usb: patches to resolve issues around Kconfig option USB_UART_CONSOLE #40220

jfischer-no · 2021-11-09T16:54:46Z

Patches to resolve issues around Kconfig option USB_UART_CONSOLE.

Alternative to #40192
Alternative to #39987

jakub-uC · 2021-11-10T13:14:53Z

@jfischer-no : May you please let me know if you have tested this fix with nrf52840 dongle?

nordic-krch · 2021-11-10T14:22:17Z

tests/bluetooth/shell/src/main.c

-	if (IS_ENABLED(CONFIG_USB_UART_CONSOLE)) {
-		usb_enable(NULL);
-		k_sleep(K_SECONDS(2));
+#if DT_NODE_HAS_COMPAT(DT_CHOSEN(zephyr_shell_uart), zephyr_cdc_acm_uart)


maybe you could extract it into a function, something like:

if (DT_NODE_HAS_COMPAT(DT_CHOSEN(zephyr_shell_uart), zephyr_cdc_acm_uart)) { wait_usb_ready(); }

That would make it obvious why it is here.

This part is removed now.

nordic-krch · 2021-11-10T14:30:45Z

subsys/usb/class/Kconfig.cdc

@@ -53,6 +53,12 @@ config CDC_ACM_DTE_RATE_CALLBACK_SUPPORT
 	  by Arduino style programmers to reset the device into the
 	  bootloader.

+config USB_UART_CONSOLE


Could we get rid of this at all? Maybe something with dt_compat_enabled could be used instead, i.e. here

zephyr/subsys/shell/backends/Kconfig.backends

Line 76 in 23ab949

default y if USB_UART_CONSOLE

No, it is there for reasons.

deprecated now

tejlmand · 2021-11-11T08:18:30Z

drivers/console/Kconfig

@@ -34,7 +34,7 @@ config CONSOLE_HANDLER

 config CONSOLE_INIT_PRIORITY
 	int "Console init priority"
-	default 95 if USB_UART_CONSOLE || UART_MUX
+	default 95 if UART_MUX


First observation, the default priority of 95 was chosen because of the USB stack initialization, see:
df2bb46

usb: console: Initialize USB console after USB Device stack

It does make sense to initialize USB console after USB Device stack.
Note that the value is selected only if we specify USB_UART_CONSOLE
in prj.conf, not in menuconfig afterwards.

Fixes #16518

So what has changed, since console can now be initialized at 60 instead ?
This commit only changed the symbol used: d8bc37b
but the default value is still 50
SERIAL_INIT_PRIORITY defaults to KERNEL_INIT_PRIORITY_DEVICE default == 50

Second observation:
The CONSOLE_INIT_PRIORITY defaults to KERNEL_INIT_PRIORITY_DEFAULT == 40 when UART_CONSOLE is not enabled, and the setting is having a prompt.
This means this setting suffers from the stuck symbols syndrome:
https://docs.zephyrproject.org/latest/guides/build/kconfig/tips.html#stuck-symbols-in-menuconfig-and-guiconfig

meaning that if anyone later enables the USB_UART_CONSOLE or the UART_CONSOLE, then the priority will not change, in which case the 40 value would risk still being active and not the preferred default of 60.

I believe majority of boards are having CONFIG_UART_CONSOLE=y in their _defconfig so it might not be a real problem, but it still looks as we might have a risk of strangely behaving system.

And it sounds to me as there is also some upper bound dependency since 95 is no longer a good default.
So should this priority in reality be a range ?
Where the upper and lower bounds might be defined by other settings.

Or should the setting be promptless in certain cases ?
Note, a promptless setting can still be controlled through a Kconfig file, for example Kconfig.board

So what has changed, since console can now be initialized at 60 instead ?
This commit only changed the symbol used: d8bc37b
but the default value is still 50
SERIAL_INIT_PRIORITY defaults to KERNEL_INIT_PRIORITY_DEVICE default == 50

CDC ACM UART driver has been reworked a bit and does not block when USB device support is not initialized, (behaves like regular UART).

Or should the setting be promptless in certain cases ?

I am in favor of this option.

CDC ACM UART driver has been reworked a bit and does not block when USB device support is not initialized, (behaves like regular UART).

I would like to see this sentence in the commit message, as that is an important piece of information in order to understand why changing the default from 95 to 60 is safe.

tejlmand · 2021-11-11T08:50:22Z

subsys/usb/class/Kconfig.cdc

+	  This option is only used to enable relevant options in backend
+	  configuration (if any), and for backward compatibility.


