[PATCH v2 0/6] NFS: Add READ_PLUS support

[PATCH v2 0/6] NFS: Add READ_PLUS support

[Anna Schumaker]

These patches add client support for the NFS v4.2 operation READ_PLUS.  This
operation is triggered by doing any kind of read on a NFS v4.2 mounted
filesystem.  `

Changes since v1:
- Add support for decoding multiple segments.

These patches and the corresponding server changes are available in the
[read_plus] branch of

	git://git.linux-nfs.org/projects/anna/linux-nfs.git

Questions?  Comments?  Thoughts?

Anna


Anna Schumaker (6):
  NFS: Add READ_PLUS support for data segments
  SUNRPC: Split out a function for setting current page
  SUNRPC: Add the ability to expand holes in data pages
  NFS: Add READ_PLUS support for hole segments
  SUNRPC: Add the ability to shift data to a specific offset
  NFS: Add support for decoding multiple segments

 fs/nfs/nfs42xdr.c          | 163 ++++++++++++++++++++++++++++++
 fs/nfs/nfs4proc.c          |  30 +++++-
 fs/nfs/nfs4xdr.c           |   1 +
 include/linux/nfs4.h       |   1 +
 include/linux/nfs_fs_sb.h  |   1 +
 include/linux/nfs_xdr.h    |   2 +-
 include/linux/sunrpc/xdr.h |   2 +
 net/sunrpc/xdr.c           | 245 ++++++++++++++++++++++++++++++++++++++++++++-
 8 files changed, 440 insertions(+), 5 deletions(-)

-- 
2.2.2

[PATCH v2 1/6] NFS: Add READ_PLUS support for data segments

[Anna Schumaker]

This patch adds support for decoding a single NFS4_CONTENT_DATA segment,
matching the behaviour of READ without introducing holes.

Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 fs/nfs/nfs42xdr.c         | 135 ++++++++++++++++++++++++++++++++++++++++++++++
 fs/nfs/nfs4proc.c         |  30 ++++++++++-
 fs/nfs/nfs4xdr.c          |   1 +
 include/linux/nfs4.h      |   1 +
 include/linux/nfs_fs_sb.h |   1 +
 include/linux/nfs_xdr.h   |   2 +-
 6 files changed, 167 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c
index 038a7e1..3d7c01b 100644
--- a/fs/nfs/nfs42xdr.c
+++ b/fs/nfs/nfs42xdr.c
@@ -13,6 +13,15 @@
 #define encode_deallocate_maxsz		(op_encode_hdr_maxsz + \
 					 encode_fallocate_maxsz)
 #define decode_deallocate_maxsz		(op_decode_hdr_maxsz)
+#define encode_read_plus_maxsz		(op_encode_hdr_maxsz + \
+					 encode_stateid_maxsz + 3)
+#define decode_read_plus_maxsz		(op_decode_hdr_maxsz + \
+					 1 /* rpr_eof */ + \
+					 1 /* rpr_contents count */ + \
+					 1 /* data_content4 */ + \
+					 2 /* data_info4.di_offset */ + \
+					 2 /* data_info4.di_length */)
+
 #define encode_seek_maxsz		(op_encode_hdr_maxsz + \
 					 encode_stateid_maxsz + \
 					 2 /* offset */ + \
@@ -35,6 +44,12 @@
 #define NFS4_dec_deallocate_sz		(compound_decode_hdr_maxsz + \
 					 decode_putfh_maxsz + \
 					 decode_deallocate_maxsz)
+#define NFS4_enc_read_plus_sz		(compound_encode_hdr_maxsz + \
+					 encode_putfh_maxsz + \
+					 encode_read_plus_maxsz)
+#define NFS4_dec_read_plus_sz		(compound_decode_hdr_maxsz + \
+					 decode_putfh_maxsz + \
+					 decode_read_plus_maxsz)
 #define NFS4_enc_seek_sz		(compound_encode_hdr_maxsz + \
 					 encode_putfh_maxsz + \
 					 encode_seek_maxsz)
@@ -67,6 +82,16 @@ static void encode_deallocate(struct xdr_stream *xdr,
 	encode_fallocate(xdr, args);
 }
 
