Building with just ESP-IDF

[oresk]

First of all, thnx for all the work on this project!

Wanted to check one thing. You say in the readme that The codebase is split into components and build on vanilla ESP-IDF. I have just the esp-idf installed on my machine. When I try to start the menuconfig I get this error CMake Error at CMakeLists.txt:5 (include): include could not find requested file: /CMakeLists.txt

Which I think is produced by root CMakeLists.txt line include($ENV{ADF_PATH}/CMakeLists.txt).

What should I do with that? Install ADF, point the ADF-PATH to the IDF_PATH, or do some other change?

---

I tried just changing the ADF-PATH to IDF-PATH, and I get this error later in the menuconfig start up

CMake Error at /home/lovro/esp/esp-idf/CMakeLists.txt:215 (idf_build_set_property): Unknown CMake command "idf_build_set_property". Call Stack (most recent call first): CMakeLists.txt:5 (include)

My ESP-IDF version is: ESP-IDF v5.0-dev-329-g4a011f3183

Thnx in advance!

[kiesel]

Stumbled over this problem as well; fixed by installing esp-adf and pointing ADF_PATH to that installation - following the steps as documented here: https://docs.espressif.com/projects/esp-adf/en/latest/get-started/#get-started-get-esp-adf

One could improve the documentation with that information, I guess:-)

[oresk]

Thnx for the info. What ESP_IDF version are you using?

[kiesel]

$ idf.py --version
ESP-IDF v4.3.1-dirty

[oresk]

A bit of a context before I close the issue.

In the end I managed to build it by using current master (06.12.2021) of the adf and using the esp-idf 4.2.1 version.

All in all it was a bit of a hit and miss. Using the esp-idf from the adf did not work, and using the current 5.0 dev version of the esp-idf did not work also.

Had to make a small change in the opus component code (initialize a variable on the start) so that the compiler does not complain.

The esp connects to the snapcast server that I am running, but it crashes afterwards trying to connect to the pcm5102A (I think I actually have pcm5100 that the aliexpress people have sanded the top of and lasered on a new inscription, but it should work for testing just fine). The pcm51xx code in the repo has a prediction that it can control the volume by using the i2c, which does not happen on my board, so next step is to figure out that bit.

[ckcr4lyf]

FYI: The error in the opus code, if someone finds this by searching, is:

/home/poiasd/Documents/snapclient/components/opus/opus/silk/quant_LTP_gains.c: In function 'silk_quant_LTP_gains':
/home/poiasd/Documents/snapclient/components/opus/opus/silk/quant_LTP_gains.c:129:21: error: 'res_nrg_Q15' may be used uninitialized in this function [-Werror=maybe-uninitialized]
         res_nrg_Q15 = silk_RSHIFT32( res_nrg_Q15, 2 );

Fixed by adding a line such as

res_nrg_Q15 = 0;

in the file