Date   

Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception #bluetoothmesh

Carles Cufi
 

Hi Vikrant,

 

Thanks for looking into this. It seems like we do not get a buffer back from the command tx queue: https://github.com/zephyrproject-rtos/zephyr/blob/master/subsys/bluetooth/host/hci_core.c#L3318

 

I have copied Luiz and Johan from Intel in case they know what the cause might be.

 

Regards,

 

Carles

 

From: users@... <users@...> On Behalf Of vikrant8051
Sent: 30 May 2018 12:31
To: Cufi, Carles <carles.cufi@...>
Cc: devel@...; users@...
Subject: Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hi Carles,

 

Input ->

/opt/zephyr-sdk/sysroots/x86_64-pokysdk-linux/usr/bin/arm-zephyr-eabi/arm-zephyr-eabi-addr2line -e /home/vikrant/projects/zephyr/zephyr/samples/bluetooth/mesh2/build/zephyr/zephyr.elf 0x5222

 

Output ->

/home/vikrant/projects/zephyr/zephyr/subsys/bluetooth/host/hci_core.c:3318

 

Thank You !!

 

On Wed, May 30, 2018 at 3:51 PM, Cufi, Carles <Carles.Cufi@...> wrote:

If you are using the Zephyr SDK:

 

path-to-zephyr-sdk/sysroots/x86_64-pokysdk-linux/usr/bin/arm-zephyr-eabi/arm-zephyr-eabi-addr2line -e zephyr.elf <addr>

 

Carles

 

 

From: Vikrant More <vikrant8051@...>
Sent: 30 May 2018 12:11
To: Cufi, Carles <carles.cufi@...>
Cc: devel@...; users@...
Subject: Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hi Carles,

 

I will but don't know, how to do that ?

 

 

On Wed, May 30, 2018 at 3:19 PM, Cufi, Carles <Carles.Cufi@...> wrote:

Hi Vikrant,

 

Can you please run addr2line with this address present in the error log: 0x5222?

 

Thanks,

 

Carles

 

From: users@... <users@...> On Behalf Of vikrant8051
Sent: 30 May 2018 11:29
To: devel@...; users@...
Subject: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hello World !!

 

Today after upgrading to v1.12.0-rc2, I'm getting following error while playing with sample code which is inspired from sample/bluetooth/mesh

 

 

[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [ERR] event_scan_prepare: assert: '!_radio.ticker_id_prepare' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x1b77e
Fatal fault in ISR! Spinning...

 

-------------------------------------------------------------------------------------------------------------------------

 

Sometime it completely hangs & nothing get displayed on debug terminal

 

-------------------------------------------------------------------------------------------------------------------------

 

Thank You !!

 

 

 

 

 

 

 


Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception #bluetoothmesh

vikrant8051 <vikrant8051@...>
 

Hi Carles,

Input ->
/opt/zephyr-sdk/sysroots/x86_64-pokysdk-linux/usr/bin/arm-zephyr-eabi/arm-zephyr-eabi-addr2line -e /home/vikrant/projects/zephyr/zephyr/samples/bluetooth/mesh2/build/zephyr/zephyr.elf 0x5222

Output ->
/home/vikrant/projects/zephyr/zephyr/subsys/bluetooth/host/hci_core.c:3318

Thank You !!

On Wed, May 30, 2018 at 3:51 PM, Cufi, Carles <Carles.Cufi@...> wrote:

If you are using the Zephyr SDK:

 

path-to-zephyr-sdk/sysroots/x86_64-pokysdk-linux/usr/bin/arm-zephyr-eabi/arm-zephyr-eabi-addr2line -e zephyr.elf <addr>

 

Carles

 

 

From: Vikrant More <vikrant8051@...>
Sent: 30 May 2018 12:11
To: Cufi, Carles <carles.cufi@...>
Cc: devel@...; users@...
Subject: Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hi Carles,

 

I will but don't know, how to do that ?

 

 

On Wed, May 30, 2018 at 3:19 PM, Cufi, Carles <Carles.Cufi@...> wrote:

Hi Vikrant,

 

Can you please run addr2line with this address present in the error log: 0x5222?

 

Thanks,

 

Carles

 

From: users@... <users@...> On Behalf Of vikrant8051
Sent: 30 May 2018 11:29
To: devel@...; users@...
Subject: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hello World !!

 

Today after upgrading to v1.12.0-rc2, I'm getting following error while playing with sample code which is inspired from sample/bluetooth/mesh

 

 

[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [ERR] event_scan_prepare: assert: '!_radio.ticker_id_prepare' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x1b77e
Fatal fault in ISR! Spinning...

