BeagleBoard-X15 - seriously? :)

7,959 views
Skip to first unread message

Maxim Podbereznyy

unread,
Nov 7, 2014, 4:05:21 AM11/7/14
to beagleboard
Hey guys!

BeagleBoard-X15 - Are you really going to release such monster? :)

Alfredo Muniz

unread,
Nov 7, 2014, 4:14:38 AM11/7/14
to beagl...@googlegroups.com

On Fri, Nov 7, 2014 at 4:05 AM, Maxim Podbereznyy <lisa...@gmail.com> wrote:
BeagleBoard-X15 - Are you really going to release such monster? :)

That sounds amazing. What are we looking at in terms of expected price?

William Hermans

unread,
Nov 7, 2014, 4:42:47 AM11/7/14
to beagl...@googlegroups.com

--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Bas Laarhoven

unread,
Nov 7, 2014, 9:10:45 AM11/7/14
to beagl...@googlegroups.com
On 7-11-2014 10:42, William Hermans wrote:

But notice that the processor still has PRUSS units integrated (4 PRUs if I'm reading it right) ! So this looks very promising :-)

-- Bas

Gerald Coley

unread,
Nov 7, 2014, 9:37:04 AM11/7/14
to beagl...@googlegroups.com
Yep. That monster is showing some life! It is actually looking good. But there is that other issue we always seem to have to deal with.

Processor availability of a new part. We plan to build  a limited number of these initially just to make sure we get it all pounded out before we build a bunch of them.

Oh and we also have a plan for a lower cost version based on a single core version.


Gerald

Message has been deleted

Drew Fustini

unread,
Nov 7, 2014, 11:39:52 AM11/7/14
to beagl...@googlegroups.com
On Fri, Nov 7, 2014 at 10:07 AM, _rh__ <richard...@lavabit.com> wrote:
> On Fri, 7 Nov 2014 12:05:13 +0300>
> Beagleboard-x15, the nexgen pseudo-opensource hardware based on

Why "pseudo-opensource hardware"?

thanks,
drew
http://keybase.io/pdp7

Robert Nelson

unread,
Nov 7, 2014, 1:12:30 PM11/7/14
to Beagle Board
On Fri, Nov 7, 2014 at 10:39 AM, Drew Fustini <pdp7...@gmail.com> wrote:
> On Fri, Nov 7, 2014 at 10:07 AM, _rh__ <richard...@lavabit.com> wrote:
>> On Fri, 7 Nov 2014 12:05:13 +0300>
>> Beagleboard-x15, the nexgen pseudo-opensource hardware based on
>
> Why "pseudo-opensource hardware"?

He's probably irritated about the 'sgx' graphics. He's more then free
to contact TI, place a $bil order and request TI use an "open source"
gpu, probably take around a year for development.

Or he could just start reverse engineering sgx or pay someone. But in
my 6 years working bb.org no user has done that and gotten anywhere
yet..

Regards,

--
Robert Nelson
http://www.rcn-ee.com/

Gerald Coley

unread,
Nov 7, 2014, 1:32:49 PM11/7/14
to beagl...@googlegroups.com
Definitely not one of those. Look at what it has on it and make a good guess.

Gerald


On Fri, Nov 7, 2014 at 10:07 AM, _rh__ <richard...@lavabit.com> wrote:
On Fri, 7 Nov 2014 12:05:13 +0300
Maxim Podbereznyy <lisa...@gmail.com> wrote:

> Hey guys!
>
> BeagleBoard-X15 <http://www.elinux.org/Beagleboard:BeagleBoard-X15> -

> Are you really going to release such monster? :)

Beagleboard-x15, the nexgen pseudo-opensource hardware based on
TI AM5728 SoC dual core 1.5GHZ A15,2GB DDR3L (dual 32bit busses),
eSATA, 3 USB3.0 ports, integrated HDMI (1920x1080@60),separate LCD
port, video In port, 4GB eMMC, microSD, analog audio i/o, dual-Gb
ethernet.

Will it fall under $100 or will it be $100-200? Make your bets.

Dr. H. Nikolaus Schaller

unread,
Nov 7, 2014, 1:39:03 PM11/7/14
to beagl...@googlegroups.com
There is even a GNU project (wiki, mailing list) with this goal. But a project
with nobody contributing recently:

http://powervr.gnu.org.ve/doku.php

>
> Regards,
>
> --
> Robert Nelson
> http://www.rcn-ee.com/
>

Maxim Podbereznyy

unread,
Nov 7, 2014, 1:48:21 PM11/7/14
to beagleboard

I believe RPi is really pseudo open source because you can't make the same board although you have all schematics. Stuff from bb.org was cloned dozens of times

07 Ноя 2014 г. 21:38 пользователь "Dr. H. Nikolaus Schaller" <h...@goldelico.com> написал:

Robert Nelson

unread,
Nov 7, 2014, 1:49:15 PM11/7/14
to Beagle Board
On Fri, Nov 7, 2014 at 12:38 PM, Dr. H. Nikolaus Schaller
<h...@goldelico.com> wrote:
>
> Am 07.11.2014 um 19:12 schrieb Robert Nelson <robert...@gmail.com>:
>
>> On Fri, Nov 7, 2014 at 10:39 AM, Drew Fustini <pdp7...@gmail.com> wrote:
>>> On Fri, Nov 7, 2014 at 10:07 AM, _rh__ <richard...@lavabit.com> wrote:
>>>> On Fri, 7 Nov 2014 12:05:13 +0300>
>>>> Beagleboard-x15, the nexgen pseudo-opensource hardware based on
>>>
>>> Why "pseudo-opensource hardware"?
>>
>> He's probably irritated about the 'sgx' graphics. He's more then free
>> to contact TI, place a $bil order and request TI use an "open source"
>> gpu, probably take around a year for development.
>>
>> Or he could just start reverse engineering sgx or pay someone. But in
>> my 6 years working bb.org no user has done that and gotten anywhere
>> yet..
>
> There is even a GNU project (wiki, mailing list) with this goal. But a project
> with nobody contributing recently:
>
> http://powervr.gnu.org.ve/doku.php


Yeah, that gnu project has been around for awhile, they have only
updated the list of boards..

I thought this we a little interesting..

http://blog.imgtec.com/powervr/the-latest-powervr-sdk-v3-4-and-tools-are-live

imgtec must be now desperate for new business, as they provide the isa
for the 6 series..

PowerVR SGX544 is used on the am57xx

Robert Nelson

unread,
Nov 7, 2014, 1:51:11 PM11/7/14
to Beagle Board
On Fri, Nov 7, 2014 at 12:48 PM, Maxim Podbereznyy <lisa...@gmail.com> wrote:
> I believe RPi is really pseudo open source because you can't make the same
> board although you have all schematics. Stuff from bb.org was cloned dozens
> of times

more like "mob" open source.. you can start making a board
(hardkernel guys), but then broadcom stops selling you chips ;)
Because they don't want you to compete with the foundation..

Laurent Desnogues

unread,
Nov 7, 2014, 4:41:07 PM11/7/14
to beagl...@googlegroups.com
On Fri, Nov 7, 2014 at 7:32 PM, Gerald Coley <ger...@beagleboard.org> wrote:
> Definitely not one of those. Look at what it has on it and make a good
> guess.

You got me wondering because Richard basically copied his info
from Nishanth Menon himself:

https://patchwork.kernel.org/patch/5245961/

"BeagleBoard-X15 is the next generation Open Source Hardware
BeagleBoard based on TI's AM5728 SoC featuring dual core 1.5GHZ A15
processor. The platform features 2GB DDR3L (w/dual 32bit busses),
eSATA, 3 USB3.0 ports, integrated HDMI (1920x108@60), separate LCD
port, video In port, 4GB eMMC, uSD, Analog audio in/out, dual 1G
Ethernet."

