drivers/qma6100p: add QST QMA6100P 3-axis accelerometer driver by leduclean · Pull Request #22352 · RIOT-OS/RIOT

leduclean · 2026-06-06T16:15:34Z

Add a driver for the QST QMA6100P 3-axis accelerometer over I2C.

This driver was needed to support the Seeed Studio SenseCAP T1000-E tracker,
which embeds a QMA6100P as its motion sensor.

Features:

Configurable full-scale range, output data rate and power mode
Acceleration readout on all three axes
Data-ready interrupt support
SAUL integration
Kconfig configuration

Includes a test application under tests/drivers/qma6100p covering basic
readout and the data-ready interrupt.

Testing procedure

Wire a QMA6100P to a board exposing an I2C bus (e.g. the Seeed Studio SenseCAP
T1000-E), then build and flash the test application (adjust BOARD and the I2C
parameters as needed):

BOARD=<your-board> make -C tests/drivers/qma6100p flash term

The test:

Initializes the device over I2C.
Runs a data-ready interrupt rate sweep at 12.5 / 25 / 100 Hz, measuring the
ISR frequency and asserting it matches the configured ODR.
Streams samples: the data-ready ISR signals a reader thread that reads each
sample over I2C.

Expected output:

=== QMA6100P accelerometer driver test ===
[init] I2C_DEV(0) addr 0x12 ... OK

--- data-ready interrupt rate sweep ---
[ODR 1/3] set rate -> 12 Hz ... OK
  try 1/3: measured 12 Hz (expect ~12) -> PASS
[ODR 1/3] PASS (~12 Hz)
[ODR 2/3] set rate -> 25 Hz ... OK
  try 1/3: measured 25 Hz (expect ~25) -> PASS
[ODR 2/3] PASS (~25 Hz)
[ODR 3/3] set rate -> 100 Hz ... OK
  try 1/3: measured 100 Hz (expect ~100) -> PASS
[ODR 3/3] PASS (~100 Hz)

=== ALL TESTS PASSED ===

--- interrupt-driven streaming ---
[data] x=... y=... z=... ug
...

Issues/PRs references

#22267

Useful link

QMA6100P spec

Declaration of AI Tools / LLMs usage

Claude Code for code review and guidance, with user authorship of the implementation

Co-authored-by: Baptiste Le Duc baptiste.leduc@etik.com
Signed-off-by: Léandre Le Duc leandre.leduc38@gmail.com

Add a driver for the QST QMA6100P 3-axis accelerometer over I2C. Features: - Configurable full-scale range, output data rate and power mode - Acceleration readout on all three axes - Data-ready interrupt support - SAUL integration - Kconfig configuration Includes a test application under tests/drivers/qma6100p covering basic readout and the data-ready interrupt. Co-authored-by: Léandre Le Duc <leandre.leduc38@gmail.com> Signed-off-by: Baptiste Le Duc <baptiste.leduc@etik.com>

github-actions · 2026-06-06T16:15:43Z

Hey @leduclean, thank you for your first contribution! We really appreciate it! If you haven't already, please take a look at our contributing guidelines before the review process starts. Also, due to how the GitHub review system works, please avoid force-pushing or squashing your commits unless asked to by a maintainer (or unless your commit is still in "draft commit" stage). Your pull request will be reviewed as soon as possible.

riot-ci · 2026-06-06T21:32:03Z

Murdock results

❌ FAILED

f13dc60 drivers/qma6100p: add QST QMA6100P 3-axis accelerometer driver

Success	Failures	Total	Runtime
3173	0	10422	04m:08s

Artifacts

AnnsAnns · 2026-06-07T15:08:54Z

First Sanji and now Foxy, seems like there is a theme here 😛

Thank you for your PR, it seems like there are still some issues you need to fix to get through our CI, see here: https://ci.riot-os.org/details/b17c1a96704744cbb4adb7763282d497

I'd presume that @baptleduc can vouch for the functionality of this?

baptleduc · 2026-06-07T15:12:13Z

First Sanji and now Foxy, seems like there is a theme here 😛

Thank you for your PR, it seems like there are still some issues you need to fix to get through our CI, see here: https://ci.riot-os.org/details/b17c1a96704744cbb4adb7763282d497

I'd presume that @baptleduc can vouch for the functionality of this?

Ahah brothers fans of one piece 😄

Yes I will fix it but @leduclean already took a look and he struggled to find what was the real issue

baptleduc · 2026-06-07T15:26:13Z

It is weird because the murdock error indicates it is because of map_regs variable that could be used uninitialized in qma6100p_set_data_ready_interrupt() while there are no branches where it could be the case.

Could you please help us to understand this ?

Thanks !

AnnsAnns · 2026-06-07T17:57:57Z

Hmmm does this also happen when you run the test locally or only on murdock?

crasbe · 2026-06-07T18:33:03Z

The compiler complains about the use of map_reg in the function qma6100p_set_data_ready_int. In fact, the function _set_intpin_conf won't set the value of map_reg in the default case.

It is possible that the compiler can't resolve the goto conditions properly and therefore assumes that map_reg might be used uninitialized.

In general, I think the use of goto and the READ_REG macros that also get the goto label as a parameter is not really a good style tbh.

AnnsAnns · 2026-06-07T18:49:23Z

It is possible that the compiler can't resolve the goto conditions properly

The compiler once it sees goto

baptleduc · 2026-06-07T19:24:11Z

In general, I think the use of goto and the READ_REG macros that also get the goto label as a parameter is not really a good style tbh.

