You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In ESP_IDF, CONFIG_FREERTOS_HZ is the tick rate at which FreeRTOS does preemptive context switching. Currently we have CONFIG_FREERTOS_HZ = 100 (10 msecs), which is the default. However, it's very often set to 1000 (1 msec):
Many examples in ESP-IDF use 1000.
MicroPython's bstacksdkconfig uses 1000.
ESP-IDF Arduino uses 1000, to allow delay() to be more accurate for small values.
People who want faster HID need to change this to achieve faster output. This is also related to the bInterval value in the HID device descriptor, which is default 8, or 8 msecs.
I haven't found a lot of discussion about any detriments of using a faster tick rate. I assume there is some increased overhead, but it may not be significant.
As an experiment, let's bump it to 1000 for 9.1.0 and see what people notice.
(Motivated by long thread with @jzhvymetal in discord and brief discussion with @tannewt there.)
The text was updated successfully, but these errors were encountered:
In ESP_IDF,
CONFIG_FREERTOS_HZ
is the tick rate at which FreeRTOS does preemptive context switching. Currently we haveCONFIG_FREERTOS_HZ = 100
(10 msecs), which is the default. However, it's very often set to1000
(1 msec):bstack
sdkconfig
uses 1000.delay()
to be more accurate for small values.bInterval
value in the HID device descriptor, which is default8
, or 8 msecs.I haven't found a lot of discussion about any detriments of using a faster tick rate. I assume there is some increased overhead, but it may not be significant.
As an experiment, let's bump it to 1000 for 9.1.0 and see what people notice.
(Motivated by long thread with @jzhvymetal in discord and brief discussion with @tannewt there.)
The text was updated successfully, but these errors were encountered: