Libre RISC-V Mailing list Archives (accessible via public-inbox)
 help / color / Atom feed
* [libre-riscv-dev] additional ddr3 interfaces
From: Jacob Lifshay @ 2020-04-01  0:44 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <3AF0E5F3-86A0-4610-82E4-AD6C5BE0E588@gatech.edu>

On Tue, Mar 31, 2020, 17:23 Immanuel, Yehowshua U <yimmanuel3@gatech.edu>
wrote:

> > the NEC processors from the powerpc-notebook project *might* be POWER9.
>
> I looked into this - these are actually NXP processors. NXP T2080 to be
> precise.
> The data sheet is a little unhelpful - not making it clear what POWER it
> is.
>
> Anyways - the processor itself has some similarities to our goals. It
> might be a good idea to browse through its specs and see if we can match
> its pinouts down the road. Hopefully, by that time, powerpc-notebook is
> ready to go, and our processor could be just plug and play…
>

that would be nice, however the t2080 appears to have a 64-bit memory
interface along with some really high-speed serdes -- both of which will be
more difficult to achieve.

However, if we get the funding required for the open-source custom DDR3
interface, we could potentially put two copies on our SoC, which would
improve GPU performance by a large factor (1.5x?) due to the additional
memory bandwidth. That would also double the maximum amount of memory we
could attach to the SoC which is another benefit. If we designed it so it
would work even with one interface disabled, that would benefit low cost
applications where the pcb designers don't want to have to pay for two dram
chips. I don't know if we could, but we might be able to sell two different
variants where one of them has a smaller package where one of the memory
interfaces is not bonded from the pins to the die, that way, we could save
money on the chips where both memory interfaces aren't needed without
needing more than one mask.

Jacob
_______________________________________________
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] additional ddr3 interfaces
From: Immanuel, Yehowshua U @ 2020-04-01  0:47 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD5p1fYgxrnOgKq4GfcejMkcqpFC-uoN23hxeQ51aiUdYg@mail.gmail.com>

> that would be nice, however the t2080 appears to have a 64-bit memory
> interface along with some really high-speed serdes -- both of which will be
> more difficult to achieve.

Yes. I was thinking skipping on the SERDES since I personally don’t care for PCIE.
(Its not even remotely Libre).

Unless we want a 10G ethernet, we don’t need high speed SERDES.

> However, if we get the funding required for the open-source custom DDR3
> interface,

Yup. Working on the funding part at the Create-X angle :)
Also, it’d be really fascinating if we hear back more from Epic Games.
_______________________________________________
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] additional ddr3 interfaces
From: Jacob Lifshay @ 2020-04-01  1:01 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <52700CC5-3847-49C1-831F-FCD760230AB4@gatech.edu>

On Tue, Mar 31, 2020, 17:48 Immanuel, Yehowshua U <yimmanuel3@gatech.edu>
wrote:

> > that would be nice, however the t2080 appears to have a 64-bit memory
> > interface along with some really high-speed serdes -- both of which will
> be
> > more difficult to achieve.
>
> Yes. I was thinking skipping on the SERDES since I personally don’t care
> for PCIE.
> (Its not even remotely Libre).
>
> Unless we want a 10G ethernet, we don’t need high speed SERDES.
>

If we do want to support multi-socket using something like OmniXtend, which
would be amazing but not strictly necessary, 10G ethernet would be
basically required, 1G would work, but would be really slow. We could also
implement something in-between such as 2.5G or 5G.

See http://bugs.libre-riscv.org/show_bug.cgi?id=70 for some previous
discussion.

Using omnixtend allows getting cache coherent memory-mapped i/o over
ethernet (standard ethernet switches work, it can be used simultaneously to
standard tcp/ip), allowing us to easily add custom devices, build giant
numa servers with thousands of cores, add pcie interfaces using a fpga,
interoperate with other processors that support omnixtend, etc.

Also, since it is based on standard ethernet, any computer with an ethernet
port could act as an emulated memory-mapped device or just additional
memory.

>
> > However, if we get the funding required for the open-source custom DDR3
> > interface,
>
> Yup. Working on the funding part at the Create-X angle :)
> Also, it’d be really fascinating if we hear back more from Epic Games.
>

Yay!