+static void encode_read_plus(struct xdr_stream *xdr,
+			     struct nfs_pgio_args *args,
+			     struct compound_hdr *hdr)
+{
+	encode_op_hdr(xdr, OP_READ_PLUS, decode_read_plus_maxsz, hdr);
+	encode_nfs4_stateid(xdr, &args->stateid);
+	encode_uint64(xdr, args->offset);
+	encode_uint32(xdr, args->count);
+}
+
 static void encode_seek(struct xdr_stream *xdr,
 			struct nfs42_seek_args *args,
 			struct compound_hdr *hdr)
@@ -114,6 +139,28 @@ static void nfs4_xdr_enc_deallocate(struct rpc_rqst *req,
 }
 
 /*
+ * Encode READ_PLUS request
+ */
+static void nfs4_xdr_enc_read_plus(struct rpc_rqst *req,
+				   struct xdr_stream *xdr,
+				   struct nfs_pgio_args *args)
+{
+	struct compound_hdr hdr = {
+		.minorversion = nfs4_xdr_minorversion(&args->seq_args),
+	};
+
+	encode_compound_hdr(xdr, req, &hdr);
+	encode_sequence(xdr, &args->seq_args, &hdr);
+	encode_putfh(xdr, args->fh, &hdr);
+	encode_read_plus(xdr, args, &hdr);
+
+	xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2,
+			 args->pages, args->pgbase, args->count);
+	req->rq_rcv_buf.flags |= XDRBUF_READ;
+	encode_nops(&hdr);
+}
+
+/*
  * Encode SEEK request
  */
 static void nfs4_xdr_enc_seek(struct rpc_rqst *req,
@@ -141,6 +188,68 @@ static int decode_deallocate(struct xdr_stream *xdr, struct nfs42_falloc_res *re
 	return decode_op_hdr(xdr, OP_DEALLOCATE);
 }
 
+static int decode_read_plus_data(struct xdr_stream *xdr, struct nfs_pgio_res *res)
+{
+	__be32 *p;
+	uint32_t count, recvd;
+	uint64_t offset;
+
+	p = xdr_inline_decode(xdr, 8 + 4);
+	if (unlikely(!p))
+		goto out_overflow;
+
+	p = xdr_decode_hyper(p, &offset);
+	count = be32_to_cpup(p);
+
+	recvd = xdr_read_pages(xdr, count);
+	if (recvd < count)
+		res->eof = 0;
+
+	res->count = recvd;
+	return 0;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+static int decode_read_plus(struct xdr_stream *xdr, struct nfs_pgio_res *res)
+{
+	__be32 *p;
+	int status, type;
+	uint32_t segments;
+
+	status = decode_op_hdr(xdr, OP_READ_PLUS);
+	if (status)
+		return status;
+
+	p = xdr_inline_decode(xdr, 4 + 4);
+	if (unlikely(!p))
+		goto out_overflow;
+
+	res->count = 0;
+	res->eof = be32_to_cpup(p++);
+	segments = be32_to_cpup(p++);
+	if (segments == 0)
+		return 0;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(!p))
+		goto out_overflow;
+
+	type = be32_to_cpup(p++);
+	if (type == NFS4_CONTENT_DATA)
+		status = decode_read_plus_data(xdr, res);
+	else
+		status = -ENOTSUPP;
+
+	if (segments > 1)
+		res->eof = 0;
+	return status;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
 static int decode_seek(struct xdr_stream *xdr, struct nfs42_seek_res *res)
 {
 	int status;
@@ -212,6 +321,32 @@ out:
 }
 
 /*
+ * Decode READ_PLUS request
+ */
+static int nfs4_xdr_dec_read_plus(struct rpc_rqst *rqstp,
+				  struct xdr_stream *xdr,
+				  struct nfs_pgio_res *res)
+{
+	struct compound_hdr hdr;
+	int status;
+
+	status = decode_compound_hdr(xdr, &hdr);
+	if (status)
+		goto out;
+	status = decode_sequence(xdr, &res->seq_res, rqstp);
+	if (status)
+		goto out;
+	status = decode_putfh(xdr);
+	if (status)
+		goto out;
+	status = decode_read_plus(xdr, res);
+	if (!status)
+		status = res->count;
+out:
+	return status;
+}
+
+/*
  * Decode SEEK request
  */
 static int nfs4_xdr_dec_seek(struct rpc_rqst *rqstp,
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 0f75b92..5887340 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -68,6 +68,10 @@
 
