Fix LPUART1 ISR priority override causing FreeRTOS assert in boost mode by Copilot · Pull Request #7 · electricalhog/PD_Charger

Copilot · 2026-03-11T19:19:51Z

HW_TRACER_EMB_StartRX() hardcodes LPUART1 NVIC priority to 3, silently overriding the compliant priority-5 set in HW_TRACER_EMB_Init(). When LPUART1 RXNE fires (e.g. STM32CubeMonitor response during active USB-PD trace traffic), GUI_CALLBACK_RX() calls osMessageQueuePut() from priority-3 context — below configLIBRARY_MAX_SYSCALL_INTERRUPT_PRIORITY = 5 — triggering vPortValidateInterruptPriority assert → taskDISABLE_INTERRUPTS() + for(;;). In boost mode this manifests as a hung default task with HRTIM still switching (priority 1 unaffected by BASEPRI=0x50); in buck mode there is no inbound LPUART traffic so the assert never fires and the system runs normally.

Changes

tracer_emb_hw.c — HW_TRACER_EMB_StartRX(): replace hardcoded priority 3 with TRACER_EMB_TX_IRQ_PRIORITY (defined as 5 in tracer_emb_conf.h), consistent with the priority already set in HW_TRACER_EMB_Init().

// Before
NVIC_SetPriority(TRACER_EMB_USART_IRQ, 3);

// After
NVIC_SetPriority(TRACER_EMB_USART_IRQ, TRACER_EMB_TX_IRQ_PRIORITY);

TRACER_EMB_TX_IRQ_PRIORITY = 5 sits exactly at the FreeRTOS syscall ceiling, so osMessageQueuePut() in GUI_CALLBACK_RX() is now called from a compliant ISR priority in both modes.

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

…, adc_monitor, pd_interface modules) Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

… constant, HRTIM reconfigure) Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

electricalhog · 2026-03-14T18:04:40Z

STM32CubeIDE/final/Core/Inc/regulator_config.h

+#define PIN_OUTPUT_DIS_PIN GPIO_PIN_9
+
+/* =========================================================================
+ * SECTION 6: HRTIM TIMING CONSTANTS


Timer constants should be compiled from time-based calculations instead of assuming a period count

electricalhog · 2026-03-14T18:23:31Z

STM32CubeIDE/final/Core/Inc/adc_monitor.h

+ * Called from HAL_ADC_ConvCpltCallback.  Reads adc1_dma_buffer and updates
+ * adc_measurements.v_out_mv, i_inductor_ma, i_out_ma, i_in_ma.
+ *
+ * May also be called directly from the PID ISR for a synchronous read.


nice feature for grabbing vales off the buffer

electricalhog · 2026-03-14T18:24:18Z

STM32CubeIDE/final/Core/Inc/pid_controller.h

+/**
+ * PidConfig — tuning coefficients and execution parameters.
+ *
+ * These values are runtime-mutable (§8.7): they may be written from the
+ * debug interface without stopping the regulator.  The PID ISR reads them
+ * on every execution cycle.
+ */
+typedef struct
+{
+    float kp;           /**< Proportional gain. Default 1.0 (§8.7). */
+    float ki;           /**< Integral gain.     Default 100.0 (§8.7). */
+    float kd;           /**< Derivative gain.   Default 0.0 (§8.7). */
+    float dt_seconds;   /**< Execution period in seconds. Default 50 µs. */
+    float output_min;   /**< Minimum clamped output (PID_OUTPUT_MIN). */
+    float output_max;   /**< Maximum clamped output (PID_OUTPUT_MAX). */
+} PidConfig;


nicely placed at the top

electricalhog · 2026-03-14T18:26:04Z

STM32CubeIDE/final/Core/Inc/regulator.h

+ *   1. Reconfigure HRTIM output fault levels to INACTIVE (all FETs off on
+ *      fault).  CubeMX sets FaultLevel = NONE; this corrects it (§5.5).
+ *   2. Enable FLT1 and FLT2 fault response on Timer A and Timer B.
+ *   3. Configure backstop Compare 1 registers on Timer A and Timer B (§10.3).


update compare for blanking and backstop

electricalhog · 2026-03-14T18:27:03Z

STM32CubeIDE/final/Core/Inc/slope_comp.h

+typedef enum
+{
+    SLOPE_COMP_MODE_BUCK  = 0,  /**< Buck: downslope = V_out / L */
+    SLOPE_COMP_MODE_BOOST = 1   /**< Boost: downslope = (V_out − V_in) / L */
+} SlopeCompMode;


could reference the other mode type

electricalhog · 2026-03-14T18:31:15Z

STM32CubeIDE/final/Core/Src/regulator.c

+    if (regulator_mode == REGULATOR_MODE_BUCK)
+    {
+        hrtim_apply_buck_mode_static_leg();
+    }
+    else
+    {
+        hrtim_apply_boost_mode_static_leg();
+    }


needs future-proofing for buck-boost

electricalhog · 2026-03-14T18:32:05Z

STM32CubeIDE/final/Core/Src/regulator.c

+        HAL_HRTIM_WaveformOutputStart(&hhrtim1,
+                                       HRTIM_OUTPUT_TB1 | HRTIM_OUTPUT_TB2);
+    }
+    else /* BOOST */


not always boost, may be 4-switch later

electricalhog · 2026-03-14T18:34:55Z

STM32CubeIDE/final/Core/Src/regulator.c

I'd like to implement external event blanking using compare 1, so increasing index of backstop and bootstrap is necessary.