Jacob
_______________________________________________
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] name change announcement on crowdsupply
From: Jacob Lifshay @ 2020-04-01  1:51 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDxPk_bMLgRgdQoo_hJEMdsKtPf9FSUQk+x=7WcEDWzWMg@mail.gmail.com>

On Tue, Mar 31, 2020 at 11:15 AM Luke Kenneth Casson Leighton
<lkcl@lkcl.net> wrote:
>
> jacob the example's a bit big, it's the kind of thing that would be
> great in its own update (which we move it to) with a full explanation.
> could you put it in an "example" file in a repo somewhere, link to it,
> and say "we'll do an update about the IR later"?

I moved the example code to the Kazan repo.

I did notice that we *forgot* to announce we changed our name from
Libre-RISCV to Libre-SOC on any crowdsupply updates -- we should do
that and change the crowdsupply description page to match.

I also updated the latest update to consistently spell Libre-SOC the
way we decided on: Libre-SOC.

Jacob

_______________________________________________
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] name change announcement on crowdsupply
From: Jacob Lifshay @ 2020-04-01  1:52 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD5yMuHBTP=wZD9G0ZDdwt4G4ZN0qczfoySQS=44C-_+Sw@mail.gmail.com>

On Tue, Mar 31, 2020 at 6:51 PM Jacob Lifshay <programmerjake@gmail.com> wrote:
> I did notice that we *forgot* to announce we changed our name from
> Libre-RISCV to Libre-SOC on any crowdsupply updates -- we should do
> that and change the crowdsupply description page to match.

we should also update our page on NLNet.

Jacob

_______________________________________________
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] submitted bugreport to upstream nmigen
From: Jacob Lifshay @ 2020-04-01  3:21 UTC (permalink / raw)
  To: Libre-RISCV General Development

This is the same issue encountered when trying to add gitlab CI support.

https://github.com/nmigen/nmigen/issues/344

_______________________________________________
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] test_decoder_gas.py still fails after update for unknown reason
From: Tobias Platen @ 2020-04-01  6:59 UTC (permalink / raw)
  To: libre-riscv-dev

I installed a working version of the GNU toolchain (via buildroot),
but the test_decoder_gas.py still fails. Both soc and nmigen have been
updated to the latest version.

ERROR: test_rot (__main__.DecoderTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "test_decoder_gas.py", line 448, in test_rot
    self.run_tst(RotateOp, "rot")
  File "test_decoder_gas.py", line 432, in run_tst
    traces=[pdecode2.ports()]):
  File "/usr/local/lib/python3.7/dist-packages/nmigen-0.3.dev12+g2d1e12d-py3.7.egg/nmigen/back/pysim.py", line 1120, in write_vcd
    vcd_file=vcd_file, gtkw_file=gtkw_file, traces=traces)
  File "/usr/local/lib/python3.7/dist-packages/nmigen-0.3.dev12+g2d1e12d-py3.7.egg/nmigen/back/pysim.py", line 95, in __init__
    if trace not in signal_names:
  File "/usr/lib/python3.7/_collections_abc.py", line 666, in __contains__
    self[key]
  File "/usr/local/lib/python3.7/dist-packages/nmigen-0.3.dev12+g2d1e12d-py3.7.egg/nmigen/hdl/ast.py", line 1392, in __getitem__
    key = None if key is None else self._map_key(key)
  File "/usr/local/lib/python3.7/dist-packages/nmigen-0.3.dev12+g2d1e12d-py3.7.egg/nmigen/hdl/ast.py", line 1582, in __init__
    raise TypeError("Object {!r} is not an nMigen signal".format(signal))
TypeError: Object [(sig raw_opcode_in), (sig bigendian), (sig opcode_in), (sig function_unit), (sig in1_sel), (sig in2_sel), (sig in3_sel), (sig out_sel), (sig ldst_len), (sig rc_sel), (sig internal_op), (sig form), (sig cr_in), (sig cr_out), (sig inv_a), (sig inv_out), (sig cry_out), (sig br), (sig sgn_ext), (sig upd), (sig rsrv), (sig is_32b), (sig sgn), (sig lk), (sig sgl_pipe), (sig valid), (sig insn_type), (sig nia), (sig lk), (sig invert_a), (sig invert_out), (sig input_carry), (sig output_carry), (sig input_cr), (sig output_cr), (sig is_32bit), (sig is_signed), (sig insn), (sig data_len), (sig byte_reverse), (sig sign_extend), (sig update), (sig oe), (sig oe_ok), (sig rc), (sig rc_ok), (sig spro), (sig spro_ok), (sig spr1), (sig spr1_ok), (sig spr2), (sig spr2_ok), (sig rego), (sig rego_ok), (sig reg1), (sig reg1_ok), (sig reg2), (sig reg2_ok), (sig reg3), (sig reg3_ok), (sig imm), (sig imm_ok)] is not an nMigen signal

