Skip to content

Installation Troubleshooting

This page provides guidance on how to troubleshoot common issues that may arise when installing ethereum/execution-spec-tests.

Problem: Failed building wheel for coincurve

Problem: Failed building wheel for coincurve

Installing EEST and its dependencies via uv sync --all-extras fails with:

Stored in directory: /tmp/...
  Building wheel for coincurve (pyproject.toml) ... error
  error: subprocess-exited-with-error

  × Building wheel for coincurve (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [27 lines of output]
      ...
        571 | #include <secp256k1_extrakeys.h>
            |          ^~~~~~~~~~~~~~~~~~~~~~~
      compilation terminated.
      error: command '/usr/bin/gcc' failed with exit code 1
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building wheel for coincurve

Solution: Install the libsecp256k1 library

On Ubuntu, you can install this library with:

sudo apt update
sudo apt-get install libsecp256k1-dev

Problem: solc Installation issues

Problem: CERTIFICATE_VERIFY_FAILED

Problem: Failed to ... CERTIFICATE_VERIFY_FAILED

When running fill you might encounter the following error:

Exit: Failed to ...: <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:992)>

Solution: Update your system’s CA certificates

On Ubuntu, run the following commands:

sudo apt-get update
sudo apt-get install ca-certificates

Solution: Update your system’s CA certificates

On macOS, Python provides a built-in script to install the required certificates:

/Applications/Python\ 3.11/Install\ Certificates.command

Problem: Exception: failed to compile yul source

Problem: Running fill on static_tests fails with tests that contain yul source code on ARM platforms

To resolve the issue you must acquire the solc binary e.g. by building solidity from source. The guide below installs v0.8.28 but you might prefer to choose a different version.

Solution: Build solc from source

The following steps have been tested on Ubuntu 24.04.2 LTS ARM64: 

git clone --branch v0.8.28 --depth 1 https://github.com/ethereum/solidity.git
cd solidity && mkdir build && cd build
sudo apt install build-essential libboost-all-dev z3
cmake ..
make
mv $HOME/Documents/execution-spec-tests/.venv/bin/solc $HOME/Documents/execution-spec-tests/.venv/bin/solc-x86-64
cp ./solc/solc $HOME/Documents/execution-spec-tests/.venv/bin/
chmod +x $HOME/Documents/execution-spec-tests/.venv/bin/solc
Running uv run solc --version should now return the expected version.

Problem: ValueError: unsupported hash type ripemd160

Problem: Running fill fails with tests that use the RIPEMD160 precompile (0x03)

When running fill, you encounter the following error:

ValueError: unsupported hash type ripemd160
# or
ValueError: [digital envelope routines] unsupported

This is due to the removal of certain cryptographic primitives in OpenSSL 3. These were re-introduced in OpenSSL v3.0.7.

Solution: Modify OpenSSL configuration

On platforms where OpenSSL v3.0.7 is unavailable (e.g., Ubuntu 22.04), modify your OpenSSL configuration to enable RIPEMD160. Make the following changes in the OpenSSL config file:

[openssl_init]
providers = provider_sect

# List of providers to load
[provider_sect]
default = default_sect
legacy = legacy_sect

[default_sect]
activate = 1

[legacy_sect]
activate = 1

This will enable the legacy cryptographic algorithms, including RIPEMD160. See ethereum/execution-specs#506 for more information.

Problem: VS Code "Autoformat on Save" with Ruff Not Working

Problem: 'Autoformat on Save' with Ruff not working as expected in VS Code

If you are using VS Code and "autoformat on save" is not working as expected, or if it produces different formatting than the official tox -e lint command, you may have a version mismatch with the ruff formatter. This problem can be confirmed if git diff shows changes to an otherwise unmodified file after you have saved it.

This issue often occurs when VS Code is not configured to use the project's virtual environment (.venv) or if the linting dependencies have not been installed. In this case, VS Code's Ruff extension falls back to a bundled version of ruff, which may not match the version pinned in the project's pyproject.toml file.

Solution: Install all required dependencies and select the correct interpreter

  1. Ensure all dependencies are installed, including the lint extras.

    uv sync --all-extras

  2. In VS Code, open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P) and select Python: Select Interpreter.

  3. Choose the interpreter located in the project's .venv directory. This ensures VS Code uses the correct ruff version from your project's environment.

Other Issues Not Listed?

If you're facing an issue that's not listed here, you can easily report it on GitHub for resolution.

Click here to report a documentation issue related to installation

Please include the following details in your report:

  1. The command that triggered the issue.
  2. Any relevant error messages or screenshots.
  3. Relevant version information from the output of uv run eest info (if running consume from within eest).