I have recently run across this exact same issue with a ESP32C3. Code that compiles and works perfectly on the ESP8266 does not work on the ESP32C3 (maybe other ESP32 versions?) if (tcpClient.availableForWrite() > 0) { tcpClient.write(sendbuf, sendlength); } This skips the write because the result f...

FWIW, the GPIO latency is a function of the bus they're on. If you need faster GPIO, we do support dedicated GPIO on most newer chips (S2 and up, iirc): https://docs.espressif.com/projects/esp-idf/en/latest/esp32s3/api-reference/peripherals/dedic_gpio.html Thanks. I would also like to know if there...

Is there going to be some concrete information about the GPIO latency on these parts? I would hope that Espressif could figure out how to do single cycle GPIO by now. Taking hundreds of nanoseconds (with a sub-5ns clock) just to change a pin state is a huge limitation that none of their competitors ...

I have this same problem as well!

int i = tcpClient.availableForWrite();

This always returns with i equal to 0. This code works perfectly when compiled for the ESP8266.

Well, that is interesting. Is that actually documented anywhere? I found this: "If the Host only writes data, the dummy bit workaround and the frequency check can be disabled by setting the bit SPI_DEVICE_NO_DUMMY in the member spi_device_interface_config_t::flags. When disabled, the output frequenc...

Bump? Doesn't anyone actually use SPI?

I have a board setup using both HSPI and VSPI. I am using the dedicated pins for both of these ports. There are two issues that I can not seem to resolve. Issue 1 - I can not make the port go any faster than 26MHz. Anything up to 26MHz works fine. The device it is connected to will handle 38MHz, and...