| Age | Commit message (Collapse) | Author |
|---|
|
Olga K. observed that rpcrdma_marsh_req() allocates sparse pages
only when it has determined that a Reply chunk is necessary. There
are plenty of cases where no Reply chunk is needed, but the
XDRBUF_SPARSE_PAGES flag is set. The result would be a crash in
rpcrdma_inline_fixup() when it tries to copy parts of the received
Reply into a missing page.
To avoid crashing, handle sparse page allocation up front.
Until XATTR support was added, this issue did not appear often
because the only SPARSE_PAGES consumer always expected a reply large
enough to always require a Reply chunk.
Reported-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
When receiving pages data, return value 'ret' when positive includes
`buf->page_base`, so we should subtract that before it is used for
changing `offset` and comparing against `want`.
This was discovered on the very rare cases where the server returned a
chunk of bytes that when added to the already received amount of bytes
for the pages happened to match the current `recv.len`, for example
on this case:
buf->page_base : 258356
actually received from socket: 1740
ret : 260096
want : 260096
In this case neither of the two 'if ... goto out' trigger, and we
continue to tail parsing.
Worth to mention that the ensuing EMSGSIZE from the continued execution of
`xs_read_xdr_buf` may be observed by an application due to 4 superfluous
bytes being added to the pages data.
Fixes: 277e4ab7d530 ("SUNRPC: Simplify TCP receive code by switching to using iterators")
Signed-off-by: Dan Aloni <dan@kernelim.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
There's no need to defer allocation of pages for the receive buffer.
- This upcall is quite infrequent
- gssp_alloc_receive_pages() can allocate the pages with GFP_KERNEL,
unlike the transport
- gssp_alloc_receive_pages() knows exactly how many pages are needed
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Olga Kornievskaia <kolga@netapp.com>
|
|
We can simplify code around cache_downcall unifying memory
allocations using kvmalloc. This has the benefit of getting rid of
cache_slow_downcall (and queue_io_mutex), and also matches userland
allocation size and limits.
Signed-off-by: Roberto Bergantinos Corpas <rbergant@redhat.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
'snprintf' returns the number of characters which would have been written
if enough space had been available, excluding the terminating null byte.
Thus, the return value of 'sizeof(buf)' means that the last character
has been dropped.
Signed-off-by: Fedor Tokarev <ftokarev@gmail.com>
Fixes: 2f34b8bfae19 ("SUNRPC: add links for all client xprts to debugfs")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
While we always want to align to the next page and/or the beginning of
the tail buffer when we call xdr_set_next_page(), the functions
xdr_align_data() and xdr_expand_hole() really want to align to the next
object in that next page or tail.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
rpc_prepare_reply_pages() currently expects the 'hdrsize' argument to
contain the length of the data that we expect to want placed in the head
kvec plus a count of 1 word of padding that is placed after the page data.
This is very confusing when trying to read the code, and sometimes leads
to callers adding an arbitrary value of '1' just in order to satisfy the
requirement (whether or not the page data actually needs such padding).
This patch aims to clarify the code by changing the 'hdrsize' argument
to remove that 1 word of padding. This means we need to subtract the
padding from all the existing callers.
Fixes: 02ef04e432ba ("NFS: Account for XDR pad of buf->pages")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Fix up xdr_read_pages() so that it can handle object lengths that are
larger than the page length, by simply aligning to the next object in
the buffer tail.
The function will continue to return the length of the truncate object
data that actually fit into the pages.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Allow xdr_set_iov() to set a base so that we can use it to set the
cursor to a specific position in the kvec buffer.
If the new base overflows the kvec/pages buffer in either xdr_set_iov()
or xdr_set_page_base(), then truncate it so that we point to the end of
the buffer.
Finally, change both function to return the number of bytes remaining to
read in their buffers.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
We already know that the head buffer and page are empty, so if there is
any data, it is in the tail.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
After we've looked up the transport module, we need to ensure it can't
go away until we've finished running the transport setup code.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
According to RFC5666, the correct netid for an IPv6 addressed RDMA
transport is "rdma6", which we've supported as a mount option since
Linux-4.7. The problem is when we try to load the module "xprtrdma6",
that will fail, since there is no modulealias of that name.
Fixes: 181342c5ebe8 ("xprtrdma: Add rdma6 option to support NFS/RDMA IPv6")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
Currently, we wake up the tasks by priority queue ordering, which means
that we ignore the batching that is supposed to help with QoS issues.
Fixes: c049f8ea9a0d ("SUNRPC: Remove the bh-safe lock requirement on the rpc_wait_queue->lock")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
There's no need to defer allocation of pages for the receive buffer.
- This upcall is quite infrequent
- gssp_alloc_receive_pages() can allocate the pages with GFP_KERNEL,
unlike the transport
- gssp_alloc_receive_pages() knows exactly how many pages are needed
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Olga Kornievskaia <aglo@umich.edu>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Convert the READ_BUF macro in nfs4xdr.c from open code to instead
use the new xdr_stream-style decoders already in use by the encode
side (and by the in-kernel NFS client implementation). Once this
conversion is done, each individual NFSv4 argument decoder can be
independently cleaned up to replace these macros with C code.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
A "permanent" struct xdr_stream is allocated in struct svc_rqst so
that it is usable by all server-side decoders. A per-rqst scratch
buffer is also allocated to handle decoding XDR data items that
cross page boundaries.
To demonstrate how it will be used, add the first call site for the
new svcxdr_init_decode() API.
As an additional part of the overall conversion, add symbolic
constants for successful and failed XDR operations. Returning "0" is
overloaded. Sometimes it means something failed, but sometimes it
means success. To make it more clear when XDR decoding functions
succeed or fail, introduce symbolic constants.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Clean up: De-duplicate some frequently-used code.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Commit c509f15a5801 ("SUNRPC: Split the xdr_buf event class") added
display of the rqst's XID to the svc_xdr_buf_class. However, when
the recvfrom tracepoint fires, rq_xid has yet to be filled in with
the current XID. So it ends up recording the previous XID that was
handled by that svc_rqst.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
An efficient way to handle multiple Read chunks is to post them all
together and then take a single completion. This is also how the
code is already structured: when the Read completion fires, all
portions of the incoming RPC message are available to be assembled.
The difficult problem is setting up the Read sink buffers so that
the server pulls the client's data into place, making subsequent
pull-up unnecessary. There are several cases:
* No Read chunks. No-op.
* One data item Read chunk. This is the fast case, where the inline
part of the RPC-over-RDMA message becomes the head and tail, and
the data item chunk is placed in buf->pages.
* A Position-zero Read chunk. Treated like TCP: the Read chunk is
pulled into contiguous pages.
+ A Position-zero Read chunk with data item chunks. Treated like
TCP: all of the Read chunks are pulled into contiguous pages.
+ Multiple data item chunks. Treated like TCP: the inline part is
copied and the data item chunks are pulled into contiguous pages.
The "*" cases are already supported. This patch adds support for the
"+" cases.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
As a pre-requisite for handling multiple Read chunks in each Read
list, convert svc_rdma_recv_read_chunk() to use the new parsed Read
chunk list.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
I'm about to change the purpose of ri_chunklen: Instead of tracking
the number of bytes in one Read chunk, it will track the total
number of bytes in the Read list. Rename it for clarity.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
We already have trace_svcrdma_decode_rseg(), which records each
ingress Read segment. Instead of reporting those again when they
are about to be posted as RDMA Reads, let's fire one tracepoint
before posting each type of chunk.
So we'll get:
nfsd-1998 [002] 321.666615: svcrdma_decode_rseg: cq.id=4 cid=42 segno=0 position=0 192@0x013ca9ebfae14000:0xb0010b05
nfsd-1998 [002] 321.666615: svcrdma_decode_rseg: cq.id=4 cid=42 segno=1 position=0 7688@0x013ca9ebf914e000:0xb0010a05
nfsd-1998 [002] 321.666615: svcrdma_decode_rseg: cq.id=4 cid=42 segno=2 position=0 28@0x013ca9ebfae15000:0xb0010905
nfsd-1998 [002] 321.666622: svcrdma_decode_rqst: cq.id=4 cid=42 xid=0x013ca9eb vers=1 credits=128 proc=RDMA_NOMSG hdrlen=100
nfsd-1998 [002] 321.666642: svcrdma_post_read_chunk: cq.id=3 cid=112 sqecount=3
kworker/2:1H-221 [002] 321.673949: svcrdma_wc_read: cq.id=3 cid=112 status=SUCCESS (0/0x0)
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Clean up: These pointers are no longer used.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor svc_rdma_send_reply_chunk() so that it Sends only the parts
of rq_res that do not contain a result payload.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor: svc_rdma_map_reply_msg() is restructured to DMA map only
the parts of rq_res that do not contain a result payload.
This change has been tested to confirm that it does not cause a
regression in the no Write chunk and single Write chunk cases.
