Trond Myklebust reports an NFSD crash in svc_rdma_sendto(). Further
investigation shows that the crash occurred while NFSD was handling
a deferred request.

This patch addresses two inter-related issues that prevent request
deferral from working correctly for RPC/RDMA requests:

1. Prevent the crash by ensuring that the original
   svc_rqst::rq_xprt_ctxt value is available when the request is
   revisited. Otherwise svc_rdma_sendto() does not have a Receive
   context available with which to construct its reply.

2. Possibly since before commit 71641d99 ("svcrdma: Properly
   compute .len and .buflen for received RPC Calls"),
   svc_rdma_recvfrom() did not include the transport header in the
   returned xdr_buf. There should have been no need for svc_defer()
   and friends to save and restore that header, as of that commit.
   This issue is addressed in a backport-friendly way by simply
   having svc_rdma_recvfrom() set rq_xprt_hlen to zero
   unconditionally, just as svc_tcp_recvfrom() does. This enables
   svc_deferred_recv() to correctly reconstruct an RPC message
   received via RPC/RDMA.
Reported-by: Trond Myklebust <trondmy@hammerspace.com>
Link: https://lore.kernel.org/linux-nfs/82662b7190f26fb304eb0ab1bb04279072439d4e.camel@hammerspace.com/Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Cc: <stable@vger.kernel.org>

Make it a little less racy, by removing the refcount_read() test. Then
remove the redundant 'is_hashed' variable.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

The call to filemap_flush() in nfsd_file_put() is there to ensure that
we clear out any writes belonging to a NFSv3 client relatively quickly
and avoid situations where the file can't be evicted by the garbage
collector. It also ensures that we detect write errors quickly.

The problem is this causes a regression in performance for some
workloads.

So try to improve matters by deferring writeback until we're ready to
close the file, and need to detect errors so that we can force the
client to resend.
Tested-by: Jan Kara <jack@suse.cz>
Fixes: b6669305 ("nfsd: Reduce the number of calls to nfsd_file_gc()")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Link: https://lore.kernel.org/all/20220330103457.r4xrhy2d6nhtouzk@quack3.lanSigned-off-by: Chuck Lever <chuck.lever@oracle.com>

Return boolean values ("true" or "false") instead of 1 or 0 from bool
functions.  This fixes the following warnings from coccicheck:

./fs/nfsd/nfs2acl.c:289:9-10: WARNING: return of 0/1 in function
'nfsaclsvc_encode_accessres' with return type bool
./fs/nfsd/nfs2acl.c:252:9-10: WARNING: return of 0/1 in function
'nfsaclsvc_encode_getaclres' with return type bool
Signed-off-by: Haowen Bai <baihaowen@meizu.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

While the original code is valid, it is not the obvious choice for the
sizeof() call and in preparation to limit the scope of the list iterator
variable the sizeof should be changed to the size of the destination.
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

When compiling with -Wformat, clang emits the following warnings:

fs/nfsd/flexfilelayout.c:120:27: warning: format specifies type 'unsigned
char' but the argument has type 'int' [-Wformat]
                         "%s.%hhu.%hhu", addr, port >> 8, port & 0xff);
                             ~~~~              ^~~~~~~~~
                             %d
fs/nfsd/flexfilelayout.c:120:38: warning: format specifies type 'unsigned
char' but the argument has type 'int' [-Wformat]
                         "%s.%hhu.%hhu", addr, port >> 8, port & 0xff);
                                  ~~~~                    ^~~~~~~~~~~
                                  %d

The types of these arguments are unconditionally defined, so this patch
updates the format character to the correct ones for ints and unsigned
ints.

Link: https://github.com/ClangBuiltLinux/linux/issues/378Signed-off-by: Bill Wendling <morbo@google.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Tom Haynes <loghyr@hammerspace.com>

On a 32 bit system, the "len * sizeof(*p)" operation can have an
integer overflow.

Cc: stable@vger.kernel.org
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Smatch complains:

	fs/nfsd/nfsxdr.c:341 nfssvc_decode_writeargs()
	warn: no lower bound on 'args->len'

Change the type to unsigned to prevent this issue.

Cc: stable@vger.kernel.org
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Update lock usage of lock_manager_operations' functions to reflect
the changes in commit 6109c850 ("locks: add a dedicated spinlock
to protect i_flctx lists").
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>

These have been incorrect since the function was introduced.

A proper kerneldoc comment is added since this function, though
static, is part of an external interface.
Reported-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

The common practice is to name function instances the same as the
method names, but with a uniquifying prefix. Commit aef9583b
("NFSD: Get reference of lockowner when coping file_lock") missed
this -- the new function names should both have been of the form
"nfsd4_lm_*".

