how to understand "When adding a new driver make sure to use the same values in user and kernel space”？

[zhenshuitieniu]

It means that when nvidia-fs.ko supports the new file system, and the new-fs-client.ko also supports RDMA and GDS-related API calls, does it need to be supported at the cuda library level?

// These are same key values as the one's in the user space.
// When adding a new driver make sure to use the same values in user and kernel space.
#define NVFS_PROC_MOD_NVME_KEY "nvme"
#define NVFS_PROC_MOD_NVME_RDMA_KEY "nvme_rdma"
#define NVFS_PROC_MOD_SCSI_KEY "scsi_mod"
#define NVFS_PROC_MOD_SCALEFLUX_CSD_KEY "sfxvdriver"
#define NVFS_PROC_MOD_NVMESH_KEY "nvmeib_common"
#define NVFS_PROC_MOD_DDN_LUSTRE_KEY "lnet"
#define NVFS_PROC_MOD_GPFS_KEY "mmfslinux"
#define NVFS_PROC_MOD_NFS_KEY "rpcrdma"
#define NVFS_PROC_MOD_WEKAFS_KEY "wekafsio"

Take nfs as an example,
First need to increase symbol item in nvidia-fs.ko
struct module_entry modules_list[] = // nvfs symbol table
{
1,
0,
NVFS_PROC_MOD_NFS_KEY,
0,
"rpcrdma_register_nvfs_dma_ops",
0,
"rpcrdma_unregister_nvfs_dma_ops",
0,
&nvfs_dev_dma_rw_ops
},
Secondly, rpcrdma_register_nvfs_dma_ops is supported in the nfs driver.
In addition to the above two steps, do I need special settings or support for the cuda library?

[fredfany]

Hi, We are also facing a similar issue. I was wondering if you have had success adding new distributed filesystems using the above method. Also, if we want to add a new file system to use GDS, do we only need to adapt it in nvidia-fs, or do we have to support it at the cuda library level. Thanks in advance!