Libre RISC-V Mailing list Archives (accessible via public-inbox)
 help / color / Atom feed
* Re: [libre-riscv-dev] memory interface diagram woes
From: Jacob Lifshay @ 2020-04-27  6:06 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDz4VDx=qWcYKRPvrorJbrA7NJKOj40WmqG2ob0phj5b5A@mail.gmail.com>

On Sun, Apr 26, 2020, 05:28 Luke Kenneth Casson Leighton <lkcl@lkcl.net>
wrote:

> On Thursday, April 23, 2020, Jacob Lifshay <programmerjake@gmail.com>
> wrote:
>
> I'll try to write it out as text on the wiki tomorrow.
>
>
> how you getting on?
>

still working on it... sorry I'm taking so long

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] memory interface diagram woes
From: Luke Kenneth Casson Leighton @ 2020-04-26 12:27 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD4xkEJFTaPo7gKG9ejMvJyNtoA2zbDEKVqHk8abFv-jVg@mail.gmail.com>

On Thursday, April 23, 2020, Jacob Lifshay <programmerjake@gmail.com> wrote:

>
>  not been able to draw it
> out adequately because I keep getting frustrated by the lack of nice,
> power-of-2 alignment features and support for structurally recursive
> diagrams in Dia and the lack of copy-paste when using physical paper.


if it makes you feel any better i'm on to the 10th redraw of the LDST
CompUnit...


> I'll try to write it out as text on the wiki tomorrow.


how you getting on?

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

* [libre-riscv-dev] [Bug 216] LOAD STORE buffer needed
From: bugzilla-daemon @ 2020-04-26 11:28 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-216-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=216

--- Comment #27 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
i may need to implement a very basic (working) of the API in
http://bugs.libre-riscv.org/show_bug.cgi?id=216#c26
effectively a thin "wrapper" around nmigen Memory rdport and wrport.

it's not quite a perfect match for nmigen rdport/wrport as the
address phase needs latching in, separately from how r/w data is handled.

i'll make sure that the (decoded) opcode is passed over because it
contains crucial information (not just ld/st-length).

-- 
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 216] LOAD STORE buffer needed
From: bugzilla-daemon @ 2020-04-25 12:40 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-216-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=216

--- Comment #26 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
Writeup of the interface / API needed for LDST to talk to the LDST
Computational Unit.  Additional signals will come through such as length, type,
atomic etc
these are in order of time sequence

(edit this comment as necessary):

LD
--

* in: is_ld
* in: address
* in: len (1/2/4/8)
* in: go_addr
* out: addr_ok (no exception will occur)
* out: addr_exc (exception type)
* in: exc_ack

* out: ld_data
* out: ld_data_ok
* in: ld_data_ack (acknowledge data)

ST
--

* is_st
* in: address
* in: len (1/2/4/8)
* in: go_addr
* out: addr_ok (no exception will occur)
* out: addr_exc (exception type)
* in: exc_ack

* in: st_data
* in: go_st
* out: st_data_ack

Notes
-----

* both identical up to address exception

* beyond addr_ok the operation MUST be successful.  L1 Cache, TLB etc MUST all
have been done

* after addr_ok sent, LD data can be placed as soon as ready on output bus
without requiring "permission" from a go_ld signal.

* LDSTCompUnit will wait for EITHER addr exception OR data out.

* addr_exc is acknowledged by exc_ack and both LDSTCompUnit and L0 Cache/Buffer
go back to reset.

* ld_data_ok MUST NOT be asserted if addr_exc was asserted

* ld_ack therefore acknowledges data and unit goes back to reset once sent.

* ST on the other hand must have the ADDR checking (and L1 cache line loading)
done separate from the write.

* ST therefore has two phases: addr ok (or not) followed by waiting for
permission to carry out the actual store.

* the ST data in *will* be valid at the clock where go_st is raised.

* go_st will *not* be raised by the LDSTCompUnit until *after* addr_ok/exc

* again: no exceptions can occur after the addr_ok signal is raised.

* data_ack indicates that the ST unit can reset

* the difference is that in LD Mode, the L0 Cache/Buffer is "in charge",
driving the go-ack

* however for ST this is the OPPOSITE, the LDSTCompUnit is "in charge", the
switchover point of who is in charge being after ADDR and TLB checking and
reservation of L1 Cache lines.

