arm: dts: remove "mclk" from AD7124 overlays #3001
Open
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
Signed-off-by: Jorge Marques <[email protected]>
Add if hdl tag is present in devicetree. Compile common defconfig and devicetrees. Signed-off-by: Jorge Marques <[email protected]>
False positives. Signed-off-by: Jorge Marques <[email protected]>
Reduce podman-bias, to avoid users installing it if already using other container-engine. Signed-off-by: Jorge Marques <[email protected]>
Add flow for rpi branches, with rpi defconfigs. For testing, it is usual to open a pr against a ci/* branch, while staging branches the pr is open against targets, in this case, just the base or source branch must have the ci files. Signed-off-by: Jorge Marques <[email protected]>
Signed-off-by: Jorge Marques <[email protected]>
fix path to point in the right folder for uploading artifacs. Signed-off-by: Andreea Andrisan <[email protected]>
create a version file that includes information useful for the boot partition folder. Signed-off-by: Andreea Andrisan <[email protected]>
Signed-off-by: Jorge Marques <[email protected]>
Signed-off-by: Jorge Marques <[email protected]>
So we can show a pretty red X on the GUI, by returning error codes from the bash methods, beyond the fail environment variable. Signed-off-by: Jorge Marques <[email protected]>
Meant to fail individual steps, can be captured with || true Signed-off-by: Jorge Marques <[email protected]>
About merging err and fail into one for individual step failure. Signed-off-by: Jorge Marques <[email protected]>
Use trap ERR to catch any exit error and set the step as fail. Signed-off-by: Jorge Marques <[email protected]>
We already have our own CI stuff in place so remove the one we get from upstream RPI. We might want to take some ideas from it and adapt to our own workflow but for now, let's just remove the whole thing. The dtoverlaycheck also looks interesting but before it's useful, we have to fix all our overlays to comply with it. Signed-off-by: Nuno Sá <[email protected]>
Clear reset status flag, to keep error status register clean after reset (ad3552r manual, rev B table 38). Reset error flag was left to 1, so debugging registers, the "Error Status Register" was dirty (0x01). It is important to clear this bit, so if there is any reset event over normal working mode, it is possible to detect it. Fixes: 8f2b548 ("drivers:iio:dac: Add AD3552R driver support") Signed-off-by: Angelo Dureghello <[email protected]> Link: https://patch.msgid.link/20250125-wip-bl-ad3552r-clear-reset-v2-1-aa3a27f3ff8c@baylibre.com Cc: <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit e17b9f2)
Apply a mask to the raw value received over the SPI bus for unsigned
direct reads. As we found recently, SPI controllers may not set unused
bits to 0 when reading with bits_per_word != {8,16,32}. The ad7944 uses
bits_per_word of 14 and 18, so we need to mask the value to be sure we
returning the correct value to userspace during a direct read.
Fixes: d1efcf8 ("iio: adc: ad7944: add driver for AD7944/AD7985/AD7986")
Signed-off-by: David Lechner <[email protected]>
Reviewed-by: Nuno Sá <[email protected]>
Link: https://patch.msgid.link/20250505-iio-adc-ad7944-max-high-bits-on-direct-read-v1-1-b173facceefe@baylibre.com
Cc: <[email protected]>
Signed-off-by: Jonathan Cameron <[email protected]>
(cherry picked from commit 7cdfbc0)
The sinc4 filter has a factor 0.23 between Output Data Rate and f_{3dB}
and for sinc3 the factor is 0.272 according to the data sheets for
ad7124-4 (Rev. E.) and ad7124-8 (Rev. F).
Fixes: cef2760 ("iio: adc: ad7124: add 3db filter")
Signed-off-by: Uwe Kleine-König <[email protected]>
Reviewed-by: Marcelo Schmitt <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jonathan Cameron <[email protected]>
(cherry picked from commit 8712e49)
There are several issues with the function that implements writing to the filter_low_pass_3db_frequency property: - The sinc3 factor should be 0.272 not 0.262 (this is fixed for the reading side in the previous patch). - For freq > 1 the if condition is always true so the sinc4 filter is hardly ever chosen. - In the nearly always taken if branch the filter is set to sinc3, but the frequency is set for sinc4. (And vice versa in the else branch.) This is broken enough to justify the claim that there isn't any serious user. Also it it counter-intuitive that setting the 3db frequency modifies the sample frequency and the filter type. So drop the ability to write that property. Signed-off-by: Uwe Kleine-König <[email protected]> Reviewed-by: Marcelo Schmitt <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit 7dd17a4)
Some drivers are tristates of other symbols, for example:
config DEVICE
tristate "Device support"
config DEVICE_SPI
depends on SPI
selects DEVICE
config DEVICE_I2C
depends on I2C
selects DEVICE
In this case, touching device.c, infers CONFIG_DEVICE, but this symbol
is selected by DEVICE_SPI | DEVICE_I2C. Extend the inferring script to
look for symbols that selects it, limited at the same Kconfig.
