Interrupt Allocation

Overview

The ESP32 has two cores, with 32 interrupts each. Each interrupt has a fixed priority, most (but not all) interrupts are connected to the interrupt matrix.

Because there are more interrupt sources than interrupts, sometimes it makes sense to share an interrupt in multiple drivers. The esp_intr_alloc() abstraction exists to hide all these implementation details.

A driver can allocate an interrupt for a certain peripheral by calling esp_intr_alloc() (or esp_intr_alloc_intrstatus()). It can use the flags passed to this function to specify the type, priority, and trigger method of the interrupt to allocate. The interrupt allocation code will then find an applicable interrupt, use the interrupt matrix to hook it up to the peripheral, and install the given interrupt handler and ISR to it.

The interrupt allocator presents two different types of interrupts, namely shared interrupts and non-shared interrupts, both of which require different handling. Non-shared interrupts will allocate a separate interrupt for every esp_intr_alloc() call, and this interrupt is use solely for the peripheral attached to it, with only one ISR that will get called. Shared interrupts can have multiple peripherals triggering them, with multiple ISRs being called when one of the peripherals attached signals an interrupt. Thus, ISRs that are intended for shared interrupts should check the interrupt status of the peripheral they service in order to check if any action is required.

Non-shared interrupts can be either level- or edge-triggered. Shared interrupts can only be level interrupts due to the chance of missed interrupts when edge interrupts are used.

To illustrate why shard interrupts can only be level-triggered, take the scenario where peripheral A and peripheral B share the same edge-triggered interrupt. Peripheral B triggers an interrupt and sets its interrupt signal high, causing a low-to-high edge, which in turn latches the CPU's interrupt bit and triggers the ISR. The ISR executes, checks that peripheral A did not trigger an interrupt, and proceeds to handle and clear peripheral B's interrupt signal. Before the ISR returns, the CPU clears its interrupt bit latch. Thus, during the entire interrupt handling process, if peripheral A triggers an interrupt, it will be missed due the CPU clearing the interrupt bit latch.

Multicore Issues

Peripherals that can generate interrupts can be divided in two types:

• External peripherals, within the ESP32 but outside the Xtensa cores themselves. Most ESP32 peripherals are of this type.

• Internal peripherals, part of the Xtensa CPU cores themselves.

Interrupt handling differs slightly between these two types of peripherals.

Internal Peripheral Interrupts

Each Xtensa CPU core has its own set of six internal peripherals:

• Three timer comparators

• A performance monitor

• Two software interrupts

Internal interrupt sources are defined in esp_intr_alloc.h as ETS_INTERNAL_*_INTR_SOURCE.

These peripherals can only be configured from the core they are associated with. When generating an interrupt, the interrupt they generate is hard-wired to their associated core; it is not possible to have, for example, an internal timer comparator of one core generate an interrupt on another core. That is why these sources can only be managed using a task running on that specific core. Internal interrupt sources are still allocatable using esp_intr_alloc() as normal, but they cannot be shared and will always have a fixed interrupt level (namely, the one associated in hardware with the peripheral).

External Peripheral Interrupts

The remaining interrupt sources are from external peripherals. These are defined in soc/soc.h as ETS_*_INTR_SOURCE.

Non-internal interrupt slots in both CPU cores are wired to an interrupt matrix, which can be used to route any external interrupt source to any of these interrupt slots.

• Allocating an external interrupt will always allocate it on the core that does the allocation.

• Freeing an external interrupt must always happen on the same core it was allocated on.

• Disabling and enabling external interrupts from another core is allowed.

• Multiple external interrupt sources can share an interrupt slot by passing ESP_INTR_FLAG_SHARED as a flag to esp_intr_alloc().

Care should be taken when calling esp_intr_alloc() from a task which is not pinned to a core. During task switching, these tasks can migrate between cores. Therefore it is impossible to tell which CPU the interrupt is allocated on, which makes it difficult to free the interrupt handle and may also cause debugging difficulties. It is advised to use xTaskCreatePinnedToCore() with a specific CoreID argument to create tasks that allocate interrupts. In the case of internal interrupt sources, this is required.

IRAM-Safe Interrupt Handlers

The ESP_INTR_FLAG_IRAM flag registers an interrupt handler that always runs from IRAM (and reads all its data from DRAM), and therefore does not need to be disabled during flash erase and write operations.

This is useful for interrupts which need a guaranteed minimum execution latency, as flash write and erase operations can be slow (erases can take tens or hundreds of milliseconds to complete).

It can also be useful to keep an interrupt handler in IRAM if it is called very frequently, to avoid flash cache misses.

Refer to the SPI flash API documentation for more details.

Multiple Handlers Sharing A Source

Several handlers can be assigned to a same source, given that all handlers are allocated using the ESP_INTR_FLAG_SHARED flag. They will all be allocated to the interrupt, which the source is attached to, and called sequentially when the source is active. The handlers can be disabled and freed individually. The source is attached to the interrupt (enabled), if one or more handlers are enabled, otherwise detached. A handler will never be called when disabled, while its source may still be triggered if any one of its handler enabled.

Sources attached to non-shared interrupt do not support this feature.