 #include "nfs4trace.h"
 
+#ifdef CONFIG_NFS_V4_2
+#include "nfs42.h"
+#endif /* CONFIG_NFS_V4_2 */
+
 #define NFSDBG_FACILITY		NFSDBG_PROC
 
 #define NFS4_POLL_RETRY_MIN	(HZ/10)
@@ -4150,9 +4154,15 @@ static bool nfs4_read_stateid_changed(struct rpc_task *task,
 
 static int nfs4_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-
+	struct nfs_server *server = NFS_SERVER(hdr->inode);
 	dprintk("--> %s\n", __func__);
 
+	if ((server->caps & NFS_CAP_READ_PLUS) && (task->tk_status == -ENOTSUPP)) {
+		server->caps &= ~NFS_CAP_READ_PLUS;
+		if (rpc_restart_call_prepare(task))
+			task->tk_status = 0;
+		return -EAGAIN;
+	}
 	if (!nfs4_sequence_done(task, &hdr->res.seq_res))
 		return -EAGAIN;
 	if (nfs4_read_stateid_changed(task, &hdr->args))
@@ -4161,12 +4171,27 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 				    nfs4_read_done_cb(task, hdr);
 }
 
+#ifdef CONFIG_NFS_V4_2
+static void nfs42_read_plus_support(struct nfs_server *server, struct rpc_message *msg)
+{
+	if (server->caps & NFS_CAP_READ_PLUS)
+		msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ_PLUS];
+	else
+		msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ];
+}
+#else
+static void nfs42_read_plus_support(struct nfs_server *server, struct rpc_message *msg)
+{
+	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ];
+}
+#endif /* CONFIG_NFS_V4_2 */
+
 static void nfs4_proc_read_setup(struct nfs_pgio_header *hdr,
 				 struct rpc_message *msg)
 {
 	hdr->timestamp   = jiffies;
 	hdr->pgio_done_cb = nfs4_read_done_cb;
-	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ];
+	nfs42_read_plus_support(NFS_SERVER(hdr->inode), msg);
 	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 0);
 }
 
@@ -8463,6 +8488,7 @@ static const struct nfs4_minor_version_ops nfs_v4_2_minor_ops = {
 		| NFS_CAP_ATOMIC_OPEN_V1
 		| NFS_CAP_ALLOCATE
 		| NFS_CAP_DEALLOCATE
+		| NFS_CAP_READ_PLUS
 		| NFS_CAP_SEEK,
 	.init_client = nfs41_init_client,
 	.shutdown_client = nfs41_shutdown_client,
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 4125339..da01f00 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -7397,6 +7397,7 @@ struct rpc_procinfo	nfs4_procedures[] = {
 	PROC(SEEK,		enc_seek,		dec_seek),
 	PROC(ALLOCATE,		enc_allocate,		dec_allocate),
 	PROC(DEALLOCATE,	enc_deallocate,		dec_deallocate),
+	PROC(READ_PLUS,		enc_read_plus,		dec_read_plus),
 #endif /* CONFIG_NFS_V4_2 */
 };
 
diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h
index 022b761..41fc2f4 100644
--- a/include/linux/nfs4.h
+++ b/include/linux/nfs4.h
@@ -492,6 +492,7 @@ enum {
 	NFSPROC4_CLNT_SEEK,
 	NFSPROC4_CLNT_ALLOCATE,
 	NFSPROC4_CLNT_DEALLOCATE,
+	NFSPROC4_CLNT_READ_PLUS,
 };
 
 /* nfs41 types */
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index ddea982..95609f2 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -236,5 +236,6 @@ struct nfs_server {
 #define NFS_CAP_SEEK		(1U << 19)
 #define NFS_CAP_ALLOCATE	(1U << 20)
 #define NFS_CAP_DEALLOCATE	(1U << 21)
+#define NFS_CAP_READ_PLUS	(1U << 22)
 
