[J-core] My trip to Japan

Rob Landley rob at landley.net
Sun Dec 29 23:07:25 UTC 2019

On 12/28/19 9:12 AM, John Paul Adrian Glaubitz wrote:
> Hi!
> On 12/26/19 10:43 PM, Joh-Tob Schäg wrote:
>> time. During the occasion i received a turtle board with some binaries
>> . I am free to talk about it, so if you have questions ask me. The
> Are Turtle boards already available for purchase for interested developers?

We're putting together a crowdsupply campaign, but we need full bitstream
support for all the hardware and pi/turtle have a lot of I/O devices. The USB,
HDMI, and audio aren't finished yet. And we're like to update the SD card from
mmc to SD 1.0 I(which is like 10x faster).

Our current round of prototypes (we fabbed 50 this time, of the LX45 variant)
have a glitch: the 3.5 audio jack have the pins in the wrong order, which
wouldn't be so bad if one of them wasn't GROUND in the wrong place. Grrr. (That
we can't fix in the bitstream, and no you can't use a normal 3.5/RCA audio
connector with ground on the wrong ring.)

Oh, and we're redoing the serial console speed from 38.4k to 250k, but that
requires reflashing the atmel boot processor's software too because they have to
match. (uartlite has a hardwired speed, which _mostly_ doesn't matter when
you're piping it into USB serial which converts it to USB packets, but the data
transmission is like 5x faster and it makes the serial console _feel_ a lot

Ideally, we'd have the crowdsupply campaign up for our LCA talk on January 13,
but it doesn't look like we'll have it ready. (Well, we might have the
_campaign_ up but not devices shipping yet if we want to preload the right
bitstream. But that's sort of cheating...)

If somebody really wants to buy one of the prototypes immediately we can
probably mail a couple more out by hand, the way we did for Joh-Toh, and you can
update the bitstream once we've got the full one ready? That's a Jeff question.

>> FPGA came flashed with a dual core J32. Jeff also showed me how to
>> reflash the FPGA. It has a dedicated flash mode which can be triggered
>> at start-up. In the flash mode the board talks over serial and has an
>> assistant for flashing an image form the µSD card.
>> The linux build that came with it did not support J32 yet which means
>> it crashes when mmap is called. (I understood from Robs explanation
>> that this expected behaviour when running a nommu kernel on a device
>> with MMU).
> J32 would be equivalent to SH4, wouldn't it?

More like sh3. I believe sh4 was dual-issue? Userspace is the same (trying for
100% compatibility in userspace) but the MMU is programmed differently.

> If yes, I'm really looking
> forward to that. Having new SH4 hardware available for purchase would
> help Debian's sh4 port (of which I am the principal maintainer of) and
> NetBSD's port for sh3/sh4.

Probably what we'll do is put the crowdsupply page up for a bit taking preorders
to see what the LX25/LX45 ratio looks like, and then kick off another batch of
boards which will probably once again take another 3 weeks for the parts going
_in_ to make it through customs and the boards going OUT to make it through
customs, on _top_ of the actual manufacturing time...

As to _when_ the page will be up, the current plan is "sometime in january".
(I'm _hoping_ this doesn't means 11:59 Jan 31 with 1/3 of it still placeholder
content, but deadlines are motivational...)

> For Debian, I'm currently building most packages for sh4 using qemu-user
> which has some bugs meaning that not all packages build fine which other-
> wise build fine on real hardware. So, I'm really looking forward for new
> sh4-compatible hardware.

To be honest we haven't talked about the turtle boards since before christmas
because I'm off for the holidays trying to get toysh usable and Jeff's been
neck-deep in researching ASIC tools (if anybody knows where
https://isi.edu/~jsondeen/magic.html went, that would be nice), but we fly back
to Japan on the 7th and turtle support is first on the agenda. (Well, probably
second after linux.conf.au which is the following monday.)

> Adrian


More information about the J-core mailing list