-- 
Tobias Platen <libre-soc[at]platen-software[dot]de>

_______________________________________________
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] extremely busy crowdsupply update started
From: Luke Kenneth Casson Leighton @ 2020-04-01 10:08 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <55435FF3-CF79-4543-B992-837CC2F6B8B9@getmailspring.com>

On Wednesday, April 1, 2020, Cole Poirier <colepoirier@gmail.com> wrote:

> On Mar 31 2020, at 11:35 am, Cole Poirier <colepoirier@gmail.com> wrote:
> > On Mar 31 2020, at 11:09 am, Luke Kenneth Casson Leighton
> > <lkcl@lkcl.net> wrote:
> >> could someone write something about the build servers?
> >>
> >> thx.
> >>
> >> l.
> >
> > Sure I'll put in a short blurb and Jacob can add to it if he thinks it
> > necessary. Do I have the necessary access to the crowdsupply git repo?
> >
> > Cole
>
> Sorry it turns out I don't have push access to the crowdsupply repo.
> Luke, may I request push access to the crowdsupply repo.


ok.


> I remember that
> in an email a few days ago you told I should give you a name for the
> repo I want created for the environment setup scripts. Can you create
> and also grant me push access to a repository named dev-env-setup?
> Unless you can think of a more appropriate name.


nope sounds great.




-- 
---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
_______________________________________________
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] test_decoder_gas.py still fails after update for unknown reason
From: Luke Kenneth Casson Leighton @ 2020-04-01 10:14 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <20200401085921.19fb39bec4ce5c52e2ea1928@platen-software.de>

On Wednesday, April 1, 2020, Tobias Platen <libre-soc@platen-software.de>
wrote:

> I installed a working version of the GNU toolchain (via buildroot),
> but the test_decoder_gas.py still fails. Both soc and nmigen have been
> updated to the latest version.


raise a bugreport tobias. ref this msg in archive.


>     raise TypeError("Object {!r} is not an nMigen signal".format(signal))
> TypeError: Object [(sig raw_opcode_in), (sig bigendian), (sig opcode_in),
> (sig function_unit), (sig in1_sel), (sig in2_sel), (sig in3_sel), (sig
> out_sel), (sig ldst_len), (sig rc_sel), (sig internal_op), (sig form), (sig
> cr_in), (sig cr_out), (sig inv_a), (sig inv_out), (sig cry_out), (sig br),
> (sig sgn_ext), (sig upd), (sig rsrv), (sig is_32b), (sig sgn), (sig lk),
> (sig sgl_pipe), (sig valid), (sig insn_type), (sig nia), (sig lk), (sig
> invert_a), (sig invert_out), (sig input_carry), (sig output_carry), (sig
> input_cr), (sig output_cr), (sig is_32bit), (sig is_signed), (sig insn),
> (sig data_len), (sig byte_reverse), (sig sign_extend), (sig update), (sig
> oe), (sig oe_ok), (sig rc), (sig rc_ok), (sig spro), (sig spro_ok), (sig
> spr1), (sig spr1_ok), (sig spr2), (sig spr2_ok), (sig rego), (sig rego_ok),
> (sig reg1), (sig reg1_ok), (sig reg2), (sig reg2_ok), (sig reg3), (sig
> reg3_ok), (sig imm), (sig imm_ok)] is not an nMigen signal


 that's a list.  ports is returning an object containing a list rather than
a flattened iterator.

l.