-- 
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] morphing 6600 code to use power decoder
From: Luke Kenneth Casson Leighton @ 2020-04-24 20:11 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDy5CCRESLD7-rT9e3pHMP8kEUQW0KpRGGwkyw=ZS6Vdkw@mail.gmail.com>

i started on adding LD/ST back in: and, argh, i spotted that the
LDSTCompUnit needs some rework / redesign.  it "worked" previously
however there were some assumptions which, because of the new
multi-signalling, don't work.

i think i will take the opportunity to remove the triple-functionality
from the LDSTCompUnit.  the original intention was to have its ADD ALU
be capable of performing standard ADD arithmetic, thus allowing us to
have dozens of parallel ADD units.

we're running out of time, and unfortunately i'm not coping with the
logic complexity involved in writing out these state-based diagrams
(last time it took 3 weeks to get right), so i'm going to cut ADD and
simply make the address-generation part of the actual LDSTCompUnit.

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 292] implement multi-way read/write 6600 signals
From: bugzilla-daemon @ 2020-04-24 14:39 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-292-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=292

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|CONFIRMED                   |RESOLVED

-- 
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 292] implement multi-way read/write 6600 signals
From: bugzilla-daemon @ 2020-04-24 14:38 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-292-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=292

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
    parent task for|                            |81
  budget allocation|                            |

-- 
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 81] implement 6600-style "precise" out-of-order scoreboard
From: bugzilla-daemon @ 2020-04-24 14:38 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-81-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=81

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   budget (EUR) for|3600                        |3000
 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 292] implement multi-way read/write 6600 signals
From: bugzilla-daemon @ 2020-04-24 14:37 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-292-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=292

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   budget (EUR) for|0                           |600
 completion of task|                            |
   (excludes budget|                            |
       allocated to|                            |
          subtasks)|                            |
    NLnet milestone|---                         |NLnet.2019.02

-- 
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 292] New: implement multi-way read/write 6600 signals
From: bugzilla-daemon @ 2020-04-24 14:36 UTC (permalink / raw)
  To: libre-riscv-dev

https://bugs.libre-soc.org/show_bug.cgi?id=292

            Bug ID: 292
           Summary: implement multi-way read/write 6600 signals
           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: ---

https://libre-soc.org/3d_gpu/architecture/6600scoreboard/600x-compunit_multi_rw.jpg
https://libre-soc.org/3d_gpu/architecture/6600scoreboard/600x-dependence_cell_multi_pending.jpg

the original 6600 scoreboard is single-signal request-ack protection
for multiple reads on the regfile.  overporting is then added so that
multiple batches of ALUs can simultaneously read (and write) to the
regfile.

we cannot do massive-way regfile porting.

additionally we need to do *different* regfiles for POWER: one of
these is the "Branch Control" Regfile, including CR0, CTR and other
conditional registers.

therefore it becomes more important to have fine-grain control over
the regfile, meaning that multiple go-read/req and go-write/req
wires are needed per FunctionUnit.

-- 
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] morphing 6600 code to use power decoder
From: Luke Kenneth Casson Leighton @ 2020-04-23 17:16 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDxKtj9=F8HAhth5S2ai+DA0BLLAo0s9tPs6jCkoEE-bHA@mail.gmail.com>

this is excruciatingly slow work. i've only just got 4 out-of-order
instructions (2 per Function Unit) operational again after changing to
use multi-control signals on the register file: one each for each
port.

that's still not out of the woods, yet: the last time i did this it
took hundreds of runs of random operations to be partially confident
of an operational OoO design.

i still have LD/ST and branch to put back 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] scammers looking for access to NLNet Funds
From: Luke Kenneth Casson Leighton @ 2020-04-23 17:04 UTC (permalink / raw)
  To: Libre-RISCV General Development

https://list.libresilicon.com/pipermail/libresilicon-developers/2020-April/001059.html

scammers are currently contacting individuals associated with NLNet
Grants, offering them access to legitimate-looking socially-engineered
websites.

i've therefore registered libre-soc.com to make sure it doesn't get
taken and used for scamming.