This sound like a setting we don't really assume is in use, but we don't know.

I think it should be investigated and marked deprecated so that it can be completely removed in two releases from now.

Else I fear this setting will just stay and never get cleaned up.

No, it is in use and is investigated, see commit message:
"Kconfig option USB_UART_CONSOLE does not have any dependencies
on console drivers, but is used to select backend specific
Kconfig option in subsys/shell/backends/Kconfig.backends.
Keep the name of the option USB_UART_CONSOLE but move it
to CDC ACM configuration and update description."

then I suggest to rework the comment, cause that is not clear to me.

And as this setting is deprecated, are there any plans / tasks to follow up on removing the dependency in subsys/shell/backends/Kconfig.backends ?

Commit was reworked. I will remove Kconfig option USB_UART_CONSOLE and dependency in subsys/shell/backends/Kconfig.backends in subsequent PR because MCUboot needs to be adapted as well.

jfischer-no · 2021-11-11T15:21:17Z

@jfischer-no : May you please let me know if you have tested this fix with nrf52840 dongle?

yes I am testing it with nrf52840dongle_nrf52840

Use same init level and priority as serial drivers. Align priority to the changes in commit ad14505 ("drivers: serial: Refactor drivers to use shared init priority Kconfig") Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Do not change init level and priority of console driver if Kconfig option CONFIG_USB_UART_CONSOLE is enabled because CDC ACM UART driver now uses the same init level and priority as regular serial driver. Replace CONFIG_USB_UART_CONSOLE with DT_NODE_HAS_COMPAT(DT_CHOSEN(zephyr_...), zephyr_cdc_acm_uart) in the source code. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Kconfig option USB_UART_CONSOLE does not have any dependencies on console drivers, but is used to select backend specific Kconfig option in subsys/shell/backends/Kconfig.backends. Move Kconfig option USB_UART_CONSOLE to CDC ACM configuration, deprecate it and update description. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Possibility to initialize USB device support is useful when only CDC ACM class is enabled and CDC ACM UART is used as backend for console, shell, or logging. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Add generic shields that provide devicetree and configuration overlays, and configure USB device stack so that CDC ACM UART can be used as backend for console, logging, or shell. This approach allows us to avoid many identical overlays in samples and tests directories, and cumbersome board configurations. These shields use Kconfig option CONFIG_USB_DEVICE_INITIALIZE_AT_BOOT and is mainly intended to be used with boards like nrf52840dongle_nrf52840 or bl654_usb. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Remove CDC ACM UART devicetree and config overlay, and explicit USB support initialization in favor of generic CDC ACM shield. For shell support via USB CDC ACM, this example is to be built as follows: west build -b nrf52840dongle_nrf52840 samples/subsys/shell/shell_module -- -DSHIELD=cdc_acm_shell Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Remove CDC ACM UART devicetree and config overlay, and explicit USB support initialization in favor of generic CDC ACM shield. For shell support via USB CDC ACM, this example is to be built as follows: west build -b nrf52840dongle_nrf52840 tests/bluetooth/shell -- -DSHIELD=cdc_acm_shell Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Remove CDC ACM UART devicetree and config overlay, and explicit USB support initialization in favor of generic CDC ACM shield. For console support via USB CDC ACM, this example is to be built as follows: west build -b reel_board tests/drivers/uart/uart_basic_api -- -DSHIELD=cdc_acm_console Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Remove CDC ACM UART devicetree overlay, and explicit USB support initialization in favor of generic CDC ACM shield. For console support via USB CDC ACM, this example is to be built as follows: west build -b sensortile_box samples/boards/sensortile_box -- -DSHIELD=cdc_acm_console Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

This sample uses console driver and therefore Kconfig option CONFIG_CONSOLE and CONFIG_UART_CONSOLE are required. Kconfig option CONFIG_USB_UART_CONSOLE no longer has any influence here and should be removed. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Remove CDC ACM UART devicetree nodes, chosen properties and explicit USB support configuration in favor of generic CDC ACM shield. For console support via USB CDC ACM, in-tree samples for this boards is to be built as follows: west build -b nrf52840dongle_nrf52840 samples/basic/threads -- -DSHIELD=cdc_acm_console For shell support via USB CDC ACM, in-tree samples for this boards is to be built as follows: west build -b bl654_usb samples/subsys/shell/shell_module -- -DSHIELD=cdc_acm_shell Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Since there are no more users and dependencies of Kconfig option USB_UART_CONSOLE in the tree, remove the remains. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Add support for chosen property "zephyr,bt-c2h-uart". Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