So where's the truth? :-)

AM5728 looks exciting, in particular the 2xC66!


Laurent

Nishanth Menon

unread,
Nov 7, 2014, 4:43:26 PM11/7/14
to BeagleBoard
On Fri, Nov 7, 2014 at 3:40 PM, Laurent Desnogues
<laurent....@gmail.com> wrote:
> On Fri, Nov 7, 2014 at 7:32 PM, Gerald Coley <ger...@beagleboard.org> wrote:
>> Definitely not one of those. Look at what it has on it and make a good
>> guess.
>
> You got me wondering because Richard basically copied his info
> from Nishanth Menon himself:
>
> https://patchwork.kernel.org/patch/5245961/
>
Sigh.. at least use the typo corrected patch V2:
https://patchwork.kernel.org/patch/5254791/ :P

--
Regards,
NM


--
---
Regards,
Nishanth Menon

sa_Penguin

unread,
Nov 7, 2014, 8:06:31 PM11/7/14
to beagl...@googlegroups.com
I must admit to a little curiosity about the design process.
Is there a particular reason to avoid DIMM socket RAM?

I find the idea of recycling old PC RAM, or buying new from a PC shop, rather appealing.
I get a cheap(er) board, can upgrade, and can even venture into overclocking...

John Syn

unread,
Nov 7, 2014, 10:56:08 PM11/7/14
to beagl...@googlegroups.com
Looking at BeagleBoard-X15 wiki, I like that the connectors are only on two sides, so now it is possible to use enclosures like this:


Pricing is pretty reasonable:


Nice work Gerald. 

Regards,
John

Kumar Abhishek

unread,
Nov 8, 2014, 5:53:54 AM11/8/14
to beagl...@googlegroups.com
I see almost nil GPIO pin headers on this board. How many GPIO pins would be there on this one, and would it include all the PRU pins?

Also, just to confirm if there's an eSATA port onboard.

Gerald Coley

unread,
Nov 8, 2014, 9:14:44 AM11/8/14
to beagl...@googlegroups.com
There is only four connectors with 66 pins each, so I could only get 157 GPIO pins on it. And yes it is eSATA.

BTW, the connectosr are on the back of the board. You can see their mounting holes from the top of the board, three on each end. The connectors are SMT.

Gerald

sixvolts

unread,
Nov 8, 2014, 11:31:25 AM11/8/14
to beagl...@googlegroups.com
Any ideas how this compares to the wandboard quad? 4x Cortex-A9 vs 2x Cortex-A15?

Also, this would be awesome to setup a small Arm cluster with since it has a decent RAM/core count ratio, gig/e and sata. 

Scott Murray

unread,
Nov 8, 2014, 2:06:33 PM11/8/14
to beagl...@googlegroups.com
Are either of the AM5278's PCIe ports run out to the connectors?

Thanks,

Scott

Robert Nelson

unread,
Nov 8, 2014, 2:15:56 PM11/8/14
to Beagle Board


On Nov 8, 2014 1:06 PM, "Scott Murray" <sc...@spiteful.org> wrote:
>
> Are either of the AM5278's PCIe ports run out to the connectors?

Yeap!

Kumar Abhishek

unread,
Nov 8, 2014, 2:27:43 PM11/8/14
to beagl...@googlegroups.com

I'm just hoping that the pricing is right - this is a killer board at around 150$.
Not sure if this can replace a desktop but it sounds like a good NAS/Media server/HTPC builds though that might not be an intended application of this beast :)

Is it the PCIe part which isn't open source on the board?

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/1t3yxkPYSB8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

Robert Nelson

unread,
Nov 8, 2014, 2:47:04 PM11/8/14
to Beagle Board


On Nov 8, 2014 1:27 PM, "Kumar Abhishek" <kumar.abhi...@gmail.com> wrote:
>
> I'm just hoping that the pricing is right - this is a killer board at around 150$.
> Not sure if this can replace a desktop but it sounds like a good NAS/Media server/HTPC builds though that might not be an intended application of this beast :)
>
> Is it the PCIe part which isn't open source on the board?

Imgtech, sgx 3d graphics core is closed..


Message has been deleted

William Hermans

unread,
Nov 8, 2014, 4:05:25 PM11/8/14
to beagl...@googlegroups.com
So, *why* does TI keep using hardware that has closed source drivers ? Does TI realize this is killing their stuff in the eyes of many ppl out there ?

On Sat, Nov 8, 2014 at 1:05 PM, rh_ <richard...@lavabit.com> wrote:
On Fri, 7 Nov 2014 12:32:19 -0600
Gerald Coley <ger...@beagleboard.org> wrote:

> Definitely not one of those. Look at what it has on it and make a good
> guess.

Ok "definitely" must mean $200-400. Is this not good guessing?
This starts to be a in different realm and the competition is more
fierce. But you know more about that than I ever will and it's
likely that I just don't understand the landscape. Dual Gb ethernet
will attract a lot of attention for server/firewall/IDS/IPS.

Robert Nelson

unread,
Nov 8, 2014, 4:22:12 PM11/8/14
to Beagle Board


On Nov 8, 2014 3:05 PM, "William Hermans" <yyr...@gmail.com> wrote:
>
> So, *why* does TI keep using hardware that has closed source drivers ? Does TI realize this is killing their stuff in the eyes of many ppl out there ?

Well it does have a Vivante GC320 2D GPU so maybe they are starting the transition....

Gerald Coley

unread,
Nov 8, 2014, 4:36:21 PM11/8/14
to beagl...@googlegroups.com
Both PCIe ports are connected to the expansion headers.

As to TI using closed source drivers,that is the result of buying IP from other sources, something that most SOC suppliers do in one area or the other.

Gerald

William Hermans

unread,
Nov 8, 2014, 5:12:44 PM11/8/14
to beagl...@googlegroups.com
Gerald, perhaps, but it is a huge disappointment to me, and I am sure to many others out there.

This board to me screams low-end x86 killer, and squarely lands in the multimedia arena. A bit much for a NAS like system if you ask me, but surely would do a great job of it.

It will be interesting to see how the USB3 drivers shape up as well.

Robert Nelson

unread,
Nov 8, 2014, 5:20:34 PM11/8/14
to Beagle Board


On Nov 8, 2014 4:12 PM, "William Hermans" <yyr...@gmail.com> wrote:
>
> Gerald, perhaps, but it is a huge disappointment to me, and I am sure to many others out there.
>
> This board to me screams low-end x86 killer, and squarely lands in the multimedia arena. A bit much for a NAS like system if you ask me, but surely would do a great job of it.

With the dual dsp's and dual m4's, video decoding should be nice..

>
> It will be interesting to see how the USB3 drivers shape up as well.

It's not the classic 'musb' IP from a completely different manufacture.

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/drivers/usb/dwc3

John Syn

unread,
Nov 8, 2014, 5:22:00 PM11/8/14
to beagl...@googlegroups.com

From: sixvolts <drew...@gmail.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Saturday, November 8, 2014 at 8:31 AM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: [beagleboard] Re: BeagleBoard-X15 - seriously? :)

Any ideas how this compares to the wandboard quad? 4x Cortex-A9 vs 2x Cortex-A15?
According to this article:


Cortex-A15 performance is double that of Cortex-A9. Given that Wandboard runs at just at 1GHz and the X15 runs at 1.5GHz, I would guess that the X15 runs about 30% faster; however, the X15 has a dual 32bit bus, so the memory bandwidth is way higher. Also, the X15 also has dual C66x DSP processors, dual Cortex M4s and dual PRUs, so this board will be a screamer compared to the wandboard.

Regards,
John

Also, this would be awesome to setup a small Arm cluster with since it has a decent RAM/core count ratio, gig/e and sata. 

