Skip to contentSkip to conversation

smb3: Use iov_iters down to the network transport and fix DIO page pinning

PATCH In review David Howells 4 participants 22 messages Jan 31, 2023 – Feb 3, 2023

Subsystems (8)
lore
Share & tools
lore

David Howells

author maintainer
#0 lore raw

cover smb3: Use iov_iters down to the network transport and fix DIO page pinning

Hi Steve,

Here's an updated version of my patchset to make the cifs/smb3 driver pass iov_iters down to the lowest layers where they can be passed directly to the network transport rather than passing lists of pages around.

I've fixed a couple of bugs and rebased on top of a merge of my two iov_iter_extract_pages() patches onto your for-next branch. The merge is so that the same commits are used as are in the linux-block tree.

The series also deals with some other issues:

(*) By pinning pages, it fixes the race between concurrent DIO read and
fork, whereby the pages containing the DIO read buffer may end up belonging to the child process and not the parent - with the result that the parent might not see the retrieved data.

(*) cifs shouldn't take refs on pages extracted from non-user-backed
iterators (eg. KVEC). With these changes, cifs will apply the appropriate cleanup. Note that there is the possibility the network transport might, but that's beyond the scope of this patchset.

(*) Making it easier to transition to using folios in cifs rather than
pages by dealing with them through BVEC and XARRAY iterators.

The first couple of patches to provide function to pin or leave unpinned the pages from an iterator (and not take a ref on them).

(1) Define qualifying flags for extraction functions.

(2) Define iov_iter_extract_pages() to do the extraction and iov_iter_extract_will_pin() to indicate how it should be cleaned up.

Then there are a couple of patches that add stuff to netfslib that I want to use there as well as in cifs:

(3) Add a netfslib function to extract and pin pages from an ITER_IOBUF or ITER_UBUF iterator into an ITER_BVEC iterator.

(4) Add a netfslib function to extract pages from an iterator that's of type ITER_UBUF/IOVEC/BVEC/KVEC/XARRAY and add them to a scatterlist. The cleanup will need to be done as for iov_iter_extract_pages().

BVEC, KVEC and XARRAY iterators can be rendered into elements that span multiple pages.

Then a fix:

(5) Fix oops due to uncleared server->smbd_conn in reconnect

Then there are some cifs helpers that work with iterators:

(6) Implement cifs_splice_read() to use an ITER_BVEC rather than an ITER_PIPE, bulk-allocating the pages, attaching them to the bvec, doing the I/O and then pushing the pages into the pipe. This avoids the problem with cifs wanting to split the pipe iterator in a later patch.

(7) Add a function to walk through an ITER_BVEC/KVEC/XARRAY iterator and add elements to an RDMA SGE list. Only the DMA addresses are stored, and an element may span multiple pages (say if an xarray contains a multipage folio).

(8) Add a function to walk through an ITER_BVEC/KVEC/XARRAY iterator and pass the contents into a shash function.

(9) Add functions to walk through an ITER_XARRAY iterator and perform various sorts of cleanup on the folios held therein, to be used on I/O completion.

(10) Add a function to read from the transport TCP socket directly into an iterator.

Then come the patches that actually do the work of iteratorising cifs:

(11) The main patch. Replace page lists with iterators. It extracts the pages from ITER_UBUF and ITER_IOVEC iterators to an ITER_BVEC iterator, pinning or getting refs on them, before passing them down as the I/O may be done from a worker thread.

The iterator is extracted into a scatterlist in order to talk to the crypto interface or to do RDMA.

(12) In the cifs RDMA code, extract the iterator into an RDMA SGE[] list, removing the scatterlist intermediate - at least for smbd_send(). There appear to be other ways for cifs to talk to the RDMA layer that don't go through that that I haven't managed to work out.

(13) Remove a chunk of now-unused code.

(14) Fix a problem with encrypted RDMA data read.

(15) Allow DIO to/from KVEC-type iterators.

