Clean up: sc_dma_used is not required for correct operation. It is
simply a debugging tool to report when svcrdma has leaked DMA maps.

However, manipulating an atomic has a measurable CPU cost, and DMA
map accounting specific to svcrdma will be meaningless once svcrdma
is converted to use the new generic r/w API.

A similar kind of debug accounting can be done simply by enabling
the IOMMU or by using CONFIG_DMA_API_DEBUG, CONFIG_IOMMU_DEBUG, and
CONFIG_IOMMU_LEAK.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

The current sendto code appears to support clients that provide only
one of a Read list, a Write list, or a Reply chunk. My reading of
that code is that it doesn't support the following cases:

 - Read list + Write list
 - Read list + Reply chunk
 - Write list + Reply chunk
 - Read list + Write list + Reply chunk

The protocol allows more than one Read or Write chunk in those
lists. Some clients do send a Read list and Reply chunk
simultaneously. NFSv4 WRITE uses a Read list for the data payload,
and a Reply chunk because the GETATTR result in the reply can
contain a large object like an ACL.

Generalize one of the sendto code paths needed to support all of
the above cases, and attempt to ensure that only one pass is done
through the RPC Call's transport header to gather chunk list
information for building the reply.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

The ctxt's count field is overloaded to mean the number of pages in
the ctxt->page array and the number of SGEs in the ctxt->sge array.
Typically these two numbers are the same.

However, when an inline RPC reply is constructed from an xdr_buf
with a tail iovec, the head and tail often occupy the same page,
but each are DMA mapped independently. In that case, ->count equals
the number of pages, but it does not equal the number of SGEs.
There's one more SGE, for the tail iovec. Hence there is one more
DMA mapping than there are pages in the ctxt->page array.

This isn't a real problem until the server's iommu is enabled. Then
each RPC reply that has content in that iovec orphans a DMA mapping
that consists of real resources.

krb5i and krb5p always populate that tail iovec. After a couple
million sent krb5i/p RPC replies, the NFS server starts behaving
erratically. Reboot is needed to clear the problem.

Fixes: 9d11b51c ("svcrdma: Fix send_reply() scatter/gather set-up")
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Clean up: Pass in just the piece of the svc_rqst that is needed
here.

While we're in the area, add an informative documenting comment.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Clean up.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Clean up: svc_rdma_get_read_chunk() already returns a pointer
to the Read list. No need to set "ch" again to the value it
already contains.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

An xdr_buf has a head, a vector of pages, and a tail. Each
RPC request is presented to the NFS server contained in an
xdr_buf.

The RDMA transport would like to supply the NFS server with only
the NFS WRITE payload bytes in the page vector. In some common
cases, that would allow the NFS server to swap those pages right
into the target file's page cache.

Have the transport's RDMA Read logic put XDR pad bytes in the tail
iovec, and not in the pages that hold the data payload.

The NFSv3 WRITE XDR decoder is finicky about the lengths involved,
so make sure it is looking in the correct places when computing
the total length of the incoming NFS WRITE request.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

The SRP initiator allows to set max_sectors to a value that exceeds
the largest amount of data that can be mapped at once with an mlx4
HCA using fast registration and a page size of 4 KB. Hence modify
ib_map_mr_sg() such that it can map partial sg-elements. If an
sg-element has been mapped partially, let the caller know
which fraction has been mapped by adjusting *sg_offset.
Signed-off-by: NBart Van Assche <bart.vanassche@sandisk.com>
Tested-by: NLaurence Oberman <loberman@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Signed-off-by: NChristoph Hellwig <hch@lst.de>
Tested-by: NSteve Wise <swise@opengridcomputing.com>
Reviewed-by: NBart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Calling ib_poll_cq() to sort through WCs during a completion is a
common pattern amongst RDMA consumers. Since commit 14d3a3b2
("IB: add a proper completion queue abstraction"), WC sorting can
be handled by the IB core.

By converting to this new API, svcrdma is made a better neighbor to
other RDMA consumers, as it allows the core to schedule the delivery
of completions more fairly amongst all active consumers.

This new API also aims each completion at a function that is
specific to the WR's opcode. Thus the ctxt->wr_op field and the
switch in process_context is replaced by a set of methods that
handle each completion type.

Because each ib_cqe carries a pointer to a completion method, the
core can now post operations on a consumer's QP, and handle the
completions itself.

The server's rdma_stat_sq_poll and rdma_stat_sq_prod metrics are no
longer updated.

