using wild cards is that a later wild card import may replace something you did import before; this is avoided by namespacing the import with 'import ... as ...'.
> it looks like the minerva team got a long way with that.  alsoprovided a wishbone master interface:
> https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/minerva/units/debug/wbmaster.py;h=db02af95b4eb3ef8ac25b348f3abaa2bcbe7d96f;hb=a54adcb65bad37b398b11e33a824c7d08c5fe509

That code seems to focus on the debugging protocol over the JTAG interface the lower level part seems to be missing.In debug/top.py:...
# FIXME: figure out where JTAGTap is
# from jtagtap import JTAGTap


class JTAGTap:
    def __init__(self):
        raise NotImplementedError(
            "jtagtap package not found: figure out where JTAGTap is")
...

My JTAG interface does actually the low level part, the protocol has to be done by the user on top of it. I do think I have a flexible way of adding extra shift registers and jtag commands. I also do provide boundary scan for the IO pins which should make PCB testing much easier (BTW, this was actually the original use case for JTAG).
greets,Staf.



[-- Attachment #2: Type: text/plain, Size: 177 bytes --]

_______________________________________________
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] RISCV-V Extension
From: Luke Kenneth Casson Leighton @ 2020-05-20 17:05 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <C8102C6D-A60C-4D99-811D-941AAC7561D0@gatech.edu>

On Wed, May 20, 2020 at 5:31 PM Yehowshua <yimmanuel3@gatech.edu> wrote:
>
> It seems pretty similar to SimpleV.

yes.  i tracked what they were doing, and based VL, SUBVL, predication
etc. and many other aspects: data-dependent fail-on-first and so on.

> Register Length agnostic vector instructions that then get unrolled onto hardware as necessary.

actually the assumption is that there are hardware "lanes" and that if
there is a mismatch between VL (=2, =3) and the "Lanes" (4) those
non-matching "Lanes" run empty.

this is a fundamental design assumption, that although setting VL
*requests* a number of elements to be processed, the hardware does not
have to do exactly what is requested: it may assign any non-zero
*actual* VL.

luckily all loop assembly code is specifically designed to be
independent of what the hardware actually sets VL to on any given
loop.

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] Debug port (was Re: minimum viable ASIC)
From: Luke Kenneth Casson Leighton @ 2020-05-20 17:22 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <803285a5aa4bed5be9c8d4bc5855f49174742f33.camel@fibraservi.eu>

On Wed, May 20, 2020 at 5:44 PM Staf Verhaegen <staf@fibraservi.eu> wrote:
>
> Luke Kenneth Casson Leighton schreef op vr 08-05-2020 om 23:19 [+0100]:
> > ---crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
> >
> > On Fri, May 8, 2020 at 6:23 PM Staf Verhaegen <staf@fibraservi.eu> wrote:
> >
> > ah.  as a software engineer, the practice of using wildcard imports isan extremely bad one.  i strongly advocate *not* getting into thehabit of doing "from nmigen import *" everywhere - it will make yourlife - and other users lives - absolute hell when it comes to tryingto track and debug code.
> > minerva has a JTAG interface as well.  i corrected the practice ofusing "import *" in this one
>
> Would 'import nmigen as nm' work for you?

and then have everything as "x = nm.Signal()...." and so on?  yes that
would be really good.  i actually worked for a company that made it
mandatory to use the full explicit module name (and used this trick to
get round the issue of ridiculously long names).

they required:

import nmigen
x = nmigen.Signal()
...

however it gets very, very tedious.

import nmigen as nm
x = nm.Signal()

is actually a really good idea.  saves space, explicitly says where
the "thing" (Signal etc.) come from.


> I did not touch code for a few months and in the mean time I do use a more PEP compliant code style.
> But I am annoyed by the individual list of things to import.

Staf: trust me on this one.  the hell it creates for people other than
yourself is beyond measure.  the imports serve a secondary critically
important purpose of documenting _where_ a given variable (class) may
be located.

"import *" completely and utterly obliterates that trail, making
maintenance and code-readability by other people an absolute
nightmare.

> For nmigen or Hurricane there is typically a few lines of these includes.

you quickly get used to cut/pasting from other files.