I've pushed the patches here also:

https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs.git/log/?h=iov-cifs

David

David Howells (12):
netfs: Add a function to extract a UBUF or IOVEC into a BVEC iterator
netfs: Add a function to extract an iterator into a scatterlist
cifs: Implement splice_read to pass down ITER_BVEC not ITER_PIPE
cifs: Add a function to build an RDMA SGE list from an iterator
cifs: Add a function to Hash the contents of an iterator
cifs: Add some helper functions
cifs: Add a function to read into an iter from a socket
cifs: Change the I/O paths to use an iterator rather than a page list
cifs: Build the RDMA SGE list directly from an iterator
cifs: Remove unused code
cifs: Fix problem with encrypted RDMA data read
cifs: DIO to/from KVEC-type iterators should now work

 fs/cifs/Kconfig       |    1 +
 fs/cifs/cifsencrypt.c |  172 +++-
 fs/cifs/cifsfs.c      |   12 +-
 fs/cifs/cifsfs.h      |    6 +
 fs/cifs/cifsglob.h    |   66 +-
 fs/cifs/cifsproto.h   |   11 +-
 fs/cifs/cifssmb.c     |   15 +-
 fs/cifs/connect.c     |   14 +
 fs/cifs/file.c        | 1848 +++++++++++++++++++----------------------
 fs/cifs/fscache.c     |   22 +-
 fs/cifs/fscache.h     |   10 +-
 fs/cifs/misc.c        |  128 +--
 fs/cifs/smb2ops.c     |  365 ++++----
 fs/cifs/smb2pdu.c     |   53 +-
 fs/cifs/smbdirect.c   |  535 +++++++-----
 fs/cifs/smbdirect.h   |    7 +-
 fs/cifs/transport.c   |   54 +-
 fs/netfs/Makefile     |    1 +
 fs/netfs/iterator.c   |  371 +++++++++
 fs/splice.c           |    1 +
 include/linux/netfs.h |    6 +
 mm/vmalloc.c          |    1 +
 22 files changed, 2013 insertions(+), 1686 deletions(-)
 create mode 100644 fs/netfs/iterator.c
author maintainer
#1 lore raw

PATCH 1/12 netfs: Add a function to extract a UBUF or IOVEC into a BVEC iterator

Add a function to extract the pages from a user-space supplied iterator (UBUF- or IOVEC-type) into a BVEC-type iterator, retaining the pages by getting a pin on them (as FOLL_PIN) as we go.

This is useful in three situations:

(1) A userspace thread may have a sibling that unmaps or remaps the process's VM during the operation, changing the assignment of the pages and potentially causing an error. Retaining the pages keeps some pages around, even if this occurs; futher, we find out at the point of extraction if EFAULT is going to be incurred.

(2) Pages might get swapped out/discarded if not retained, so we want to retain them to avoid the reload causing a deadlock due to a DIO from/to an mmapped region on the same file.

(3) The iterator may get passed to sendmsg() by the filesystem. If a fault occurs, we may get a short write to a TCP stream that's then tricky to recover from.

We don't deal with other types of iterator here, leaving it to other mechanisms to retain the pages (eg. PG_locked, PG_writeback and the pipe lock).

---
 fs/netfs/Makefile     |   1 +
 fs/netfs/iterator.c   | 102 ++++++++++++++++++++++++++++++++++++++++++
 include/linux/netfs.h |   2 +
 3 files changed, 105 insertions(+)
 create mode 100644 fs/netfs/iterator.c
fs/netfs/Makefile +1 -0
index f684c0cd1ec5..386d6fb92793 100644
@@ -3,6 +3,7 @@
 netfs-y := \
 	buffered_read.o \
 	io.o \
+	iterator.o \
 	main.o \
 	objects.o
 
fs/netfs/iterator.c new file +102 -0

102 lines not on this page · raw message

include/linux/netfs.h +2 -0
index 4c76ddfb6a67..e8c560131170 100644
@@ -296,6 +296,8 @@ void netfs_get_subrequest(struct netfs_io_subrequest *subreq,
 void netfs_put_subrequest(struct netfs_io_subrequest *subreq,
 			  bool was_async, enum netfs_sreq_ref_trace what);
 void netfs_stats_show(struct seq_file *);
+ssize_t netfs_extract_user_iter(struct iov_iter *orig, size_t orig_len,
+				struct iov_iter *new, unsigned int extract_flags);
 
 /**
  * netfs_inode - Get the netfs inode context from the inode
author maintainer
#2 lore raw

PATCH 2/12 netfs: Add a function to extract an iterator into a scatterlist

In reply to David Howells

Provide a function for filling in a scatterlist from the list of pages contained in an iterator.

If the iterator is UBUF- or IOBUF-type, the pages have a pin taken on them (as FOLL_PIN).

If the iterator is BVEC-, KVEC- or XARRAY-type, no pin is taken on the pages and it is left to the caller to manage their lifetime. It cannot be assumed that a ref can be validly taken, particularly in the case of a KVEC iterator.

