Re: [Machinekit] help upgrading to latest 4.14 kernel machinekit image

[Robert Nelson]

On Sun, Dec 17, 2017 at 2:41 PM, 'Ric W' via Machinekit
 wrote:
> Hi,
> not wanting to hijack the thread or anything, but I've also been looking at
> updating the kernel to the latest version for machinekit
> I'm currently in the process of figuring out how to setup a new machinekit
> setup
> I'm using a BeagleBone Green (v1) With the latest debian stretch image from
> the beaglebone site
> using the guide from
> https://machinekoder.com/machinekit-debian-stretch-beaglebone-black/
>
> I've been using the commands you've mentioned to test out some of the newer
> kernel releases
> (also doing a git pull in that directory to update /opt/scripts/tools/)
>
> For version 4.4
> I think this works okay so far, but I'd prefer to use a later version.
>
> For version 4.9
> I think there's a glitch in the kernel when I tested it with realtime
> ```
> realtime start
> halcmd show pin
> realtime stop
> ```
>
> This resulted in dmesg
> ```
> [  160.328837] IPv6: ADDRCONF(NETDEV_UP): usb1: link is not ready
> [  163.498868] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
> [  168.193722] random: crng init done
> [  290.635798] [ cut here ]
> [  290.635849] WARNING: CPU: 0 PID: 816 at fs/sysfs/dir.c:31
> sysfs_warn_dup+0x78/0x88
> [  290.635857] sysfs: cannot create duplicate filename
> '/bus/platform/devices/4a30.pruss'

With v4.9.x-ti you only get ti's remote_proc, ^ this error occrus when

uboot_overlay_pru=/lib/firmware/AM335X-PRU-UIO-00A0.dtbo

is set..

(i haven't been able to isolate it like i did in v4.4.x-ti where we
swap between ti's and communities pruss driver..)


> [  290.635863] Modules linked in: pruss_soc_bus(+) evdev uio_pdrv_genirq uio
> usb_f_mass_storage usb_f_acm u_serial usb_f_ecm usb_f_rndis u_ether
> libcomposite iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat
> nf_conntrack iptable_mangle iptable_filter spidev pru_rproc pruss_intc pruss
> tieqep ip_tables x_tables
> [  290.635938] CPU: 0 PID: 816 Comm: systemd-udevd Not tainted
> 4.9.69-ti-rt-r84 #1
> [  290.635942] Hardware name: Generic AM33XX (Flattened Device Tree)
> [  290.635980] [] (unwind_backtrace) from []
> (show_stack+0x20/0x24)
> [  290.635993] [] (show_stack) from []
> (dump_stack+0x80/0x94)
> [  290.636006] [] (dump_stack) from []
> (__warn+0xf8/0x110)
> [  290.636016] [] (__warn) from []
> (warn_slowpath_fmt+0x58/0x74)
> [  290.636026] [] (warn_slowpath_fmt) from []
> (sysfs_warn_dup+0x78/0x88)
> [  290.636038] [] (sysfs_warn_dup) from []
> (sysfs_do_create_link_sd+0xc4/0xcc)
> [  290.636046] [] (sysfs_do_create_link_sd) from []
> (sysfs_create_link+0x38/0x44)
> [  290.636063] [] (sysfs_create_link) from []
> (bus_add_device+0x11c/0x1c0)
> [  290.636074] [] (bus_add_device) from []
> (device_add+0x2e0/0x5e0)
> [  290.636086] [] (device_add) from []
> (of_device_add+0x44/0x4c)
> [  290.636097] [] (of_device_add) from []
> (of_platform_device_create_pdata+0x94/0xcc)
> [  290.636108] [] (of_platform_device_create_pdata) from
> [] (of_platform_bus_create+0x18c/0x33c)
> [  290.636116] [] (of_platform_bus_create) from []
> (of_platform_populate+0x9c/0x13c)
> [  290.636146] [] (of_platform_populate) from []
> (pruss_soc_bus_probe+0x12c/0x290 [pruss_soc_bus])
>290.636191] [] (pruss_soc_bus_probe [pruss_soc_bus]) from
> [] (platform_drv_probe+0x60/0xc0)
> [  290.636203] [] (platform_drv_probe) from []
> (driver_probe_device+0x27c/0x46c)
> [  290.636213] [] (driver_probe_device) from []
> (__driver_attach+0x10c/0x128)
> [  290.636222] [] (__driver_attach) from []
> (bus_for_each_dev+0x8c/0xd0)
> [  290.636231] [] (bus_for_each_dev) from []
> (driver_attach+0x2c/0x30)
> [  290.636241] [] (driver_attach) from []
> (bus_add_driver+0x1b8/0x278)
> [  290.636250] [] (bus_add_driver) from []
> (driver_register+0x88/0x104)
> [  290.636261] [] (driver_register) from []
> (__platform_driver_register+0x50/0x58)
> [  290.636275] [] (__platform_driver_register) from []
> (pruss_soc_bus_driver_init+0x20/0xc0 [pruss_soc_bus])
> [  290.636293] [] (pruss_soc_bus_driver_init [pruss_soc_bus]) from
> [] (do_one_initcall+0x64/0x1a0)
> [  290.636313] [] (do_one_initcall) from []
> (do_init_module+0x74/0x3d8)
> [  290.636329] [] (do_init_module) from []
> (load_module+0x1fd4/0x26d8)
> [  290.636341] [] (load_module) from []
> (SyS_finit_module+0xf0/0x120)
> [  290.636356] [] (SyS_finit_module) from []
> (__sys_trace_return+0x0/0x10)
> [  290.636362] ---[ end trace 0002 ]---
> ```
>
> With version 4.14
> I think this works okay (no kernel messages when using testing with
> realtime)
> But I've noticed there's no linux-firmware-image for 4.14.6-ti-rt-r17
> https://rcn-ee.com/repos/debian/pool/main/l/linux-upstream/
> I'm not sure if it's missing or perhaps not needed with this kernel version