erwango · 2021-11-15T09:34:22Z

tests/drivers/uart/uart_basic_api/testcase.yaml

@@ -19,8 +19,7 @@ tests:
    filter: CONFIG_UART_CONSOLE
    harness: keyboard
  drivers.uart.cdc_acm:
-    extra_args: OVERLAY_CONFIG="overlay-usb.conf"
-                DTC_OVERLAY_FILE="usb.overlay"
+    extra_args: SHIELD="cdc_acm_console"


A question here, I like the general direction, but I'd like to know how tests will work with boards that already declare usb for value of zephyr,console chosen.

AFAIK last property definition wins. With this PR, there are no more boards that define chosen property with reference to "zephyr,cdc-acm-uart" compatible node. But it works even if it is the case and shield is used, then there would be two CDC ACM devices, one of them would be used by the test.

carlescufi

I am not against this and I have tested it and works fine. but maybe we should rethink the word SHIELD here. Perhaps we could come up with something that describes both meta-configurations and actual hardware shields. Some new name, like METACFG?

Imagine we used it for other purposes too, like:

west build samples/hello_world -- -DMETACFG=usb_cdc_acm -DMETACFG=use_mcuboot -DMETACFG=my_hw_shield

tejlmand

Main issue I see is the SHIELD=cdc_acm_console, please see detailed comments.

The reason for this PR is stated to be

Patches to resolve issues around Kconfig option USB_UART_CONSOLE.
#40220 (comment)

but it is not clear to me why this series of commits are needed to solve the issue.

Is is possible to split this PR into the part that is needed to solve the particular issue, and then a second part with the cdc acm console enhancement ?

If not, then please make a proper description why the big rewrite to the system is needed.

tejlmand · 2021-11-16T12:24:14Z

drivers/console/Kconfig

@@ -34,7 +34,7 @@ config CONSOLE_HANDLER

 config CONSOLE_INIT_PRIORITY
 	int "Console init priority"
-	default 95 if USB_UART_CONSOLE || UART_MUX
+	default 95 if UART_MUX


CDC ACM UART driver has been reworked a bit and does not block when USB device support is not initialized, (behaves like regular UART).

I would like to see this sentence in the commit message, as that is an important piece of information in order to understand why changing the default from 95 to 60 is safe.

tejlmand · 2021-11-16T12:25:42Z

subsys/usb/class/Kconfig.cdc

+	  This option is only used to enable relevant options in backend
+	  configuration (if any), and for backward compatibility.


then I suggest to rework the comment, cause that is not clear to me.

And as this setting is deprecated, are there any plans / tasks to follow up on removing the dependency in subsys/shell/backends/Kconfig.backends ?

tejlmand · 2021-11-16T12:40:10Z

boards/shields/cdc_acm/doc/index.rst

@@ -0,0 +1,64 @@
+.. _cdc_acm_shield:
+
+Generic shields for CDC ACM UART


Shields today are a piece of hardware you may attach to another board.
file:///projects/github/ncs/zephyr/doc/_doc/html/boards/shields/index.html

This commit 4d53a79 uses the shield feature for something that is not a shield.

That is a direction I dislike.

It's another approach of #14740 and #33656, just through shield this time.

Using shield this way makes it unclear what a shield is, and it also suffers the same disadvantage / complexity as mentioned here: #14740 (comment)

If you would like to propose a solution like the shield then please make a dedicated PR with such proposal the can be discussed, also at dev-review. For example the METACFG name as mentioned here: #40220 (review)

Not being against such idea, my concern is still related to:

How does end-user know when this meta-config can be used.

One more place (in an already complex system) from where files are picked up

What would be the threshold when new meta-config should be introduced.

