v4.4.x-ti changes (r25+)

[Robert Nelson]

Just a heads up, TI just landed two large changes to the v4.4.x tree today:

https://github.com/RobertCNelson/ti-linux-kernel/compare/584991ea2c87ead7ce1e0ef145a9e66a4408a9d5...ee87bb1741b3e2f05a5ff46b1ad59ca82b8f2514

remoteproc_pruss & sgx

remoteproc_pruss: this should be compatible with v4.1.x-ti

sgx: the "modules" should build and enter the archive when the kernel *.deb get's built..

Here is the user space:

http://git.ti.com/gitweb/?p=graphics/omap5-sgx-ddk-um-linux.git;a=shortlog;h=refs/heads/ti-img-sgx/1.14.3699939

Regards,

--

Robert Nelson
https://rcn-ee.com/

[tz.s...@gmail.com]

Hello Robert,

This is excellent news! I hope this fixes the installation issue I just reported for the BB xm, will give it a try when the deb become available.

Best regards,

Tziang

[Robert Nelson]

Hi Tziang,

On Thu, May 5, 2016 at 1:22 PM, <tz.s...@gmail.com> wrote:

Hello Robert,

This is excellent news! I hope this fixes the installation issue I just reported for the BB xm, will give it a try when the deb become available.

Un-fortunately, this will not help you with the xM.  For your issue, you should also retry with:

[--use-testing-kernel] which will bring in: [4.6.0-rc6-armv7-x2]

Back to v4.4.x-ti, so TI dropped "dm37xx" support back in v3.14.x-ti time frame, thus we also don't enable any omap3 devices in v4.1.x-ti or v4.4.x-ti.

Just am335x/am43xx/am57xx devices..

Regards,

[tz.s...@gmail.com]

Hello Robert,

This is really useful, thanks. I'll consider 3.14 instead as you suggest, and 4.6 too.

Best regards,

Tziang

[Robert Nelson]

On Thu, May 5, 2016 at 1:18 PM, Robert Nelson <robert...@gmail.com> wrote:

Just a heads up, TI just landed two large changes to the v4.4.x tree today:

https://github.com/RobertCNelson/ti-linux-kernel/compare/584991ea2c87ead7ce1e0ef145a9e66a4408a9d5...ee87bb1741b3e2f05a5ff46b1ad59ca82b8f2514

remoteproc_pruss & sgx

remoteproc_pruss: this should be compatible with v4.1.x-ti

sgx: the "modules" should build and enter the archive when the kernel *.deb get's built..

The sgx modules for stretch were successfully built, the build farm is currently working on all of them, so they should get pushed to the deb repo later tonight..

Thus, the sgx flag in the ./update_kernel.sh script is now set, so it'll install the sgx modules by default for v4.4.x-ti kernel's going forward..

Regards,

[Charles Steinkuehler]

On 5/5/2016 4:27 PM, Robert Nelson wrote:
>
> The sgx modules for stretch were successfully built, the build farm is currently
> working on all of them, so they should get pushed to the deb repo later tonight..
>
> Thus, the sgx flag in the ./update_kernel.sh script is now set, so it'll install
> the sgx modules by default for v4.4.x-ti kernel's going forward..

Great news!

Any progress on getting an sgx accelerated desktop, or are we still
waiting on wayland?

--
Charles Steinkuehler
cha...@steinkuehler.net

[Robert Nelson]

From there release notes, it needs wayland 1.9..

jessie: 1.6.0

stretch: 1.10 

Not a lot depends on wayland, so i think it would be a safe backport.. ;)

Regards,

[William Hermans]

From there release notes, it needs wayland 1.9..

jessie: 1.6.0

stretch: 1.10 

Not a lot depends on wayland, so i think it would be a safe backport.. ;)

Regards,

Looking forward to reading that guide when and if ones made. And watching a demo video ;)

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard/CAOCHtYjm%3Dy1fjSz%3DdoJHpVOTOoTVrp9-E4Ls9cTWubV-KJ26aQ%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Baozhu Zuo]

I think there are some wrong with pwm driver on linux-v4.4 kernel.
run:
var b = require('bonescript');
b.pinMode('P9_14', b.OUTPUT);
b.analogWrite('P9_14', 0.7, 2000.0, printStatus);
function printStatus(x) {
    console.log('x.value = ' + x.value);
    console.log('x.err = ' + x.err);
}
output:
x.err = error updating PWM freq and value: undefined/pwm0, Error: ENOENT, no such file or directory 'undefined/pwm0/period'

root@beaglebone:~/workspace# ls /sys/class/pwm/
pwmchip0  pwmchip2  pwmchip4  pwmchip5    pwmchip6

the export and unexport file disappeared.