Signed-off-by: Jorge Marques <[email protected]>
Collect compile_gcc_fanalyzer errors to warnings bucket, to not fail the CI, mostly for existing errors not worth fixing. Signed-off-by: Jorge Marques <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace direct access with the iio_get_masklength() helper function to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace direct access with the iio_get_masklength() helper function to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace direct access with the iio_get_masklength() helper function to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace direct access with the iio_get_masklength() helper function to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace the for_each_set_bit loop with the iio_for_each_active_channel() helper function and direct access to indio_dev->masklength with iio_get_masklength() to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace the for_each_set_bit loop with the iio_for_each_active_channel() helper function to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace the for_each_set_bit loop with the iio_for_each_active_channel() helper function and direct access to indio_dev->masklength with iio_get_masklength() to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
Upstream has removed direct access to indio_dev->masklength. Replace direct access with the iio_get_masklength() helper function to align with this change and to avoid warnings from the sparse checker. Signed-off-by: Nuno Sá <[email protected]>
The format property set to 'spdif' has no meaning on these DTs as that interface type is handled by the generic hdmi-sound driver. Signed-off-by: Nuno Sá <[email protected]>
Add a boolean debugfs attribute to allow disabling the SINGLE_CYCLE bit in the FILTER registers. This causes data to be read on every conversion instead of doing the usual 3 or 4 conversions per sample (depending on the filter). This is only needed for very specific use cases, such as validating the performance of the ADC. So we just expose this feature through debugfs for the rare cases where it is needed by people who really know what they are doing. Signed-off-by: David Lechner <[email protected]> Reviewed-by: Nuno Sá <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit 017c7fa)
Inline ad7124_enable_channel() at the only call site. This simplifies the code by avoiding a bit of extra indirection. ch->nr is replaced by address as that is the same value and avoids more indirection. Signed-off-by: David Lechner <[email protected]> Reviewed-by: Nuno Sá <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit 021eef2)
Remove the unused `nr` field from the `ad7124_channel` struct. There are no more users of this field (it is only assigned to but never read) so can be removed. Signed-off-by: David Lechner <[email protected]> Reviewed-by: Nuno Sá <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit bd93170)
Use AD7124_MAX_CHANNELS macro instead of hardcoding 16 in ad7124_disable_all(). We already have the macro, so we should use it. Signed-off-by: David Lechner <[email protected]> Reviewed-by: Nuno Sá <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit f6fbdbc)
Use devm_mutex_init() to initialize the mutex to handle automatically freeing in debug builds. Signed-off-by: David Lechner <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit 9bbf038)
Remove __ad7124_set_channel() wrapper function. This just added an unnecessary layer of indirection with an extra call to container_of(). Signed-off-by: David Lechner <[email protected]> Reviewed-by: Nuno Sá <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit 6e83fb6)
Use the same .ext_info for the temperature channel as for the voltage channels. In the ADC, these all go though a mux to select the source and otherwise operate the same. These attributes probably won't be used much, but since it is trivial to add this, we might as well include them. Signed-off-by: David Lechner <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit bcdb38b)
Change the allocation strategy of the 8 SETUP registers from a least- recently-used (LRU) to a first-come-first-served basis. The AD7124 chips can have up to 16 channels enabled at a time in the sequencer for buffered reads, but only have 8 SETUP configurations (namely the OFFSET, GAIN, CONFIG and FILTER registers) that must be shared among the 16 channels. This means some of the channels must use the exact same configuration parameters so that they can share a single SETUP group of registers. The previous LRU strategy did not keep track of how many different configurations were requested at the same time, so if there were more than 8 different configurations requested, some channels would end up using the incorrect configuration