As a clean up, the cq_event_handler, the dto_tasklet, and all
associated locking is removed, as they are no longer referenced or
used.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Tested-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Clean up: close_out is reached only when ctxt == NULL and XPT_CLOSE
is already set.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Tested-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

RFC 5666 Section 4.2 states:

> When the peer detects an RPC-over-RDMA header version that it does
> not support (currently this document defines only version 1), it
> replies with an error code of ERR_VERS, and provides the low and
> high inclusive version numbers it does, in fact, support.

And:

> When other decoding errors are detected in the header or chunks,
> either an RPC decode error MAY be returned or the RPC/RDMA error
> code ERR_CHUNK MUST be returned.

The Linux NFS server does throw ERR_VERS when a client sends it
a request whose rdma_version is not "one." But it does not return
ERR_CHUNK when a header decoding error occurs. It just drops the
request.

To improve protocol extensibility, it should reject invalid values
in the rdma_proc field instead of treating them all like RDMA_MSG.
Otherwise clients can't detect when the server doesn't support
new rdma_proc values.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Tested-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

When constructing an error reply, svc_rdma_xdr_encode_error()
needs to view the client's request message so it can get the
failing request's XID.

svc_rdma_xdr_decode_req() is supposed to return a pointer to the
client's request header. But if it fails to decode the client's
message (and thus an error reply is needed) it does not return the
pointer. The server then sends a bogus XID in the error reply.

Instead, unconditionally generate the pointer to the client's header
in svc_rdma_recvfrom(), and pass that pointer to both functions.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Tested-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Fix several issues with svc_rdma_send_error():

 - Post a receive buffer to replace the one that was consumed by
   the incoming request
 - Posting a send should use DMA_TO_DEVICE, not DMA_FROM_DEVICE
 - No need to put_page _and_ free pages in svc_rdma_put_context
 - Make sure the sge is set up completely in case the error
   path goes through svc_rdma_unmap_dma()
 - Replace the use of ENOSYS, which has a reserved meaning

Related fixes in svc_rdma_recvfrom():

 - Don't leak the ctxt associated with the incoming request
 - Don't close the connection after sending an error reply
 - Let svc_rdma_send_error() figure out the right header error code

As a last clean up, move svc_rdma_send_error() to svc_rdma_sendto.c
with other similar functions. There is some common logic in these
functions that could someday be combined to reduce code duplication.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Tested-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Clean up: Most svc_rdma_post_recv() call sites close the transport
connection when a receive cannot be posted. Wrap that in a common
helper.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Tested-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

We now alwasy have a per-PD local_dma_lkey available.  Make use of that
fact in svc_rdma and stop registering our own MR.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NJason Gunthorpe <jgunthorpe@obsidianresearch.com>
Reviewed-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Acked-by: NJ. Bruce Fields <bfields@redhat.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

To support the server-side of an NFSv4.1 backchannel on RDMA
connections, add a transport class that enables backward
direction messages on an existing forward channel connection.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Acked-by: NBruce Fields <bfields@fieldses.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Instead of maintaining a fastreg page list, keep an sg table
and convert an array of pages to a sg list. Then call ib_map_mr_sg
and construct ib_reg_wr.
Signed-off-by: NSagi Grimberg <sagig@mellanox.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Tested-by: NSteve Wise <swise@opengridcomputing.com>
Tested-by: NSelvin Xavier <selvin.xavier@avagotech.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Now that the NFS server advertises a maximum payload size of 1MB
for RPC/RDMA again, it crashes in svc_process_common() when NFS
client sends a 1MB NFS WRITE on an NFS/RDMA mount.

The server has set up a 259 element array of struct page pointers
in rq_pages[] for each incoming request. The last element of the
array is NULL.

When an incoming request has been completely received,
rdma_read_complete() attempts to set the starting page of the
incoming page vector:

  rqstp->rq_arg.pages = &rqstp->rq_pages[head->hdr_count];

and the page to use for the reply:

  rqstp->rq_respages = &rqstp->rq_arg.pages[page_no];

But the value of page_no has already accounted for head->hdr_count.
Thus rq_respages now points past the end of the incoming pages.

For NFS WRITE operations smaller than the maximum, this is harmless.
But when the NFS WRITE operation is as large as the server's max
payload size, rq_respages now points at the last entry in rq_pages,
which is NULL.