 

-------------------------------------------------------------------------------------------------------------------------

 

Sometime it completely hangs & nothing get displayed on debug terminal

 

-------------------------------------------------------------------------------------------------------------------------

 

Thank You !!

 

 

 

 

 

 



Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception #bluetoothmesh

Carles Cufi
 

If you are using the Zephyr SDK:

 

path-to-zephyr-sdk/sysroots/x86_64-pokysdk-linux/usr/bin/arm-zephyr-eabi/arm-zephyr-eabi-addr2line -e zephyr.elf <addr>

 

Carles

 

 

From: Vikrant More <vikrant8051@...>
Sent: 30 May 2018 12:11
To: Cufi, Carles <carles.cufi@...>
Cc: devel@...; users@...
Subject: Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hi Carles,

 

I will but don't know, how to do that ?

 

 

On Wed, May 30, 2018 at 3:19 PM, Cufi, Carles <Carles.Cufi@...> wrote:

Hi Vikrant,

 

Can you please run addr2line with this address present in the error log: 0x5222?

 

Thanks,

 

Carles

 

From: users@... <users@...> On Behalf Of vikrant8051
Sent: 30 May 2018 11:29
To: devel@...; users@...
Subject: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hello World !!

 

Today after upgrading to v1.12.0-rc2, I'm getting following error while playing with sample code which is inspired from sample/bluetooth/mesh

 

 

[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [ERR] event_scan_prepare: assert: '!_radio.ticker_id_prepare' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x1b77e
Fatal fault in ISR! Spinning...

 

-------------------------------------------------------------------------------------------------------------------------

 

Sometime it completely hangs & nothing get displayed on debug terminal

 

-------------------------------------------------------------------------------------------------------------------------

 

Thank You !!

 

 

 

 

 

 


Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception #bluetoothmesh

vikrant8051 <vikrant8051@...>
 

Hi Carles,

I will but don't know, how to do that ?


On Wed, May 30, 2018 at 3:19 PM, Cufi, Carles <Carles.Cufi@...> wrote:

Hi Vikrant,

 

Can you please run addr2line with this address present in the error log: 0x5222?

 

Thanks,

 

Carles

 

From: users@... <users@...> On Behalf Of vikrant8051
Sent: 30 May 2018 11:29
To: devel@...; users@...
Subject: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hello World !!

 

Today after upgrading to v1.12.0-rc2, I'm getting following error while playing with sample code which is inspired from sample/bluetooth/mesh

 

 

[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [ERR] event_scan_prepare: assert: '!_radio.ticker_id_prepare' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x1b77e
Fatal fault in ISR! Spinning...

 

-------------------------------------------------------------------------------------------------------------------------

 

Sometime it completely hangs & nothing get displayed on debug terminal

 

-------------------------------------------------------------------------------------------------------------------------

 

Thank You !!

 

 

 

 

 



Re: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception #bluetoothmesh

Carles Cufi
 

Hi Vikrant,

 

Can you please run addr2line with this address present in the error log: 0x5222?

 

Thanks,

 

Carles

 

From: users@... <users@...> On Behalf Of vikrant8051
Sent: 30 May 2018 11:29
To: devel@...; users@...
Subject: [Zephyr-users] #BluetoothMesh ...about latest kernel OOPS & exception

 

Hello World !!

 

Today after upgrading to v1.12.0-rc2, I'm getting following error while playing with sample code which is inspired from sample/bluetooth/mesh

 

 

[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [ERR] event_scan_prepare: assert: '!_radio.ticker_id_prepare' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x1b77e
Fatal fault in ISR! Spinning...

 

-------------------------------------------------------------------------------------------------------------------------

 

Sometime it completely hangs & nothing get displayed on debug terminal

 

-------------------------------------------------------------------------------------------------------------------------

 

Thank You !!

 

 

 

 

 


#BluetoothMesh ...about latest kernel OOPS & exception #bluetoothmesh

vikrant8051 <vikrant8051@...>
 

Hello World !!

Today after upgrading to v1.12.0-rc2, I'm getting following error while playing with sample code which is inspired from sample/bluetooth/mesh


[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [WRN] process_events: Unexpected k_poll event state 7
[bt] [WRN] process_events: Unexpected k_poll event state 8
[bt] [ERR] hci_tx_thread: assert: 'err == 0' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x5222
Fatal fault in thread 0x20000484! Aborting.
[bt] [ERR] event_scan_prepare: assert: '!_radio.ticker_id_prepare' failed
***** Kernel OOPS! *****
Current thread ID = 0x20000484
Faulting instruction address = 0x1b77e
Fatal fault in ISR! Spinning...