 #endif
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index e5c3b62..f09f023 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -513,7 +513,7 @@ struct nfs_pgio_args {
 struct nfs_pgio_res {
 	struct nfs4_sequence_res	seq_res;
 	struct nfs_fattr *	fattr;
-	__u32			count;
+	__u64			count;
 	int			eof;		/* used by read */
 	struct nfs_writeverf *	verf;		/* used by write */
 	const struct nfs_server *server;	/* used by write */
-- 
2.2.2

Re: [PATCH v2 1/6] NFS: Add READ_PLUS support for data segments

[Christoph Hellwig]

On Wed, Jan 28, 2015 at 03:43:45PM -0500, Anna Schumaker wrote:
> +	if (type == NFS4_CONTENT_DATA)
> +		status = decode_read_plus_data(xdr, res);
> +	else
> +		status = -ENOTSUPP;
> +
> +	if (segments > 1)
> +		res->eof = 0;

Given that the server can return any amount of data and hole
segments it doesn't make any sense to separate this patch from patches 4
and 6.

Re: [PATCH v2 1/6] NFS: Add READ_PLUS support for data segments

[Anna Schumaker]

On 02/05/2015 09:10 AM, Christoph Hellwig wrote:
> On Wed, Jan 28, 2015 at 03:43:45PM -0500, Anna Schumaker wrote:
>> +	if (type == NFS4_CONTENT_DATA)
>> +		status = decode_read_plus_data(xdr, res);
>> +	else
>> +		status = -ENOTSUPP;
>> +
>> +	if (segments > 1)
>> +		res->eof = 0;
> 
> Given that the server can return any amount of data and hole
> segments it doesn't make any sense to separate this patch from patches 4
> and 6.
> 

I had them split up to make development easier, but I can squash them together if that would make more sense.

[PATCH v2 2/6] SUNRPC: Split out a function for setting current page

[Anna Schumaker]

I'm going to need this bit of code in a few places for READ_PLUS
decoding, so let's make it a helper function.

Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 net/sunrpc/xdr.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c
index 4439ac4..b1c4ffb 100644
--- a/net/sunrpc/xdr.c
+++ b/net/sunrpc/xdr.c
@@ -759,6 +759,12 @@ static int xdr_set_page_base(struct xdr_stream *xdr,
 	return 0;
 }
 
+static void xdr_set_page(struct xdr_stream *xdr, unsigned int base)
+{
+	if (xdr_set_page_base(xdr, base, PAGE_SIZE) < 0)
+		xdr_set_iov(xdr, xdr->buf->tail, xdr->buf->len);
+}
+
 static void xdr_set_next_page(struct xdr_stream *xdr)
 {
 	unsigned int newbase;
@@ -766,8 +772,7 @@ static void xdr_set_next_page(struct xdr_stream *xdr)
 	newbase = (1 + xdr->page_ptr - xdr->buf->pages) << PAGE_SHIFT;
 	newbase -= xdr->buf->page_base;
 
-	if (xdr_set_page_base(xdr, newbase, PAGE_SIZE) < 0)
-		xdr_set_iov(xdr, xdr->buf->tail, xdr->buf->len);
+	xdr_set_page(xdr, newbase);
 }
 
 static bool xdr_set_next_buffer(struct xdr_stream *xdr)
-- 
2.2.2

[PATCH v2 3/6] SUNRPC: Add the ability to expand holes in data pages

[Anna Schumaker]

This patch adds the ability to "read a hole" into a set of XDR data
pages by taking the following steps:

1) Shift all data after the current xdr->p to the right, possibly into
   the tail,
2) Zero the specified range, and
3) Update xdr->p to point beyond the hole.

Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 include/linux/sunrpc/xdr.h |   1 +
 net/sunrpc/xdr.c           | 104 +++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 105 insertions(+)

diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h
index 70c6b92..81c5a3f 100644
--- a/include/linux/sunrpc/xdr.h
+++ b/include/linux/sunrpc/xdr.h
@@ -229,6 +229,7 @@ extern __be32 *xdr_inline_decode(struct xdr_stream *xdr, size_t nbytes);
 extern unsigned int xdr_read_pages(struct xdr_stream *xdr, unsigned int len);
 extern void xdr_enter_page(struct xdr_stream *xdr, unsigned int len);
 extern int xdr_process_buf(struct xdr_buf *buf, unsigned int offset, unsigned int len, int (*actor)(struct scatterlist *, void *), void *data);
+extern size_t xdr_expand_hole(struct xdr_stream *, size_t, size_t);
 