Fixes: cc9a903d ('svcrdma: Change maximum server payload . . .')
BugLink: https://bugzilla.linux-nfs.org/show_bug.cgi?id=270Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSagi Grimberg <sagig@dev.mellanox.co.il>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Reviewed-by: NShirley Ma <shirley.ma@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

This patch split up struct ib_send_wr so that all non-trivial verbs
use their own structure which embedds struct ib_send_wr.  This dramaticly
shrinks the size of a WR for most common operations:

sizeof(struct ib_send_wr) (old):	96

sizeof(struct ib_send_wr):		48
sizeof(struct ib_rdma_wr):		64
sizeof(struct ib_atomic_wr):		96
sizeof(struct ib_ud_wr):		88
sizeof(struct ib_fast_reg_wr):		88
sizeof(struct ib_bind_mw_wr):		96
sizeof(struct ib_sig_handover_wr):	80

And with Sagi's pending MR rework the fast registration WR will also be
down to a reasonable size:

sizeof(struct ib_fastreg_wr):		64
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com> [srp, srpt]
Reviewed-by: Chuck Lever <chuck.lever@oracle.com> [sunrpc]
Tested-by: NHaggai Eran <haggaie@mellanox.com>
Tested-by: NSagi Grimberg <sagig@mellanox.com>
Tested-by: NSteve Wise <swise@opengridcomputing.com>

The server rdma_read_chunk_lcl() and rdma_read_chunk_frmr() functions
were not taking into account the initial page_offset when determining
the rdma read length.  This resulted in a read who's starting address
and length exceeded the base/bounds of the frmr.

The server gets an async error from the rdma device and kills the
connection, and the client then reconnects and resends.  This repeats
indefinitely, and the application hangs.

Most work loads don't tickle this bug apparently, but one test hit it
every time: building the linux kernel on a 16 core node with 'make -j
16 O=/mnt/0' where /mnt/0 is a ramdisk mounted via NFSRDMA.

This bug seems to only be tripped with devices having small fastreg page
list depths.  I didn't see it with mlx4, for instance.

Fixes: 0bf48289 ('svcrdma: refactor marshalling logic')
Signed-off-by: NSteve Wise <swise@opengridcomputing.com>
Tested-by: NChuck Lever <chuck.lever@oracle.com>
Cc: stable@vger.kernel.org
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Signed-off-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Fields in struct rpcrdma_msg are __be32. Don't byte-swap these
fields when decoding RPC calls and then swap them back for the
reply. For the most part, they can be left alone.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Introduce helper rdma_cap_read_multi_sge() to help us check if the port of an
IB device support RDMA Read Multiple Scatter-Gather Entries.
Signed-off-by: NMichael Wang <yun.wang@profitbricks.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Tested-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NSean Hefty <sean.hefty@intel.com>
Reviewed-by: NJason Gunthorpe <jgunthorpe@obsidianresearch.com>
Tested-by: NDoug Ledford <dledford@redhat.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Use raw management helpers to reform IB-ulp xprtrdma.
Signed-off-by: NMichael Wang <yun.wang@profitbricks.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Tested-by: NIra Weiny <ira.weiny@intel.com>
Reviewed-by: NSean Hefty <sean.hefty@intel.com>
Reviewed-by: NJason Gunthorpe <jgunthorpe@obsidianresearch.com>
Tested-by: NDoug Ledford <dledford@redhat.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Most NFS RPCs place their large payload argument at the end of the
RPC header (eg, NFSv3 WRITE). For NFSv3 WRITE and SYMLINK, RPC/RDMA
sends the complete RPC header inline, and the payload argument in
the read list. Data in the read list is the last part of the XDR
stream.

One important case is not like this, however. NFSv4 COMPOUND is a
counted array of operations. A WRITE operation, with its large data
payload, can appear in the middle of the compound's operations
array. Thus NFSv4 WRITE compounds can have header content after the
WRITE payload.

The Linux client, for example, performs an NFSv4 WRITE like this:

  { PUTFH, WRITE, GETATTR }

Though RFC 5667 is not precise about this, the proper way to convey
this compound is to place the GETATTR inline, _after_ the front of
the RPC header. The receiver inserts the read list payload into the
XDR stream after the initial WRITE arguments, and before the GETATTR
operation, thanks to the value of the read list "position" field.

The Linux client currently sends the GETATTR at the end of the
RPC/RDMA read list, which is incorrect. It will be corrected in the
future.

