Ludovic,

I'm having trouble getting the experimental-esp-idf-v5 branch to even configure.

Configuring and building the "hello_world" example with idf5.0 works without any issues, so it's related to our sources.

I've tried removing the build directory, doing "idf.py python-clean", doing "idf.py clean" "idf.py fullclean", but I always get this, already when doing set-target:

balzer@leela:~/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3> idf.py set-target esp32
Adding "set-target"'s dependency "fullclean" to list of commands with default set of options.
Executing action: fullclean
Build directory '/home/balzer/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3/build' not found. Nothing to clean.
Executing action: set-target
Set Target to: esp32, new sdkconfig created. Existing sdkconfig renamed to sdkconfig.old.
Running cmake in directory /home/balzer/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3/build
Executing "cmake -G Ninja -DPYTHON_DEPS_CHECKED=1 -DESP_PLATFORM=1 -DIDF_TARGET=esp32 -DCCACHE_ENABLE=0 /home/balzer/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3"...
CMake Warning (dev) at CMakeLists.txt:8 (message):
WARNING - OVMSv3 build is only supported on ESP-IDF 3.3.x
This warning is for project developers. Use -Wno-dev to suppress it.

-- Found Git: /usr/bin/git (found version "2.35.3")
-- The C compiler identification is GNU 11.2.0
-- The CXX compiler identification is GNU 11.2.0
-- The ASM compiler identification is GNU
-- Found assembler: /home/balzer/.espressif/tools/xtensa-esp32-elf/esp-2022r1-11.2.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /home/balzer/.espressif/tools/xtensa-esp32-elf/esp-2022r1-11.2.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /home/balzer/.espressif/tools/xtensa-esp32-elf/esp-2022r1-11.2.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-g++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Building ESP-IDF components for target esp32
Processing 2 dependencies:
[1/2] espressif/mdns (1.0.7)
[2/2] idf (5.0.0)
-- Found BISON: /usr/bin/bison (found version "3.0.4")
-- Found FLEX: /usr/bin/flex (found version "2.6.4")
-- IDF Version: 5.0.0 - NOT enabling strverscmp
CMake Error at /home/balzer/esp/esp-idf/tools/cmake/component.cmake:224 (message):
CMake Warning (dev) at build_properties.temp.cmake:8:

    Syntax Warning in cmake code at column 51

    Argument not separated from preceding token by whitespace.

Call Stack (most recent call first):

    /home/balzer/esp/esp-idf/tools/cmake/scripts/component_get_requirements.cmake:3 (include)

This warning is for project developers. Use -Wno-dev to suppress it.

CMake Error: Error required internal CMake variable not set, cmake may not
be built correctly.

Missing variable is:

CMAKE_FIND_LIBRARY_PREFIXES

CMake Error: Error required internal CMake variable not set, cmake may not
be built correctly.

Missing variable is:

CMAKE_FIND_LIBRARY_SUFFIXES

Call Stack (most recent call first):
/home/balzer/esp/esp-idf/tools/cmake/build.cmake:553 (__component_get_requirements)
/home/balzer/esp/esp-idf/tools/cmake/project.cmake:440 (idf_build_process)
CMakeLists.txt:18 (project)

-- Configuring incomplete, errors occurred!
See also "/home/balzer/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3/build/CMakeFiles/CMakeOutput.log".
cmake failed with exit code 1, output of the command is in the /home/balzer/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3/build/log/idf_py_stderr_output_8211 and /home/balzer/esp/Open-Vehicle-Monitoring-System-3/vehicle/OVMS.V3/build/log/idf_py_stdout_output_8211

Any idea about this?

Regards,
Michael

Am 02.02.23 um 22:38 schrieb Ludovic LANGE:

Hi list,

I started to split my big list of patches into smaller, more manageable, easier to review PRs:

missing headers #825

ESP-IDF v5+: dependency on ovms_webserver.h #824

ESP-IDF v5+: change type of TaskHandle_t #823

changes to (long deprecated) ADC constants #822

ESP-IDF v5+: static assertion failed: std::vector must have the same value_type as its allocator #821

ESP-IDF v5+: static assertion failed: comparison object must be invocable as const #820

ESP-IDF v4+ : remove deprecated and unused esp_event_loop.h #819

remove warning by replacing rom/queue.h with sys/queue.h #818

remove warning : '_GNU_SOURCE' redefined #817

ESP-IDF v4+ : esp_intr.h is deprecated, replaced by esp_intr_alloc.h #816

Add header and fix function signature #815

replace some constants with their more adapted counterpart #813

Let me know if it's the proper approach ; and if some of you can review those changes and comment on them, it'll help merging those - and progress in reaching ESP-IDFv5 compatibility.

Best regards,

Le 31/01/2023 à 01:49, Ludovic LANGE a écrit :

Hi list,

Another quick post to share with you the progress on the ESP-IDFv5 + cmake endeavour:

ESP-IDF v5 version of OVMSv3 is now officially booting and seems to work fine (then again, I'm not doing any real-world testing for lack of time / car at the moment - it's here that you can help me) with a few caveats detailed in the PR or the README.md.

The branch https://github.com/llange/Open-Vehicle-Monitoring-System-3/tree/experimental-esp-idf-v5 is now ready for testing, sharing, studying, ...

I created a draft PR https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/pull/810 (just like for ESP-IDFv4), in order to show the diff and gather feedback. Not to be merged as-is.

I'll try to break it into smaller low-impact / easy-to-review PR's that could be merged as-is in current master (warning corrections, header definitions, xxprintf type impact, etc...), so that the big pieces could be more easily reviewed.

Please note that those branches / PRs are still Work-In-Progress, and that I may certainly (and wrongly) `push --force` some changes as I rebase, update or split the big PR into smaller, easy to study PRs. I'll try not to but can't promise :-)

So if you checkout the branch, please be warned and know your git commands to either reset or rebase your working copy...

Next on my list are, in no particular order:

Have a few patches/bugs fixed in upstream wolfSSL (in progress) in order to enable

Check to see if we could upgrade mongoose (not absolutely necessary because it works / compile OK with a small patch)

Fix the remaining warnings

And change the ADC implementation that is deprecated now

deprecation of esp_vfs_fat_spiflash_mount / esp_vfs_fat_spiflash_unmount

Fix the WIP WIP WIP commits

Resume work on https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/752

Boot time warnings to study and check if OK:

W (1091) esp_psram: Virtual address not enough for PSRAM, map as much as we can. 4MB is mapped

W (2132) ADC: legacy driver is deprecated, please migrate to `esp_adc/adc_oneshot.h` (<- this one is easy to understand...)

Additional instructions:

When first building, your sdkconfig will be converted as some config options did change betwen ESP-IDF v3.x and v5.x.

Run `menuconfig` and :

Select FREERTOS_ENABLE_BACKWARD_COMPATIBILITY(=y) "configENABLE_BACKWARD_COMPATIBILITY"

Unselect MG_ENABLE_SSL(=n) "Enable MONGOOSE SSL/TLS support"

If any of you with some free time could be nice enough to try to reproduce such a build and share the experience here. Let me know how the instructions could be updated and what issues you are facing.

Best regards,
_______________________________________________
OvmsDev mailing list
OvmsDev@lists.openvehicles.com
http://lists.openvehicles.com/mailman/listinfo/ovmsdev

-- 
Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
Fon 02333 / 833 5735 * Handy 0176 / 206 989 26