--

William Hermans

unread,
Nov 8, 2014, 5:46:21 PM11/8/14
to beagl...@googlegroups.com
I did not notice the eSATA on the specs that I've read so far. However assuming this board does have an eSATA port on it, does this port also support hardware, or software port multipliers ?

Gerald Coley

unread,
Nov 8, 2014, 5:48:22 PM11/8/14
to beagl...@googlegroups.com
There is eSATA on the board. I do not know if it supports port multipliers. That would be a TI question.

Gerald

William Hermans

unread,
Nov 8, 2014, 6:07:15 PM11/8/14
to beagl...@googlegroups.com
I'd be surprised if it did. I've been keeping up with the technology for many years now, and mostly only Silicon image ( SIL3132 most notably ) chip sets support port multiplication. However, in the last year or so I have not been keeping as close an eye on this stuff.

Id be pleasantly surprised though !

robert.berger

unread,
Nov 9, 2014, 12:16:09 AM11/9/14
to beagl...@googlegroups.com
Hi,


On Saturday, November 8, 2014 6:31:25 PM UTC+2, sixvolts wrote:
Any ideas how this compares to the wandboard quad? 4x Cortex-A9 vs 2x Cortex-A15

Well the Cortex-A15 includes full hardware visualization, which the Cortex-A9 lacks. If this new board with multiple Cortex-A15 and dual Cortex-M4 is reasonably priced it might used as a references board for an open source Linux/RTOS solution.

*) FreeRTOS on one Cortex-A15, GNU/Linux on the other
*) FreeRTOS on one Cortex-M4, GNU/Linux on the Cortex-A15 cores,...

anyone interested in something like this?

Regards,

Robert

Maxim Podbereznyy

unread,
Nov 9, 2014, 1:13:42 AM11/9/14
to beagleboard

John Syn,
Wandboard Quad does have 64 bit memory bus.

P.S. I don't know who needs dual DSP onboard because TI definitely will not support them as should like it was for omap3/dm37. I'd better have well supported hardware video encoder/decoder rather than double general purpose DSPs without any software support

09 Ноя 2014 г. 1:21 пользователь "John Syn" <john...@gmail.com> написал:

Dr. H. Nikolaus Schaller

unread,
Nov 9, 2014, 5:00:40 AM11/9/14
to beagl...@googlegroups.com

Am 08.11.2014 um 21:05 schrieb rh_ <richard...@lavabit.com>:

> On Fri, 7 Nov 2014 12:32:19 -0600
> Gerald Coley <ger...@beagleboard.org> wrote:
>
>> Definitely not one of those. Look at what it has on it and make a good
>> guess.
>
> Ok "definitely" must mean $200-400. Is this not good guessing?
> This starts to be a in different realm and the competition is more
> fierce. But you know more about that than I ever will and it's
> likely that I just don't understand the landscape. Dual Gb ethernet
> will attract a lot of attention for server/firewall/IDS/IPS.

I think the feature set of the X15 is almost comparable to the
OMAP5432EVM (which AFAIR was originally planned to become
a PandaBoard 5):

http://www.svtronics.com/5432

So this seems to confirm the price range $200-400...

Laurent Desnogues

unread,
Nov 9, 2014, 5:17:48 AM11/9/14
to beagl...@googlegroups.com
On Sun, Nov 9, 2014 at 7:13 AM, Maxim Podbereznyy <lisa...@gmail.com> wrote:
>
> P.S. I don't know who needs dual DSP onboard because TI definitely will not
> support them as should like it was for omap3/dm37. I'd better have well
> supported hardware video encoder/decoder rather than double general purpose
> DSPs without any software support

Do you mean there won't be any way to access and use the DSP
from Linux, or just that TI won't provide us with codecs and other
software goodies?

My understanding is that DSP Bridge and DSP Link have been
replaced by IPC 3.x (which has OMAP5 support).

http://processors.wiki.ti.com/index.php/IPC_Software_Options

I hope we'll have support for that on the new shiny board!


Laurent

Robert Nelson

unread,
Nov 9, 2014, 9:42:37 AM11/9/14
to Beagle Board

The am57xx uses the omap4 interface which is mainline as rpmsg.

>
>
> Laurent

Chris Morgan

unread,
Nov 9, 2014, 9:56:28 AM11/9/14
to beagl...@googlegroups.com
The bbb sgfx is one reason why I've found the minnowboard is so
appealing. While we (my company) are using the bbb, and its a great
board, if we needed something in a higher performance range and could
float the cost we've been eyeing the minnowboard due to its fully open
graphics stack.

If the sgx was easier to get working with yocto or mainline kernels we
might think differently but we've already sunk a few weeks of effort
into getting accelerated graphics working without much success.

Chris

John Syn

unread,
Nov 9, 2014, 10:27:38 AM11/9/14
to beagl...@googlegroups.com

From: Robert Nelson <robert...@gmail.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Sunday, November 9, 2014 at 6:42 AM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: Re: [beagleboard] Re: BeagleBoard-X15 - seriously? :)


On Nov 9, 2014 4:17 AM, "Laurent Desnogues" <laurent....@gmail.com> wrote:
>
> On Sun, Nov 9, 2014 at 7:13 AM, Maxim Podbereznyy <lisa...@gmail.com> wrote:
> >
> > P.S. I don't know who needs dual DSP onboard because TI definitely will not
> > support them as should like it was for omap3/dm37. I'd better have well
> > supported hardware video encoder/decoder rather than double general purpose
> > DSPs without any software support
>
> Do you mean there won't be any way to access and use the DSP
> from Linux, or just that TI won't provide us with codecs and other
> software goodies?
>
> My understanding is that DSP Bridge and DSP Link have been
> replaced by IPC 3.x (which has OMAP5 support).
>
> http://processors.wiki.ti.com/index.php/IPC_Software_Options
>
> I hope we'll have support for that on the new shiny board!

The am57xx uses the omap4 interface which is mainline as rpmsg.

Did you mean OMAP5 interface? The last I looked, OMAP5 support wasn’t complete and as you know I’ve been pushing for RPMSG/REMOTEPROC support on this processor and the only feedback I received was that support wasn’t going to happened anytime soon. To me, RPMSG/REMOTEPROC and SGX should be on TI’s high priority list or as Maxim said, this processor will end up just like the DM3730; great processor without any ability to use all the features. BTW, I believe IPC3.x is SYSLINK3 which is RPMSG. Once we have RPMSG/REMOTEPROC, all the video encoding/decoding together with codecs etc should pretty simple for TI to add and that would make this a fantastic product. 

Regards,
John

manu....@gmail.com

unread,
Nov 9, 2014, 6:36:08 PM11/9/14
to beagl...@googlegroups.com
Please tell me that the whole MCASP port 1 & 2 are exposed on the connectors (or port 3 to 8 ...)

Emmanuel.

Gerald Coley

unread,
Nov 9, 2014, 8:17:48 PM11/9/14
to beagl...@googlegroups.com
They are not. Ports 5 and 6 should be, but I have not tried counting all the pins of the McASP that are out there lately. Been a little busy. 


Gerald

Jason Kridner

unread,
Nov 10, 2014, 4:15:28 AM11/10/14
to beagl...@googlegroups.com
On Sun, Nov 9, 2014 at 1:13 AM, Maxim Podbereznyy <lisa...@gmail.com> wrote:
> John Syn,
> Wandboard Quad does have 64 bit memory bus.
>
> P.S. I don't know who needs dual DSP onboard because TI definitely will not
> support them as should like it was for omap3/dm37. I'd better have well
> supported hardware video encoder/decoder rather than double general purpose
> DSPs without any software support

When OMAP3 came out, the only compilers for C6000 were expensive,
closed-source compilers. Now, there is support in mainline GCC for
C6000. As the BeagleBoard.org community, we have to work together to
enable use of the DSPs if they are of interest to the
community---there aren't any barriers in our way.

