Skip to content

编译错误修复 #21

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
EthanPeng-2048 merged 1 commit into from
Oct 19, 2026
Merged

编译错误修复 #21

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 14 additions & 8 deletions CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -391,6 +391,13 @@ option(NN_ENABLE_PCH "Precompile nn.hpp once and reuse it across targets" ON)
function(nn_add_executable name)
add_executable(${name} ${ARGN})
target_link_libraries(${name} PRIVATE nn_core)
# GPU 块按这份名单统一挂 NN_HAS_VULKAN / Vulkan::Vulkan。名单与创建点
# 同源，避免"测试目标只在 NN_ENABLE_TESTS=ON 时才进 GPU 名单"这类漏配：
# 编译门是 NN_BUILD_APPS OR NN_ENABLE_TESTS（默认 Release 也会编出全部
# 测试程序），漏挂时共享 PCH 里的 NN_HAS_VULKAN 会泄漏进 TU，表现为
# 链接期整屏 `undefined reference to vk*`。
list(APPEND NN_EXECUTABLE_TARGETS ${name})
set(NN_EXECUTABLE_TARGETS ${NN_EXECUTABLE_TARGETS} PARENT_SCOPE)
if(NN_ENABLE_PCH)
set(NN_PCH_CANDIDATES ${NN_PCH_CANDIDATES} ${name} PARENT_SCOPE)
endif()
Expand Down Expand Up @@ -419,6 +426,7 @@ option(NN_ENABLE_PROBES "Compile diagnostic probes (hot-path trace hooks + probe
# 收集实际构建的目标，供 GPU 统一启用（避免引用未定义目标）
set(NN_APP_TARGETS "")
set(NN_TEST_TARGETS "")
set(NN_EXECUTABLE_TARGETS "") # 所有经 nn_add_executable 创建的可执行目标
set(NN_BENCH_TARGETS "") # 性能基准已移除（bench 工具不注册 ctest 且无法在 -Werror 下编译）

# ── 应用 / 探针 / 基准目标（NN_BUILD_APPS，默认 = 是否顶层工程）───────────
Expand Down Expand Up @@ -1127,14 +1135,12 @@ if(Vulkan_FOUND)
target_link_libraries(neuralnet_cpp INTERFACE Vulkan::Vulkan)
add_dependencies(neuralnet_cpp compile_gpu_shaders)

# 为所有实际构建的目标启用 Vulkan（含测试目标：它们由上方
# NN_BUILD_APPS OR NN_ENABLE_TESTS 门创建）
set(NN_GPU_TARGETS ${NN_APP_TARGETS})
if(NN_ENABLE_TESTS)
list(APPEND NN_GPU_TARGETS ${NN_TEST_TARGETS})
endif()

foreach(TARGET ${NN_GPU_TARGETS})
# 为所有实际构建的目标启用 Vulkan。名单来自 nn_add_executable 的
# 创建点本身（应用 + 探针 + 测试），**不按 NN_ENABLE_TESTS 拼**：
# 测试目标在默认 Release（NN_BUILD_APPS=ON / NN_ENABLE_TESTS=OFF）
# 下同样会被编译，漏挂宏/链接会让 gpu_f16_test 等目标在链接期报
# `undefined reference to vk*`（详见 nn_add_executable 处注释）。
foreach(TARGET ${NN_EXECUTABLE_TARGETS})
target_compile_definitions(${TARGET} PRIVATE NN_HAS_VULKAN)
target_include_directories(${TARGET} PRIVATE ${CMAKE_CURRENT_BINARY_DIR}/generated)
target_link_libraries(${TARGET} PRIVATE Vulkan::Vulkan)
Expand Down
Loading