I am convinced that shield approach is a better fit for CDC ACM UART, similar to how one would connect a real UART controller through I2C/SPI using a shield.

tejlmand · 2021-11-16T12:49:46Z

samples/boards/sensortile_box/app.overlay

@@ -1,18 +0,0 @@
-/*


Removing this means that users are now forced to specify the -DSHIELD=cdc_acm_console or they'll see those warnings:

Parsing /projects/github/ncs/zephyr/Kconfig Loaded configuration '/projects/github/ncs/zephyr/boards/arm/nrf52840dk_nrf52840/nrf52840dk_nrf52840_defconfig' Merged configuration '/projects/github/ncs/zephyr/samples/boards/sensortile_box/prj.conf' Configuration saved to '/projects/github/ncs/zephyr/samples/boards/sensortile_box/build/zephyr/.config' Kconfig header saved to '/projects/github/ncs/zephyr/samples/boards/sensortile_box/build/zephyr/include/generated/autoconf.h' warning: USB_DEVICE_VID (defined at subsys/usb/Kconfig:21) was assigned the value '0x0483' but got the value ''. Check these unsatisfied dependencies: USB_DEVICE_STACK (=n). See http://docs.zephyrproject.org/latest/reference/kconfig/CONFIG_USB_DEVICE_VID.html and/or look up USB_DEVICE_VID in the menuconfig/guiconfig interface. The Application Development Primer, Setting Configuration Values, and Kconfig - Tips and Best Practices sections of the manual might be helpful too. warning: USB_DEVICE_PID (defined at subsys/usb/Kconfig:27) was assigned the value '0x1234' but got the value ''. Check these unsatisfied dependencies: USB_DEVICE_STACK (=n). See http://docs.zephyrproject.org/latest/reference/kconfig/CONFIG_USB_DEVICE_PID.html and/or look up USB_DEVICE_PID in the menuconfig/guiconfig interface. The Application Development Primer, Setting Configuration Values, and Kconfig - Tips and Best Practices sections of the manual might be helpful too. warning: USB_DEVICE_PRODUCT (defined at subsys/usb/Kconfig:39) was assigned the value 'Zephyr CDC SensorTile.box' but got the value ''. Check these unsatisfied dependencies: USB_DEVICE_STACK (=n). See http://docs.zephyrproject.org/latest/reference/kconfig/CONFIG_USB_DEVICE_PRODUCT.html and/or look up USB_DEVICE_PRODUCT in the menuconfig/guiconfig interface. The Application Development Primer, Setting Configuration Values, and Kconfig - Tips and Best Practices sections of the manual might be helpful too.

that's not a good user experience.

The user must be very familiar with Zephyr to spot that this line

zephyr/samples/boards/sensortile_box/sample.yaml

Line 10 in 461b174

extra_args: SHIELD="cdc_acm_console"

indicates this sample must be build with:

-DSHIELD=cdc_acm_console

No, they would look in samples/boards/sensortile_box/README.rst, preferably before, and see how to build it. 461b174#diff-9b13c471892c9de2d5bb33dce8bfaf55dd08edb938cb2ed9642f77bb32496000R40

but having to apply an extra argument, in a place where app.overlay used to work, seems to be a step backward.

A sample should preferable be able to built in normal mode without extra args.

tejlmand · 2021-11-16T13:00:48Z

boards/arm/arduino_nano_33_ble/Kconfig.board

@@ -5,16 +5,6 @@ config BOARD_ARDUINO_NANO_33_BLE
 	bool "Arduino Nano 33 BLE board"
 	depends on SOC_NRF52840_QIAA

-config BOARD_ARDUINO_NANO_33_BLE_EN_USB_CONSOLE


so all boards cleaned up in this commit 9c03497 must now use -DSHIELD=cdc_acm_console

That sounds like a step backward in user-friendliness if those boards are expected to always use the usb as console.

For example, this board bl654_usb doesn't seem to have any other option than using USB for console:
https://docs.zephyrproject.org/latest/boards/arm/bl654_usb/doc/bl654_usb.html

@kieran-mackey FYI.

No, it is a step forward, only with these patches it becomes possible to use CDC ACM UART (not usb) as backend for console. The configuration of this board in the tree is not usable.

The configuration of this board in the tree is not usable.

Which board or all boards ?

Read my comment:

so all boards cleaned up in this commit 9c03497 must now use -DSHIELD=cdc_acm_console

i'm refering to the entire commit 9c03497, not a single board, but github requires me to put the comment on a single line.
That is, my comment covers all those boards you are modifying in that commit:

arduino_nano_33_ble

bl654_usb

degu_evk

nrf52840dongle_nrf52840

are you saying that the current configuration in Zephyr of those boards are not working for any of them ?

as far as I'm aware, the general problem started with this cleanup:
d12331e
which again was fixing:
#38403
caused by this:
#37512

So maybe we should go back to the origin of the problem, instead of completely changing how users must use the boards.

jfischer-no · 2021-11-16T15:55:42Z

If not, then please make a proper description why the big rewrite to the system is needed.