 #endif /* __KERNEL__ */
 
diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c
index b1c4ffb..f71e227 100644
--- a/net/sunrpc/xdr.c
+++ b/net/sunrpc/xdr.c
@@ -219,6 +219,40 @@ _shift_data_right_pages(struct page **pages, size_t pgto_base,
 	} while ((len -= copy) != 0);
 }
 
+static void
+_shift_data_right_tail(struct xdr_buf *buf, size_t pgfrom_base, size_t len)
+{
+	struct kvec *tail = buf->tail;
+
+	/* Make room for new data. */
+	if (tail->iov_len > 0)
+		memmove((char *)tail->iov_base + len, tail->iov_base, len);
+
+	_copy_from_pages((char *)tail->iov_base,
+			 buf->pages,
+			 buf->page_base + pgfrom_base,
+			 len);
+
+	tail->iov_len += len;
+}
+
+static void
+_shift_data_right(struct xdr_buf *buf, size_t to, size_t from, size_t len)
+{
+	size_t shift = len;
+
+	if ((to + len) > buf->page_len) {
+		shift = (to + len) - buf->page_len;
+		_shift_data_right_tail(buf, (from + len) - shift, shift);
+		shift = len - shift;
+	}
+
+	_shift_data_right_pages(buf->pages,
+				buf->page_base + to,
+				buf->page_base + from,
+				shift);
+}
+
 /**
  * _copy_to_pages
  * @pages: array of pages
@@ -304,6 +338,38 @@ _copy_from_pages(char *p, struct page **pages, size_t pgbase, size_t len)
 EXPORT_SYMBOL_GPL(_copy_from_pages);
 
 /**
+ * _zero_data_pages
+ * @pages: array of pages
+ * @pgbase: beginning page vector address
+ * @len: length
+ */
+static void
+_zero_data_pages(struct page **pages, size_t pgbase, size_t len)
+{
+	struct page **page;
+	char *vpage;
+	size_t zero;
+
+	page = pages + (pgbase >> PAGE_CACHE_SHIFT);
+	pgbase &= ~PAGE_CACHE_MASK;
+
+	do {
+		zero = len;
+		if (pgbase + zero > PAGE_SIZE)
+			zero = PAGE_SIZE - pgbase;
+
+		vpage = kmap_atomic(*page);
+		memset(vpage + pgbase, 0, zero);
+		flush_dcache_page(*page);
+		kunmap_atomic(vpage);
+
+		page++;
+		pgbase = 0;
+
+	} while ((len -= zero) != 0);
+}
+
+/**
  * xdr_shrink_bufhead
  * @buf: xdr_buf
  * @len: bytes to remove from buf->head[0]
@@ -445,6 +511,24 @@ unsigned int xdr_stream_pos(const struct xdr_stream *xdr)
 EXPORT_SYMBOL_GPL(xdr_stream_pos);
 
 /**
+ * xdr_page_pos - Return the current offset from the start of the xdr->buf->pages
+ * @xdr: pointer to struct xdr_stream
+ */
+static size_t xdr_page_pos(const struct xdr_stream *xdr)
+{
+	unsigned int offset;
+	unsigned int base = xdr->buf->page_len;
+	void *kaddr = xdr->buf->tail->iov_base;;
+
+	if (xdr->page_ptr) {
+		base   = (xdr->page_ptr - xdr->buf->pages) * PAGE_SIZE;
+		kaddr  = page_address(*xdr->page_ptr);
+	}
+	offset = xdr->p - (__be32 *)kaddr;
+	return base + (offset * sizeof(__be32));
+}
+
+/**
  * xdr_init_encode - Initialize a struct xdr_stream for sending data.
  * @xdr: pointer to xdr_stream struct
  * @buf: pointer to XDR buffer in which to encode data
@@ -976,6 +1060,26 @@ unsigned int xdr_read_pages(struct xdr_stream *xdr, unsigned int len)
 }
 EXPORT_SYMBOL_GPL(xdr_read_pages);
 