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] memory interface diagram woes
From: Luke Kenneth Casson Leighton @ 2020-04-23 14:37 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDzzfQ3PHgc40PnZrs+KxBovFp+gYEBEV7TjiJ8SGJfLrw@mail.gmail.com>

illustration, jacob.

with 8 LDST FUs, each with 2 64-bit data ports, how will those 16
entries be inserted into MemoryQueue?
https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/memory_pipe_experiment/memory_queue.py;hb=HEAD

as in, not so much "how" (because in "coding terms" it's a "simple"
matter of writing the code).

i mean "how many wires will be required in order to multiplex up to
16x simultaneous 64-bit chunks into the same (one) MemoryQueue"?

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] memory interface diagram woes
From: Luke Kenneth Casson Leighton @ 2020-04-23 10:29 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAPweEDy7dXU7c-Rd==bwQEcuBdhHJ8PQNSX-sfBnCXXeJw3_gw@mail.gmail.com>

btw a word to the wise jacob, the current code is even though not complete,
ridiculously comprehensive and tied directly into the Dependency Matrix
system.

* LDSTs have regs therefore they need standard FU-Regs and FU-FU presence.

* FU-Regs takes care of the Reg RaW and WaR dependencies

* FU-FU takes care (in the case of LD) of the *order* of which FU is
waiting for which result, it is like a linked list using unary bits and
consequently is a square matrix

on *top* of that we have an additional problem to solve:

* preserving the order of memory accesses (if they matter)

* ensuring that overlaps result in order preservation but that when there
are no overlaps at least respect Weak Memory Order (LDs separate from STs)

on top of *that* we have an *additional* requirement to merge multi-issued
LD/STs that can fit into a single cache line hit, because of the enormous
bandwidth requirements of 3D Vectors.

it is ridiculous but necessary that even in a 180nm single core we need a
staggering 256 bit wide data path to the L1 cache and that this width will,
when we go to faster speeds, also be needed to the L2 cache as well.

the overlapping address conflict and Weak Order Preservation is achieved
with an addressmatch Matrix, last year, thanks to Mitch Alsup.

however Mitch's scheme works by comparing bits 4 thru 11 or 5 thru 11
(avoiding the VM bits and aldo avoiding bits 0 to 3) and ending up with
0.5% additional overzealous hits which make *no significant performance
degradation* in real world usage.

but his scheme was *not* possible, as-is, to use for our needs, because it
missed opportunities to merge byte level accesses because of only comparing
bits 4 thru 11 of the address.

i therefore expanded the LDST LENGTH and bits 0 thru 3 into a 24 bit
bytemask / bitmap, then split that into 2 requests.

why 2 requests, because of misaligned LDST, the 1st 16bits of the bytemap
are 1 cache line and the remaining 8 bits the 2nd cache line.

with "Weak" memory rules in place, and with the other job of the
addressmatch matrix taking care of separating LDs and STs into mutually
exclusive batches, splitting into multiple operations is perfectly "fine".

now with these bytemasks available, and the remaining bits of the binary
form of the address (bits 4 thru 48) being aligned to a cache line
boundary, all we need do is identify which requests have the exact same
upper bits and we can merge the bytemasks for all of those and thus make
only the one single L1 cache line request for all of them.

this is the task of the L0 Cache/Buffer and as you can see it very closely
and very specifically depends on the criteria being satisfied by earlier
components.

the datapath requirements are also absolutely MENTAL and we really,
*really* cannot have 8 or 16 way to 4 or 8 way multiplexers, it is just too
much as it will result in literally thousands of wires because we have 48
bit addresses, 64 bit data, and additional control signals.

assume 128 bits per LDST Function Unit, if we have 8 LDST FUs, each with 2
ports (one for aligned, one for misaligned), that is 128 x 2 x 8 bits wide
i.e. 2048 bits datapath!

now imagine we design an algorithm that needs a 4 or 8 way multiplexer on
that, to get 4 or 8 simultaneous requests into even just an 8 entry
"Queue", implemented as a standard 4R4W (or 8R8 W) Memory or registers, and
that is a staggering 8192 or 16384 wires through multiplexers!