---
 fs/netfs/iterator.c   | 269 ++++++++++++++++++++++++++++++++++++++++++
 include/linux/netfs.h |   4 +
 mm/vmalloc.c          |   1 +
 3 files changed, 274 insertions(+)
fs/netfs/iterator.c +269 -0
index 7b7cdc2863fc..9d198d0a5503 100644
@@ -7,7 +7,9 @@
 
 #include <linux/export.h>
 #include <linux/slab.h>
+#include <linux/mm.h>
 #include <linux/uio.h>
+#include <linux/scatterlist.h>
 #include <linux/netfs.h>
 #include "internal.h"
 
@@ -100,3 +102,270 @@ ssize_t netfs_extract_user_iter(struct iov_iter *orig, size_t orig_len,
 	return npages;
 }
 EXPORT_SYMBOL_GPL(netfs_extract_user_iter);
+
+/*
+ * Extract and pin a list of up to sg_max pages from UBUF- or IOVEC-class
+ * iterators, and add them to the scatterlist.
+ */
+static ssize_t netfs_extract_user_to_sg(struct iov_iter *iter,
+					ssize_t maxsize,
+					struct sg_table *sgtable,
+					unsigned int sg_max,
+					unsigned int extract_flags)
+{
+	struct scatterlist *sg = sgtable->sgl + sgtable->nents;
+	struct page **pages;
+	unsigned int npages;
+	ssize_t ret = 0, res;
+	size_t len, off;
+
+	/* We decant the page list into the tail of the scatterlist */
+	pages = (void *)sgtable->sgl + array_size(sg_max, sizeof(struct scatterlist));
+	pages -= sg_max;
+
+	do {
+		res = iov_iter_extract_pages(iter, &pages, maxsize, sg_max,
+					     extract_flags, &off);
+		if (res < 0)
+			goto failed;
+
+		len = res;
+		maxsize -= len;
+		ret += len;
+		npages = DIV_ROUND_UP(off + len, PAGE_SIZE);
+		sg_max -= npages;
+
+		for (; npages < 0; npages--) {
+			struct page *page = *pages;
+			size_t seg = min_t(size_t, PAGE_SIZE - off, len);
+
+			*pages++ = NULL;
+			sg_set_page(sg, page, len, off);
+			sgtable->nents++;
+			sg++;
+			len -= seg;
+			off = 0;
+		}
+	} while (maxsize > 0 && sg_max > 0);
+
+	return ret;
+
+failed:
+	while (sgtable->nents > sgtable->orig_nents)
+		put_page(sg_page(&sgtable->sgl[--sgtable->nents]));
+	return res;
+}
+
+/*
+ * Extract up to sg_max pages from a BVEC-type iterator and add them to the
+ * scatterlist.  The pages are not pinned.
+ */
+static ssize_t netfs_extract_bvec_to_sg(struct iov_iter *iter,
+					ssize_t maxsize,
+					struct sg_table *sgtable,
+					unsigned int sg_max,
+					unsigned int extract_flags)
+{
+	const struct bio_vec *bv = iter->bvec;
+	struct scatterlist *sg = sgtable->sgl + sgtable->nents;
+	unsigned long start = iter->iov_offset;
+	unsigned int i;
+	ssize_t ret = 0;
+
+	for (i = 0; i < iter->nr_segs; i++) {
+		size_t off, len;
+
+		len = bv[i].bv_len;
+		if (start >= len) {
+			start -= len;
+			continue;
+		}
+
+		len = min_t(size_t, maxsize, len - start);
+		off = bv[i].bv_offset + start;
+
+		sg_set_page(sg, bv[i].bv_page, len, off);
+		sgtable->nents++;
+		sg++;
+		sg_max--;
+
+		ret += len;
+		maxsize -= len;
+		if (maxsize <= 0 || sg_max == 0)
+			break;
+		start = 0;
+	}
+
+	if (ret > 0)
+		iov_iter_advance(iter, ret);
+	return ret;
+}
+
+/*
+ * Extract up to sg_max pages from a KVEC-type iterator and add them to the
+ * scatterlist.  This can deal with vmalloc'd buffers as well as kmalloc'd or
+ * static buffers.  The pages are not pinned.
+ */
+static ssize_t netfs_extract_kvec_to_sg(struct iov_iter *iter,
+					ssize_t maxsize,
+					struct sg_table *sgtable,
+					unsigned int sg_max,
+					unsigned int extract_flags)
+{
+	const struct kvec *kv = iter->kvec;
+	struct scatterlist *sg = sgtable->sgl + sgtable->nents;
+	unsigned long start = iter->iov_offset;
+	unsigned int i;
+	ssize_t ret = 0;
+
+	for (i = 0; i < iter->nr_segs; i++) {
+		struct page *page;
+		unsigned long kaddr;
+		size_t off, len, seg;
+
+		len = kv[i].iov_len;
+		if (start >= len) {
+			start -= len;
+			continue;
+		}
+
+		kaddr = (unsigned long)kv[i].iov_base + start;
+		off = kaddr & ~PAGE_MASK;
+		len = min_t(size_t, maxsize, len - start);
+		kaddr &= PAGE_MASK;
+
+		maxsize -= len;
+		ret += len;
+		do {
+			seg = min_t(size_t, len, PAGE_SIZE - off);
+			if (is_vmalloc_or_module_addr((void *)kaddr))
+				page = vmalloc_to_page((void *)kaddr);
+			else
+				page = virt_to_page(kaddr);
+
+			sg_set_page(sg, page, len, off);