-------------------------------------------------------------------------------------------------------------------------

Sometime it completely hangs & nothing get displayed on debug terminal

-------------------------------------------------------------------------------------------------------------------------

Thank You !!






Re: How to support multiple defconfigs to be merged?

Sebastian Boe
 

CONF_FILE is a space-separated list instead of the normal CMake semicolon-separated list.

So doing

set(CONF_FILE "defconfig_1 defconfig_2 defconfig_3")

is expected to work without any upstream modifications.

I can't remember the reason it was decided to be space-separated.

It could have been to support the user specifying

cmake -DCONF_FILE="a b c"

instead of

cmake -DCONF_FILE="a;b;c"

________________________________________
From: devel@lists.zephyrproject.org <devel@lists.zephyrproject.org> on behalf of Li, Jun R <jun.r.li@intel.com>
Sent: Wednesday, 30 May 2018 1:24:31 AM
To: Rosen, Michael R; devel@lists.zephyrproject.org
Subject: Re: [Zephyr-devel] How to support multiple defconfigs to be merged?

Thank you, Mike! Yeah, it is better to have a PR in Zephyr to support this.

Regards,
Jun


From: <devel@lists.zephyrproject.org> on behalf of Michael Rosen <michael.r.rosen@intel.com>
Date: Tuesday, May 29, 2018 at 16:12
To: Jun Li <jun.r.li@intel.com>, "devel@lists.zephyrproject.org" <devel@lists.zephyrproject.org>
Subject: Re: [Zephyr-devel] How to support multiple defconfigs to be merged?

Jun,

As we discussed, you can use the following to do what you have below in the new cmake system:

set(CONF_FILE “defconfig_1”)
list(APPEND CONF_FILE “defconfig_2”)
list(APPEND CONF_FILE “defconfig_3”)

However, you need to update cmake/kconfig.cmake:13 to be:

string(REPLACE " " ";" CONF_FILE_AS_LIST "${CONF_FILE}")

ie, add quotes to CONF_FILE. Unless the Zephyr team has another suggestion for how to achieve multiple defconfig files; it might be good to add these quotes to the actual release (I can make a pull request if this is the suggestions approach)

Mike

From: devel@lists.zephyrproject.org [mailto:devel@lists.zephyrproject.org] On Behalf Of Li, Jun R
Sent: Tuesday, May 29, 2018 2:22 PM
To: devel@lists.zephyrproject.org
Subject: [Zephyr-devel] How to support multiple defconfigs to be merged?

Hi everyone,

Before Zephyr adopted cmake as the building tool, it supported to merge multiple defconfig files in project’s makefile by the following way:

CONF_FILE = defconfig_1
CONF_FILE += defconfig_2
CONF_FILE += defconfig_3

The three “defconfig_x” files will be merged together when being built.

However, in the current version with CMake, only one defconfig is acceptable by the following way in the project’s CMakeLists.txt:

set(CONF_FILE defconfig_1)

Other “set” operations on “CONF_FILE” will make previous ones to be ignored, meaning only the last “set(CONF_FILE defconfig_x)” works.

So, is there still a way to merge multiple defconfigs by using CMake?

Thank you!

Jun Li


Zephyr 1.12.0-rc2 tagged

Maureen Helm
 

Hi all,

 

We have just tagged Zephyr 1.12.0-rc2.

 

The merge window for new features has been closed since rc1, and we will continue to merge only bug fixes and documentation until the release is finalized. As a reminder, we must have zero open high priority bugs and less than 20 open medium priority bugs before we can release. Please review any open bugs that have been assigned to you.

https://github.com/zephyrproject-rtos/zephyr/issues?q=is%3Aopen+is%3Aissue+label%3Abug+label%3A%22priority%3A+high%22

https://github.com/zephyrproject-rtos/zephyr/issues?utf8=%E2%9C%93&q=is%3Aopen+is%3Aissue+label%3Abug+label%3A%22priority%3A+medium%22+-milestone%3Av1.13.0

 

The full release log can be found here:

https://github.com/zephyrproject-rtos/zephyr/releases/tag/v1.12.0-rc2

 

Thanks to everybody who contributed to this release!

 

Maureen


Re: How to support multiple defconfigs to be merged?