"in-kernel" firmware got nuked in v4.14.x mainline.. So no more
special package with blobs we don't use..

> I was going to ask which is better to use bone-rt or ti-rt
> I'm not sure 

Re: [Machinekit] help upgrading to latest 4.14 kernel machinekit image

['Ric W' via Machinekit]

Hi,
not wanting to hijack the thread or anything, but I've also been looking at 
updating the kernel to the latest version for machinekit
I'm currently in the process of figuring out how to setup a new machinekit 
setup
I'm using a BeagleBone Green (v1) With the latest debian stretch image from 
the beaglebone site
using the guide from 
https://machinekoder.com/machinekit-debian-stretch-beaglebone-black/

I've been using the commands you've mentioned to test out some of the newer 
kernel releases
(also doing a git pull in that directory to update /opt/scripts/tools/)

For version 4.4
I think this works okay so far, but I'd prefer to use a later version.

For version 4.9
I think there's a glitch in the kernel when I tested it with realtime
```
realtime start
halcmd show pin
realtime stop
```

This resulted in dmesg
```
[  160.328837] IPv6: ADDRCONF(NETDEV_UP): usb1: link is not ready
[  163.498868] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[  168.193722] random: crng init done
[  290.635798] [ cut here ]
[  290.635849] WARNING: CPU: 0 PID: 816 at fs/sysfs/dir.c:31 
sysfs_warn_dup+0x78/0x88
[  290.635857] sysfs: cannot create duplicate filename 
'/bus/platform/devices/4a30.pruss'
[  290.635863] Modules linked in: pruss_soc_bus(+) evdev uio_pdrv_genirq 
uio usb_f_mass_storage usb_f_acm u_serial usb_f_ecm usb_f_rndis u_ether 
libcomposite iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 
nf_nat nf_conntrack iptable_mangle iptable_filter spidev pru_rproc 
pruss_intc pruss tieqep ip_tables x_tables
[  290.635938] CPU: 0 PID: 816 Comm: systemd-udevd Not tainted 
4.9.69-ti-rt-r84 #1
[  290.635942] Hardware name: Generic AM33XX (Flattened Device Tree)
[  290.635980] [] (unwind_backtrace) from [] 
(show_stack+0x20/0x24)
[  290.635993] [] (show_stack) from [] 
(dump_stack+0x80/0x94)
[  290.636006] [] (dump_stack) from [] 
(__warn+0xf8/0x110)
[  290.636016] [] (__warn) from [] 
(warn_slowpath_fmt+0x58/0x74)
[  290.636026] [] (warn_slowpath_fmt) from [] 
(sysfs_warn_dup+0x78/0x88)
[  290.636038] [] (sysfs_warn_dup) from [] 
(sysfs_do_create_link_sd+0xc4/0xcc)
[  290.636046] [] (sysfs_do_create_link_sd) from [] 
(sysfs_create_link+0x38/0x44)
[  290.636063] [] (sysfs_create_link) from [] 
(bus_add_device+0x11c/0x1c0)
[  290.636074] [] (bus_add_device) from [] 
(device_add+0x2e0/0x5e0)
