dxgkrnlを有効にしてkernelビルド(5.4.y)

カーネルビルド

WSL2上で新しめのカーネルにてdxgkrnlを使う方法
なお、Windowsのバージョンが下記のため、動作確認が取れていない
Microsoft Windows [Version 10.0.21327.1000]
Microsoft Windows [Version 10.0.21332.1000]で動作を確認
linux-msft-wsl-5.10.yのdxgkrnlを使う方法はここ

stableカーネルリポジトリをcloneする

$ git clone https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git
$ cd linux
$ git checkout origin/linux-5.11.y

WSL2のカーネルリポジトリをリモートに追加し、fetchする

$ git remote add wsl-origin https://github.com/microsoft/WSL2-Linux-Kernel.git
$ git fetch wsl-origin linux-msft-wsl-5.4.y

dxgkrnドライバの変更点をcherry-pick

$ git log wsl-origin/linux-msft-wsl-5.4.y --oneline -- drivers/hyperv/dxgkrnl/ | awk '{print $1}' | tac > dxgkrnl_id
$ git cherry-pick `cat dxgkrnl_id`

コンパイルが通るようにソースコードを修正

他にも修正する必要があるかもしれない

$ vi drivers/hyperv/dxgkrnl/dxgvmbus.c
$ git add drivers/hyperv/dxgkrnl/dxgvmbus.c
$ git commit -m "change name from mmap_sem to mmap_lock"

ソースコード修正点

$ git diff HEAD~
diff --git a/drivers/hyperv/dxgkrnl/dxgvmbus.c b/drivers/hyperv/dxgkrnl/dxgvmbus.c
index 0a22d0315d56..76ec1e6bedfb 100644
--- a/drivers/hyperv/dxgkrnl/dxgvmbus.c
+++ b/drivers/hyperv/dxgkrnl/dxgvmbus.c
@@ -550,7 +550,7 @@ static u8 *dxg_map_iospace(u64 iospace_address, u32 size,
                return NULL;
        }

-       down_read(&current->mm->mmap_sem);
+       down_read(&current->mm->mmap_lock);
        vma = find_vma(current->mm, (unsigned long)va);
        if (vma) {
                pgprot_t prot = vma->vm_page_prot;
@@ -568,7 +568,7 @@ static u8 *dxg_map_iospace(u64 iospace_address, u32 size,
                pr_err("failed to find vma: %p %lx", vma, va);
                ret = -ENOMEM;
        }
-       up_read(&current->mm->mmap_sem);
+       up_read(&current->mm->mmap_lock);

        if (ret) {
                dxg_unmap_iospace((void *)va, size);

カーネルビルド&インストール

make oldconfigの設定は適宜行う。

$ make mrproper
$ git show wsl-origin/linux-msft-wsl-5.4.y:Microsoft/config-wsl > .config
$ make oldconfig
$ make
$ cp arch/x86/boot/bzImage /mnt/d/kernel-5.11.4-microsoft-standard-WSL2+

.wslconfigの設定

$ vi /mnt/c/Users/[USERNAME]/.wslconfig
$ cat /mnt/c/Users/[USERNAME]/.wslconfig
[wsl2]
kernel=d:\\kernel-5.11.4-microsoft-standard-WSL2+

動作確認

$ ls /dev/dxg
/dev/dxg
$ /usr/local/cuda-11.2/samples/bin/x86_64/linux/release/deviceQuery
/usr/local/cuda-11.2/samples/bin/x86_64/linux/release/deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "GeForce GTX 1070"
  CUDA Driver Version / Runtime Version          11.3 / 11.2
  CUDA Capability Major/Minor version number:    6.1
  Total amount of global memory:                 8192 MBytes (8589934592 bytes)
  (15) Multiprocessors, (128) CUDA Cores/MP:     1920 CUDA Cores
  GPU Max Clock rate:                            1709 MHz (1.71 GHz)
  Memory Clock rate:                             4004 Mhz
  Memory Bus Width:                              256-bit
  L2 Cache Size:                                 2097152 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
  Maximum Layered 1D Texture Size, (num) layers  1D=(32768), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(32768, 32768), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total shared memory per multiprocessor:        98304 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  2048
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device supports Managed Memory:                Yes
  Device supports Compute Preemption:            Yes
  Supports Cooperative Kernel Launch:            Yes
  Supports MultiDevice Co-op Kernel Launch:      Yes
  Device PCI Domain ID / Bus ID / location ID:   0 / 1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.3, CUDA Runtime Version = 11.2, NumDevs = 1
Result = PASS