> When I change for example a Pad to a Rectilinear I may have to update the import list.

and other people when they read that code will be able to track that
change.  i have absolutely no idea what Pad is, nor Rectilinear, for
example.

where should i start looking for it?

which module?

which *package* even?

are they in... nmigen?

are they in Hurricane?

are they in something else that has been wild-card imported?

i *literally* cannot tell, and that makes everyone's life absolute
hell.  we - as users of that code - *literally* have to do "grep -r
Pad" in MULTIPLE source code directories - for completely different
packages (QTY 2 in this one example alone) - to track down where the
definition of "Pad" and "Rectilinear" exist!


> I don't see value add of this and it annoys me a lot.

this is down to inexperience with collaborative python development.  i
say that with no disrespect: it's just that i've used python for 20
years and have seen a lot.

> So I am thinking of switching to 'import Hurricane as hur' and 'import nmigen as nm'

> To me the only valid reason for not using wild cards is that a later wild
> card import may replace something you did import before;

two things (both equally as important):

1. that is down to inexperience from developing python.  *nobody* who
is an experienced python developer uses wildcard imports except in
very, very special mitigating cases

2. you are absolutely correct: this is *another* important reason.

there are many more.

> this is avoided by namespacing the import with 'import ... as ...'.

yes, it is a good trick.

> > also provided a wishbone master interface:
> > https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/minerva/units/debug/wbmaster.py;h=db02af95b4eb3ef8ac25b348f3abaa2bcbe7d96f;hb=a54adcb65bad37b398b11e33a824c7d08c5fe509
>
> That code seems to focus on the debugging protocol over the JTAG interface the lower level part seems to be missing.In debug/top.py:...
> # FIXME: figure out where JTAGTap is
> # from jtagtap import JTAGTap
>
>
> class JTAGTap:
>     def __init__(self):
>         raise NotImplementedError(
>             "jtagtap package not found: figure out where JTAGTap is")
> ...
>
> My JTAG interface does actually the low level part,

ahh that's very interesting to note.  ah that's then a really good
combination, of the two.  one is the external interface, the other
more internal.

> the protocol has to be done by the user on top of it. I do think I have a flexible way of adding extra shift registers and jtag commands. I also do provide boundary scan for the IO pins which should make PCB testing much easier (BTW, this was actually the original use case for JTAG).

ah, really appreciated.

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] [Bug 316] bperm TODO
From: bugzilla-daemon @ 2020-05-20 17:34 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-316-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=316

--- Comment #69 from Cole Poirier <colepoirier@gmail.com> ---
I have made the small code additions necessary to connect the Bperm module to
the fu/logical pipeline, and added the test for this to test_pipe_caller.py.
Important question, has soc.decoder.isa.all.ISA() been moved to
soc.decoder.pseudo.pagereader.ISA()?

-- 
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 316] bperm TODO
From: bugzilla-daemon @ 2020-05-20 17:48 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-316-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=316

--- Comment #70 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Cole Poirier from comment #69)
> I have made the small code additions necessary to connect the Bperm module
> to the fu/logical pipeline, and added the test for this to
> test_pipe_caller.py. 

great.  if it passes, commit it, git pull, then push it.  let's take a look.

> Important question,

important _non-sequitur_ question.

> has soc.decoder.isa.all.ISA() been
> moved to soc.decoder.pseudo.pagereader.ISA()?

pagereader.ISA is for reading the ISA markdown files.  that is its job.
it creates a data structure that conveniently allows access in python
to the contents of those markdown files.

decoder.isa.all.ISA() is an auto-generated representation *of* the
pseudo-code - converted to python - for use by the simulator.

there is absolutely no relation whatsoever between the two tasks that
these two completely different modules perform, other than that, by
coincidence, pagereader.ISA happens to read the pseudocode that is
auto-oonverted to python and contained in decoder.isa.all.ISA()

hence why i am left puzzled and wondering - out of fascinated curiosity -
why there would be any reason why you imagine that one would be moved
to the other.

same name, perhaps?  was that it?

-- 
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 70] evaluate Bus Architectures
From: bugzilla-daemon @ 2020-05-20 17:49 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-70-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=70

--- Comment #13 from Jacob Lifshay <programmerjake@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #12)
> (In reply to Jacob Lifshay from comment #11)
> > discovered that there are already quite a few protocols that are much more
> > widely used than omnixtend that support cache coherent memory access over a
> > network: google "rdma cache coherent"
> 
> adding "wishbone" to that and opensparc T1 comes up
> https://www.oracle.com/technetwork/systems/opensparc/opensparc-internals-
book-1500271.pdf

neat! note that wishbone is not designed to run over ethernet or similar,
unlike most other rdma protocols.