that is completely mental and there is no way we can do that, the layout
would be far too complex even if it was practical.

this is just one of the things that we have to be extremely careful about
and it is why i am taking time to draw out these diagram.

so the current proposed L0 Cache/Buffer has *independent* ports on each
row, for use by each Function Unit.

and onto the L1 side the connection is 1R *or* 1W at 128 bit (cache line)
width.

yesterday i made some corrections that merged the pair of L0 Cache/Buffers
into a single dual ported one.  i explain why on the 6600 scoreboard page.

thus each FU produces up to 2 requests, one goes left, the other goes
right, into one of two "lanes", mutually exclusively separated by address
bit 4.

there are therefore *two* 128 bit wide ports to *two* separate L1 caches,
one for addr[4] = 0 and the other for = 1 and they interleave alternate
cache lines nicely.

the number of ways is a standard L1 cache decision as is everything else
about the L1 cache.

so please do take into account, Jacob, that the data path sizes are
completely mental, and that multi way multiplexing is not an option.

the multiplexing for the design that i did is only a 2x2 crossbar for each
FU, with 192 bit datapaths on each (might be able to get that down to 128
or slightly less).  that is *all* because on the other side, after the
bytemask merge, it is straight 1R or 1W to the L1 cache.

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] memory interface diagram woes
From: Luke Kenneth Casson Leighton @ 2020-04-23  9:02 UTC (permalink / raw)
  To: Libre-RISCV General Development
In-Reply-To: <CAC2bXD4xkEJFTaPo7gKG9ejMvJyNtoA2zbDEKVqHk8abFv-jVg@mail.gmail.com>

On Thursday, April 23, 2020, Jacob Lifshay <programmerjake@gmail.com> wrote:

> I had an idea for a memory interface design that I've been slowly
> working on for more than a week, though have not been able to draw it
> out adequately because I keep getting frustrated by the lack of nice,
> power-of-2 alignment features and support for structurally recursive
> diagrams in Dia and the lack of copy-paste when using physical paper.


doh :)

there is a reason why i write things out as diagrams (and why i look at
graphviz all the time).  they map easier in my head to gates, and as a
result it becomes possible to check that it hasn't any surprise pathways in
HW


> I'll try to write it out as text on the wiki tomorrow.


ok.


> It is a bit different than the design Luke came up with and I think it
> addresses several of that design's shortcomings.


ok i can grok that.

thanks for explaining jacob.




-- 
---
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

* [libre-riscv-dev] memory interface diagram woes
From: Jacob Lifshay @ 2020-04-23  6:49 UTC (permalink / raw)
  To: Libre-RISCV General Development

I had an idea for a memory interface design that I've been slowly
working on for more than a week, though have not been able to draw it
out adequately because I keep getting frustrated by the lack of nice,
power-of-2 alignment features and support for structurally recursive
diagrams in Dia and the lack of copy-paste when using physical paper.
I'll try to write it out as text on the wiki tomorrow.

It is a bit different than the design Luke came up with and I think it
addresses several of that design's shortcomings.

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] [Bug 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 19:42 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

--- Comment #8 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Luke Kenneth Casson Leighton from comment #6)

> full bootloader and on-board boot firmware source.

on-board boot ROM, sorry.  shorter.  if everyone's good with it
i'll write to roberto.

next we need to work out what (and where) to put something
about these:
https://www.powerprogress.org/en/
https://www.powerpc-notebook.org/en/

this is a good clean design, i like it
https://i0.wp.com/www.powerpc-notebook.org/wp-content/uploads/2020/04/ppc_bloc_diagram_05_04_2020_s.png?ssl=1

it's actually pretty similar to what would go on our SoCs,
particularly if we could get hold of a PCIe PHY.

however... 2x DDR3, that's 250-300 pins, right there.
(actually just found the datasheet, it's 1x 32/64 so will
be around... 200-250).  3 PCIE x4
plus a PCIe x2, is... a looot more.  300+ i think.

datasheet says it's 780 pins FP-PBGA, 23x23mm

-- 
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 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 19:14 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

--- Comment #7 from Yehowshua <yimmanuel3@gatech.edu> ---
Sounds solid.

small typo on my part:

> that have been know to break in the past.

should be ``known"

Final revision?:
Libre-SOC is a Libre Hardware-Software project that aims to deliver a physical 
POWER compliant SOC that comes complete with a CPU, GPU, VPU, and DDR
controller. All the software and hardware from the drivers down to the RTL and
VLSI cells are libre-licensed. Libre-SOC is also providing the necessary
drivers amongst which include Kazan (a Vulkan 3D driver) and the full
bootloader and on-board boot firmware source.

