ggml: ggml-cpu: force-no-lto-for-cpu-feats (llama/19609)

When LTO enabled in build environments it forces all builds to have LTO
in place. But feature detection logic is fragile, and causing Illegal
instruction errors with lto. This disables LTO for the feature
detection code to prevent cross-module optimization from inlining
architecture-specific instructions into the score function. Without this,
LTO can cause SIGILL when loading backends on older CPUs (e.g., loading
power10 backend on power9 crashes before feature check runs).

diff --git a/ggml/src/ggml-cpu/CMakeLists.txt b/ggml/src/ggml-cpu/CMakeLists.txt
index 43d6f7f54f7a61d4ec062e8675dca8336c7481a9..3dc948e4d8e651a8042bfe54aa0174d014f71954 100644 (file)
--- a/ggml/src/ggml-cpu/CMakeLists.txt
+++ b/ggml/src/ggml-cpu/CMakeLists.txt
@@ -9,6 +9,11 @@ function(ggml_add_cpu_backend_features cpu_name arch)
     target_compile_definitions(${GGML_CPU_FEATS_NAME} PRIVATE ${ARGN})
     target_compile_definitions(${GGML_CPU_FEATS_NAME} PRIVATE GGML_BACKEND_DL GGML_BACKEND_BUILD GGML_BACKEND_SHARED)
     set_target_properties(${GGML_CPU_FEATS_NAME} PROPERTIES POSITION_INDEPENDENT_CODE ON)
+    # Disable LTO for the feature detection code to prevent cross-module optimization
+    # from inlining architecture-specific instructions into the score function.
+    # Without this, LTO can cause SIGILL when loading backends on older CPUs
+    # (e.g., loading power10 backend on power9 crashes before feature check runs).
+    target_compile_options(${GGML_CPU_FEATS_NAME} PRIVATE -fno-lto)
     target_link_libraries(${cpu_name} PRIVATE ${GGML_CPU_FEATS_NAME})
 endfunction()