Each commit has a description, please read through it. I see no reason to copy it again in PR description.

mbolivar-nordic · 2021-11-16T21:54:18Z

Each commit has a description, please read through it. I see no reason to copy it again in PR description.

A pull request description in GitHub is like the cover letter for a patch series sent via email. It is an overall justification for the work, along with an overview of the individual commits. It clearly has a purpose.

mbolivar-nordic

I understand that this is probably the easiest way to accomplish the goal of adding some Kconfig and DT overlays into the build using a single variable setting (-DSHIELD=...), but this is really a confusing abuse of the system.

I think what's going on here is that a "shield" is a special case of something more general that has what you want (@carlescufi 's "meta-config"), and you are using the existing mechanism because we haven't solved the real problem.

As discussed offline I think this is similar to the problem @JordanYates was trying to solve for mcuboot, which he discussed at the Zephyr Developers Summit (ZDS). We should try to solve the real problem, not just abuse shields, especially since there is also ongoing work to try to solve the multi-image build problem for the special case of mcuboot.

At ZDS, we discussed having an "include path" or a list of search directories for places that would contain both Kconfig and DTS files to apply to the build. Everyone agreed about the general idea, but nobody could agree what the call that thing, and we never agreed on anything. Let's try again; we have enough interested people that I hope we can do it this time.

tejlmand · 2021-11-17T08:53:24Z

If not, then please make a proper description why the big rewrite to the system is needed.

Each commit has a description, please read through it. I see no reason to copy it again in PR description.

I did, and for example I see this in one of the commit messages:

Add generic shields that provide devicetree and configuration
overlays, and configure USB device stack so that CDC ACM UART
can be used as backend for console, logging, or shell.
This approach allows us to avoid many identical overlays in
samples and tests directories, and cumbersome board configurations.

This talks about what you do, and a little why.
But it doesn't tell me how this relates to the problem this PR is solving, which is:
#40220 (comment)

Patches to resolve issues around Kconfig option USB_UART_CONSOLE.

To me the particular commit message I just posted seems to solve a very different issue imho.

jfischer-no · 2021-11-17T15:42:49Z

I understand that this is probably the easiest way to accomplish the goal of adding some Kconfig and DT overlays into the build using a single variable setting (-DSHIELD=...), but this is really a confusing abuse of the system.

I do not abuse anything, I am adding a shield for an emulated UART controller. It is also not an abuse to have emulated boards like qemu_cortex_m0.

I think what's going on here is that a "shield" is a special case of something more general that has what you want (@carlescufi 's "meta-config"), and you are using the existing mechanism because we haven't solved the real problem.

As discussed offline I think this is similar to the problem @JordanYates was trying to solve for mcuboot, which he discussed at the Zephyr Developers Summit (ZDS). We should try to solve the real problem, not just abuse shields, especially since there is also ongoing work to try to solve the multi-image build problem for the special case of mcuboot.

@mbolivar-nordic, @tejlmand, @carlescufi
You are now connecting this with something that is not even defined. Sorry, but this sounds like long discourse with little output, especially when it comes to multi-image or mcuboot. I have a concrete issue and a solution now. The alternative is then to change the boards without debug probe to minimal configuration (remove all USB device support configuration option) which is even a kind of tacit rule.

I opened #40428 to remove USB_UART_CONSOLE option.

jfischer-no requested review from carlescufi, anangl, jakub-uC and eobalski November 9, 2021 16:54

jfischer-no requested review from jhedberg, joerchan, nashif, nordic-krch and Vudentz as code owners November 9, 2021 16:54

github-actions bot added area: Documentation area: Samples Samples area: Tests Issues related to a particular existing or missing test labels Nov 9, 2021

jfischer-no self-assigned this Nov 9, 2021

carlescufi requested a review from tejlmand November 10, 2021 10:46

nordic-krch reviewed Nov 10, 2021

View reviewed changes

tejlmand reviewed Nov 11, 2021

View reviewed changes

jfischer-no force-pushed the pr-usb_uart_console branch from 8ba5c06 to 58060d9 Compare November 12, 2021 16:58

jfischer-no requested review from erwango, galak and ioannisg as code owners November 12, 2021 16:58