The intended market includes customers who desire acceleration in the embedded
space without relying on ARM or 3rd party proprietary drivers that have been
known to break in the past.

The first iteration of Libre-SOC targets a single-core at 180nm. Subsequent
generations target SMP cores at a smaller node size, for typical use in SBC
designs.

-- 
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 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 19:00 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

--- Comment #6 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Jacob Lifshay from comment #5)
> (In reply to Luke Kenneth Casson Leighton from comment #4)
> > > The first iteration of LibreSOC targets a single-core at 180nm. Subsequent
> > > generations target more cores at a smaller node size.
> > 
> > looks good.  i'd put "target multiple SMP cores" though.  "more cores"
> > could mean "NUMA" or "SIMT".  SIMT is almost impossible to program for
> > general-purpose, and NUMA is a royal pain, memory-wise.
> 
> Sounds good, though, it would be NUMA if we have multiple chips wired
> together using OmniXpress or similar.

true


> Also, we should be consistent spelling Libre-SOC with the hyphen like we
> agreed.

good catch.

Libre-SOC is a Libre Hardware-Software project that aims to deliver a physical 
POWER compliant SOC that comes complete with a CPU, GPU, VPU, and DDR
controller. All the software and hardware from the drivers down to the RTL and
VLSI cells are libre-licensed. Libre-SOC is also providing the necessary
drivers amongst which include Kazan (a Vulkan 3D driver) and the full
bootloader and on-board boot firmware source.

The intended market includes customers who desire acceleration in the embedded
space without relying on ARM or 3rd party proprietary drivers that have been
know to break in the past.

The first iteration of Libre-SOC targets a single-core at 180nm. Subsequent
generations target SMP cores at a smaller node size, for typical use in SBC
designs.

-- 
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 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 16:16 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

Jacob Lifshay <programmerjake@gmail.com> changed:

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

--- Comment #5 from Jacob Lifshay <programmerjake@gmail.com> ---
(In reply to Luke Kenneth Casson Leighton from comment #4)
> > The first iteration of LibreSOC targets a single-core at 180nm. Subsequent
> > generations target more cores at a smaller node size.
> 
> looks good.  i'd put "target multiple SMP cores" though.  "more cores"
> could mean "NUMA" or "SIMT".  SIMT is almost impossible to program for
> general-purpose, and NUMA is a royal pain, memory-wise.

Sounds good, though, it would be NUMA if we have multiple chips wired together
using OmniXpress or similar.


Also, we should be consistent spelling Libre-SOC with the hyphen like we
agreed.

-- 
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 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 16:00 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

--- Comment #4 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Yehowshua from comment #3)
> Ah I see. That makes sense to me.
> 
> So this:
> 
> LibreSOC is an Open Hardware and Open Software project that aims to deliver
> a physical  POWER compliant SOC that comes complete with a CPU, GPU, VPU,
> and DDR controller. All the software and hardware from the drivers down to
> the RTL and VLSI cells are open. 

libre-licensed.  "open" leaves the possibility of accusations of "fake open
source".

"we'll release the code when it's revision 1.0... but it's open!".

"we'll release the code when it's ready... but it's open!"

"we'll release the code... under a non-commercial license... but it's open!"

> LibreSOC is also providing the necessary
> drivers amongst which include Vulkan and OpenCL drivers.

no, not OpenCL.  that has to come later as we'll need an entire new
grant to do it.

> The intended market includes customers who desire acceleration in the
> embedded space without relying on ARM's proprietary drivers that have been
> know to break in the past.

and piss people off :)

> The first iteration of LibreSOC targets a single-core at 180nm. Subsequent
> generations target more cores at a smaller node size.

looks good.  i'd put "target multiple SMP cores" though.  "more cores"
could mean "NUMA" or "SIMT".  SIMT is almost impossible to program for
general-purpose, and NUMA is a royal pain, memory-wise.