John Syn

unread,
Nov 10, 2014, 12:54:41 PM11/10/14
to beagl...@googlegroups.com

On 11/10/14, 1:15 AM, "Jason Kridner" <jkri...@beagleboard.org> wrote:

>On Sun, Nov 9, 2014 at 1:13 AM, Maxim Podbereznyy <lisa...@gmail.com>
>wrote:
>> John Syn,
>> Wandboard Quad does have 64 bit memory bus.
>>
>> P.S. I don't know who needs dual DSP onboard because TI definitely will
>>not
>> support them as should like it was for omap3/dm37. I'd better have well
>> supported hardware video encoder/decoder rather than double general
>>purpose
>> DSPs without any software support
>
>When OMAP3 came out, the only compilers for C6000 were expensive,
>closed-source compilers. Now, there is support in mainline GCC for
>C6000. As the BeagleBoard.org community, we have to work together to
>enable use of the DSPs if they are of interest to the
>community---there aren't any barriers in our way.
The TI C6000 does some amazing pipeline optimization, which seems to be
missing from the GCC compiler. Anyway, using CCSV6 is no big deal, but
support for RPMSG/REMOTEPROC on this processor is a big issue. The source
is difficult to follow and my guess is we would need input from the
original authors to do this work.

Regards,
John

Nishanth Menon

unread,
Nov 10, 2014, 1:58:35 PM11/10/14
to BeagleBoard, Anna, Suman, Ohad Ben-Cohen
On Mon, Nov 10, 2014 at 11:54 AM, John Syn <john...@gmail.com> wrote:
>
> On 11/10/14, 1:15 AM, "Jason Kridner" <jkri...@beagleboard.org> wrote:
>
>>On Sun, Nov 9, 2014 at 1:13 AM, Maxim Podbereznyy <lisa...@gmail.com>
>>wrote:
>>> John Syn,
>>> Wandboard Quad does have 64 bit memory bus.
>>>
>>> P.S. I don't know who needs dual DSP onboard because TI definitely will
>>>not
>>> support them as should like it was for omap3/dm37. I'd better have well
>>> supported hardware video encoder/decoder rather than double general
>>>purpose
>>> DSPs without any software support
>>
>>When OMAP3 came out, the only compilers for C6000 were expensive,
>>closed-source compilers. Now, there is support in mainline GCC for
>>C6000. As the BeagleBoard.org community, we have to work together to
>>enable use of the DSPs if they are of interest to the
>>community---there aren't any barriers in our way.
> The TI C6000 does some amazing pipeline optimization, which seems to be
> missing from the GCC compiler. Anyway, using CCSV6 is no big deal, but
> support for RPMSG/REMOTEPROC on this processor is a big issue. The source
> is difficult to follow and my guess is we would need input from the
> original authors to do this work.

Why dont we get involved in linux-omap discussions on the topic? most
of the rpmsg and remote proc discussions do take place in kernel
mailing list. usually discussing in context helps move patches forward
since it makes it clear to certain maintainers that these things are
important and help community.


Do you have anything specific that you are concerned about?

--
---
Regards,
Nishanth Menon

Robert Nelson

unread,
Nov 10, 2014, 2:13:16 PM11/10/14
to Beagle Board, Anna, Suman, Ohad Ben-Cohen
I think the biggest concern of bb.org, is that we don't end up in a
similar situation to the omap3. Where dsplink/etc never went mainline,
thus 5 years later, we still need to run the 2.6.32 psp to even
utilize the omap3's dsp.

Just the classic.. burn me once... ;)

For the am57x, since the foundation rpmsg/remoteproc is mainline, we
should be in good shape... (fingers crossed)

Regards,

--
Robert Nelson
http://www.rcn-ee.com/

Thaddeus Woskowiak

unread,
Nov 10, 2014, 2:45:31 PM11/10/14
to beagl...@googlegroups.com
From the datasheet, it looks like the SoC does not have dual ethernet MAC's. Instead it has an internal 3 port switch, two run off-chip to the ethernet ports and one to the SoC's MAC. This is a bit of a disappointment as I was hoping for true dual GbE MAC's.

On Friday, November 7, 2014 4:05:21 AM UTC-5, lisarden wrote:
Hey guys!

BeagleBoard-X15 - Are you really going to release such monster? :)

John Syn

unread,
Nov 10, 2014, 3:25:14 PM11/10/14
to beagl...@googlegroups.com, Anna, Suman, Ohad Ben-Cohen, linux-om...@vger.kernel.org
Looking at git.ti.com/rpmsg/rpmsg, I don¹t see any support for AM572x
processors. Also, Beagleboard-X15 is to be released with Kernel V3.18 but
I don¹t see support for this kernel versions. Last I heard, RPMSG was
working on OMAP4, but not fully implemented on OMAP5, but this was a while
ago and perhaps this has changed. Perhaps Suman can give us an update.
Similar concerns about REMOTEPROC. What I know is that I have been pushing
this issue on the beta list and the only feedback I received was that
"RPMSG/REMOTEPROC was in a SW blackhole².

Based on the TRM, this processor looks extremely attractive, and we need
RPMSG/REMOTEPROC to take advantage of the powerful dual DSPs and dual
CortexM4s.

Regards,
John
>
>--
>---
>Regards,
>Nishanth Menon
>

John Syn

unread,
Nov 10, 2014, 4:18:22 PM11/10/14
to Suman Anna, beagl...@googlegroups.com, Ohad Ben-Cohen, linux-om...@vger.kernel.org

On 11/10/14, 1:06 PM, "Suman Anna" <s-a...@ti.com> wrote:

>Hi John,
>One should be able to directly download the C6000 compilers at
>http://software-dl.ti.com/codegen/non-esd/downloads/download.htm#C6000
>
>>>> support for RPMSG/REMOTEPROC on this processor is a big issue. The
>>>> source
>>>> is difficult to follow and my guess is we would need input from the
>>>> original authors to do this work.
>>>
>>> Why dont we get involved in linux-omap discussions on the topic? most
>>> of the rpmsg and remote proc discussions do take place in kernel
>>> mailing list. usually discussing in context helps move patches forward
>>> since it makes it clear to certain maintainers that these things are
>>> important and help community.
>>>
>>>
>>> Do you have anything specific that you are concerned about?
>
>> Looking at git.ti.com/rpmsg/rpmsg, I don¹t see any support for AM572x
>> processors. Also, Beagleboard-X15 is to be released with Kernel V3.18
>>but
>> I don¹t see support for this kernel versions. Last I heard, RPMSG was
>> working on OMAP4, but not fully implemented on OMAP5, but this was a
>>while
>> ago and perhaps this has changed. Perhaps Suman can give us an update.
>> Similar concerns about REMOTEPROC. What I know is that I have been
>>pushing
>> this issue on the beta list and the only feedback I received was that
>> "RPMSG/REMOTEPROC was in a SW blackhole².
>
>The rpmsg-ti-linux-3.14.y branch in the above tree is the feature
>integration branch for rpmsg/remoteproc and does support all the
>processors on OMAP4, OMAP5 and AM572x/DRA7x. The AM572x support should
>be present through the am57xx-beagle-x15.dts file (its been sometime
>since I pulled the required platform branch with any updates to this).
>
>I am in the process of pushing all these features/patches upstream, but
>it will mostly be sometime next year before all the patches and their
>dependencies will make it into the upstream kernel, so until then have
>to rely on a TI tree.
Hi Suman,

That is really good news. I’m guessing that V3.18 support will occur with
the push to mainline? Any chance that we will see support when
Beagleboard-X15 launches in Feb 2015?