-- 
---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
_______________________________________________
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] submitted bugreport to upstream nmigen
From: Luke Kenneth Casson Leighton @ 2020-04-01 10:16 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD4BG7ymNZFpD-64F3yvcNEMaYEzjGfzgBFTh4nE6+hbew@mail.gmail.com>

On Wednesday, April 1, 2020, Jacob Lifshay <programmerjake@gmail.com> wrote:

> This is the same issue encountered when trying to add gitlab CI support.
>
> https://github.com/nmigen/nmigen/issues/344


we may just need to write our own run_simulation function for now.

l.



-- 
---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
_______________________________________________
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] extremely busy crowdsupply update started
From: Luke Kenneth Casson Leighton @ 2020-04-01 10:54 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDweoAka6QZ-rfrMY6yJ6N9M7LG39Ef2QJFMohVRi-0f8w@mail.gmail.com>

On Wed, Apr 1, 2020 at 11:08 AM Luke Kenneth Casson Leighton
<lkcl@lkcl.net> wrote:

>>
>> Sorry it turns out I don't have push access to the crowdsupply repo.
>> Luke, may I request push access to the crowdsupply repo.
>
>
> ok.


done just edit the .git/config file change the url to one that
includes gitolite3.

>>
>> I remember that
>> in an email a few days ago you told I should give you a name for the
>> repo I want created for the environment setup scripts. Can you create
>> and also grant me push access to a repository named dev-env-setup?


done also.

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] name change announcement on crowdsupply
From: Luke Kenneth Casson Leighton @ 2020-04-01 10:57 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD5yMuHBTP=wZD9G0ZDdwt4G4ZN0qczfoySQS=44C-_+Sw@mail.gmail.com>

On Wed, Apr 1, 2020 at 2:51 AM Jacob Lifshay <programmerjake@gmail.com> wrote:

> On Tue, Mar 31, 2020 at 11:15 AM Luke Kenneth Casson Leighton
> <lkcl@lkcl.net> wrote:
> >
> > jacob the example's a bit big, it's the kind of thing that would be
> > great in its own update (which we move it to) with a full explanation.
> > could you put it in an "example" file in a repo somewhere, link to it,
> > and say "we'll do an update about the IR later"?
>
> I moved the example code to the Kazan repo.

great

> I did notice that we *forgot* to announce we changed our name from
> Libre-RISCV to Libre-SOC on any crowdsupply updates -- we should do
> that and change the crowdsupply description page to match.

that'll be tricky, the url has to stay the same.  changing
description, yes should be ok.

> I also updated the latest update to consistently spell Libre-SOC the
> way we decided on: Libre-SOC.

ah good.

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] PPC on Talos and Playstation 3
From: Luke Kenneth Casson Leighton @ 2020-04-01 11:00 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <3AF0E5F3-86A0-4610-82E4-AD6C5BE0E588@gatech.edu>

On Wed, Apr 1, 2020 at 1:23 AM Immanuel, Yehowshua U
<yimmanuel3@gatech.edu> wrote:
>
> > the NEC processors from the powerpc-notebook project *might* be POWER9.
>
> I looked into this - these are actually NXP processors. NXP T2080 to be precise.
> The data sheet is a little unhelpful - not making it clear what POWER it is.
>
> Anyways - the processor itself has some similarities to our goals. It might be a good idea to browse through its specs and see if we can match its pinouts down the road.

with USD $20 million just for that effort alone, yes that would be possible.

it's a massive thousand-pin monster with multi-gigabit interfaces.
packaging costs alone will far exceed the cost of the appx-300-pin SoC
we're currently aiming for.

> Hopefully, by that time, powerpc-notebook is ready to go, and our processor could be just plug and play…
>
> And I wanted to attach an image from the data sheet, but… too large at 88KB :(

send link.