-- 
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 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 15:41 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

--- Comment #3 from Yehowshua <yimmanuel3@gatech.edu> ---
Ah I see. That makes sense to me.

So this:

LibreSOC is an Open Hardware and Open Software project that aims to deliver a
physical  POWER compliant SOC that comes complete with a CPU, GPU, VPU, and DDR
controller. All the software and hardware from the drivers down to the RTL and
VLSI cells are open. LibreSOC is also providing the necessary drivers amongst
which include Vulkan and OpenCL drivers.

The intended market includes customers who desire acceleration in the embedded
space without relying on ARM's proprietary drivers that have been know to break
in the past.

The first iteration of LibreSOC targets a single-core at 180nm. Subsequent
generations target more cores at a smaller node size.

-- 
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 257] Implement demo Load/Store queueing algorithm
From: bugzilla-daemon @ 2020-04-22 13:38 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-257-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=257

--- Comment #27 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Luke Kenneth Casson Leighton from comment #26)

> another "easier" way may be to simply have *FOUR* 64-bit Wishbone
> interfaces:
> 
> * L1 left LO-64
> * L1 left HI-64
> * L1 right LO-64
> * L1 right HI-64
> 
> then we "jam" those down onto a single 64-bit Wishbone interface
> using a standard Wishbone Bus arbiter

i was going to say that those could all be 4 independent L1 caches,
however both left ones and both right ones have the exact same address,
so it would be a waste of power.

*however*... the 16-bit byte read/write-enable lines back at L0 can be
split into two then ORed together to produce a *single* "LO-64" and "HI-64"
L1 cache line read/write.

-- 
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 257] Implement demo Load/Store queueing algorithm
From: bugzilla-daemon @ 2020-04-22 13:06 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-257-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=257

--- Comment #26 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
L1 cache concept, go for it - that's really straightforward and standard,
although important to appreciate that 16 *byte* level read/write-enable lines
will come through from the Function Units.

if you can also do a non-standard-wishbone-128 to standard-wishbone-64
bit "funnel", that analyses the 16 byte-level read/write-enable lines,
splitting them into two halves covering the LO-64-bit and HI-64-bit
that would fit well.

actually to be honest at the 128-bit-level it doesn't even need to be
Wishbone, which may make design easier.

another "easier" way may be to simply have *FOUR* 64-bit Wishbone
interfaces:

* L1 left LO-64
* L1 left HI-64
* L1 right LO-64
* L1 right HI-64

then we "jam" those down onto a single 64-bit Wishbone interface
using a standard Wishbone Bus arbiter


L0 cache/buffer, give it a little time for some review / thought
as i just found a mistake in the L0 cache/buffer design and had
to adjust it:
https://libre-soc.org/3d_gpu/twin_l0_cache_buffer.jpg

-- 
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 288] intro post for powerpc-notebook.org and https://www.powerprogress.org/en/
From: bugzilla-daemon @ 2020-04-22 12:49 UTC (permalink / raw)
  To: libre-riscv-dev
In-Reply-To: <bug-288-13@https.bugs.libre-soc.org/>

https://bugs.libre-soc.org/show_bug.cgi?id=288

--- Comment #2 from Luke Kenneth Casson Leighton <lkcl@lkcl.net> ---
(In reply to Yehowshua from comment #1)
> Something Like This:
> 
> LibreSOC is an Open Hardware and Open Software project that aims to deliver
> a physical quad core 180nm POWER compliant SOC that comes complete with a

no, single-core.  we're not doing quad-core, it will require something
like 160 mm^2 and that would cost around USD $60,000 just on its own.
not to mention, the power consumption would be immense, well beyond
what a standard low-cost JEDEC QFP package could handle.

we're keeping it to single-core, no SMP, no L2 cache.
https://libre-soc.org/3d_gpu/180nm_single_core_testasic_memlayout.jpg

the *next* version is quad-core, 28-45nm, full SMP.

-- 
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