Regards,
John
>
>regards
>Suman

Graham

unread,
Nov 10, 2014, 4:54:49 PM11/10/14
to beagl...@googlegroups.com
Gerald:

The picture on  
http://www.elinux.org/Beagleboard:BeagleBoard-X15
looks great. 

But, I note that it does not show a heatsink or fan on the Sitara.

Is this just some "photographic license" or is this Sitara going to need
some proactive cooling?

Thanks,
--- Graham

==

John Syn

unread,
Nov 10, 2014, 4:58:10 PM11/10/14
to beagl...@googlegroups.com

From: Graham <gra...@flex-radio.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Monday, November 10, 2014 at 1:54 PM

To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: [beagleboard] Re: BeagleBoard-X15 - seriously? :)

Gerald:

The picture on  
http://www.elinux.org/Beagleboard:BeagleBoard-X15
looks great. 

But, I note that it does not show a heatsink or fan on the Sitara.

Is this just some "photographic license" or is this Sitara going to need
some proactive cooling?
Gerald added two layers to the PCB to act as a heatsink, so no additional heatsink or fan will be necessary. At least that is the plan.

Regards,
John


Thanks,
--- Graham

==


On Friday, November 7, 2014 3:05:21 AM UTC-6, lisarden wrote:
Hey guys!

BeagleBoard-X15 - Are you really going to release such monster? :)

--

Gerald Coley

unread,
Nov 10, 2014, 5:00:19 PM11/10/14
to beagl...@googlegroups.com
TBD. We have a heatsink and a fan option

Gerald

Graham Haddock

unread,
Nov 10, 2014, 5:19:34 PM11/10/14
to beagl...@googlegroups.com
Gerald:
Thanks. Let us know when you know.

As a comment, I could not find any on-die temperature sensor.
In dealing with FPGA's and other high power chips, having a way to
directly read die temperature has helped us solve all kinds of problems.
Please pass on to TI marketing, if you get a chance.

--- Graham

==

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/1t3yxkPYSB8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

Gerald Coley

unread,
Nov 10, 2014, 5:21:17 PM11/10/14
to beagl...@googlegroups.com
There are on die and onboard temperature sensors. Current boards have a heatsink only.

Gerald

John Syn

unread,
Nov 10, 2014, 5:26:09 PM11/10/14
to beagl...@googlegroups.com


From: Gerald Coley <ger...@beagleboard.org>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Monday, November 10, 2014 at 2:21 PM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: Re: [beagleboard] BeagleBoard-X15 - seriously? :)

There are on die and onboard temperature sensors. Current boards have a heatsink only.
Did the heatsink layers work as expected? 

Regards,
John

Gerald Coley

unread,
Nov 10, 2014, 5:29:11 PM11/10/14
to beagl...@googlegroups.com
Yes. They helped a lot. But crank everything up and it may still need some help.

Nishanth Menon

unread,
Nov 10, 2014, 5:29:43 PM11/10/14
to BeagleBoard, Suman Anna, Ohad Ben-Cohen, linux-omap
Fixing up linux-omap mailing list

Thaddeus Woskowiak

unread,
Nov 10, 2014, 5:30:09 PM11/10/14
to beagl...@googlegroups.com
Graham,
I had a quick look through the manual: http://www.ti.com/lit/ug/spruhz6/spruhz6.pdf
On page 4007 it states: "There are five temperature sensors on the device die."

Robert Nelson

unread,
Nov 10, 2014, 5:31:51 PM11/10/14
to Beagle Board
On Mon, Nov 10, 2014 at 4:29 PM, Thaddeus Woskowiak
<tswos...@gmail.com> wrote:
> Graham,
> I had a quick look through the manual:
> http://www.ti.com/lit/ug/spruhz6/spruhz6.pdf
> On page 4007 it states: "There are five temperature sensors on the device
> die."

For each core...

MPU/GPU/CORE/IVA/DSPEVE

Graham Haddock

unread,
Nov 10, 2014, 6:34:27 PM11/10/14
to beagl...@googlegroups.com
All:
Excellent. I was not searching for the right terms.

That will prove to be most helpful.

Thanks,
--- Graham     

liyaoshi

unread,
Nov 10, 2014, 9:39:29 PM11/10/14
to beagl...@googlegroups.com
Another Version of OMAP5 uEVM ?
can  I want a quad core version ?

You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.

Gerald Coley

unread,
Nov 10, 2014, 9:42:17 PM11/10/14
to beagl...@googlegroups.com
Different processor. Dual core only.

Gerald

liyaoshi

unread,
Nov 10, 2014, 9:51:23 PM11/10/14
to beagl...@googlegroups.com
Hello Gerald
Seems am5728 compatible with dra74x ,Ti also called Jcinto 6 
Will you release the scheme ?
This is very interesting , when and where can I buy it ?

Gerald Coley

unread,
Nov 10, 2014, 9:57:06 PM11/10/14
to beagl...@googlegroups.com
Will go on sale the end of February. We are in beta. No schematics will be released until we have the final version,

liyaoshi

unread,
Nov 10, 2014, 10:00:10 PM11/10/14
to beagl...@googlegroups.com
Thanks 
And can you consider of add another I2S connector in your next version ?
In CAR infotainment environment , we always want more audio suource .

Gerald Coley

unread,
Nov 10, 2014, 10:02:11 PM11/10/14
to beagl...@googlegroups.com
There are two. That is the best we can do,

liyaoshi

unread,
Nov 10, 2014, 10:12:26 PM11/10/14
to beagl...@googlegroups.com
Thanks 
Another question , Is this the dual channel version of EMIF ?
And with LPDDR3 1333 @533Mhz right ? 

Gerald Coley

unread,
Nov 10, 2014, 10:14:15 PM11/10/14
to beagl...@googlegroups.com
Dual 32b busses using DDR3L @ 533

liyaoshi

unread,
Nov 10, 2014, 10:23:07 PM11/10/14
to beagl...@googlegroups.com
Thanks 
And waiting for your another great success 
330.gif

Gerald Coley

unread,
Nov 10, 2014, 10:24:43 PM11/10/14
to beagl...@googlegroups.com
I hope so!
330.gif

liyaoshi

unread,
Nov 10, 2014, 10:47:13 PM11/10/14
to beagl...@googlegroups.com
GC320 can do 8 surface hardware blit .

2014-11-09 5:21 GMT+08:00 Robert Nelson <robert...@gmail.com>:


On Nov 8, 2014 3:05 PM, "William Hermans" <yyr...@gmail.com> wrote:
>
> So, *why* does TI keep using hardware that has closed source drivers ? Does TI realize this is killing their stuff in the eyes of many ppl out there ?

Well it does have a Vivante GC320 2D GPU so maybe they are starting the transition....

>
> On Sat, Nov 8, 2014 at 1:05 PM, rh_ <richard...@lavabit.com> wrote:
>>
>> On Fri, 7 Nov 2014 12:32:19 -0600
>> Gerald Coley <ger...@beagleboard.org> wrote:
>>
>> > Definitely not one of those. Look at what it has on it and make a good
>> > guess.
>>
>> Ok "definitely" must mean $200-400. Is this not good guessing?
>> This starts to be a in different realm and the competition is more
>> fierce. But you know more about that than I ever will and it's
>> likely that I just don't understand the landscape. Dual Gb ethernet
>> will attract a lot of attention for server/firewall/IDS/IPS.


>>
>> --
>> For more options, visit http://beagleboard.org/discuss
>> ---
>> You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

Message has been deleted

Gerald Coley

unread,
Nov 11, 2014, 11:22:12 AM11/11/14
to beagl...@googlegroups.com
I have no set a price range so how do you know?