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] additional ddr3 interfaces
From: Luke Kenneth Casson Leighton @ 2020-04-01 11:10 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD5p1fYgxrnOgKq4GfcejMkcqpFC-uoN23hxeQ51aiUdYg@mail.gmail.com>

On Wed, Apr 1, 2020 at 1:44 AM Jacob Lifshay <programmerjake@gmail.com> wrote:

> that would be nice, however the t2080 appears to have a 64-bit memory
> interface along with some really high-speed serdes -- both of which will be
> more difficult to achieve.
>
> However, if we get the funding required for the open-source custom DDR3
> interface, we could potentially put two copies on our SoC,

each DDR3/4 RAM interface adds around 0.4 watts per 800mhz clock rate
which, due to power being a square law, ramps up 1.6 watts by 1600mhz.

additionally, they take 150 pins *each* including 20 pins for VDD and GND.

two such high-speed DDR3/4 RAM interfaces not only therefore increases
the power consumption to more than DOUBLE the entire budget allocated
for the SoC, it also puts it into the approx 450-pin category.

with that many pins we might need to use a 0.5 or smaller pitch grid
(otherwise it would be a 20 mm on a side BGA rather than a 15 mm one),
the packaging costs go up, the yields go down...

that said, there's no reason why we can't put the interface on... and
not bring it out....

> which would
> improve GPU performance by a large factor (1.5x?) due to the additional
> memory bandwidth. That would also double the maximum amount of memory we
> could attach to the SoC which is another benefit. If we designed it so it
> would work even with one interface disabled, that would benefit low cost
> applications where the pcb designers don't want to have to pay for two dram
> chips.

four.  designs with only one DRAM chip are extremely rare
(ultra-ultra-low-cost embedded).  typical DRAMs are x16 so you need
two.  smartphones (etc.) typically use special "stacked" ASICs with
eMMC and 1x or 2x 32-bit DRAM interfaces.  they're well north of 250
pins.

> I don't know if we could, but we might be able to sell two different
> variants where one of them has a smaller package where one of the memory
> interfaces is not bonded from the pins to the die, that way, we could save
> money on the chips where both memory interfaces aren't needed without
> needing more than one mask.

yep, that's how it's done.  ST and ATMEL both actually design the
exact same embedded controller for 10 different purposes, some of
which they discover, during testing, that parts of them fail.

all that happens is: they get shoved in a package, sold as lower-cost
items, with the pins not brought out.

it's a really good way to make sure that yields are kept high, nothing wasted.

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 269] auto-conversion / parser of POWER ISA Spec v3.0B
From: bugzilla-daemon @ 2020-04-01 13:42 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-269-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=269

--- Comment #4 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
aaaaa it's a horrible mess however it's functional.

i need to do a parser for the actual pages in openpower/isa, next.
btw i don't know why this didn't occur to me earlier, i've added
the wiki as a submodule into soc.

this allows us to treat the isa files as "local" (in the repo).
no need to try to grab them by wget.

-- 
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 273] New: replace use of wget in soc with use of git submodule
From: bugzilla-daemon @ 2020-04-01 13:43 UTC (permalink / raw)
  To: libre-riscv-dev

http://bugs.libre-riscv.org/show_bug.cgi?id=273

            Bug ID: 273
           Summary: replace use of wget in soc with use of git submodule
           Product: Libre-SOC's first SoC
           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: ---

it occurred to me that we can treat the wiki git repository as a git submodule.
this means that all use of wget to grab POWER spec files can be replaced with
path-relative file reads.

-- 
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 269] auto-conversion / parser of POWER ISA Spec v3.0B
From: bugzilla-daemon @ 2020-04-01 13:48 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-269-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=269

--- Comment #5 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
the way it works is:

* lexers (PowerLexer) and filters sort out python-style indentation
* parser (PowerParser) turns pseudo-code into an AST
* as part of that parsing, registers that were read and written to
  are noted
* all reg index numbers (fields - RA, RB) that were "read" from GPR
  are "yield" obtained from the opcode.
* a variable named after the field is dropped into the dictionary
  of code to be executed
* code is executed
* all reg index numbers that were "written" in the code are "yield"
  obtained (again), and, from the dictionary of exec()'d code,
  the value is obtained, then stored in the "regfile".

