Build notes — poolside llama.cpp laguna on DGX Spark

Independent Spark build notes.

Engine

git clone https://github.com/poolsideai/llama.cpp.git ~/src/llama.cpp-laguna
cd ~/src/llama.cpp-laguna
git checkout laguna   # pin: 04b2b72cb54048ead292884adbe11f284e3ec950

Patch (host-measured · Spark / GNU 13.3)

common/speculative.cpp calls std::isfinite. On this Spark host (GNU 13.3) that can fail to compile even with #include <cmath> depending on header order / libstdc++.

This is a local host build fix, not an upstream laguna behavioral change and not part of the public claim surface. Apply only if stock checkout fails to build the same way.

Measured fix that builds clean here:

Ensure #include <cmath> and #include <math.h> near the top system includes.
Replace the call site:

// before
if (!std::isfinite(v)) {

// after (Spark / GNU 13.3 host build)
if (!::isfinite(static_cast<double>(v))) /* spark-isfinite-fix */ {

Pin checkout before patching: git checkout 04b2b72cb54048ead292884adbe11f284e3ec950.
Record git rev-parse HEAD → pack results/engine_sha.txt.

Do not treat a failed UI-asset HF fetch during build as fatal — llama-server still links; only the optional embedded UI may stay stale.

Configure

export PATH=/usr/local/cuda/bin:/usr/bin:$PATH
export CUDA_HOME=/usr/local/cuda
export CC=/usr/bin/gcc CXX=/usr/bin/g++ CUDAHOSTCXX=/usr/bin/g++
# GB10: sm_121a
cmake -S . -B build \
  -DGGML_CUDA=ON \
  -DCMAKE_CUDA_ARCHITECTURES=121a \
  -DCMAKE_BUILD_TYPE=Release
cmake --build build -j"$(nproc)" --target llama-server llama-cli llama-bench

If CMake translates 121 poorly, force 121a.

Verify

export LD_LIBRARY_PATH=$HOME/src/llama.cpp-laguna/build/bin:$LD_LIBRARY_PATH
~/src/llama.cpp-laguna/build/bin/llama-server --version
# version: 1 (04b2b72)
# built with GNU 13.3.0 for Linux aarch64

Measured: binaries present after this patch (llama-server, llama-cli, llama-bench).