+size_t xdr_expand_hole(struct xdr_stream *xdr, size_t offset, size_t length)
+{
+	struct xdr_buf *buf = xdr->buf;
+	size_t from = 0;
+
+	if ((offset + length) > buf->page_len)
+		length = buf->page_len - offset;
+
+	if (offset == 0)
+		xdr_align_pages(xdr, xdr->nwords << 2);
+	else
+		from = xdr_page_pos(xdr);
+
+	_shift_data_right(buf, offset + length, from, xdr->nwords << 2);
+	_zero_data_pages(buf->pages, buf->page_base + offset, length);
+	xdr_set_page(xdr, offset + length);
+	return length;
+}
+EXPORT_SYMBOL_GPL(xdr_expand_hole);
+
 /**
  * xdr_enter_page - decode data from the XDR page
  * @xdr: pointer to xdr_stream struct
-- 
2.2.2

[PATCH v2 4/6] NFS: Add READ_PLUS support for hole segments

[Anna Schumaker]

This patch adds support for decoding a single hole segment returned by
the server.

Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 fs/nfs/nfs42xdr.c | 26 +++++++++++++++++++++++++-
 1 file changed, 25 insertions(+), 1 deletion(-)

diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c
index 3d7c01b..17126e8 100644
--- a/fs/nfs/nfs42xdr.c
+++ b/fs/nfs/nfs42xdr.c
@@ -212,6 +212,30 @@ out_overflow:
 	return -EIO;
 }
 
+static int decode_read_plus_hole(struct xdr_stream *xdr, struct nfs_pgio_res *res)
+{
+	__be32 *p;
+	uint64_t offset, length;
+	size_t recvd;
+
+	p = xdr_inline_decode(xdr, 8 + 8);
+	if (unlikely(!p))
+		goto out_overflow;
+
+	p = xdr_decode_hyper(p, &offset);
+	p = xdr_decode_hyper(p, &length);
+
+	recvd = xdr_expand_hole(xdr, 0, length);
+	if (recvd < length)
+		res->eof = 0;
+
+	res->count = recvd;
+	return 0;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
 static int decode_read_plus(struct xdr_stream *xdr, struct nfs_pgio_res *res)
 {
 	__be32 *p;
@@ -240,7 +264,7 @@ static int decode_read_plus(struct xdr_stream *xdr, struct nfs_pgio_res *res)
 	if (type == NFS4_CONTENT_DATA)
 		status = decode_read_plus_data(xdr, res);
 	else
-		status = -ENOTSUPP;
+		status = decode_read_plus_hole(xdr, res);
 
 	if (segments > 1)
 		res->eof = 0;
-- 
2.2.2

[PATCH v2 5/6] SUNRPC: Add the ability to shift data to a specific offset

[Anna Schumaker]

Expanding holes tends to put the data content a few bytes to the right
of where we want it.  This patch implements a left-shift operation to
line everything up properly.

Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 include/linux/sunrpc/xdr.h |   1 +
 net/sunrpc/xdr.c           | 132 +++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 133 insertions(+)

diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h
index 81c5a3f..3670bf6 100644
--- a/include/linux/sunrpc/xdr.h
+++ b/include/linux/sunrpc/xdr.h
@@ -230,6 +230,7 @@ extern unsigned int xdr_read_pages(struct xdr_stream *xdr, unsigned int len);
 extern void xdr_enter_page(struct xdr_stream *xdr, unsigned int len);
 extern int xdr_process_buf(struct xdr_buf *buf, unsigned int offset, unsigned int len, int (*actor)(struct scatterlist *, void *), void *data);
 extern size_t xdr_expand_hole(struct xdr_stream *, size_t, size_t);
+extern uint64_t xdr_align_data(struct xdr_stream *, uint64_t, uint64_t);
 
 #endif /* __KERNEL__ */
 
diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c
index f71e227..372a53c 100644
--- a/net/sunrpc/xdr.c
+++ b/net/sunrpc/xdr.c
@@ -16,6 +16,9 @@
 #include <linux/sunrpc/xdr.h>
 #include <linux/sunrpc/msg_prot.h>
 