-- 
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 333] New: investigate why CR pipeline code took 100% CPU and locked up generating ILANG
From: bugzilla-daemon @ 2020-05-20 17:55 UTC (permalink / raw)
  To: libre-riscv-dev

https://bugs.libre-soc.org/show_bug.cgi?id=333

            Bug ID: 333
           Summary: investigate why CR pipeline code took 100% CPU and
                    locked up generating ILANG
           Product: Libre-SOC's first SoC
           Version: unspecified
          Hardware: PC
                OS: Mac OS
            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: ---

problem is "fixed" by this commit:
https://git.libre-soc.org/?p=soc.git;a=commitdiff;h=60f06a9eb815419f0477a5501c627824da07e078

however should not have occurred in the first place.  needs investigating.

-- 
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 316] bperm TODO
From: bugzilla-daemon @ 2020-05-20 18:07 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-316-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=316

--- Comment #71 from Cole Poirier <colepoirier@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #70)
> (In reply to Cole Poirier from comment #69)
> > I have made the small code additions necessary to connect the Bperm module
> > to the fu/logical pipeline, and added the test for this to
> > test_pipe_caller.py. 
> 
> great.  if it passes, commit it, git pull, then push it.  let's take a look.

I can't run the test, see below.

> > Important question,
> 
> important _non-sequitur_ question.
> 
> > has soc.decoder.isa.all.ISA() been
> > moved to soc.decoder.pseudo.pagereader.ISA()?
> 
> pagereader.ISA is for reading the ISA markdown files.  that is its job.
> it creates a data structure that conveniently allows access in python
> to the contents of those markdown files.
> 
> decoder.isa.all.ISA() is an auto-generated representation *of* the
> pseudo-code - converted to python - for use by the simulator.
> 
> there is absolutely no relation whatsoever between the two tasks that
> these two completely different modules perform, other than that, by
> coincidence, pagereader.ISA happens to read the pseudocode that is
> auto-oonverted to python and contained in decoder.isa.all.ISA()
> 
> hence why i am left puzzled and wondering - out of fascinated curiosity -
> why there would be any reason why you imagine that one would be moved
> to the other.
> 
> same name, perhaps?  was that it?

Yes, has caused an hour of headache. decoder/all.py is needed in order to run
the tests. all.py is auto-generated, so is not present in the directory. How am
I supposed to auto-generate it? Is it supposed to be autogenerated by the test
itself? This would be odd as the fu/logical/tests/test_pipe_caller.py file
imports this at the top of the file. Essentially, help I am quite stuck after
trying to solve this for an hour.

-- 
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 316] bperm TODO
From: bugzilla-daemon @ 2020-05-20 18:13 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-316-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=316

--- Comment #72 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Cole Poirier from comment #71)
> (In reply to Luke Kenneth Casson Leighton from comment #70)
> > (In reply to Cole Poirier from comment #69)
> > > I have made the small code additions necessary to connect the Bperm module
> > > to the fu/logical pipeline, and added the test for this to
> > > test_pipe_caller.py. 
> > 
> > great.  if it passes, commit it, git pull, then push it.  let's take a look.
> 
> I can't run the test, see below.
> 
> > > Important question,
> > 
> > important _non-sequitur_ question.
> > 
> > > has soc.decoder.isa.all.ISA() been
> > > moved to soc.decoder.pseudo.pagereader.ISA()?
> > 
> > pagereader.ISA is for reading the ISA markdown files.  that is its job.
> > it creates a data structure that conveniently allows access in python
> > to the contents of those markdown files.
> > 
> > decoder.isa.all.ISA() is an auto-generated representation *of* the
> > pseudo-code - converted to python - for use by the simulator.
> > 
> > there is absolutely no relation whatsoever between the two tasks that
> > these two completely different modules perform, other than that, by
> > coincidence, pagereader.ISA happens to read the pseudocode that is
> > auto-oonverted to python and contained in decoder.isa.all.ISA()
> > 
> > hence why i am left puzzled and wondering - out of fascinated curiosity -
> > why there would be any reason why you imagine that one would be moved
> > to the other.
> > 
> > same name, perhaps?  was that it?
> 
> Yes, has caused an hour of headache. decoder/all.py is needed in order to
> run the tests. all.py is auto-generated, so is not present in the directory.
> How am I supposed to auto-generate it?

ah - you should have asked earlier :)  look in the Makefile:
    python3 src/soc/decoder/pseudo/pywriter.py


> Is it supposed to be autogenerated by
> the test itself?

no because it takes a good couple of minutes to go over the 300+ pseudo-code
code-fragments.