Gerald

On Tuesday, November 11, 2014, __rh___ <richard...@lavabit.com> wrote:
On Sun, 9 Nov 2014 11:00:21 +0100
"Dr. H. Nikolaus Schaller" <h...@goldelico.com> wrote:

>
> Am 08.11.2014 um 21:05 schrieb rh_
> <richard...@lavabit.com>:

>
> > On Fri, 7 Nov 2014 12:32:19 -0600
> > Gerald Coley <ger...@beagleboard.org> wrote:
> >
> >> Definitely not one of those. Look at what it has on it and make a
> >> good guess.
> >
> > Ok "definitely" must mean $200-400. Is this not good guessing?
> > This starts to be a in different realm and the competition is more
> > fierce. But you know more about that than I ever will and it's
> > likely that I just don't understand the landscape. Dual Gb ethernet
> > will attract a lot of attention for server/firewall/IDS/IPS.
>
> I think the feature set of the X15 is almost comparable to the
> OMAP5432EVM (which AFAIR was originally planned to become
> a PandaBoard 5):
>
> http://www.svtronics.com/5432
>
> So this seems to confirm the price range $200-400...

X15 not interesting at this price range. Plus some of the features
are really vaporware as it's likely they remain in their software
blackhole, as mentioned earlier.


--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--

Robert Nelson

unread,
Nov 11, 2014, 11:38:46 AM11/11/14
to Beagle Board
On Tue, Nov 11, 2014 at 10:19 AM, __rh___ <richard...@lavabit.com> wrote:
> On Sun, 9 Nov 2014 11:00:21 +0100
> "Dr. H. Nikolaus Schaller" <h...@goldelico.com> wrote:
>
>>
>> Am 08.11.2014 um 21:05 schrieb rh_
>> <richard...@lavabit.com>:
>>
>> > On Fri, 7 Nov 2014 12:32:19 -0600
>> > Gerald Coley <ger...@beagleboard.org> wrote:
>> >
>> >> Definitely not one of those. Look at what it has on it and make a
>> >> good guess.
>> >
>> > Ok "definitely" must mean $200-400. Is this not good guessing?
>> > This starts to be a in different realm and the competition is more
>> > fierce. But you know more about that than I ever will and it's
>> > likely that I just don't understand the landscape. Dual Gb ethernet
>> > will attract a lot of attention for server/firewall/IDS/IPS.
>>
>> I think the feature set of the X15 is almost comparable to the
>> OMAP5432EVM (which AFAIR was originally planned to become
>> a PandaBoard 5):
>>
>> http://www.svtronics.com/5432
>>
>> So this seems to confirm the price range $200-400...
>
> X15 not interesting at this price range. Plus some of the features
> are really vaporware as it's likely they remain in their software
> blackhole, as mentioned earlier.

At least vent your frustration to the correct people:

http://www.imgtec.com/

They are the ones who don't want to support their ip found in TI products..

Gerald Coley

unread,
Nov 11, 2014, 11:50:53 AM11/11/14
to beagl...@googlegroups.com
This true!

Gerald
--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

John Syn

unread,
Nov 11, 2014, 1:58:36 PM11/11/14
to beagl...@googlegroups.com

On 11/11/14, 8:19 AM, "__rh___" <richard...@lavabit.com> wrote:

>On Sun, 9 Nov 2014 11:00:21 +0100
>"Dr. H. Nikolaus Schaller" <h...@goldelico.com> wrote:
>
>>
>> Am 08.11.2014 um 21:05 schrieb rh_
>> <richard...@lavabit.com>:
>>
>> > On Fri, 7 Nov 2014 12:32:19 -0600
>> > Gerald Coley <ger...@beagleboard.org> wrote:
>> >
>> >> Definitely not one of those. Look at what it has on it and make a
>> >> good guess.
>> >
>> > Ok "definitely" must mean $200-400. Is this not good guessing?
>> > This starts to be a in different realm and the competition is more
>> > fierce. But you know more about that than I ever will and it's
>> > likely that I just don't understand the landscape. Dual Gb ethernet
>> > will attract a lot of attention for server/firewall/IDS/IPS.
>>
>> I think the feature set of the X15 is almost comparable to the
>> OMAP5432EVM (which AFAIR was originally planned to become
>> a PandaBoard 5):
>>
>> http://www.svtronics.com/5432
>>
>> So this seems to confirm the price range $200-400...
>
>X15 not interesting at this price range. Plus some of the features
>are really vaporware as it's likely they remain in their software
>blackhole, as mentioned earlier.
Maybe not as bad as you think. I had a discussion on the linux-omap
mailing list and TI have RPMSG/REMOTEPROC working on the AM572x processor
and will push this to mainline early 2015.

Regards,
John

John Syn

unread,
Nov 11, 2014, 6:39:58 PM11/11/14
to beagl...@googlegroups.com
I guess if enough users way in and convince IMGTEC to support Linux
mainline, they may rethink their reluctance to support the open source
community. Post your comments on:

http://www.imgtec.com/community/


If you don¹t stand up and be counted, then you have no right to complain.

Regards,
John
>
>Regards,
>
>--
>Robert Nelson
>http://www.rcn-ee.com/
>

Alfredo Muniz

unread,
Nov 11, 2014, 7:32:35 PM11/11/14
to beagl...@googlegroups.com

On Tue, Nov 11, 2014 at 6:39 PM, John Syn <john...@gmail.com> wrote:
If you don¹t stand up and be counted, then you have no right to complain.

Nishanth Menon

unread,
Nov 11, 2014, 11:11:27 PM11/11/14
to BeagleBoard
On 11/07/2014 03:43 PM, Nishanth Menon wrote:
> On Fri, Nov 7, 2014 at 3:40 PM, Laurent Desnogues
> <laurent....@gmail.com> wrote:
>> On Fri, Nov 7, 2014 at 7:32 PM, Gerald Coley <ger...@beagleboard.org> wrote:
>>> Definitely not one of those. Look at what it has on it and make a good
>>> guess.
>>
>> You got me wondering because Richard basically copied his info
>> from Nishanth Menon himself:
>>
>> https://patchwork.kernel.org/patch/5245961/
>>
> Sigh.. at least use the typo corrected patch V2:
> https://patchwork.kernel.org/patch/5254791/ :P


just in case folks are interested:
https://git.kernel.org/cgit/linux/kernel/git/tmlind/linux-omap.git/commit/?h=omap-for-v3.19/dt&id=80c4955b7ad9daaf34a46f47f7cb556ef2728af7

queued for 3.19-rc1


--
Regards,
Nishanth Menon

Robert Nelson

unread,
Nov 11, 2014, 11:23:53 PM11/11/14
to Beagle Board
Nice work Nishanth!

This has to be one of the fastest mainline enabled arm kits ever!

William Hermans

unread,
Nov 12, 2014, 12:00:29 AM11/12/14
to beagl...@googlegroups.com
Kind of tells me a "few" ppl want it ;)

Emmanuel Fusté

unread,
Nov 12, 2014, 10:03:34 AM11/12/14
to beagl...@googlegroups.com

Look at the recently submited mail box subsystem.