Li, Jun R
 

Thank you, Mike! Yeah, it is better to have a PR in Zephyr to support this.

 

Regards,

Jun

 

 

From: <devel@...> on behalf of Michael Rosen <michael.r.rosen@...>
Date: Tuesday, May 29, 2018 at 16:12
To: Jun Li <jun.r.li@...>, "devel@..." <devel@...>
Subject: Re: [Zephyr-devel] How to support multiple defconfigs to be merged?

 

Jun,

 

As we discussed, you can use the following to do what you have below in the new cmake system:

 

set(CONF_FILE “defconfig_1”)

list(APPEND CONF_FILE “defconfig_2”)

list(APPEND CONF_FILE “defconfig_3”)

 

However, you need to update cmake/kconfig.cmake:13 to be:

 

string(REPLACE " " ";" CONF_FILE_AS_LIST "${CONF_FILE}")

 

ie, add quotes to CONF_FILE. Unless the Zephyr team has another suggestion for how to achieve multiple defconfig files; it might be good to add these quotes to the actual release (I can make a pull request if this is the suggestions approach)

 

Mike

 

From: devel@... [mailto:devel@...] On Behalf Of Li, Jun R
Sent: Tuesday, May 29, 2018 2:22 PM
To: devel@...
Subject: [Zephyr-devel] How to support multiple defconfigs to be merged?

 

Hi everyone,

 

Before Zephyr adopted cmake as the building tool, it supported to merge multiple defconfig files in project’s makefile by the following way:

 

CONF_FILE = defconfig_1

CONF_FILE += defconfig_2

CONF_FILE += defconfig_3

 

The three “defconfig_x” files will be merged together when being built.

 

However, in the current version with CMake, only one defconfig is acceptable by the following way in the project’s CMakeLists.txt:

 

set(CONF_FILE defconfig_1)

 

Other “set” operations on “CONF_FILE” will make previous ones to be ignored, meaning only the last “set(CONF_FILE defconfig_x)” works.

 

So, is there still a way to merge multiple defconfigs by using CMake?

 

Thank you!

 

Jun Li

 

 


Re: How to support multiple defconfigs to be merged?

Michael Rosen
 

Jun,

 

As we discussed, you can use the following to do what you have below in the new cmake system:

 

set(CONF_FILE “defconfig_1”)

list(APPEND CONF_FILE “defconfig_2”)

list(APPEND CONF_FILE “defconfig_3”)

 

However, you need to update cmake/kconfig.cmake:13 to be:

 

string(REPLACE " " ";" CONF_FILE_AS_LIST "${CONF_FILE}")

 

ie, add quotes to CONF_FILE. Unless the Zephyr team has another suggestion for how to achieve multiple defconfig files; it might be good to add these quotes to the actual release (I can make a pull request if this is the suggestions approach)

 

Mike

 

From: devel@... [mailto:devel@...] On Behalf Of Li, Jun R
Sent: Tuesday, May 29, 2018 2:22 PM
To: devel@...
Subject: [Zephyr-devel] How to support multiple defconfigs to be merged?

 

Hi everyone,

 

Before Zephyr adopted cmake as the building tool, it supported to merge multiple defconfig files in project’s makefile by the following way:

 

CONF_FILE = defconfig_1

CONF_FILE += defconfig_2

CONF_FILE += defconfig_3

 

The three “defconfig_x” files will be merged together when being built.

 

However, in the current version with CMake, only one defconfig is acceptable by the following way in the project’s CMakeLists.txt:

 

set(CONF_FILE defconfig_1)

 

Other “set” operations on “CONF_FILE” will make previous ones to be ignored, meaning only the last “set(CONF_FILE defconfig_x)” works.

 

So, is there still a way to merge multiple defconfigs by using CMake?

 

Thank you!

 

Jun Li

 

 


How to support multiple defconfigs to be merged?

Li, Jun R
 

Hi everyone,

 

Before Zephyr adopted cmake as the building tool, it supported to merge multiple defconfig files in project’s makefile by the following way:

 

CONF_FILE = defconfig_1

CONF_FILE += defconfig_2

CONF_FILE += defconfig_3

 

The three “defconfig_x” files will be merged together when being built.

 

However, in the current version with CMake, only one defconfig is acceptable by the following way in the project’s CMakeLists.txt:

 

set(CONF_FILE defconfig_1)

 

Other “set” operations on “CONF_FILE” will make previous ones to be ignored, meaning only the last “set(CONF_FILE defconfig_x)” works.