The Linux server currently rejects NFSv4 compounds with inline
content after the read list. For the above NFSv4 WRITE compound, the
NFS compound header indicates there are three operations, but the
server finds nonsense when it looks in the XDR stream for the third
operation, and the compound fails with OP_ILLEGAL.

Move trailing inline content to the end of the XDR buffer's page
list. This presents incoming NFSv4 WRITE compounds to NFSD in the
same way the socket transport does.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

This is a pre-requisite for a subsequent patch.

Read list XDR round-up needs to be done _before_ additional inline
content is copied to the end of the XDR buffer's page list. Move
the logic added by commit e560e3b5 ("svcrdma: Add zero padding
if the client doesn't send it").
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Currently the Linux server can not decode RDMA_NOMSG type requests.
Operations whose length exceeds the fixed size of RDMA SEND buffers,
like large NFSv4 CREATE(NF4LNK) operations, must be conveyed via
RDMA_NOMSG.

For an RDMA_MSG type request, the client sends the RPC/RDMA, RPC
headers, and some or all of the NFS arguments via RDMA SEND.

For an RDMA_NOMSG type request, the client sends just the RPC/RDMA
header via RDMA SEND. The request's read list contains elements for
the entire RPC message, including the RPC header.

NFSD expects the RPC/RMDA header and RPC header to be contiguous in
page zero of the XDR buffer. Add logic in the RDMA READ path to make
the read list contents land where the server prefers, when the
incoming message is a type RDMA_NOMSG message.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

An RPC/RDMA client may send large RPC arguments via a read
list. This is a list of scatter/gather elements which convey
RPC call arguments too large to fit in a small RDMA SEND.

Each entry in the read list has a "position" field, whose value is
the byte offset in the XDR stream where the data in that entry is to
be inserted. Entries which share the same "position" value make up
the same RPC argument. The receiver inserts entries with the same
position field value in list order into the XDR stream.

Currently the Linux NFS/RDMA server cannot handle receiving read
chunks in more than one position, mostly because no current client
sends read lists with elements in more than one position. As a
sanity check, ensure that all received chunks have the same
"rc_position."
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

The RDMA reader function doesn't change once an svcxprt_rdma is
instantiated. Instead of checking sc_devcap during every incoming
RPC, set the reader function once when the connection is accepted.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Current convention is to avoid using BUG_ON() in places where an
oops could cause complete system failure.

Replace BUG_ON() call sites in svcrdma with an assertion error
message and allow execution to continue safely.

Some BUG_ON() calls are removed because they have never fired in
production (that we are aware of).

Some WARN_ON() calls are also replaced where a back trace is not
helpful; e.g., in a workqueue task.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

The byte_count argument is not used, and the function is called
only from one place.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Nit: Fix inconsistent white space in dprintk messages.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Reviewed-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

See RFC 5666 section 3.7: clients don't have to send zero XDR
padding.

BugLink: https://bugzilla.linux-nfs.org/show_bug.cgi?id=246Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Fencing forces the invalidate to only happen after all prior send
work requests have been completed.
Signed-off-by: NSteve Wise <swise@opengridcomputing.com>
Reported by : Devesh Sharma <Devesh.Sharma@Emulex.Com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

This patch refactors the NFSRDMA server marshalling logic to
remove the intermediary map structures.  It also fixes an existing bug
where the NFSRDMA server was not minding the device fast register page
list length limitations.
Signed-off-by: NTom Tucker <tom@opengridcomputing.com>
Signed-off-by: NSteve Wise <swise@opengridcomputing.com>

The server regression was caused by the addition of rq_next_page
(afc59400). There were a few places that
were missed with the update of the rq_respages array.
Signed-off-by: NTom Tucker <tom@ogc.us>
Tested-by: NSteve Wise <swise@ogc.us>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

It may be a matter of personal taste, but I find this makes the code
clearer.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

The svcrdma transport was un-marshalling requests in-place. This resulted
in sparse warnings due to __beXX data containing both NBO and HBO data.

The code has been restructured to do byte-swapping as the header is
parsed instead of when the header is validated immediately after receipt.

Also moved extern declarations for the workqueue and memory pools to the
private header file.
Signed-off-by: NTom Tucker <tom@ogc.us>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

There are several error paths in the code that do not unmap DMA. This
patch adds calls to svc_rdma_unmap_dma to free these DMA contexts.
Signed-off-by: NTom Tucker <tom@opengridcomputing.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>