Before more lock manager operations are added in NFSD, rename these
two functions for consistency.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

CONFIG_NFSD_V3 has been removed. NFSD support for NFSv3 can no
longer be disabled.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Eventually support for NFSv2 in the Linux NFS server is to be
deprecated and then removed.

However, NFSv2 is the "always supported" version that is available
as soon as CONFIG_NFSD is set.  Before NFSv2 support can be removed,
we need to choose a different "always supported" version.

This patch removes CONFIG_NFSD_V3 so that NFSv3 is always supported,
as NFSv2 is today. When NFSv2 support is removed, NFSv3 will become
the only "always supported" NFS version.

The defconfigs still need to be updated to remove CONFIG_NFSD_V3=y.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

The nfsd file cache table can be pretty large and its allocation
may require as many as 80 contigious pages.

Employ the same fix that was employed for similar issue that was
reported for the reply cache hash table allocation several years ago
by commit 8f97514b ("nfsd: more robust allocation failure handling
in nfsd_reply_cache_init").

Fixes: 65294c1f ("nfsd: add a new struct file caching facility to nfsd")
Link: https://lore.kernel.org/linux-nfs/e3cdaeec85a6cfec980e87fc294327c0381c1778.camel@kernel.org/Suggested-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Amir Goldstein <amir73il@gmail.com>

Initial support for the RPC_AUTH_TLS authentication flavor enables
NFSD to eventually accept an RPC_AUTH_TLS probe from clients. This
patch simply prevents NFSD from rejecting these probes completely.

In the meantime, graft this support in now so that RPC_AUTH_TLS
support keeps up with generic code and API changes in the RPC
server.

Down the road, server-side transport implementations will populate
xpo_start_tls when they can support RPC-with-TLS. For example, TCP
will eventually populate it, but RDMA won't.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Hoist svo_function back into svc_serv and remove struct
svc_serv_ops, since the struct is now devoid of fields.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

struct svc_serv_ops is about to be removed.

Neil Brown says:
> I suspect svo_module can go as well - I don't think the thread is
> ever the thing that primarily keeps a module active.

A random sample of kthread_create() callers shows sunrpc is the only
one that manages module reference count in this way.
Suggested-by: Neil Brown <neilb@suse.de>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up: svc_shutdown_net() now does nothing but call
svc_close_net(). Replace all external call sites.

svc_close_net() is renamed to be the inverse of svc_xprt_create().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up: Use the "svc_xprt_<task>" function naming convention as
is used for other external APIs.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up: Use the "svc_xprt_<task>" function naming convention as
is used for other external APIs.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up. Neil observed that "any code that calls svc_shutdown_net()
knows what the shutdown function should be, and so can call it
directly."
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: NeilBrown <neilb@suse.de>

Neil says:
"These functions were separated in commit 0971374e ("SUNRPC:
Reduce contention in svc_xprt_enqueue()") so that the XPT_BUSY check
happened before taking any spinlocks.

We have since moved or removed the spinlocks so the extra test is
fairly pointless."

I've made this a separate patch in case the XPT_BUSY change has
unexpected consequences and needs to be reverted.
Suggested-by: Neil Brown <neilb@suse.de>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

We have never been able to track down and address the underlying
cause of the performance issues with workqueue-based service
support. svo_enqueue_xprt is called multiple times per RPC, so
it adds instruction path length, but always ends up at the same
function: svc_xprt_do_enqueue(). We do not anticipate needing
this flexibility for dynamic nfsd thread management support.

As a micro-optimization, remove .svo_enqueue_xprt because
Spectre/Meltdown makes virtual function calls more costly.

This change essentially reverts commit b9e13cdf ("nfsd/sunrpc:
turn enqueueing a svc_xprt into a svc_serv operation").
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

To make server-side trace events more useful in container-ized
environments, capture not just the remote's IP address, but the
local IP address and network namespace as well.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up: Use the new __sockaddr field to record the socket address.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up.

The PROC_ARGS macros were added when I thought that NFSD tracepoints
would be reporting endpoint information. However, tracepoints in the
RPC server now report transport endpoint information, so in general
there's no need for the upper layers to do that any more, and these
macros can be retired.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

As an example usage of the new __sockaddr field, convert some NFSD
trace points to use it.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

A helper macro was added to make reading socket addresses easier in trace
events. It pairs %pISpc with __get_sockaddr() that reads the socket
address from the ring buffer into a human readable format.

The boot up check that makes sure that trace events do not reference
pointers to memory that can later be freed when the trace event is read,
incorrectly flagged this as a delayed reference. Update the check to
handle "__get_sockaddr" and not report an error on it.

