nRF GPIO and SPI PM_DEVICE_RUNTIME #89602
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Implement PM_DEVICE_RUNTIME for gpio_nrfx.c PM is required to request/release the GPD domain if such exists for a given GPIO instance. Without this, every single API call will result in a requesting the GPD domain, which makes it slow and unusable from ISRs, given GPD requires IPC. Now, user can simply get the GPIO before using it, and put it once done, calling the APIs from ISR while resumed. Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
github-actions
bot
requested review from
anangl,
henrikbrixandersen,
jaz1-nordic,
kl-cruz,
magp-nordic,
masz-nordic,
mnkp,
nika-nordic,
tbursztyka and
teburd
Introduce spi_context_cs_get() and spi_context_cs_put() which shall be used from drivers to get/put the GPIO port the CS GPIO belongs to before and after a transaction, in line with the SPI drivers pm action hook being called. Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
Implement pm_device_runtime support to spi_loopback test suite. Specifically: - call pm_device_runtime_get() on SPI bus before using SPI bus - call pm_device_runtime_put() once SPI bus is no longer needed. This is after transceive has been performed, or after calling spi_release_dt() after a transceive has been performed. This addition has no affect if PM_DEVICE_RUNTIME is not enabled on a board/test case basis. Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
bjarki-andreasen
force-pushed
the
gpio-nrfx-pm-device
branch
from
a6d1c41 to
eb3d8e3
Compare
|
ping @anangl |
anangl
reviewed
May 14, 2025
| @@ -402,7 +401,7 @@ static void finish_transaction(const struct device *dev, int error) | |||
| spi_context_complete(ctx, dev, error); | |||
| dev_data->busy = false; | |||
|
|
|||
| finalize_spi_transaction(dev, true); | |||
| finalize_spi_transaction(dev, (dev_data->ctx.config->operation & SPI_HOLD_ON_CS) > 0); | |||
There was a problem hiding this comment.
Changes related to pm_device_runtime_put_async() should be done in a separate commit, with some explanation in the commit message.
There was a problem hiding this comment.
deal, I already forgot exactly why it was moved here :D
There was a problem hiding this comment.
moved to separate commit and added comment to driver
Implement calling spi_context_cs_get_all() and spi_context_cs_put_all() in line with pm resume/suspend. Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
In case the SPI transaction has SPI_HOLD_ON_CS set, we need to keep SPI resumed until spi_release() is called. This is required as we now need to keep the CS GPIO port resumed until transaction is complete. Suspending CS GPIO is not allowed from ISR in some cases (H20 fast GPIO instance) so we have to defer CS GPIO suspend to some thread context (put_async or spi_release()). Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
Add CS GPIO to nrf54h20dk/nrf54h20/cpuapp which is connected to PORT which is in the "fast domain". The fast domain needs to be resumed from thread context given it requires IPC communication. This thus ensures that GPIO can be used with and without PM_DEVICE_RUNTIME enabled. Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
bjarki-andreasen
force-pushed
the
gpio-nrfx-pm-device
branch
from
eb3d8e3 to
1e71b19
Compare
|
anangl
approved these changes
May 14, 2025
|
ping @masz-nordic :) |
carlescufi
approved these changes
May 14, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Problem description:
On the nRFH20, some GPIOs are on a power domain. This power domain has to be resumed to use the GPIOs, and resuming them has to be done through IPC, thus, not from ISRs. The current solution just to use the GPIOs is to internally resume the domain from the driver, set/clear the pin, then suspend it again. This is slow and has to be done from thread context.
So, if we use SPI, and the CS is a GPIO on one of these special ports, we can't control the CS GPIO from the SPI ISR.
The solution:
Implement PM_DEVICE_RUNTIME in GPIO driver, so port can be resumed once, from thread, then used from ISR with no IPC calls needed, then suspend port. Additonally, extend SPI driver to resume the port when it is resumed, and suspend once suspended, like a bus. Lastly, extend SPI test suite to support PM_DEVICE_RUNTIME.
Now, if PM_DEVICE_RUNTIME is not used, GPIO power domain is resumed on init, and kept on. if PM_DEVICE_RUNTIME is enabled, user can get the SPI, which gets the GPIO, which gets the GPIO power domain, before the transaction.
This touches the spi_context and spi_loopback test hence why the PR is not isolated to nordic :)