 

So, is there still a way to merge multiple defconfigs by using CMake?

 

Thank you!

 

Jun Li

 

 


POSIX clock_gettime() has meaningless 'tv_nsec' field.

David Brown
 

I’ve created Issue 8009 to track this.

 

The current implementation of clock_gettime() attempts to use the k_cycle_get_32() counter to provide a more accurate notion of time (filling in the tv_nsec field). Unfortunately, on targets where the cycle counter is not a power-of-two division of seconds, after the first time the counter wraps, the resulting value becomes meaningless.

 

For example, on the frdm-k64f, after 35 seconds, the time will appear to jump backwards a bit (when the counter wraps), and then once a second, the time will jump backward a bit, because the cycle counter is out of phase with seconds. Every roughly 35 seconds, the jump in the time will shift. The clock will shift around in the sub-second value, only returning to the correct offset for a 35 second period about every 94 days.

 

The reason for this email, though, is to try and determine what the best way of fixing this is. I see two ways to fix it:

 

  1. Don’t use k_cycle_get_32(), and just use the ms accuracy of the system timer. This is an easier solution, and having ms accuracy of the sub-second field is certainly more useful than having a meaningless number.
  2. Implement a 64-bit counter to detect wrap around of the cycle counter. One way to do this would be to check the cycle counter on each clock interrupt, and when it has wrapped, increment an upper 32-bit value counter.

 

Any thoughts on what would be the best way to fix this?

 

Thanks,

David


[Bluetooth pairing] psskey_confirm callback is not executing even after successful pairing

vikrant8051 <vikrant8051@...>
 

Hi,

static void auth_passkey_confirm(struct bt_conn *conn, unsigned int passkey)
{
    printk("\n\rauth_passkey_confirm !!\n\r");
}

static struct bt_conn_auth_cb auth_cb_display = {
    .passkey_display = auth_passkey_display,
    .passkey_entry = NULL,
    .cancel = auth_cancel,
    .passkey_confirm = auth_passkey_confirm,
};

I've edited samples/bluetooth/peripheral  as shown above.
But auth_passkey_confirm is not executing even after successful pairing.

Am I missing something ?

Thank You !!


Re: BLE synchronization

Carles Cufi
 

Hi Daniel,

 

Whether a reply is sent in the same connection event really depends on timing between the RX and TX paths, and right now I doubt we can process it in the TIFS 150us window as you point out.

I will defer this question to Vinayak who should be able to give you more info in the matter. He is on holiday for the next week or so so please be patient.

 

Regarding the callback on TX/ACK, this has been asked in GitHub issues, so please add your comments there:

https://github.com/zephyrproject-rtos/zephyr/issues/7659

https://github.com/zephyrproject-rtos/zephyr/issues/7654

 

Regards,

 

Carles

 

From: devel@... <devel@...> On Behalf Of Daniel Widmann
Sent: 28 May 2018 18:27
To: devel@...
Subject: [Zephyr-devel] BLE synchronization

 

Hello there,

 

I want to synchronize an event between a BLE Central and Peripheral with an accuracy of ~1ms. Both devices use a NRF52832 with zephyr. 

 

Using the normal GATT API, there is a significant delay between the API call and the callback on the other side. I believe this happens because zephyr has to wait until the next connection event to send/receive a packet. Even the shortest connection interval (7.5 ms) is too long for what I need.

 

Is it possible for a zephyr peripheral to send a reply within the same connection event? The standard seems to allow this and I think it would solve my problem. But I am not sure if zephyr can reply within 150 us (see https://stackoverflow.com/questions/42560906/).

 

Is it possible to get a callback when a BLE request was sent or the ack was received?

 

 

Thanks,

 

Daniel 


BLE synchronization

Daniel Widmann
 

Hello there,

I want to synchronize an event between a BLE Central and Peripheral with an accuracy of ~1ms. Both devices use a NRF52832 with zephyr. 

Using the normal GATT API, there is a significant delay between the API call and the callback on the other side. I believe this happens because zephyr has to wait until the next connection event to send/receive a packet. Even the shortest connection interval (7.5 ms) is too long for what I need.

Is it possible for a zephyr peripheral to send a reply within the same connection event? The standard seems to allow this and I think it would solve my problem. But I am not sure if zephyr can reply within 150 us (see https://stackoverflow.com/questions/42560906/).

Is it possible to get a callback when a BLE request was sent or the ack was received?


Thanks,

Daniel 


Re: Device Tree on separated builds

Sebastian Boe
 

"I have an app that is built on many boards, and I'd like to build the
app only once and the dts for each board."

