Unable to compile substrate template `error: failed to run custom build command for `librocksdb-sys v0.8.0+7.4.4`

Question

First, I have looked at several previous posts regarding similar build errors but have yet to get the substrate template to build locally successfully. The exact error I encounter is:

error: failed to run custom build command for `librocksdb-sys v0.8.0+7.4.4`

Caused by:
  process didn't exit successfully: `/home/brandon/workspace/substrate-node-template/target/release/build/librocksdb-sys-c7956af65347884b/build-script-build` (exit status: 101)
  --- stderr
  thread 'main' panicked at '"enum_(unnamed_at_rocksdb/include/rocksdb/c_h_981_1)" is not a valid Ident', /home/brandon/.cargo/registry/src/GitHub.com-1ecc6299db9ec823/proc-macro2-1.0.47/src/fallback.rs:756:9
  note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
warning: build failed, waiting for other jobs to finish...

I do have clang and llvm installed. I have even compiled clang + llvm from sources and installed them on the computer in an attempt to make sure the build system should work. I have the librocksdb-dev package installed as well. I am running on Debian Sid and have followed all the steps in the Linux Development Environment guide.

The output of rustup +nightly show:

Default host: x86_64-unknown-Linux-gnu
rustup home:  /home/brandon/.rustup

installed toolchains
--------------------

stable-x86_64-unknown-Linux-gnu (default)
nightly-x86_64-unknown-Linux-gnu

installed targets for active toolchain
--------------------------------------

wasm32-unknown-unknown
x86_64-unknown-linux-gnu

active toolchain
----------------

nightly-x86_64-unknown-Linux-gnu (overridden by +toolchain on the command line)
rustic 1.67.0-nightly (b3bc6bf31 2022-11-24)

I noticed some people reported issues with the M1 processor; However, I am running a 12900K Intel Processor, so I doubt that is the issue.

I have entirely uninstalled librocksdb-dev llvm clang and even rust and then reinstalled them. I am unsure how to continue as I am brand new to the substrate platform.

Answer 1

I have exactly the same error while building Substrate. The issue is not in Substrate, but rust-rocksdb.

More: https://github.com/rust-rocksdb/rust-rocksdb/issues/713

Answer 2

I recently came across this issue on a fresh Mac with an M2 chip.

After running the following commands, it seems to have been resolved:

• cargo clean
• cargo update
• rustup update

Answer 3

What solved my problem related with build of frontier, but I believe that is related with librocksdb-sys too, was when I ran these commands which update system packages and add g++ that is a open source version of c++:

sudo apt-get update
sudo apt-get install g++

Full solution
https://stackoverflow.com/questions/66583720/some-error-failed-to-run-custom-build-command-for-librocksdb-sys-v6-11-4/76791499#76791499

Answer 4

That's an issue I came across in Windows 11. And turns out it might have something to do with your machine running out of RAM.

In my windows machine I was building projects using WSL and it was capped at 2,5GB to avoid WSL from allocating all the RAM for itself. When trying to build a substrate node, the process needed more RAM to compile large dependencies like librocksdb-sys and would exit with and error.

If anyone is running into this issue while using Windows, check the RAM cap of WSL.

Answer 5

had exactly the same problem with opensuse tumbleweed because of clang and llvm, you actually need an older version not a new one, try with installing v14 of clang and llvm

Answer 6

The error message "failed to run custom build command for librocksdb-sys" usually indicates that there is a problem with the RocksDB library or its dependencies. I had the same error while building Substrate, since I changed the linux distro it's done, for example in Fedora it works ok.

Answer 7

I ran into this error on M1 and was able to solve it by downgrading to clang 14 as suggested by @kevindong on GitHub

brew install llvm@14
export PATH="/opt/homebrew/opt/llvm@14/bin:$PATH" 
clang --version # Should return 14

See: https://github.com/rust-rocksdb/rust-rocksdb/issues/768#issuecomment-1767143727

Answer 8

I use wsl to build the node in release mode and get this error. I got over this issue by creating a file that is configured to act as a swap space for my system. Normally I do something like.

sudo allocate -l 4G /swapfile

Adjust the size according to your needs(not necessarily 4GB). File name need not be swapfile

sudo mkswap /swapfile

sudo swapon /swapfile

Now you can check the configured swapfile using

sudo swapon --show

Try building the node again.