Le 10 nov. 2014 21:25, "John Syn" <john...@gmail.com> a écrit :

On 11/10/14, 10:58 AM, "Nishanth Menon" <n...@ti.com> wrote:

>On Mon, Nov 10, 2014 at 11:54 AM, John Syn <john...@gmail.com> wrote:
>>
>> On 11/10/14, 1:15 AM, "Jason Kridner" <jkri...@beagleboard.org> wrote:
>>
>>>On Sun, Nov 9, 2014 at 1:13 AM, Maxim Podbereznyy <lisa...@gmail.com>
>>>wrote:
>>>> John Syn,
>>>> Wandboard Quad does have 64 bit memory bus.
>>>>
>>>> P.S. I don't know who needs dual DSP onboard because TI definitely
>>>>will
>>>>not
>>>> support them as should like it was for omap3/dm37. I'd better have
>>>>well
>>>> supported hardware video encoder/decoder rather than double general
>>>>purpose
>>>> DSPs without any software support
>>>
>>>When OMAP3 came out, the only compilers for C6000 were expensive,
>>>closed-source compilers. Now, there is support in mainline GCC for
>>>C6000. As the BeagleBoard.org community, we have to work together to
>>>enable use of the DSPs if they are of interest to the
>>>community---there aren't any barriers in our way.
>> The TI C6000 does some amazing pipeline optimization, which seems to be
>> missing from the GCC compiler. Anyway, using CCSV6 is no big deal, but
>> support for RPMSG/REMOTEPROC on this processor is a big issue. The
>>source
>> is difficult to follow and my guess is we would need input from the
>> original authors to do this work.
>
>Why dont we get involved in linux-omap discussions on the topic? most
>of the rpmsg and remote proc discussions do take place in kernel
>mailing list. usually discussing in context helps move patches forward
>since it makes it clear to certain maintainers that these things are
>important and help community.
>
>
>Do you have anything specific that you are concerned about?
Looking at git.ti.com/rpmsg/rpmsg, I don¹t see any support for AM572x
processors. Also, Beagleboard-X15 is to be released with Kernel V3.18 but
I don¹t see support for this kernel versions. Last I heard, RPMSG was
working on OMAP4, but not fully implemented on OMAP5, but this was a while
ago and perhaps this has changed. Perhaps Suman can give us an update.
Similar concerns about REMOTEPROC. What I know is that I have been pushing
this issue on the beta list and the only feedback I received was that
"RPMSG/REMOTEPROC was in a SW blackhole².

Based on the TRM, this processor looks extremely attractive, and we need
RPMSG/REMOTEPROC to take advantage of the powerful dual DSPs and dual
CortexM4s.

Regards,
John
>
>--
>---
>Regards,
>Nishanth Menon

>
>--
>For more options, visit http://beagleboard.org/discuss
>---
>You received this message because you are subscribed to the Google Groups
>"BeagleBoard" group.
>To unsubscribe from this group and stop receiving emails from it, send an
>email to beagleboard...@googlegroups.com.
>For more options, visit https://groups.google.com/d/optout.


--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/1t3yxkPYSB8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

Suman Anna

unread,
Nov 12, 2014, 10:03:37 AM11/12/14
to John Syn, beagl...@googlegroups.com, Ohad Ben-Cohen, linux-om...@vger.kernel.org
Hi John,

On 11/10/2014 02:24 PM, John Syn wrote:
>
> On 11/10/14, 10:58 AM, "Nishanth Menon" <n...@ti.com> wrote:
>
>> On Mon, Nov 10, 2014 at 11:54 AM, John Syn <john...@gmail.com> wrote:
>>>
>>> On 11/10/14, 1:15 AM, "Jason Kridner" <jkri...@beagleboard.org> wrote:
>>>
>>>> On Sun, Nov 9, 2014 at 1:13 AM, Maxim Podbereznyy <lisa...@gmail.com>
>>>> wrote:
>>>>> John Syn,
>>>>> Wandboard Quad does have 64 bit memory bus.
>>>>>
>>>>> P.S. I don't know who needs dual DSP onboard because TI definitely
>>>>> will
>>>>> not
>>>>> support them as should like it was for omap3/dm37. I'd better have
>>>>> well
>>>>> supported hardware video encoder/decoder rather than double general
>>>>> purpose
>>>>> DSPs without any software support
>>>>
>>>> When OMAP3 came out, the only compilers for C6000 were expensive,
>>>> closed-source compilers. Now, there is support in mainline GCC for
>>>> C6000. As the BeagleBoard.org community, we have to work together to
>>>> enable use of the DSPs if they are of interest to the
>>>> community---there aren't any barriers in our way.
>>> The TI C6000 does some amazing pipeline optimization, which seems to be
>>> missing from the GCC compiler. Anyway, using CCSV6 is no big deal, but

One should be able to directly download the C6000 compilers at
http://software-dl.ti.com/codegen/non-esd/downloads/download.htm#C6000

>>> support for RPMSG/REMOTEPROC on this processor is a big issue. The
>>> source
>>> is difficult to follow and my guess is we would need input from the
>>> original authors to do this work.
>>
>> Why dont we get involved in linux-omap discussions on the topic? most
>> of the rpmsg and remote proc discussions do take place in kernel
>> mailing list. usually discussing in context helps move patches forward
>> since it makes it clear to certain maintainers that these things are
>> important and help community.
>>
>>
>> Do you have anything specific that you are concerned about?

> Looking at git.ti.com/rpmsg/rpmsg, I don¹t see any support for AM572x
> processors. Also, Beagleboard-X15 is to be released with Kernel V3.18 but
> I don¹t see support for this kernel versions. Last I heard, RPMSG was
> working on OMAP4, but not fully implemented on OMAP5, but this was a while
> ago and perhaps this has changed. Perhaps Suman can give us an update.
> Similar concerns about REMOTEPROC. What I know is that I have been pushing
> this issue on the beta list and the only feedback I received was that
> "RPMSG/REMOTEPROC was in a SW blackhole².

The rpmsg-ti-linux-3.14.y branch in the above tree is the feature
integration branch for rpmsg/remoteproc and does support all the
processors on OMAP4, OMAP5 and AM572x/DRA7x. The AM572x support should
be present through the am57xx-beagle-x15.dts file (its been sometime
since I pulled the required platform branch with any updates to this).

I am in the process of pushing all these features/patches upstream, but
it will mostly be sometime next year before all the patches and their
dependencies will make it into the upstream kernel, so until then have
to rely on a TI tree.

regards
Suman

Suman Anna

unread,
Nov 12, 2014, 10:03:44 AM11/12/14
to John Syn, beagl...@googlegroups.com, Ohad Ben-Cohen, linux-om...@vger.kernel.org
On 11/10/2014 03:18 PM, John Syn wrote:
> Hi Suman,
>
> That is really good news. I’m guessing that V3.18 support will occur with
> the push to mainline? Any chance that we will see support when
> Beagleboard-X15 launches in Feb 2015?

Feb 2015 is too early to have all the remoteproc/rpmsg pieces make it to
upstream, so for now the 3.14-LTS based TI kernel is the latest where
everything is in place.

regards
Suman

>
> Regards,
> John
Message has been deleted

Robert Nelson

unread,
Nov 12, 2014, 10:33:27 AM11/12/14
to Beagle Board
> Ha! I'm not definitely not frustrated. And as seen in other recent
> posts there are other features that may not get needed TI support.

Seriously, spit it out, what feature?

dsp: c6000 gcc port

m4: ducati: firmware here
http://git.omapzoom.org/?p=device/ti/proprietary-open.git;a=tree;f=jacinto6;h=6203e0dfe7cf89456517a1b877610f8ac2b4fd86;hb=refs/heads/d-lollipop-release

pru:
ti c-complier:

wget http://software-dl.ti.com/codegen/esd/cgt_public_sw/PRU/2.1.0/ti_cgt_pru_2.1.0_armlinuxa8hf_busybox_installer.sh
chmod +x ti_cgt_pru_2.1.0_armlinuxa8hf_busybox_installer.sh
sudo ./ti_cgt_pru_2.1.0_armlinuxa8hf_busybox_installer.sh

+ gcc port:



>> http://www.imgtec.com/
>
> And I don't care about the graphics. But since you're projecting
> I will say that the issue is with TI not imgtec. No one forced
> TI to use imgtec. Just like no one forced intel to use it when
> they decided on GMA500/Poulsbo. But instead of crying about it
> an intel engineer decided to do a little heavy lifting and wrote
> a driver for GMA500.

Umm, slightly different. Poulsbo also included the 2d engine, so with
out GMA500 driver, no "DISPLAY PERIOD..", in our case we can get video
working on mainline, it's just the 3d offload we need the sgx bits
for..

