github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[PR #8539] [MERGED] next build #75017

New Issue
Closed
opened 2026-05-05 07:22:01 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-05-05 07:22:01 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/8539
Author: @mxyng
Created: 1/22/2025
Status: ✅ Merged
Merged: 1/29/2025
Merged by: @mxyng

Base: main ← Head: mxyng/next-build

📝 Commits (10+)

📊 Changes

542 files changed (+5778 additions, -11451 deletions)

View changed files

📝 .dockerignore (+3 -1)
📝 .gitattributes (+9 -0)
📝 .github/workflows/release.yaml (+266 -631)
📝 .github/workflows/test.yaml (+88 -264)
📝 .gitignore (+3 -2)
➕ CMakeLists.txt (+112 -0)
➕ CMakePresets.json (+110 -0)
📝 Dockerfile (+117 -190)
➖ Makefile (+0 -103)
➕ Makefile.sync (+56 -0)
📝 discover/amd_common.go (+4 -9)
📝 discover/amd_linux.go (+2 -4)
📝 discover/amd_windows.go (+3 -7)
📝 discover/gpu.go (+24 -64)
📝 discover/gpu_darwin.go (+0 -3)
➕ discover/path.go (+53 -0)
📝 discover/types.go (+1 -2)
📝 docs/development.md (+63 -108)
📝 envconfig/config.go (+0 -9)
📝 go.mod (+2 -1)

...and 80 more files

📄 Description

split from #7913

this changes updates the directory structure splitting llama.cpp and ggml into separate, reusable packages. as a result the build has also changed significantly. the build now uses cmake to build dependencies as shared objects which will be dynamically loaded when necessary.

current (work in progress) build instructions:

  • go build . to build ollama. this includes a default, basic cpu runner
  • cmake --preset Default; cmake --build --preset Default to configure and build the default targets. this will configure and build cuda and rocm if those are available
  • cmake --preset CPU; cmake --build --preset CPU to configure and build only CPU variants
  • cmake --preset CUDA; cmake --build --preset CUDA to configure and build only CUDA
  • cmake --preset ROCm; cmake --build --preset ROCm to configure and build only ROCm

TODO:

  • Windows CI
  • Build docs
  • Update CMake output directory

🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/ollama/ollama/pull/8539 **Author:** [@mxyng](https://github.com/mxyng) **Created:** 1/22/2025 **Status:** ✅ Merged **Merged:** 1/29/2025 **Merged by:** [@mxyng](https://github.com/mxyng) **Base:** `main` ← **Head:** `mxyng/next-build` --- ### 📝 Commits (10+) - [`144f63e`](https://github.com/ollama/ollama/commit/144f63e2fbd349f529ec0c71aef0ee1636a05780) next build - [`09320e8`](https://github.com/ollama/ollama/commit/09320e8e3df5e146bb597223b016342659375d3b) add build to .dockerignore - [`988de7b`](https://github.com/ollama/ollama/commit/988de7b7356025afdd4d0d5d3f19bc9a5f94eb1c) test: only build one arch - [`303248d`](https://github.com/ollama/ollama/commit/303248d031b083a47ca125b74c40c25436895ffd) add build to .gitignore - [`214eb1a`](https://github.com/ollama/ollama/commit/214eb1a2e8c1215f0278b2d978fb9af6dad87e0e) fix ccache path - [`25bf875`](https://github.com/ollama/ollama/commit/25bf87545ae47048e1a15637a507229c4caeb7a1) filter amdgpu targets - [`861e17a`](https://github.com/ollama/ollama/commit/861e17a0377d03f0269aba138eca10d30c080c04) only filter if autodetecting - [`6a75786`](https://github.com/ollama/ollama/commit/6a75786a451be2b4a9932806b455527e10235bd9) Don't clobber gpu list for default runner - [`4ec1765`](https://github.com/ollama/ollama/commit/4ec1765e78d10550fea2c7e1173b555c7aad220e) explicitly set CXX compiler for HIP - [`5a9c704`](https://github.com/ollama/ollama/commit/5a9c704c77509c4e4b09609139beeebd58880150) Update build_windows.ps1 ### 📊 Changes **542 files changed** (+5778 additions, -11451 deletions) <details> <summary>View changed files</summary> 📝 `.dockerignore` (+3 -1) 📝 `.gitattributes` (+9 -0) 📝 `.github/workflows/release.yaml` (+266 -631) 📝 `.github/workflows/test.yaml` (+88 -264) 📝 `.gitignore` (+3 -2) ➕ `CMakeLists.txt` (+112 -0) ➕ `CMakePresets.json` (+110 -0) 📝 `Dockerfile` (+117 -190) ➖ `Makefile` (+0 -103) ➕ `Makefile.sync` (+56 -0) 📝 `discover/amd_common.go` (+4 -9) 📝 `discover/amd_linux.go` (+2 -4) 📝 `discover/amd_windows.go` (+3 -7) 📝 `discover/gpu.go` (+24 -64) 📝 `discover/gpu_darwin.go` (+0 -3) ➕ `discover/path.go` (+53 -0) 📝 `discover/types.go` (+1 -2) 📝 `docs/development.md` (+63 -108) 📝 `envconfig/config.go` (+0 -9) 📝 `go.mod` (+2 -1) _...and 80 more files_ </details> ### 📄 Description split from #7913 this changes updates the directory structure splitting `llama.cpp` and `ggml` into separate, reusable packages. as a result the build has also changed significantly. the build now uses `cmake` to build dependencies as shared objects which will be dynamically loaded when necessary. current (work in progress) build instructions: - `go build .` to build ollama. this includes a default, basic cpu runner - `cmake --preset Default; cmake --build --preset Default` to configure and build the default targets. this will configure and build cuda and rocm if those are available - `cmake --preset CPU; cmake --build --preset CPU` to configure and build _only_ CPU variants - `cmake --preset CUDA; cmake --build --preset CUDA` to configure and build _only_ CUDA - `cmake --preset ROCm; cmake --build --preset ROCm` to configure and build _only_ ROCm TODO: - [x] Windows CI - [x] Build docs - [x] Update CMake output directory --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
GiteaMirror added the pull-request label 2026-05-05 07:22:01 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label pull-request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#75017
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?