[  290.636086] [] (device_add) from [] 
(of_device_add+0x44/0x4c)
[  290.636097] [] (of_device_add) from [] 
(of_platform_device_create_pdata+0x94/0xcc)
[  290.636108] [] (of_platform_device_create_pdata) from 
[] (of_platform_bus_create+0x18c/0x33c)
[  290.636116] [] (of_platform_bus_create) from [] 
(of_platform_populate+0x9c/0x13c)
[  290.636146] [] (of_platform_populate) from [] 
(pruss_soc_bus_probe+0x12c/0x290 [pruss_soc_bus])
   290.636191] [] (pruss_soc_bus_probe [pruss_soc_bus]) from 
[] (platform_drv_probe+0x60/0xc0)
[  290.636203] [] (platform_drv_probe) from [] 
(driver_probe_device+0x27c/0x46c)
[  290.636213] [] (driver_probe_device) from [] 
(__driver_attach+0x10c/0x128)
[  290.636222] [] (__driver_attach) from [] 
(bus_for_each_dev+0x8c/0xd0)
[  290.636231] [] (bus_for_each_dev) from [] 
(driver_attach+0x2c/0x30)
[  290.636241] [] (driver_attach) from [] 
(bus_add_driver+0x1b8/0x278)
[  290.636250] [] (bus_add_driver) from [] 
(driver_register+0x88/0x104)
[  290.636261] [] (driver_register) from [] 
(__platform_driver_register+0x50/0x58)
[  290.636275] [] (__platform_driver_register) from [] 
(pruss_soc_bus_driver_init+0x20/0xc0 [pruss_soc_bus])
[  290.636293] [] (pruss_soc_bus_driver_init [pruss_soc_bus]) 
from [] (do_one_initcall+0x64/0x1a0)
[  290.636313] [] (do_one_initcall) from [] 
(do_init_module+0x74/0x3d8)
[  290.636329] [] (do_init_module) from [] 
(load_module+0x1fd4/0x26d8)
[  290.636341] [] (load_module) from [] 
(SyS_finit_module+0xf0/0x120)
[  290.636356] [] (SyS_finit_module) from [] 
(__sys_trace_return+0x0/0x10)
[  290.636362] ---[ end trace 0002 ]---
```

With version 4.14
I think this works okay (no kernel messages when using testing with 
realtime)
But I've noticed there's no linux-firmware-image for 4.14.6-ti-rt-r17
https://rcn-ee.com/repos/debian/pool/main/l/linux-upstream/
I'm not sure if it's missing or perhaps not needed with this kernel version


I was going to ask which is better to use bone-rt or ti-rt
I'm not sure what the differences are, but looking at the thread it looks 
like ti-rt is preferred
although bone-rt does have a firmware image for 4.13
linux-firmware-image-4.13.16-bone-rt-r3_1stretch_armhf.deb
(perhaps these are not needed for 4.14 I'm unsure)

Many Thanks
Richard


On Friday, December 15, 2017 at 3:04:29 PM UTC, Robert Nelson wrote:
>
> On Fri, Dec 15, 2017 at 8:55 AM,   
> wrote: 
> > I'm flashing the 4.14 TI RT kernel that Robert linked earlier. I'll let 
> you 
> > know how it goes. Where is the image of the 4.4 TI RT kernel that you'd 
> like 
> > me to try? Or is there an easy what to switch