+static void _copy_to_pages(struct page **, size_t, const char *, size_t);
+
+
 /*
  * XDR functions for basic NFS types
  */
@@ -253,6 +256,117 @@ _shift_data_right(struct xdr_buf *buf, size_t to, size_t from, size_t len)
 				shift);
 }
 
+
+/**
+ * _shift_data_left_pages
+ * @pages: vector of pages containing both the source and dest memory area.
+ * @pgto_base: page vector address of destination
+ * @pgfrom_base: page vector address of source
+ * @len: number of bytes to copy
+ *
+ * Note: the addresses pgto_base and pgfrom_base are both calculated in
+ *       the same way:
+ *            if a memory area starts at byte 'base' in page 'pages[i]',
+ *            then its address is given as (i << PAGE_CACHE_SHIFT) + base
+ * Alse note: pgto_base must be < pgfrom_base, but the memory areas
+ * 	they point to may overlap.
+ */
+static void
+_shift_data_left_pages(struct page **pages, size_t pgto_base,
+			size_t pgfrom_base, size_t len)
+{
+	struct page **pgfrom, **pgto;
+	char *vfrom, *vto;
+	size_t copy;
+
+	BUG_ON(pgfrom_base <= pgto_base);
+
+	pgto = pages + (pgto_base >> PAGE_CACHE_SHIFT);
+	pgfrom = pages + (pgfrom_base >> PAGE_CACHE_SHIFT);
+
+	pgto_base = pgto_base % PAGE_CACHE_SIZE;
+	pgfrom_base = pgfrom_base % PAGE_CACHE_SIZE;
+
+	do {
+		if (pgto_base >= PAGE_CACHE_SIZE) {
+			pgto_base = 0;
+			pgto++;
+		}
+		if (pgfrom_base >= PAGE_CACHE_SIZE){
+			pgfrom_base = 0;
+			pgfrom++;
+		}
+
+		copy = len;
+		if (copy > (PAGE_CACHE_SIZE - pgto_base))
+			copy = PAGE_CACHE_SIZE - pgto_base;
+		if (copy > (PAGE_CACHE_SIZE - pgfrom_base))
+			copy = PAGE_CACHE_SIZE - pgfrom_base;
+
+		if (pgto_base == 131056)
+			break;
+
+		vto = kmap_atomic(*pgto);
+		if (*pgto != *pgfrom) {
+			vfrom = kmap_atomic(*pgfrom);
+			memcpy(vto + pgto_base, vfrom + pgfrom_base, copy);
+			kunmap_atomic(vfrom);
+		} else
+			memmove(vto + pgto_base, vto + pgfrom_base, copy);
+		flush_dcache_page(*pgto);
+		kunmap_atomic(vto);
+
+		pgto_base += copy;
+		pgfrom_base += copy;
+
+	} while ((len -= copy) != 0);
+}
+
+static void
+_shift_data_left_tail(struct xdr_buf *buf, size_t pgto_base,
+		      size_t tail_from, size_t len)
+{
+	struct kvec *tail = buf->tail;
+	size_t shift = len;
+
+	if (len == 0)
+		return;
+	if (pgto_base + len > buf->page_len)
+		shift = buf->page_len - pgto_base;
+
+	_copy_to_pages(buf->pages,
+			buf->page_base + pgto_base,
+			(char *)(tail->iov_base + tail_from),
+			shift);
+
+	memmove((char *)tail->iov_base, tail->iov_base + tail_from + shift, shift);
+	tail->iov_len -= (tail_from + shift);
+}
+
+static void
+_shift_data_left(struct xdr_buf *buf, size_t to, size_t from, size_t len)
+{
+	size_t shift = len;
+
+	if (from < buf->page_len) {
+		shift = min(len, buf->page_len - from);
+		_shift_data_left_pages(buf->pages,
+					buf->page_base + to,
+					buf->page_base + from,
+					shift);
+		to += shift;
+		from += shift;
+		shift = len - shift;
+	}
+
+	if (shift == 0)
+		return;
+	if (from >= buf->page_len)
+		from -= buf->page_len;
+
+	_shift_data_left_tail(buf, to, from, shift);
+}
+
 /**
  * _copy_to_pages
  * @pages: array of pages
@@ -1080,6 +1194,24 @@ size_t xdr_expand_hole(struct xdr_stream *xdr, size_t offset, size_t length)
 }
 EXPORT_SYMBOL_GPL(xdr_expand_hole);
 
