Upgrade llama.cpp from b8808 to b8831#88
Merged
bernardladenthin merged 4 commits intomasterfrom Apr 17, 2026
Merged
Conversation
Breaking changes fixed: - Rename CMake link target: common → llama-common (jllama and jllama_test) - Replace removed `build_info` string with `llama_build_info()` function in server.hpp (6 sites), jllama.cpp (1 site), utils.hpp (1 site) - Add `#include "build-info.h"` in server.hpp and utils.hpp https://claude.ai/code/session_01ABjsNTGnSH3WJvCGx7JbWB
… 32-bit MSVC b8831 added ggml_graph_next_uid() in ggml.c which uses the MSVC compiler intrinsic __InterlockedIncrement64. This intrinsic only exists on x64/ARM64; on 32-bit x86 it causes LNK2019 at link time. Map it to the Win32 API InterlockedIncrement64 (same signature, available on x86 via CMPXCHG8B) for the ggml-base target when building with MSVC on a 32-bit host. https://claude.ai/code/session_01ABjsNTGnSH3WJvCGx7JbWB
Add three new rows to the known breaking changes table: - CMake target common → llama-common - build_info string → llama_build_info() in build-info.h - __InterlockedIncrement64 unavailable on Windows x86 Update range header from b8808 to b8831. https://claude.ai/code/session_01ABjsNTGnSH3WJvCGx7JbWB
The compile-definition approach failed because <intrin.h> defines #define __InterlockedIncrement64 _InterlockedIncrement64 after our /D flag, so the compiler still emits an extern call to _InterlockedIncrement64 (linker symbol __InterlockedIncrement64) with no definition. Provide the definition in a new compat TU (ggml_x86_compat.c) added to ggml-base via target_sources, guarded by MSVC AND CMAKE_SIZEOF_VOID_P EQUAL 4. The wrapper calls Win32 InterlockedIncrement64 (CMPXCHG8B, available on all x86 Windows) satisfying <intrin.h>'s extern __cdecl declaration at link time. https://claude.ai/code/session_01ABjsNTGnSH3WJvCGx7JbWB
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Upgrade llama.cpp from b8808 to b8831
Breaking changes fixed:
common→llama-common(jllama and jllama_test)build_infostring withllama_build_info()function in server.hpp (6 sites), jllama.cpp (1 site), utils.hpp (1 site)#include "build-info.h"in server.hpp and utils.hppggml_graph_next_uid()uses_InterlockedIncrement64via<intrin.h>, an MSVC intrinsic unavailable on 32-bit x86; provide an implementation insrc/main/cpp/compat/ggml_x86_compat.c(wrapping Win32InterlockedIncrement64) added toggml-baseviatarget_sourcesguarded byMSVC AND CMAKE_SIZEOF_VOID_P EQUAL 4Documentation:
CLAUDE.mdknown-breaking-changes table to cover b8808→b8831 (3 new rows)README.mdbadge andCLAUDE.mdpinned version to b8831https://claude.ai/code/session_01ABjsNTGnSH3WJvCGx7JbWB