John Syn

unread,
Nov 12, 2014, 1:53:21 PM11/12/14
to beagl...@googlegroups.com

From: Emmanuel Fusté <manu....@gmail.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Monday, November 10, 2014 at 12:29 PM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: Re: [beagleboard] Re: BeagleBoard-X15 - seriously? :)

Look at the recently submited mail box subsystem.

Please share a link

menon.n...@gmail.com

unread,
Nov 12, 2014, 1:54:24 PM11/12/14
to BeagleBoard
On Wed, Nov 12, 2014 at 12:52 PM, John Syn <john...@gmail.com> wrote:
>
> From: Emmanuel Fusté <manu....@gmail.com>
> Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
> Date: Monday, November 10, 2014 at 12:29 PM
> To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
> Subject: Re: [beagleboard] Re: BeagleBoard-X15 - seriously? :)
>
> Look at the recently submited mail box subsystem.
>
> Please share a link


http://marc.info/?l=linux-omap&w=2&r=1&s=mailbox&q=b

Emmanuel Fusté

unread,
Nov 12, 2014, 5:56:31 PM11/12/14
to beagl...@googlegroups.com
Yes thank you, I did not saw my mail reaching the mailing list.
The recently merged generic mailbox subsystem and the omap mailbox
driver are the foundations for remoteproc.
All in on the way. This is a big step forward for all past and future ti
soc.
http://marc.info/?l=linux-omap&m=141505601225611&w=2

Emmanuel.

Nishanth Menon

unread,
Nov 12, 2014, 9:11:31 PM11/12/14
to beagl...@googlegroups.com
Emanuel,
Posted on mailing is just step #1.
here are other steps needed:
- Device Tree maintainers need to ack - that takes a bunch of time and
sometimes depends on the polarity of solar flares and grace of a bunch
of overworked devicetree maintainers dragged in all directions.
- Mailbox maintainer need to ack as well - which means the mailbox and
rproc framework cannot conflict against other SoCs..
- All of these must finally be accepted by Linus.

Just coz it is posted around 3.18-rc4 requesting dt reviewers indicates
it probably will miss 3.19 window, 3.20-rc1 is the next opportunity (umm
5 months away)? yeah - it will be nice to get additional review help on
the linux-omap mailing list so that maintainer jobs are a little
easier.. all contributions welcome. we will continue to push from our
side ofcourse.. but kernel.org support takes time...


---
Regards,
Nishanth Menon

danesh....@gmail.com

unread,
Nov 13, 2014, 3:25:32 PM11/13/14
to beagl...@googlegroups.com
Was checking the AM572x TRM and noticed that the PCIe Controller section (24.9) is blank. Any particular reason why the information on this block cannot be shared in the public domain?

Thanks

On Friday, November 7, 2014 1:05:21 AM UTC-8, lisarden wrote:
Hey guys!

BeagleBoard-X15 - Are you really going to release such monster? :)

Gerald Coley

unread,
Nov 13, 2014, 4:15:53 PM11/13/14
to beagl...@googlegroups.com
TI is waiting for approval from the IP owner as to what can be put into the document. This was a way to get it out sooner and not have to wait on the IP owner.
It will be included in the final release.


Gerald


--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--

danesh....@gmail.com

unread,
Nov 13, 2014, 5:03:23 PM11/13/14
to beagl...@googlegroups.com
Thanks for clarifying!
Message has been deleted

John Syn

unread,
Nov 16, 2014, 11:01:45 PM11/16/14
to beagl...@googlegroups.com

On 11/16/14, 5:56 PM, "+_+_rh_o_o" <richard...@lavabit.com> wrote:

>On Wed, 12 Nov 2014 09:33:20 -0600
>Robert Nelson <robert...@gmail.com> wrote:
>
>> > Ha! I'm not definitely not frustrated. And as seen in other recent
>> > posts there are other features that may not get needed TI support.
>>
>> Seriously, spit it out, what feature?
>
>Re-read this thread.
>
>
>> > And I don't care about the graphics. But since you're projecting
>> > I will say that the issue is with TI not imgtec. No one forced
>> > TI to use imgtec. Just like no one forced intel to use it when
>> > they decided on GMA500/Poulsbo. But instead of crying about it
>> > an intel engineer decided to do a little heavy lifting and wrote
>> > a driver for GMA500.
>>
>> Umm, slightly different. Poulsbo also included the 2d engine, so with
>> out GMA500 driver, no "DISPLAY PERIOD..", in our case we can get video
>> working on mainline, it's just the 3d offload we need the sgx bits
>> for..
>>
>
>The blame goes to TI. They chose imgtec technology. The best and
>most powerful message to imgtec would have been for TI to say
>"no thanks". But now any BBB owner pays imgtec wittingly or
>unwittingly, but they still pay imgtec indirectly.
>
>I guess it's another sign of the technological dark age where we all
>now preside.
If you feel so strongly about this issue, then direct your efforts to make
a difference. Express your views on IMGTEC developer forums or contact
your TI rep or express your views on TI¹s E2E. We all agree that the lack
of Linux support for IMGTEC IP is a big issue but no one here can make a
difference.

http://forum.imgtec.com/discussion/3394/linux-and-open-source


Regards,
John

Ken

unread,
Nov 22, 2014, 5:40:29 AM11/22/14
to beagl...@googlegroups.com
Nice :)
Hopefully I can do something with those USB towers for a slimmer case.

plug...@p10link.net

unread,
Nov 29, 2014, 9:20:44 AM11/29/14
to beagl...@googlegroups.com


On Saturday, 8 November 2014 16:31:25 UTC, sixvolts wrote:
Any ideas how this compares to the wandboard quad? 4x Cortex-A9 vs 2x Cortex-A15?

 A15 should be faster than A9 clock for clock. http://www.itproportal.com/2012/10/31/arm-cortex-a57-and-a53-vs-cortex-a8-a9-a15-and-a7-a-performance-analysis/ claims 2.5 DMIPs/clk for the A9 and 3.5 for the A15. Also the beagleboard-x15 is supposed to be clocked at 1.5GHZ while the wandboard quad only runs at 1GHz. So that would make the each core on the beaglboard x15 2.1 times as fast as on the wandboard quad.

Of course DMIPs are not the be all and end all of CPU performance but i'd expect the beagleboard X15 and wandboard quad to be broadly comparable in highly multithreaded workloads and the beagleboard x15 to be the clear winner in workloads with low thread count sections.

This is also good for buildboxes because it means you can use a lower -j value which will result in less memory pressure.

matthij...@gmail.com

unread,
Jan 10, 2015, 6:32:09 AM1/10/15
to beagl...@googlegroups.com
On Thursday, 13 November 2014 22:15:53 UTC+1, Gerald wrote:
TI is waiting for approval from the IP owner as to what can be put into the document. This was a way to get it out sooner and not have to wait on the IP owner.
It will be included in the final release.

On Thu, Nov 13, 2014 at 2:25 PM, <danesh....@gmail.com> wrote:
Was checking the AM572x TRM and noticed that the PCIe Controller section (24.9) is blank. Any particular reason why the information on this block cannot be shared in the public domain?

It's worth mentioning that various other reasonably recent TI SoCs have a PCIe subsystem with public documentation, e.g. Netra (DM816x / AM389x), Centaurus (DM814x / AM387x) and the Keystone series. They all have slightly different versions of the same Synopsys DesignWare Core, so it seems likely the one in AM572x will be similar too.  (Of course I could be wrong there)

Gerald Coley

unread,
Jan 12, 2015, 1:32:36 PM1/12/15
to beagl...@googlegroups.com
You are correct. And all of those that were made public, evidently, should not have been.

Gerald


--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
It is loading more messages.
0 new messages