…log ring buffer Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

…LPUART1/DMA 3/0→5 Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

… immediate VIN_RANGE fault Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Initial plan

5712cfc

Copilot AI assigned Copilot and electricalhog Mar 11, 2026

Copilot started work on behalf of electricalhog March 11, 2026 19:19 View session

Copilot AI and others added 2 commits March 11, 2026 19:46

feat: implement NLSpec regulator firmware (regulator, pid, slope_comp…

ad35e86

…, adc_monitor, pd_interface modules) Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

fix: address code review issues (slope arg, mode transition, inductor…

739d6c2

… constant, HRTIM reconfigure) Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] Implement NLSpec based on generated code and project spec~~ NLSpec Implementation: Buck-Boost Regulator Firmware Mar 11, 2026

Copilot finished work on behalf of electricalhog March 11, 2026 19:51

electricalhog requested a review from TG-Techie March 11, 2026 20:32

fixes for build

892ab17

Copilot started work on behalf of electricalhog March 13, 2026 18:13 View session

Bootstrap reset design, TIM7 IRQ fix, default regulation test task

51528f1

Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Copilot AI changed the title ~~NLSpec Implementation: Buck-Boost Regulator Firmware~~ Bootstrap reset, TIM7 IRQ fix, HRTIM bit corrections, default test task Mar 13, 2026

Copilot AI requested a review from electricalhog March 13, 2026 18:36

Copilot finished work on behalf of electricalhog March 13, 2026 18:36

electricalhog added 3 commits March 13, 2026 17:20

debug

ed42e93

switch to cmake

6ee9b46

timer blanking and backstop

7ed4761

electricalhog reviewed Mar 14, 2026

View reviewed changes

Copilot started work on behalf of electricalhog March 14, 2026 18:44 View session

Reassign HRTIM CMP1=blanking/CMP2=backstop/CMP3=bootstrap; add debug_…

8207232

…log ring buffer Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Copilot AI changed the title ~~Bootstrap reset, TIM7 IRQ fix, HRTIM bit corrections, default test task~~ HRTIM CMP reassignment (blanking/backstop/bootstrap), debug_log ring buffer, fix CMake sources Mar 14, 2026

Copilot AI requested a review from electricalhog March 14, 2026 19:10

Copilot finished work on behalf of electricalhog March 14, 2026 19:10

Copilot started work on behalf of electricalhog March 14, 2026 21:19 View session

Fix FreeRTOS ISR priority violations: UCPD1 priority 4→5, TRACER_EMB …

724d785

…LPUART1/DMA 3/0→5 Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Copilot AI changed the title ~~HRTIM CMP reassignment (blanking/backstop/bootstrap), debug_log ring buffer, fix CMake sources~~ Fix FreeRTOS ISR priority violations causing configASSERT at port.c:754 Mar 14, 2026

Copilot finished work on behalf of electricalhog March 14, 2026 21:30

Copilot started work on behalf of electricalhog March 15, 2026 00:36 View session

Fix stale v_in_mv=0 in regulator_start() causing wrong BOOST mode and…

7bca63d

… immediate VIN_RANGE fault Co-authored-by: electricalhog <15219301+electricalhog@users.noreply.github.com>

Copilot AI changed the title ~~Fix FreeRTOS ISR priority violations causing configASSERT at port.c:754~~ Fix wrong BOOST mode selection on first regulator_start() due to stale v_in_mv=0 Mar 15, 2026

Copilot finished work on behalf of electricalhog March 15, 2026 00:43

electricalhog added 2 commits March 14, 2026 21:11

interrupt issues

c71e863

copilot comment of intent

47e6dff

Copilot started work on behalf of electricalhog March 15, 2026 01:15 View session

Copilot AI changed the title ~~Fix wrong BOOST mode selection on first regulator_start() due to stale v_in_mv=0~~ Fix LPUART1 ISR priority override causing FreeRTOS assert in boost mode Mar 15, 2026

Copilot finished work on behalf of electricalhog March 15, 2026 02:08

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix LPUART1 ISR priority override causing FreeRTOS assert in boost mode#7

Fix LPUART1 ISR priority override causing FreeRTOS assert in boost mode#7
Copilot wants to merge 13 commits intospecfrom
copilot/nl-spec-implementation

Copilot AI commented Mar 11, 2026 •

edited

Loading

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

electricalhog Mar 14, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

Copilot AI commented Mar 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Changes

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!

2 participants

Copilot AI commented Mar 11, 2026 •

edited

Loading