AUR (en) - llama.cpp-gfx1151

Search Criteria

Package Details: llama.cpp-gfx1151 b11580-1

Git Clone URL:	https://aur.archlinux.org/llama.cpp-gfx1151.git (read-only, click to copy)
Package Base:	llama.cpp-gfx1151
Description:	Port of Facebook's LLaMA model in C/C++ (Optimized for gfx1151, ROCm + Vulkan)
Upstream URL:	https://github.com/ggml-org/llama.cpp
Licenses:	MIT
Conflicts:	ggml, ggml-rocm, ggml-vulkan, libggml, llama-cpp, llama.cpp, llama.cpp-hip, llama.cpp-vulkan, stable-diffusion.cpp
Provides:	ggml, ggml-rocm, ggml-vulkan, libggml, llama-cpp, llama.cpp, llama.cpp-hip, llama.cpp-vulkan
Submitter:	Orion-zhen
Maintainer:	Orion-zhen (darkbasic)
Last Packager:	Orion-zhen
Votes:	3
Popularity:	0.23
First Submitted:	2026-04-03 06:56 (UTC)
Last Updated:	2026-10-13 04:03 (UTC)

Git Clone URL:

https://aur.archlinux.org/llama.cpp-gfx1151.git (read-only, click to copy)

Package Base:

llama.cpp-gfx1151

Description:

Port of Facebook's LLaMA model in C/C++ (Optimized for gfx1151, ROCm + Vulkan)

Upstream URL:

https://github.com/ggml-org/llama.cpp

Licenses:

MIT

Conflicts:

ggml, ggml-rocm, ggml-vulkan, libggml, llama-cpp, llama.cpp, llama.cpp-hip, llama.cpp-vulkan, stable-diffusion.cpp

Provides:

ggml, ggml-rocm, ggml-vulkan, libggml, llama-cpp, llama.cpp, llama.cpp-hip, llama.cpp-vulkan

Submitter:

Orion-zhen

Maintainer:

Orion-zhen (darkbasic)

Last Packager:

Orion-zhen

Votes:

Popularity:

0.23

First Submitted:

2026-04-03 06:56 (UTC)

Last Updated:

2026-10-13 04:03 (UTC)

Latest Comments

darkbasic commented on 2026-08-29 14:21 (UTC)

If someone is interested in the unmerged mtp patches for qwen3.8 flash you can find them in this experimental PKGBUILD: https://aur.archlinux.org/packages/llama.cpp-hip-gfx1151 There is a README.md with some explanations.

Orion-zhen commented on 2026-08-29 07:40 (UTC)

Hi, @jyaworski, thank you for your report! Hi, @darkbasic, thank you for your timely fixes!

I’ve already synced these fixes to the upstream aur-packages. However, I consider aur-packages the single source of truth, so fixing them only here is pointless, since the once-daily sync workflow will overwrite them. If you have any fixes, please prioritize submitting a PR to aur-packages. Thanks.

darkbasic commented on 2026-08-28 17:19 (UTC)

I've fixed the error plus some other minor improvements.

jyaworski commented on 2026-08-28 12:36 (UTC)

That warning turned into an error for me, and it no longer builds: clang: error: unknown argument: '-famd-opt'

Orion-zhen commented on 2026-07-18 10:43 (UTC)

@s-k, don't worry about that, it's OK :)

s-k commented on 2026-07-18 08:58 (UTC)

clang: warning: argument unused during compilation: '-famd-opt' [-Wunused-command-line-argument]

darkbasic commented on 2026-05-11 17:08 (UTC)

-  # patch -d "${srcdir}/llama.cpp" -Np1 -i "${srcdir}/llama-gfx1151-41ebffac.patch"
+  patch -d "${srcdir}/llama.cpp" -Np1 -i "${srcdir}/llama-gfx1151.patch"

I would personally put the commit hash in the file name because otherwise when the patch gets updated AUR helpers will simply throw a checksum error instead of trying to re-download it.

ltj16cn commented on 2026-04-20 10:50 (UTC)

yay -S llama.cpp-gfx1151 AUR Explicit (1): llama.cpp-gfx1151-b8851-2 :: PKGBUILD up to date, skipping download: llama.cpp-gfx1151 1 llama.cpp-gfx1151 (Build Files Exist) ==> Packages to cleanBuild? ==> [N]one [A]ll [Ab]ort [I]nstalled [No]tInstalled or (1 2 3, 1-3, ^4) ==> 1 llama.cpp-gfx1151 (Build Files Exist) ==> Diffs to show? ==> [N]one [A]ll [Ab]ort [I]nstalled [No]tInstalled or (1 2 3, 1-3, ^4) ==> ==> Making package: llama.cpp-gfx1151 b8851-2 (Mon 20 Apr 2026 06:47:45 PM CST) ==> Retrieving sources... -> Found llama.cpp-gfx1151-b8851.tar.gz -> Found llama.cpp.service -> Found llama.cpp.conf ==> WARNING: Skipping verification of source file PGP signatures. ==> Validating source files with sha256sums... llama.cpp-gfx1151-b8851.tar.gz ... Passed llama.cpp.service ... Passed llama.cpp.conf ... Passed :: (1/1) Parsing SRCINFO: llama.cpp-gfx1151 ==> Making package: llama.cpp-gfx1151 b8851-2 (Mon 20 Apr 2026 06:47:47 PM CST) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Found llama.cpp-gfx1151-b8851.tar.gz -> Found llama.cpp.service -> Found llama.cpp.conf ==> Validating source files with sha256sums... llama.cpp-gfx1151-b8851.tar.gz ... Passed llama.cpp.service ... Passed llama.cpp.conf ... Passed ==> Removing existing $srcdir/ directory... ==> Extracting sources... -> Extracting llama.cpp-gfx1151-b8851.tar.gz with bsdtar ==> Starting prepare()... patch: **** Can't open patch file /home/sonic/.cache/yay/llama.cpp-gfx1151/src/llama-gfx1151-41ebffac.patch : No such file or directory ==> ERROR: A failure occurred in prepare(). Aborting... -> error making: llama.cpp-gfx1151-exit status 4 -> Failed to install the following packages. Manual intervention is required: llama.cpp-gfx1151 - exit status 4

moparisthebest commented on 2026-04-15 03:46 (UTC)

This is missing spirv-headers in makedepends, fails to build without it.

Arch Linux User Repository