> This would be odd as the
> fu/logical/tests/test_pipe_caller.py file imports this at the top of the
> file. Essentially, help I am quite stuck after trying to solve this for an
> hour.

should have asked earlier :)

-- 
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 333] investigate why CR pipeline code took 100% CPU and locked up generating ILANG
From: bugzilla-daemon @ 2020-05-20 18:16 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-333-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=333

Michael Nolan <mtnolan2640@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mtnolan2640@gmail.com

--- Comment #1 from Michael Nolan <mtnolan2640@gmail.com> ---
It seems we both stumbled on this issue at about the same time, and arrived at
similar solutions. I narrowed down the commit that introduced it to this one: 

commit 7638005f8577ba6545a65a0c6e31c5419f1f684e
Author: Luke Kenneth Casson Leighton <lkcl@lkcl.net>
Date:   Sun May 17 17:15:58 2020 +0100

    use slightly more elegant way to access CR lookup table

diff --git a/src/soc/cr/main_stage.py b/src/soc/cr/main_stage.py
index e29d5ee..23eba32 100644
--- a/src/soc/cr/main_stage.py
+++ b/src/soc/cr/main_stage.py
@@ -67,9 +67,10 @@ class CRMainStage(PipeModBase):
         # the CR to determine what the resulting bit should be.

         # Grab the lookup table for cr_op type instructions
-        lut = Signal(4, reset_less=True)
+        lut = Array([Signal(name=f"lut{i}") for i in range(4)])
         # There's no field, just have to grab it directly from the insn
-        comb += lut.eq(self.i.ctx.op.insn[6:10])
+        for i in range(4):
+            comb += lut[i].eq(self.i.ctx.op.insn[6+i])

         # Generate the mask for mtcrf, mtocrf, and mfocrf
         fxm = Signal(xfx_fields['FXM'][0:-1].shape())
@@ -108,28 +109,15 @@ class CRMainStage(PipeModBase):
                 comb += ba.eq(xl_fields['BA'][0:-1])
                 comb += bb.eq(xl_fields['BB'][0:-1])

-                # Extract the two input bits from the CR
-                bit_a = Signal(reset_less=True)
-                bit_b = Signal(reset_less=True)
-                comb += bit_a.eq(cr_arr[ba])
-                comb += bit_b.eq(cr_arr[bb])
-
-                # Use the two input bits to look up the result in the
-                # lookup table
-                bit_out = Signal(reset_less=True)
-                comb += bit_out.eq(Mux(bit_b,
-                                       Mux(bit_a, lut[3], lut[1]),
-                                       Mux(bit_a, lut[2], lut[0])))
-                # Set the output to the result above
-                comb += cr_out_arr[bt].eq(bit_out)
+                # Use the two input bits to look up the result in the LUT
+                comb += cr_out_arr[bt].eq(lut[Cat(cr_arr[bb], cr_arr[ba])])

             ##### mtcrf #####
             with m.Case(InternalOp.OP_MTCRF):
                 # mtocrf and mtcrf are essentially identical
                 # put input (RA) - mask-selected - into output CR, leave
                 # rest of CR alone.
-                comb += cr_o.eq((self.i.a[0:32] & mask) |
-                                     (self.i.cr & ~mask))
+                comb += cr_o.eq((self.i.a[0:32] & mask) | (self.i.cr & ~mask))

             ##### mfcr #####
             with m.Case(InternalOp.OP_MFCR):


I think what happens is nmigen generates something really big for arr[index].
Since that gets used to index lut, and then gets assigned to cr_out_arr[idx],
it generates a gigantic expression. By assigning the temporary expressions to
signals, this doesn't happen. Just a guess

-- 
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 333] investigate why CR pipeline code took 100% CPU and locked up generating ILANG
From: bugzilla-daemon @ 2020-05-20 18:37 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-333-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=333

--- Comment #2 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Michael Nolan from comment #1)

> I think what happens is nmigen generates something really big for
> arr[index]. Since that gets used to index lut, and then gets assigned to
> cr_out_arr[idx], it generates a gigantic expression. By assigning the
> temporary expressions to signals, this doesn't happen. Just a guess

that's twice we've encountered problems related to pmux (Array-indexing).
the other one being PriorityEncoder.

ah hang on.  i think i know.  if the result of the lut is assigned to
a temporary signal *before* going in as the assignment...

yeah.

-- 
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