That is not supported if I am understanding you correctly. Your options are
to either link in a prebuilt 'app' or hope/ensure that the compiler command
is the same for the app so that ccache caches the redundant compiler invocations.

I assume this is for build-time performance reasons?

________________________________________
From: devel@lists.zephyrproject.org <devel@lists.zephyrproject.org> on behalf of Inaki Malerba <inaki@malerba.space>
Sent: Monday, 28 May 2018 4:24:44 PM
To: users@lists.zephyrproject.org; devel@lists.zephyrproject.org
Subject: [Zephyr-devel] Device Tree on separated builds

Hi there !

Is there any way to build the device tree on a different bin than the
kernel?

I have an app that is built on many boards, and I'd like to build the
app only once and the dts for each board.


Thanks!

--
- ina


Device Tree on separated builds

Inaki Malerba <inaki@...>
 

Hi there !

Is there any way to build the device tree on a different bin than the
kernel?

I have an app that is built on many boards, and I'd like to build the
app only once and the dts for each board.


Thanks!

--
- ina


Fixing issues

Nashif, Anas
 

Hi everyone,

 

When fixing bugs, please make sure to use the right keywords to enable auto-close when the pull request is merged. Please see

 

https://help.github.com/articles/closing-issues-using-keywords/

 

for more details. In short, try to use the “Fixes #1234” or any supported combination as explain above. This will help us keep the bug count down, i.e. bugs that are fixed are also closed.

 

Also, when submitting a multi-commit PR, make sure you have a summary in the PR explain the whole changeset, do not leave the PR description empty, this will allow for faster review, the reviewer will have an idea about the change set without having to go through the commits.

 

Thank you,

 

Anas


What end user can't do if have only NET & APP key for #BluetoothMesh n/w ? #bluetoothmesh

vikrant8051 <vikrant8051@...>
 

Hi,

Assume
1) a company don't wanna go with any cloud service to store & upgrade #BluetoothMesh .json file for his products.

2) And no #OOB authentication during provisioning

3) On other side, user provisioned many NODEs using a Provisioner App where NET & APP Key get created using his credentials which he/she can't forget in any case.  (No random keys)

In this case, even App get uninstall or smartphone get damage, user can reinstall App & add his known credentials to generate NET & APP key

So my question is, what user can do or can't do if he has only NET & APP key of his Mesh Network ?

Device key gonna random & different for every NODE.

Can he/she re-create entire .json file as it is by connecting with one by one NODE using only NET & APP Key ? Is it possible to connect with only one NODE & fetch data of all NODEs in vicinity ?

Excuse me, if it is heavily wrong concept.

Thank You !!





Re: Mailing list archives before 2017

Erwan Gouriou
 

Indeed, I've found what I was looking for.

Thanks Brett

On 23 May 2018 at 19:30, Brett Preston <bpreston@...> wrote:
All,

Just an update that the older archives (2016) are now available on the mail lists -

Thank you

On Wed, May 16, 2018 at 1:16 PM, Brett Preston <bpreston@...> wrote:
All - quick update in that we have identified what caused not all of the needed archives to be migrated over. Resolving this is both possible and should not take long. Task should be completed by EOD Thursday, May 17.

Thank you

On Wed, May 16, 2018 at 7:11 AM, Brett Preston <bpreston@...> wrote:
Thanks Carles. Checking with our IT team and will revert back once I have an update.

On Wed, May 16, 2018 at 7:10 AM, Cufi, Carles <Carles.Cufi@...> wrote:

+ Brett

 

From: devel@... <devel@...> On Behalf Of Erwan Gouriou
Sent: 16 May 2018 14:59
To: zephyr-devel <zephyr-devel@...ect.org>
Subject: [Zephyr-devel] Mailing list archives before 2017

 

Hi all,

 

I'm looking for mail archives of this mainling list before 2017.

Looking in https://lists.zephyrproject.org/g/devel, it only starts on Jan '17.

This might be because we changed mailinglist service recently.

 

Are there archives of archives?

 

Erwan

 

 




--
Brett Preston
The Linux Foundation
+1 (971) 303-9030
bpreston@...

Skype: bprestoncf



--
Brett Preston
The Linux Foundation
+1 (971) 303-9030
bpreston@...

Skype: bprestoncf



--
Brett Preston
The Linux Foundation
+1 (971) 303-9030
bpreston@...

Skype: bprestoncf

3401 - 3420 of 8033