+			sgtable->nents++;
+			sg++;
+			sg_max--;
+
+			len -= seg;
+			kaddr += PAGE_SIZE;
+			off = 0;
+		} while (len > 0 && sg_max > 0);
+
+		if (maxsize <= 0 || sg_max == 0)
+			break;
+		start = 0;
+	}
+
+	if (ret > 0)
+		iov_iter_advance(iter, ret);
+	return ret;
+}
+
+/*
+ * Extract up to sg_max folios from an XARRAY-type iterator and add them to
+ * the scatterlist.  The pages are not pinned.
+ */
+static ssize_t netfs_extract_xarray_to_sg(struct iov_iter *iter,
+					  ssize_t maxsize,
+					  struct sg_table *sgtable,
+					  unsigned int sg_max,
+					  unsigned int extract_flags)
+{
+	struct scatterlist *sg = sgtable->sgl + sgtable->nents;
+	struct xarray *xa = iter->xarray;
+	struct folio *folio;
+	loff_t start = iter->xarray_start + iter->iov_offset;
+	pgoff_t index = start / PAGE_SIZE;
+	ssize_t ret = 0;
+	size_t offset, len;
+	XA_STATE(xas, xa, index);
+
+	rcu_read_lock();
+
+	xas_for_each(&xas, folio, ULONG_MAX) {
+		if (xas_retry(&xas, folio))
+			continue;
+		if (WARN_ON(xa_is_value(folio)))
+			break;
+		if (WARN_ON(folio_test_hugetlb(folio)))
+			break;
+
+		offset = offset_in_folio(folio, start);
+		len = min_t(size_t, maxsize, folio_size(folio) - offset);
+
+		sg_set_page(sg, folio_page(folio, 0), len, offset);
+		sgtable->nents++;
+		sg++;
+		sg_max--;
+
+		maxsize -= len;
+		ret += len;
+		if (maxsize <= 0 || sg_max == 0)
+			break;
+	}
+
+	rcu_read_unlock();
+	if (ret > 0)
+		iov_iter_advance(iter, ret);
+	return ret;
+}
+
+/**
+ * netfs_extract_iter_to_sg - Extract pages from an iterator and add ot an sglist
+ * @iter: The iterator to extract from
+ * @maxsize: The amount of iterator to copy
+ * @sgtable: The scatterlist table to fill in
+ * @sg_max: Maximum number of elements in @sgtable that may be filled
+ * @extract_flags: Flags to qualify the request
+ *
+ * Extract the page fragments from the given amount of the source iterator and
+ * add them to a scatterlist that refers to all of those bits, to a maximum
+ * addition of @sg_max elements.
+ *
+ * The pages referred to by UBUF- and IOVEC-type iterators are extracted and
+ * pinned; BVEC-, KVEC- and XARRAY-type are extracted but aren't pinned; PIPE-
+ * and DISCARD-type are not supported.
+ *
+ * No end mark is placed on the scatterlist; that's left to the caller.
+ *
+ * @extract_flags can have ITER_ALLOW_P2PDMA set to request peer-to-peer DMA be
+ * allowed on the pages extracted.
+ *
+ * If successul, @sgtable->nents is updated to include the number of elements
+ * added and the number of bytes added is returned.  @sgtable->orig_nents is
+ * left unaltered.
+ *
+ * The iov_iter_extract_mode() function should be used to query how cleanup
+ * should be performed.
+ */
+ssize_t netfs_extract_iter_to_sg(struct iov_iter *iter, size_t maxsize,
+				 struct sg_table *sgtable, unsigned int sg_max,
+				 unsigned int extract_flags)
+{
+	if (maxsize == 0)
+		return 0;
+
+	switch (iov_iter_type(iter)) {
+	case ITER_UBUF:
+	case ITER_IOVEC:
+		return netfs_extract_user_to_sg(iter, maxsize, sgtable, sg_max,
+						extract_flags);
+	case ITER_BVEC:
+		return netfs_extract_bvec_to_sg(iter, maxsize, sgtable, sg_max,
+						extract_flags);
+	case ITER_KVEC:
+		return netfs_extract_kvec_to_sg(iter, maxsize, sgtable, sg_max,
+						extract_flags);
+	case ITER_XARRAY:
+		return netfs_extract_xarray_to_sg(iter, maxsize, sgtable, sg_max,
+						  extract_flags);
+	default:
+		pr_err("netfs_extract_iter_to_sg(%u) unsupported\n",
+		       iov_iter_type(iter));
+		WARN_ON_ONCE(1);
+		return -EIO;
+	}
+}
+EXPORT_SYMBOL_GPL(netfs_extract_iter_to_sg);
include/linux/netfs.h +4 -0
index e8c560131170..955225be6b75 100644
@@ -298,6 +298,10 @@ void netfs_put_subrequest(struct netfs_io_subrequest *subreq,
 void netfs_stats_show(struct seq_file *);
 ssize_t netfs_extract_user_iter(struct iov_iter *orig, size_t orig_len,
 				struct iov_iter *new, unsigned int extract_flags);
+struct sg_table;
+ssize_t netfs_extract_iter_to_sg(struct iov_iter *iter, size_t len,
+				 struct sg_table *sgtable, unsigned int sg_max,
+				 unsigned int extract_flags);
 