because the slot assigned to them would also be assigned to a different configuration that wrote over it later. Adding such tracking to solve this would make an already complex algorithm even more complex. Instead we can replace it with a simpler first-come-first-serve strategy. This makes it easy to track how many different configurations are being requested at the same time. This comes at the expense of slightly longer setup times for buffered reads since all setup registers must be written each time when a buffered read is enabled. But this is generally not considered a hot path where performance is critical, so should be acceptable. This new strategy also makes hardware debugging easier since SETUPs are now assigned in a deterministic manner and in a logical order. Signed-off-by: David Lechner <[email protected]> Reviewed-by: Nuno Sá <[email protected]> Signed-off-by: Jonathan Cameron <[email protected]> (cherry picked from commit 0a77002)
Add a new .dts file for the Analog Devices EVAL-AD7124-4ASDZ board connected to a DE10-Nano board. Signed-off-by: David Lechner <[email protected]> (cherry picked from commit a3463e1)
- Fix channel matching pattern. - Fix adi,power-up-frequency data type. - Remove unnecessary maxItems constraints. Signed-off-by: David Lechner <[email protected]> (cherry picked from commit d58704c)
Add guard(mutex) to ad9081_status_show() and ad9081_fir_bin_write() to ensure proper locking when accessing converter state. This prevents potential race conditions during debugfs interactions. Signed-off-by: Michael Hennerich <[email protected]>
Ensure the return value of sysfs_emit() is properly checked in ad9081_ext_info_read() to catch cases where no data is written or an error occurs. This prevents misleading success returns when the buffer emission fails. Set ret = -EINVAL if sysfs_emit() returns zero or a negative value, aligning with expected error handling practices in sysfs attribute reads. Signed-off-by: Michael Hennerich <[email protected]>
Convert all instances of for_each_child_of_node() to for_each_child_of_node_scoped() in the ad9081_parse_dt_tx() and ad9081_parse_dt_rx() functions. This ensures proper reference counting and automatic cleanup of child nodes, reducing the risk of memory leaks or dangling references. This change also removes redundant of_node_put() calls for child nodes, as for_each_child_of_node_scoped() handles cleanup automatically. Signed-off-by: Michael Hennerich <[email protected]>
The sample rate configuration logic following the call to ad9081_set_sample_rate() was unreachable due to an early return. This patch removes the dead code block that attempted to round and set the clock rate, which is now obsolete and no longer executed. Cleaning up this unreachable code improves readability and avoids confusion about the actual sample rate handling path. Signed-off-by: Michael Hennerich <[email protected]>
The intention of the commit 3580444 was to also upload the .dtb files present in overlays/ folder. By using .dtb* regex, unnecessary files found their way into the artifacts. Use explicit regex in order to upload only files ending in .dtb or .dtbo. Signed-off-by: Alisa-Dariana Roman <[email protected]>
Add item to emphasise that the user should compile their changes, including documentation. Remove "(if there is the case)", rely on common-sense. Add item to require to add links for lore, when applicable. Signed-off-by: Jorge Marques <[email protected]>
The upstream devicetree binding deprecated using `clock-names = "mclk";` as this is an internal clock, not an external one. Remove this from the AD7124 overlays since the ad7124 driver no longer requires it. Also use generic node names for the adc nodes while we are cleaning things up. Signed-off-by: David Lechner <[email protected]>
dlech
requested review from
CiprianRegus,
amiclaus,
btogorean,
buha,
chegbeli,
danmois,
dbogdan,
machschmitt,
mhennerich,
nunojsa and
rbolboac
as code owners
|
@dlech, Unfortunately we had to force push the rpi branch (I know, I know...) . Could you please update this PR on top of it? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
16 participants
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.
We recently did some backports from upstream for the ad7124 driver. These got automatically copied to the rpi-6.12.y branch. However, there are some devicetree overlays that only exist in the rpi branch. So we need one more patch to fix those up.
PR Description
The upstream devicetree binding deprecated using
clock-names = "mclk";as this is an internal clock, not an external one. Remove this from the AD7124 overlays since the ad7124 driver no longer requires it.Also use generic node names for the adc nodes while we are cleaning things up.
PR Type
PR Checklist