Link: https://lore.kernel.org/all/20220125160505.068dbb52@canb.auug.org.au/Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Enable a struct sockaddr to be stored in a trace record as a
dynamically-sized field. The common cases are AF_INET and AF_INET6
which are different sizes, and are vastly smaller than a struct
sockaddr_storage.

These are safer because, when used properly, the size of the
sockaddr destination field in each trace record is now guaranteed
to be the same as the source address that is being copied into it.

Link: https://lore.kernel.org/all/164182978641.8391.8277203495236105391.stgit@bazille.1015granger.net/Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Move a rarely called function call site out of the hot path.

This is an exceptionally small improvement because the compiler
inlines most of the functions that nfsd_cache_lookup() calls.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Force the compiler to skip unneeded initialization for cases that
don't need those values. For example, NFSv4 COMPOUND operations are
RC_NOCACHE.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Clean up: The details of finding the right hash bucket are exactly
the same in both nfsd_cache_lookup() and nfsd_cache_update().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

For filesystems that supports "btime" timestamp (i.e. most modern
filesystems do) we share it via kernel nfsd. Btime support for NFS
client has already been added by Trond recently.
Suggested-by: Bruce Fields <bfields@fieldses.org>
Signed-off-by: Ondrej Valousek <ondrej.valousek.xm@renesas.com>
[ cel: addressed some whitespace/checkpatch nits ]
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Pull irq fix from Thomas Gleixner:
 "A single fix for a regression caused by the recent PCI/MSI rework
  which resulted in a recursive locking problem in the VMD driver.

  The cure is to cache the relevant information upfront instead of
  retrieving it at runtime"

* tag 'irq-urgent-2022-02-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  PCI: vmd: Prevent recursive locking on interrupt allocation

Pull dma-mapping fix from Christoph Hellwig:

 - fix a swiotlb info leak (Halil Pasic)

* tag 'dma-mapping-5.17-1' of git://git.infradead.org/users/hch/dma-mapping:
  swiotlb: fix info leak with DMA_FROM_DEVICE

Pull pin control fixes from Linus Walleij:

 - Fix some drive strength and pull-up code in the K210 driver.

 - Add the Alder Lake-M ACPI ID so it starts to work properly.

 - Use a static name for the StarFive GPIO irq_chip, forestalling an
   upcoming fixes series from Marc Zyngier.

 - Fix an ages old bug in the Tegra 186 driver where we were indexing at
   random into struct and being lucky getting the right member.

* tag 'pinctrl-v5-17-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  gpio: tegra186: Fix chip_data type confusion
  pinctrl: starfive: Use a static name for the GPIO irq_chip
  pinctrl: tigerlake: Revert "Add Alder Lake-M ACPI ID"
  pinctrl: k210: Fix bias-pull-up
  pinctrl: fix loop in k210_pinconf_get_drive()

Pull tracing fixes from Steven Rostedt:

 - rtla (Real-Time Linux Analysis tool):
    - fix typo in man page
    - Update API -e to -E before it is released
    - Error message fix and memory leak fix

 - Partially uninline trace event soft disable to shrink text

 - Fix function graph start up test

 - Have triggers affect the trace instance they are in and not top level

 - Have osnoise sleep in the units it says it uses

 - Remove unused ftrace stub function

 - Remove event probe redundant info from event in the buffer

 - Fix group ownership setting in tracefs

 - Ensure trace buffer is minimum size to prevent crashes

* tag 'trace-v5.17-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
  rtla/osnoise: Fix error message when failing to enable trace instance
  rtla/osnoise: Free params at the exit
  rtla/hist: Make -E the short version of --entries
  tracing: Fix selftest config check for function graph start up test
  tracefs: Set the group ownership in apply_options() not parse_options()
  tracing/osnoise: Make osnoise_main to sleep for microseconds
  ftrace: Remove unused ftrace_startup_enable() stub
  tracing: Ensure trace buffer is at least 4096 bytes large
  tracing: Uninline trace_trigger_soft_disabled() partly
  eprobes: Remove redundant event type information
  tracing: Have traceon and traceoff trigger honor the instance
  tracing: Dump stacktrace trigger to the corresponding instance
  rtla: Fix systme -> system typo on man page

Pull memblock fix from Mike Rapoport:
 "Use kfree() to release kmalloced memblock regions

  memblock.{reserved,memory}.regions may be allocated using kmalloc()
  in memblock_double_array(). Use kfree() to release these kmalloced
  regions"

* tag 'fixes-2022-02-26' of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock:
  memblock: use kfree() to release kmalloced memblock regions