在 2016年5月6日星期五 UTC+8上午5:27:39，robert nelson写道：

[Mark Yoder]

Hmmmm.. the pwm path appears to have changed for the 4.4 kernel.  Bonescript uses

/sys/devices/platform/ocp/48304000.epwmss/48304200.ehrpwm/pwm/pwmchip6, but the new 4.4 kernel path is

/sys/devices/platform/ocp/48304000.epwmss/48304200.pwm/pwm/pwmchip6  (the ehr is missing).

A quick hack is to edit : /usr/local/lib/node_modules/bonescript/src/hw_mainline.js and remove the ehr from line 95.

--Mark

[Robert Nelson]

On Wed, May 25, 2016 at 1:52 PM, Mark Yoder <mark.a...@gmail.com> wrote:

Hmmmm.. the pwm path appears to have changed for the 4.4 kernel.  Bonescript uses

/sys/devices/platform/ocp/48304000.epwmss/48304200.ehrpwm/pwm/pwmchip6, but the new 4.4 kernel path is

/sys/devices/platform/ocp/48304000.epwmss/48304200.pwm/pwm/pwmchip6  (the ehr is missing).

A quick hack is to edit : /usr/local/lib/node_modules/bonescript/src/hw_mainline.js and remove the ehr from line 95.

This got changed in:

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/arch/arm/boot/dts/am33xx.dtsi?id=dce2a6524963f58e7fd0b8888b2d3791c150f42c

and then backported to ti's v4.4.x branch..

Since the 0.5.0 branch of bonescript became v4.4.x-ti + with:

https://github.com/jadonk/bonescript/commit/afc63224575b014b8661ada52d553ac78140e6b6

This should pwm change would be safe to push in..

Regards,

[Mark Yoder]

I'm rusting on this git stuff with remote repo's, but I think I've just put in a pull request with the fix.

--Mark

[Robert Nelson]

On Wed, May 25, 2016 at 2:24 PM, Mark Yoder <mark.a...@gmail.com> wrote:

I'm rusting on this git stuff with remote repo's, but I think I've just put in a pull request with the fix.

Close, you want to merge your 0.5.0 into jason's 0.5.0.. (looks like github defaulted to your "master" for some odd reason...)

Regards,

[Mark Yoder]

Well, I better pull Pro GIt off the shelf and look up how to do it right...

--Mark

[Robert Nelson]

On Wed, May 25, 2016 at 2:32 PM, Mark Yoder <mark.a...@gmail.com> wrote:

Well, I better pull Pro GIt off the shelf and look up how to do it right...

github need's there own book for the web interface, although it would change weekly..

staged it up for you: ;)

https://github.com/jadonk/bonescript/pull/130

Regards,

[Mark Yoder]

Thanks..., but what did you do differently?

--Mark

[Robert Nelson]

On Wed, May 25, 2016 at 2:40 PM, Mark Yoder <mark.a...@gmail.com> wrote:

Thanks..., but what did you do differently?

So after you click the pull request button, you need to manually change the branch..

github assumes everything get's merged into master..  (which for bonescript isn't true right now, as 0.5.0 is a seperate development branch)

Regards,

[Mark Yoder]

Looks like there is an error in the documentation.  analogWrite's callback only sets x.err, not x.value.  This worked for me:

(Note:  I'm using pinMode b.ANALOG_OUTPUT rather than b.OUTPUT.)

var b = require('bonescript');

b.pinMode('P9_14', b.ANALOG_OUTPUT);

b.analogWrite('P9_14', 0.1, 2000.0, printStatus);

function printStatus(x) {

    if(x.err) {

        console.log('x.err = ' + x.err);

    } else {

        console.log('No errors');

    }

}

--Mark

On Wednesday, May 25, 2016 at 11:59:02 PM UTC-4, Baozhu Zuo wrote:

I have tested bonescript 0.5.0-beta7. 

run:

var b = require('bonescript');

b.pinMode('P9_14', b.OUTPUT);

b.analogWrite('P9_14', 0.7, 2000.0, printStatus);

function printStatus(x) {

    console.log('x.value = ' + x.value);

    console.log('x.err = ' + x.err);

}

output:

Bonescript: initialized

x.value = undefined

x.err = error updating PWM freq and value: undefined, Error: ENOENT, no such file or directory 'undefined/duty_cycle'

PWM still now work.
在 2016年5月26日星期四 UTC+8上午3:42:30，robert nelson写道：

[Baozhu Zuo]

I have tested bonescript 0.5.0-beta7. 

run:

var b = require('bonescript');

b.pinMode('P9_14', b.OUTPUT);

b.analogWrite('P9_14', 0.7, 2000.0, printStatus);

function printStatus(x) {

    console.log('x.value = ' + x.value);

    console.log('x.err = ' + x.err);

}

output:

Bonescript: initialized

x.value = undefined

x.err = error updating PWM freq and value: undefined, Error: ENOENT, no such file or directory 'undefined/duty_cycle'

PWM still now work.
在 2016年5月26日星期四 UTC+8上午3:42:30，robert nelson写道：