mbox series

[v3,0/5] nfsd/sunrpc: allow starting/stopping pooled NFS server via netlink

Message ID 20240613-nfsd-next-v3-0-3b51c3c2fc59@kernel.org (mailing list archive)
Headers show
Series nfsd/sunrpc: allow starting/stopping pooled NFS server via netlink | expand

Message

Jeff Layton June 13, 2024, 6:34 p.m. UTC 
This is a resend of the patchset I sent a little over a week ago, with
a couple of new patches that allow setting the pool-mode via netlink.

This patchset first attempts to detangle the pooled/non-pooled service
handling in the sunrpc layer, unifies the codepaths that start the
pooled vs. non-pooled nfsd, and then wires up the new netlink threads
interface to allow you to start a pooled server by specifying an
array of thread counts.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
Changes in v3:
- better announce the subtle change to the /sys/module interface
- add more kerneldoc comments
- Link to v2: https://lore.kernel.org/r/20240613-nfsd-next-v2-0-20bf690d65fb@kernel.org

Changes in v2:
- add new pool-mode set/get netlink calls

---
Jeff Layton (5):
      sunrpc: fix up the special handling of sv_nrpools == 1
      nfsd: make nfsd_svc take an array of thread counts
      nfsd: allow passing in array of thread counts via netlink
      sunrpc: refactor pool_mode setting code
      nfsd: new netlink ops to get/set server pool_mode

 Documentation/netlink/specs/nfsd.yaml |  27 +++++++++
 fs/nfsd/netlink.c                     |  17 ++++++
 fs/nfsd/netlink.h                     |   2 +
 fs/nfsd/nfsctl.c                      | 100 ++++++++++++++++++++++++++----
 fs/nfsd/nfsd.h                        |   3 +-
 fs/nfsd/nfssvc.c                      |  59 +++++++++++-------
 include/linux/sunrpc/svc.h            |   3 +
 include/uapi/linux/nfsd_netlink.h     |  10 +++
 net/sunrpc/svc.c                      | 111 ++++++++++++++++++++++------------
 9 files changed, 256 insertions(+), 76 deletions(-)
---
base-commit: fec4124bac55ad92c47585fe537e646fe108b8fa
change-id: 20240604-nfsd-next-b04c0d2d89a9

Best regards,