+uint64_t xdr_align_data(struct xdr_stream *xdr, uint64_t offset, uint64_t length)
+{
+	struct xdr_buf *buf = xdr->buf;
+
+	if (offset + length > buf->page_len)
+		length = buf->page_len - offset;
+
+	if (offset == 0)
+		xdr_align_pages(xdr, xdr->nwords << 2);
+	else
+		_shift_data_left(buf, offset, xdr_page_pos(xdr), xdr->nwords << 2);
+
+	xdr->nwords -= XDR_QUADLEN(length);
+	xdr_set_page(xdr, offset + length);
+	return length;
+}
+EXPORT_SYMBOL_GPL(xdr_align_data);
+
 /**
  * xdr_enter_page - decode data from the XDR page
  * @xdr: pointer to xdr_stream struct
-- 
2.2.2

[PATCH v2 6/6] NFS: Add support for decoding multiple segments

[Anna Schumaker]

We now have everything we need to read holes and then shift data to
where it's supposed to be.

Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 fs/nfs/nfs42xdr.c | 40 ++++++++++++++++++++++------------------
 1 file changed, 22 insertions(+), 18 deletions(-)

diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c
index 17126e8..010af55 100644
--- a/fs/nfs/nfs42xdr.c
+++ b/fs/nfs/nfs42xdr.c
@@ -201,11 +201,11 @@ static int decode_read_plus_data(struct xdr_stream *xdr, struct nfs_pgio_res *re
 	p = xdr_decode_hyper(p, &offset);
 	count = be32_to_cpup(p);
 
-	recvd = xdr_read_pages(xdr, count);
+	recvd = xdr_align_data(xdr, res->count, count);
 	if (recvd < count)
 		res->eof = 0;
 
-	res->count = recvd;
+	res->count += recvd;
 	return 0;
 out_overflow:
 	print_overflow_msg(__func__, xdr);
@@ -225,11 +225,11 @@ static int decode_read_plus_hole(struct xdr_stream *xdr, struct nfs_pgio_res *re
 	p = xdr_decode_hyper(p, &offset);
 	p = xdr_decode_hyper(p, &length);
 
-	recvd = xdr_expand_hole(xdr, 0, length);
+	recvd = xdr_expand_hole(xdr, res->count, length);
 	if (recvd < length)
 		res->eof = 0;
 
-	res->count = recvd;
+	res->count += recvd;
 	return 0;
 out_overflow:
 	print_overflow_msg(__func__, xdr);
@@ -240,7 +240,7 @@ static int decode_read_plus(struct xdr_stream *xdr, struct nfs_pgio_res *res)
 {
 	__be32 *p;
 	int status, type;
-	uint32_t segments;
+	uint32_t i, segments;
 
 	status = decode_op_hdr(xdr, OP_READ_PLUS);
 	if (status)
@@ -253,20 +253,24 @@ static int decode_read_plus(struct xdr_stream *xdr, struct nfs_pgio_res *res)
 	res->count = 0;
 	res->eof = be32_to_cpup(p++);
 	segments = be32_to_cpup(p++);
-	if (segments == 0)
-		return 0;
 
-	p = xdr_inline_decode(xdr, 4);
-	if (unlikely(!p))
-		goto out_overflow;
-
-	type = be32_to_cpup(p++);
-	if (type == NFS4_CONTENT_DATA)
-		status = decode_read_plus_data(xdr, res);
-	else
-		status = decode_read_plus_hole(xdr, res);
-
-	if (segments > 1)
+	for (i = 0; i < segments; i++) {
+		p = xdr_inline_decode(xdr, 4);
+		if (unlikely(!p))
+			goto out_overflow;
+
+		type = be32_to_cpup(p);
+		if (type == NFS4_CONTENT_DATA)
+			status = decode_read_plus_data(xdr, res);
+		else
+			status = decode_read_plus_hole(xdr, res);
+		if (status)
+			break;
+		if (res->count == xdr->buf->page_len)
+			break;
+	}
+
+	if (i < segments)
 		res->eof = 0;
 	return status;
 out_overflow:
-- 
2.2.2