-- 
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] test_decoder_gas.py still fails after update for unknown reason
From: Michael Nolan @ 2020-04-01 14:54 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <CAPweEDx9EaD_HdOhikpHuzqRp2HVnaxF=XM00e27MOuNbhvBZw@mail.gmail.com>

On 4/1/20 6:14 AM, Luke Kenneth Casson Leighton wrote:
> On Wednesday, April 1, 2020, Tobias Platen <libre-soc@platen-software.de>
> wrote:
>
>> I installed a working version of the GNU toolchain (via buildroot),
>> but the test_decoder_gas.py still fails. Both soc and nmigen have been
>> updated to the latest version.
> raise a bugreport tobias. ref this msg in archive.
This is fixed in soc/HEAD. My nmigen was a bit old so it didn't error on 
my machine. Updating to latest nmigen let me reproduce the error.



_______________________________________________
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] test_decoder_gas.py still fails after update for unknown reason
From: Luke Kenneth Casson Leighton @ 2020-04-01 15:16 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <588a9207-0504-90dd-3e9f-3f463867c237@gmail.com>

On Wed, Apr 1, 2020 at 3:54 PM Michael Nolan <mtnolan2640@gmail.com> wrote:

> This is fixed in soc/HEAD. My nmigen was a bit old so it didn't error on
> my machine. Updating to latest nmigen let me reproduce the error.

great.  no need for a bugreport, then, tobias.

yeh tracking latest nmigen is a bit of a pain.

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] additional ddr3 interfaces
From: Staf Verhaegen @ 2020-04-01 15:19 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <CAPweEDxnFyjAC65kvkN5nVYoscL2Ks-791q6pWwpx623Pc0KRA@mail.gmail.com>

[-- Attachment #1.1: Type: text/plain, Size: 885 bytes --]

Luke Kenneth Casson Leighton schreef op wo 01-04-2020 om 11:10 [+0000]:
> On Wed, Apr 1, 2020 at 1:44 AM Jacob Lifshay <programmerjake@gmail.com> wrote:
> > that would be nice, however the t2080 appears to have a 64-bit memoryinterface along with some really high-speed serdes -- both of which will bemore difficult to achieve.
> > However, if we get the funding required for the open-source custom DDR3interface, we could potentially put two copies on our SoC,
> 
> each DDR3/4 RAM interface adds around 0.4 watts per 800mhz clock ratewhich, due to power being a square law, ramps up 1.6 watts by 1600mhz.

Power consumption is linear with frequency and square with voltage (P = f * C * V^2).
For overclocking increasing frequency also likely means you need to increase voltage so you need to increase power consumption more than linear with frequency.

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

* [libre-riscv-dev] [Bug 272] functions needed in POWER simulator which match 3.0B spec
From: bugzilla-daemon @ 2020-04-01 15:35 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-272-13@http.bugs.libre-riscv.org/>

http://bugs.libre-riscv.org/show_bug.cgi?id=272

--- Comment #12 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Luke Kenneth Casson Leighton from comment #11)

> i will attempt a pre and post process tomorrow which basically duplicates
> what you did to get src1 2 etc out of the opcode.

done - it works (although the code's a dreadful mess)

if you'd like to do that Int class i can use it immediately.
or, if you want to give it a shot, knock yourself out.

        for i in range(32):
            self[i] = [0] * 64 # TODO: needs to be IntClass(value=0, len=64)

the class needs to be something like:

class IntClass
    def __init__(self, value, len):
        self.value = value
        self.len = len

then __add__ etc. all have to return an object which first asserts that
the len is the same for both operands.

the one thing, assign (now in soc/decoder/pseudo/parser.py) will, instead
of returning ast.Assign(name, expr) will have to return something like
this:

ast.Call(func=ast.Attribute(value=ast.Name(id='x'), attr='eq'),
                args=[Num(n=5)],
                keywords=[],
                starargs=None,
                kwargs=None))])

i got that from a little test program, parsing this:

x.eq(5)

so when we see "x <- 5"...

i'll add it to the parser, commented-out.