github-actions bot added the area: Devicetree label Nov 12, 2021

jfischer-no added 13 commits November 15, 2021 08:27

usb: add option to initialize USB device support at boot

9b88cba

Possibility to initialize USB device support is useful when only CDC ACM class is enabled and CDC ACM UART is used as backend for console, shell, or logging. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

usb: remove the remains of USB_UART_CONSOLE

105e3da

Since there are no more users and dependencies of Kconfig option USB_UART_CONSOLE in the tree, remove the remains. Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

shields: cdc_acm: add support for zephyr,bt-c2h-uart property

6c32698

Add support for chosen property "zephyr,bt-c2h-uart". Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

jfischer-no force-pushed the pr-usb_uart_console branch from fa8076d to 6c32698 Compare November 15, 2021 08:29

erwango reviewed Nov 15, 2021

View reviewed changes

carlescufi reviewed Nov 15, 2021

View reviewed changes

tejlmand requested changes Nov 16, 2021

View reviewed changes

joerchan removed their request for review November 16, 2021 13:07

mbolivar-nordic suggested changes Nov 16, 2021

View reviewed changes

jfischer-no mentioned this pull request Nov 17, 2021

usb: remove Kconfig option USB_UART_CONSOLE #40428

Merged

jfischer-no closed this Nov 17, 2021

This was referenced Nov 24, 2021

shields: add generic shield to use CDC ACM UART as backend #40645

Closed

boards: arm: bl654_usb: Fix USB enablement issues #39987

Closed

jfischer-no deleted the pr-usb_uart_console branch November 24, 2021 21:18

carlescufi mentioned this pull request Nov 25, 2021

cmake: introduce snippets for common configuration #40669

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

usb: patches to resolve issues around Kconfig option USB_UART_CONSOLE #40220

usb: patches to resolve issues around Kconfig option USB_UART_CONSOLE #40220

jfischer-no commented Nov 9, 2021 •

edited

Loading

jakub-uC commented Nov 10, 2021

nordic-krch Nov 10, 2021

jfischer-no Nov 12, 2021

nordic-krch Nov 10, 2021

jfischer-no Nov 11, 2021

jfischer-no Nov 12, 2021

tejlmand Nov 11, 2021

jfischer-no Nov 11, 2021

tejlmand Nov 16, 2021

tejlmand Nov 11, 2021

jfischer-no Nov 11, 2021

tejlmand Nov 16, 2021

jfischer-no Nov 16, 2021

jfischer-no commented Nov 11, 2021

erwango Nov 15, 2021

jfischer-no Nov 15, 2021

carlescufi left a comment •

edited

Loading

tejlmand left a comment •

edited

Loading

tejlmand Nov 16, 2021

tejlmand Nov 16, 2021

tejlmand Nov 16, 2021

jfischer-no Nov 16, 2021

tejlmand Nov 16, 2021

jfischer-no Nov 16, 2021

tejlmand Nov 17, 2021

tejlmand Nov 16, 2021

jfischer-no Nov 16, 2021

tejlmand Nov 17, 2021

tejlmand Nov 17, 2021

jfischer-no commented Nov 16, 2021

mbolivar-nordic commented Nov 16, 2021

mbolivar-nordic left a comment

tejlmand commented Nov 17, 2021 •

edited

Loading

jfischer-no commented Nov 17, 2021 •

edited

Loading

		This option is only used to enable relevant options in backend
		configuration (if any), and for backward compatibility.

		@@ -0,0 +1,64 @@
		.. _cdc_acm_shield:

		Generic shields for CDC ACM UART

usb: patches to resolve issues around Kconfig option USB_UART_CONSOLE #40220

usb: patches to resolve issues around Kconfig option USB_UART_CONSOLE #40220

Conversation

jfischer-no commented Nov 9, 2021 • edited Loading

jakub-uC commented Nov 10, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jfischer-no commented Nov 11, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

carlescufi left a comment • edited Loading

Choose a reason for hiding this comment

tejlmand left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jfischer-no commented Nov 16, 2021

mbolivar-nordic commented Nov 16, 2021

mbolivar-nordic left a comment

Choose a reason for hiding this comment

tejlmand commented Nov 17, 2021 • edited Loading

jfischer-no commented Nov 17, 2021 • edited Loading

jfischer-no commented Nov 9, 2021 •

edited

Loading

carlescufi left a comment •

edited

Loading

tejlmand left a comment •

edited

Loading

tejlmand commented Nov 17, 2021 •

edited

Loading

jfischer-no commented Nov 17, 2021 •

edited

Loading