Differences

This shows you the differences between two versions of the page.

Link to this comparison view

--- en:iot-open:practical:hardware:sut:esp32:adv1_1 [2024/03/25 20:14] – [Steps] pczekalski
+++ en:iot-open:practical:hardware:sut:esp32:adv1_1 [2025/10/03 11:22] (current) – [FAQ] pczekalski
@@ Line 1: / Line 1: @@
 ====== ADV1: Using timers to execute code asynchronously ======
-It is advised to use timers that periodically execute a function to handle repeating tasks. Hardware timers work parallel to the CPU and consume few CPU resources. ESP32-S3 has 4 hardware timers, but each timer can execute multiple handlers.
+It is advised to use timers that periodically execute a function to handle repeating tasks. Hardware timers work parallel to the CPU and consume few CPU resources. ESP32-S3 has 4 hardware timers, but each timer can execute multiple handlers. You can think about these handlers as they are interrupted handling functions, but instead of externally triggered interrupts, those are initiated internally by the hardware timer.\\
-The idea of using the timer is to encapsulate a piece of compact code that can be run virtually asynchronously and executed is a precisely-defined time manner. In this scenario, we use a timer to update the LCD screen periodically. We choose a dummy example where Timer 1 is used to increment a value of the ''byte'' type, and Timer 2 reads this value and writes it on the LCD. Naturally, a collision may occur whenever two processes are to access a single memory block (variable). It is critical when both processes (tasks, asynchronous functions) are writing to it. However, in our example, the handler executed by Timer 1 only writes to the variable, while Timer 2 only reads from it. In this scenario, there is no need to use mutexes (semaphores). A scenario with detailed description of when to use mutexes is beyond the scope of this example.
+The idea of using the timer is to encapsulate a piece of compact code that can be run virtually asynchronously and executed is a precisely-defined time manner. In this scenario, we use a timer to update the LCD screen periodically. We choose a dummy example where Timer 1 is used to increment a value of the ''byte'' type, and Timer 2 reads this value and writes it on the LCD. Naturally, a collision may occur whenever two processes are to access a single memory block (variable). It is critical when both processes (tasks, asynchronous functions) are writing to it. However, in our example, the handler executed by Timer 1 only writes to the variable, while Timer 2 only reads from it. In this scenario, there is no need to use mutexes (semaphores). A scenario with a detailed description of when to use mutexes is beyond the scope of this example.
@@ Line 117: / Line 117: @@
 === Step 7 ===
-This way, a main loop is empty: everything runs asynchronously, thanks to the timers.
+This way, a main loop is empty: everything runs asynchronously, thanks to the timers. As suggested above, when timer handlers require long or unpredictable execution time (e.g. external communication, waiting for the reply), handlers should set a flag that is read in the main loop to execute the appropriate task and then clear the flag.
 <code c>
 void loop()
@@ Line 128: / Line 128: @@
 ===== FAQ =====
-**How many timers can I use?**: ESP32-S3 has 4 hardware timers. You may trick this limitation by using smart handlers that have, e.g., an internal counter and internally execute every N-th cycle. This helps to simulate more timers in a software way.
+**How many timers can I use?**: ESP32-S3 has 4 hardware timers. You can circumvent this limitation by using smart handlers that have, for example, an internal counter and execute every Nth cycle internally. This helps simulate multiple timers in a software-based manner.
 <WRAP noprint>

en/iot-open/practical/hardware/sut/esp32/adv1_1.1711397663.txt.gz · Last modified: 2024/03/25 20:14 by pczekalski