-- 
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] additional ddr3 interfaces
From: Luke Kenneth Casson Leighton @ 2020-04-01 15:42 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <9f72940ad53e20b7015c45771177e6dff55ae6ab.camel@fibraservi.eu>

On Wed, Apr 1, 2020 at 4:19 PM Staf Verhaegen <staf@fibraservi.eu> wrote:

> Power consumption is linear with frequency and square with voltage (P = f * C * V^2).

ah thank you.  forgot about that.  the figures i got were from a
special TI video processing chip, with dual 32-bit DDR3 1600mhz
interfaces.  the power consumption figure was 1.6 watts per channel.

> For overclocking increasing frequency also likely means you need to increase voltage so you need to increase power consumption more than linear with frequency.

yes.  i think that's where i was confused (overclocking almost always
involves increasing voltage).  actually it's a huge source of
confusion across the internet, that 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

* [libre-riscv-dev] POWER simulator
From: lkcl . @ 2020-04-01 15:55 UTC (permalink / raw)
  To: Libre-RISCV General Development

i just wanted to write a bit to everyone about what it's like when
using parsers like this.  it's... very weird :)  you create all these
pieces, you don't know if any of them will work, and then all of a
sudden the work's entirely done and it all works and you can go home.

it's horribly confusing.  the effectiveness is so high that it's
like... "is that it??  surely it had to be harder than that??"

the last time i did something like this it was with Webkit, first
adding gobject bindings and then also adding python bindings  i think
i mentioned before: it took me under 10 days to add support for python
bindings, by taking:

* Mozilla's HTML5 IDL compiler and getting it to understand the 350
Webkit-formatted HTML5 IDL files (1 day)
* python-gobject's compiler and getting it to understand the AST that
came out of the Mozilla IDL compiler (1 day)
* writing some code-generators for creating python c-based modules (3 days)
* writing some python "recognisers" for Webkit c++ datatypes (2-3 days)
* debugging (2-3 days)

that was *it*.  350 classes, 2,000 functions, and over TWENTY thousand
properties - all auto-generated - done in 10 days.  full python
bindings, a peer of what is available for javascript, to HTML5.

this is no different, just on a smaller scale.  it means placing quite
a lot of trust in auto-generated code, which is where the unit tests
come critically into play.

the code-fragments are actually compiled into a python AST, however i
discovered a library that can turn those back into actual python code.
this can be used to actually *output* python source code, which means
that we have something to examine (for bugs).

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] crowdsupply updates
From: Luke Kenneth Casson Leighton @ 2020-04-01 16:02 UTC (permalink / raw)
  To: Libre-RISCV General Development

https://libre-riscv.org/index/Updates/March_2020/

yehowshua i removed the update because it wasn't discussed.  the
consequences of pre-publishing an update via our own website are very
simple: Crowdsupply will not publish it at all.

therefore, publishing it via our website will be detrimental to the project.

also i specifically mentioned that Crowdsupply has over 600
subscribers that the updates go out to.

i now have to spend time working out how to revert gitolite3
repositories in order to remove the update, which will involve a force
push-master.

when i have done so please can everyone destroy their working copy of
the crowdsupply repository, re-checking-out a new one.

thanks.

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] IMPORTANT - PLEASE REMOVE AND RE-CHECKOUT CROWDSUPPLY REPOSITORY Re: crowdsupply updates
From: Luke Kenneth Casson Leighton @ 2020-04-01 16:08 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDz9NCTWQp1ZGpuKWejdaGvM6Y=fHzzdk4HNs13SPMmzYQ@mail.gmail.com>

On Wed, Apr 1, 2020 at 5:02 PM Luke Kenneth Casson Leighton
<lkcl@lkcl.net> wrote:

> when i have done so please can everyone destroy their working copy of
> the crowdsupply repository, re-checking-out a new one.

ok that's done: i had to edit the bare repository master file,
refs/head/master, to get it to "listen".  this could have resulted in
corruption of the wiki.

if you have write access, please ensure that you check out a NEW copy
of the crowdsupply git repository, otherwise if you did a "git pull"
recently it will be the *wrong master* and will cause massive
problems.

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

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