You are right, but where do you see such usage ? You mean that we shouldn't pass the goto label as parameter of READ_REG macros ? So it will make the READ_REG/WRITE_REG macro useless

benpicco · 2026-06-08T13:02:44Z

+    (void)args;
+    irq_count++;
+    if (reader_pid != KERNEL_PID_UNDEF) {
+        thread_flags_set(thread_get(reader_pid), FLAG_DATA_READY);


Why not just use a mutex?

benpicco · 2026-06-08T13:06:00Z

+ * @param[in]  dev          device descriptor of accelerometer
+ * @param[out] data         raw ADC counts per axis
+ *
+ * @return                  QMA6100P_NODATA if nothing has changed and keep data unchanged.


QMA6100P_NODATA is not defined

benpicco · 2026-06-08T13:11:09Z

+    if (res == -ENXIO) {
+        return QMA6100P_NODEV;
+    }
+    if (res == -EINVAL) {
+        return QMA6100P_INVALID_ARG;


Why use custom error codes here?

It is a wrapper as one of the maintainer suggest when I did the ads1x1x driver

benpicco · 2026-06-08T13:11:49Z

+    /* Wait for OTP to load */
+    do {
+        READ_REG(QMA6100P_REG_NVM, nvm_status, out);
+    } while ((nvm_status & (QMA6100P_NVM_LOAD_DONE | QMA6100P_NVM_RDY)) !=


better add a limit on the number of retries

benpicco · 2026-06-08T13:20:27Z

+ * @brief   Interrupt configuration parameters
+ */
+typedef struct {
+    gpio_t interrupt_pin;                         /**< MCU GPIO connected to the QMA6100P INT pin */


This should be part of qma6100p_params_t

benpicco · 2026-06-08T13:20:49Z

+typedef struct {
+    gpio_t interrupt_pin;                         /**< MCU GPIO connected to the QMA6100P INT pin */
+    qma6100p_int_active_level_t active_level_int; /**< active level of INT pin */
+    qma6100p_int_pin_mode_t pin_mode_int;         /**< INT pin output mode */


Why would you need this to be configurable?

benpicco · 2026-06-08T13:21:00Z

+    gpio_t interrupt_pin;                         /**< MCU GPIO connected to the QMA6100P INT pin */
+    qma6100p_int_active_level_t active_level_int; /**< active level of INT pin */
+    qma6100p_int_pin_mode_t pin_mode_int;         /**< INT pin output mode */
+    qma6100p_int_shadow_t interrupt_shadow;       /**< shadow mode */


What does this mean?

benpicco · 2026-06-08T13:21:41Z

+    qma6100p_int_active_level_t active_level_int; /**< active level of INT pin */
+    qma6100p_int_pin_mode_t pin_mode_int;         /**< INT pin output mode */
+    qma6100p_int_shadow_t interrupt_shadow;       /**< shadow mode */
+    qma6100p_int_pin_num_t interrupt_pin_num;     /**< QMA6100P INT pin routed on the board */


I guess that means you can have multiple interrupt pins for different events?

benpicco · 2026-06-08T13:26:05Z

+ *
+ * Work mode is controlled by PM_REGISTER (0x11) and can be set through I2C commands
+ *
+ * @warning By default, QMA6100P is in intermediate state after power on but it shouldn't be keep for any applications


Then there should only be a qma6100p_set_low_power(bool) right?

benpicco · 2026-06-08T13:29:15Z

+ *
+ * @warning The callback is invoked from interrupt context, keep it short
+ */
+int qma6100p_set_data_ready_int(qma6100p_t *dev, const qma6100p_int_t *interrupt);


I guess the more idiomatic way would be

Suggested change

int qma6100p_set_data_ready_int(qma6100p_t *dev, const qma6100p_int_t *interrupt);

int qma6100p_set_data_ready_int(qma6100p_t *dev, const qma6100p_int_params_t *config, qma6100p_int_cb_t cb, void *arg);

benpicco · 2026-06-08T13:30:09Z

+ */
+typedef enum {
+    QMA6100P_INT_CFG_SHADOW_EN = 0,  /**< shadowing enabled (default) */
+    QMA6100P_INT_CFG_SHADOW_DIS = 1, /**< shadowing disabled */


What does that even mean?

leduclean requested review from MichelRottleuthner, MrKevinWeiss, aabadie, jia200x and leandrolanzieri as code owners June 6, 2026 16:15

github-actions Bot added Area: tests Area: tests and testing framework Area: drivers Area: Device drivers Area: SAUL Area: Sensor/Actuator Uber Layer Area: Kconfig Area: Kconfig integration labels Jun 6, 2026

crasbe added Type: new feature The issue requests / The PR implemements a new feature for RIOT CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR labels Jun 6, 2026

benpicco reviewed Jun 8, 2026

View reviewed changes

	int qma6100p_set_data_ready_int(qma6100p_t dev, const qma6100p_int_t interrupt);
	int qma6100p_set_data_ready_int(qma6100p_t dev, const qma6100p_int_params_t config, qma6100p_int_cb_t cb, void *arg);

Conversation

leduclean commented Jun 6, 2026

Testing procedure

Issues/PRs references

Useful link

Declaration of AI Tools / LLMs usage

Uh oh!

github-actions Bot commented Jun 6, 2026

Uh oh!

riot-ci commented Jun 6, 2026

Murdock results

Artifacts

Uh oh!

AnnsAnns commented Jun 7, 2026

Uh oh!

baptleduc commented Jun 7, 2026

Uh oh!

baptleduc commented Jun 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

AnnsAnns commented Jun 7, 2026

Uh oh!

crasbe commented Jun 7, 2026

Uh oh!

AnnsAnns commented Jun 7, 2026

Uh oh!

baptleduc commented Jun 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

baptleduc commented Jun 7, 2026 •

edited

Loading

baptleduc commented Jun 7, 2026 •

edited

Loading