By default, when ESP_INTR_FLAG_SHARED flag is specified, the interrupt allocator will allocate only priority level 1 interrupts. Use ESP_INTR_FLAG_SHARED | ESP_INTR_FLAG_LOWMED to also allow allocating shared interrupts at priority levels 2 and 3.

Though the framework supports this feature, you have to use it very carefully. There usually exist two ways to stop an interrupt from being triggered: disable the source or mask peripheral interrupt status. ESP-IDF only handles enabling and disabling of the source itself, leaving status and mask bits to be handled by users.

Status bits shall either be masked before the handler responsible for it is disabled, or be masked and then properly handled in another enabled interrupt.

Note

Leaving some status bits unhandled without masking them, while disabling the handlers for them, will cause the interrupt(s) to be triggered indefinitely, resulting therefore in a system crash.

Troubleshooting Interrupt Allocation

On most Espressif SoCs, CPU interrupts are a limited resource. Therefore it is possible for a program to run out of CPU interrupts, for example by initializing several peripheral drivers. Typically, this will result in the driver initialization function returning ESP_ERR_NOT_FOUND error code.

If this happens, you can use esp_intr_dump() function to print the list of interrupts along with their status. The output of this function typically looks like this:

CPU 0 interrupt status:
Int  Level  Type   Status
0     1    Level  Reserved
1     1    Level  Reserved
2     1    Level  Used: RTC_CORE
3     1    Level  Used: TG0_LACT_LEVEL
...

The columns of the output have the following meaning:

• Int: CPU interrupt input number. This is typically not used in software directly, and is provided for reference only.

• Level: Interrupt priority (1-7) of the CPU interrupt. This priority is fixed in hardware, and cannot be changed.

• Type: Interrupt type (Level or Edge) of the CPU interrupt. This type is fixed in hardware, and cannot be changed.

• Status: One of the possible statuses of the interrupt:

  • Reserved: The interrupt is reserved either at hardware level, or by one of the parts of ESP-IDF. It can not be allocated using esp_intr_alloc().

  • Used: <source>: The interrupt is allocated and connected to a single peripheral.

  • Shared: <source1> <source2> ...: The interrupt is allocated and connected to multiple peripherals. See Multiple Handlers Sharing A Source above.

  • Free: The interrupt is not allocated and can be used by esp_intr_alloc().

• Free (not general-use): The interrupt is not allocated, but is either a high-priority interrupt (priority 4-7) or an edge-triggered interrupt. High-priority interrupts can be allocated using esp_intr_alloc() but requires the handlers to be written in Assembly, see High Priority Interrupts. Edge-triggered low- and medium-priority interrupts can also be allocated using esp_intr_alloc(), but are not used often since most peripheral interrupts are level-triggered.

If you have confirmed that the application is indeed running out of interrupts, a combination of the following suggestions can help resolve the issue:

• On multi-core SoCs, try initializing some of the peripheral drivers from a task pinned to the second core. Interrupts are typically allocated on the same core where the peripheral driver initialization function runs. Therefore by running the initialization function on the second core, more interrupt inputs can be used.

• Determine the interrupts which can tolerate higher latency, and allocate them using ESP_INTR_FLAG_SHARED flag (optionally ORed with ESP_INTR_FLAG_LOWMED). Using this flag for two or more peripherals will let them use a single interrupt input, and therefore save interrupt inputs for other peripherals. See Multiple Handlers Sharing A Source above.

• Some peripheral driver may default to allocating interrupts with ESP_INTR_FLAG_LEVEL1 flag, so priority 2 and 3 interrupts do not get used by default. If esp_intr_dump() shows that some priority 2 or 3 interrupts are available, try changing the interrupt allocation flags when initializing the driver to ESP_INTR_FLAG_LEVEL2 or ESP_INTR_FLAG_LEVEL3.

• Check if some of the peripheral drivers do not need to be used all the time, and initialize or deinitialize them on demand. This can reduce the number of simultaneously allocated interrupts.

API Reference

Header File

• components/esp_hw_support/include/esp_intr_types.h

• This header file can be included with:

#include "esp_intr_types.h"

Macros

ESP_INTR_CPU_AFFINITY_TO_CORE_ID(cpu_affinity)

Convert esp_intr_cpu_affinity_t to CPU core ID.

Type Definitions

typedef void (*intr_handler_t)(void *arg)

Function prototype for interrupt handler function

typedef struct intr_handle_data_t *intr_handle_t

Handle to an interrupt handler

Enumerations

enum esp_intr_cpu_affinity_t

Interrupt CPU core affinity.

This type specify the CPU core that the peripheral interrupt is connected to.

Values:

enumerator ESP_INTR_CPU_AFFINITY_AUTO

Install the peripheral interrupt to ANY CPU core, decided by on which CPU the interrupt allocator is running.

enumerator ESP_INTR_CPU_AFFINITY_0

Install the peripheral interrupt to CPU core 0.

enumerator ESP_INTR_CPU_AFFINITY_1

Install the peripheral interrupt to CPU core 1.

Header File

• components/esp_hw_support/include/esp_intr_alloc.h

• This header file can be included with:

#include "esp_intr_alloc.h"