Multiple Write chunks have not been tested.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
When counting the number of SGEs needed to construct a Send request,
do not count result payloads. And, when copying the Reply message
into the pull-up buffer, result payloads are not to be copied to the
Send buffer.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor: Instead of re-parsing the ingress RPC Call transport
header when constructing the egress RPC Reply transport header, use
the new parsed Write list and Reply chunk, which are version-
agnostic and already XDR decoded.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor: Instead of re-parsing the ingress RPC Call transport
header when constructing RDMA Writes, use the new parsed chunk lists
for the Write list and Reply chunk, which are version-agnostic and
already XDR-decoded.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor: Don't duplicate header decoding smarts here. Instead, use
the new parsed chunk lists.
Note that the XID sanity test is also removed. The XID is already
looked up by the cb handler, and is rejected if it's not recognized.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor: Don't duplicate header decoding smarts here. Instead, use
the new parsed chunk lists.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
This simple data structure binds the location of each data payload
inside of an RPC message to the chunk that will be used to push it
to or pull it from the client.
There are several benefits to this small additional overhead:
* It enables support for more than one chunk in incoming Read and
Write lists.
* It translates the version-specific on-the-wire format into a
generic in-memory structure, enabling support for multiple
versions of the RPC/RDMA transport protocol.
* It enables the server to re-organize a chunk list if it needs to
adjust where Read chunk data lands in server memory without
altering the contents of the XDR-encoded Receive buffer.
Construction of these lists is done while sanity checking each
incoming RPC/RDMA header. Subsequent patches will make use of the
generated data structures.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor: Match the control flow of svc_rdma_encode_write_list().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
The only RPC/RDMA ordering requirement between RDMA Writes and RDMA
Sends is that the responder must post the Writes on the Send queue
before posting the Send that conveys the RPC Reply for that Write
payload.
The Linux NFS server implementation now has a transport method that
can post result Payload Writes earlier than svc_rdma_sendto:
->xpo_result_payload()
This gets RDMA Writes going earlier so they are more likely to be
complete at the remote end before the Send completes.
Some care must be taken with pulled-up Replies. We don't want to
push the Write chunk and then send the same payload data via Send.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Have the NFSD encoders annotate the boundaries of every
direct-data-placement eligible result data payload. Then change
svcrdma to use that annotation instead of the xdr->page_len
when handling Write chunks.
For NFSv4 on RDMA, that enables the ability to recognize multiple
result payloads per compound. This is a pre-requisite for supporting
multiple Write chunks per RPC transaction.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Clean up: "result payload" is a less confusing name for these
payloads. "READ payload" reflects only the NFS usage.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Refactor for subsequent changes.
Constify the xdr_buf argument to ensure the code here does not
modify it, and to enable callers to pass in a
"const struct xdr_buf *".
At the same time, rename the helper functions, which emit RDMA
Writes, not RDMA Sends, and add documenting comments.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Clean up: Ensure the code in rw.c does not modify the argument, and
enable callers to also use "const struct xdr_buf *".
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Clean up: This enables xdr_buf_subsegment()'s callers to pass in a
const pointer to that buffer.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
When space in the Reply chunk runs out in the middle of a segment,
we end up passing a zero-length SGL to rdma_rw_ctx_init(), and it
oopses.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Some identifiers have different names between their prototypes
and the kernel-doc markup.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Reviewed-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Now that rpcrdma_ep is no longer part of rpcrdma_xprt, there are
four or five serial address dereferences needed to get to the
IB device needed for DMA unmapping.
Instead, let's use the same pattern that regbufs use: cache a
pointer to the device in the MR, and use that as the indication
that unmapping is necessary.
This also guarantees that the exact same device is used for DMA
mapping and unmapping, even if the r_xprt's ep has been replaced. I
don't think this can happen today, but future changes might break
this assumption.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Clean up: This function is now invoked only in frwr_ops.c. The move
enables deduplication of the trace_xprtrdma_mr_unmap() call site.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
->buf_free is called nearly once per RPC. Only rarely does
xprt_rdma_free() have to do anything, thus tracing every one of
these calls seems unnecessary. Instead, just throw a trace event
when that one occasional RPC still has MRs that need to be
released.
xprt_rdma_free() is further micro-optimized to reduce the amount of
work done in the common case.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Tie each MR event to the requesting rpc_task to make it easier to
follow MR ownership and control flow.
MR unmapping and recycling can happen in the background, after an
MR's mr_req field is stale, so set up a separate tracepoint class
for those events.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
- Rename it following the "_err" suffix convention
- Replace display of kernel memory addresses
- Tie MR exhaustion to a peer IP address, similar to the createmrs
tracepoint
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
- Replace displayed kernel memory addresses
- Tie the XID and event with the peer's IP address
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
