Skip to content

SUNRPC: remainder of srv queueing work

NeilBrown linux-nfs 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest10 patches 6 repliesIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/10SQUASH: SUNRPC: rename and refactor svc_get_next_xprt()net/sunrpc/svc_xprt.c+1 -10 replies
2/10SUNRPC: add list of idle threadsinclude/linux/sunrpc/svc.h include/trace/events/sunrpc.h net/sunrpc/svc.c +1 more+44 -110 replies
3/10SUNRPC: discard SP_CONGESTEDinclude/linux/sunrpc/svc.h net/sunrpc/svc.c net/sunrpc/svc_xprt.c+1 -50 replies
4/10SUNRPC: change service idle list to be an llistinclude/linux/sunrpc/svc.h net/sunrpc/svc.c net/sunrpc/svc_xprt.c+40 -373 replies
5/10SUNRPC: only have one thread waking up at a timeinclude/linux/llist.h include/linux/sunrpc/svc.h lib/llist.c +2 more+64 -520 replies
6/10SUNRPC/svc: add light-weight queuing mechanism.include/linux/sunrpc/svc_lwq.h net/sunrpc/Kconfig net/sunrpc/Makefile +1 more+221 -13 replies
7/10SUNRPC: use lwq for sp_sockets - renamed to sp_xprtsinclude/linux/sunrpc/svc.h include/linux/sunrpc/svc_xprt.h net/sunrpc/svc.c +1 more+21 -430 replies
8/10SUNRPC: change sp_nrthreads to atomic_tfs/nfsd/nfssvc.c include/linux/sunrpc/svc.h net/sunrpc/svc.c+23 -190 replies
9/10SUNRPC: discard sp_lockinclude/linux/sunrpc/svc.h net/sunrpc/svc.c+5 -60 replies
10/10SUNRPC: change the back-channel queue to lwqinclude/linux/sunrpc/svc.h include/linux/sunrpc/xprt.h net/sunrpc/backchannel_rqst.c +3 more+9 -210 replies

Reviews trailers per patch

No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.

Read the v1 thread

Cover letter [PATCH 00/10] SUNRPC: remainder of srv queueing work

This series is on top of topic-sunrpc-thread-scheduling (4fa42add1a7)

It includes a fixup for an earlier patch in that topic, revised versions of patches that have been sent but are yet to land, and the remainder of what I have been working on.

With the full set enqueueing and dequeueing of the list of idle threads is completely lockless following a LIFO discipline. Threads are woken gradually without the possibility of a thundering herd. If lots of work becomes ready only one thread is woken and once it dequeues the first work item, it will dequeue another thread if there is more work to do.

Queueing and dequeueing for work-items for these threads is partially lockless following a FIFO discipline. Enqueueing, which can happen in bh context, is lockless. Dequeueing in the server threads still requires a spinlock, but does not need to disable 'bh'.

NeilBrown

[PATCH 01/10] SQUASH: SUNRPC: rename and refactor svc_get_next_xprt() [PATCH 02/10] SUNRPC: add list of idle threads
[PATCH 03/10] SUNRPC: discard SP_CONGESTED
[PATCH 04/10] SUNRPC: change service idle list to be an llist [PATCH 05/10] SUNRPC: only have one thread waking up at a time [PATCH 06/10] SUNRPC/svc: add light-weight queuing mechanism. [PATCH 07/10] SUNRPC: use lwq for sp_sockets - renamed to sp_xprts [PATCH 08/10] SUNRPC: change sp_nrthreads to atomic_t [PATCH 09/10] SUNRPC: discard sp_lock
[PATCH 10/10] SUNRPC: change the back-channel queue to lwq