 /**
  * netfs_inode - Get the netfs inode context from the inode
mm/vmalloc.c +1 -0
index ca71de7c9d77..61f5bec0f2b6 100644
@@ -656,6 +656,7 @@ int is_vmalloc_or_module_addr(const void *x)
 #endif
 	return is_vmalloc_addr(x);
 }
+EXPORT_SYMBOL_GPL(is_vmalloc_or_module_addr);
 
 /*
  * Walk a vmap address to the struct page it maps. Huge vmap mappings will
author maintainer
#3 lore raw

PATCH 3/12 cifs: Implement splice_read to pass down ITER_BVEC not ITER_PIPE

In reply to David Howells

Provide cifs_splice_read() to use a bvec rather than an pipe iterator as the latter cannot so easily be split and advanced, which is necessary to pass an iterator down to the bottom levels. Upstream cifs gets around this problem by using iov_iter_get_pages() to prefill the pipe and then passing the list of pages down.

This is done by:

(1) Bulk-allocate a bunch of pages to carry as much of the requested amount of data as possible, but without overrunning the available slots in the pipe and add them to an ITER_BVEC.

(2) Synchronously call ->read_iter() to read into the buffer.

(3) Discard any unused pages.

(4) Load the remaining pages into the pipe in order and advance the head pointer.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Jeff Layton <[email protected]>
cc Al Viro <[email protected]>
---
 fs/cifs/cifsfs.c | 12 +++----
 fs/cifs/cifsfs.h |  3 ++
 fs/cifs/file.c   | 92 ++++++++++++++++++++++++++++++++++++++++++++++++
 fs/splice.c      |  1 +
 4 files changed, 102 insertions(+), 6 deletions(-)
fs/cifs/cifsfs.c +6 -6
index 10e00c624922..3c57e8b11692 100644
@@ -1358,7 +1358,7 @@ const struct file_operations cifs_file_ops = {
 	.fsync = cifs_fsync,
 	.flush = cifs_flush,
 	.mmap  = cifs_file_mmap,
-	.splice_read = generic_file_splice_read,
+	.splice_read = cifs_splice_read,
 	.splice_write = iter_file_splice_write,
 	.llseek = cifs_llseek,
 	.unlocked_ioctl	= cifs_ioctl,
@@ -1378,7 +1378,7 @@ const struct file_operations cifs_file_strict_ops = {
 	.fsync = cifs_strict_fsync,
 	.flush = cifs_flush,
 	.mmap = cifs_file_strict_mmap,
-	.splice_read = generic_file_splice_read,
+	.splice_read = cifs_splice_read,
 	.splice_write = iter_file_splice_write,
 	.llseek = cifs_llseek,
 	.unlocked_ioctl	= cifs_ioctl,
@@ -1398,7 +1398,7 @@ const struct file_operations cifs_file_direct_ops = {
 	.fsync = cifs_fsync,
 	.flush = cifs_flush,
 	.mmap = cifs_file_mmap,
-	.splice_read = generic_file_splice_read,
+	.splice_read = cifs_splice_read,
 	.splice_write = iter_file_splice_write,
 	.unlocked_ioctl  = cifs_ioctl,
 	.copy_file_range = cifs_copy_file_range,
@@ -1416,7 +1416,7 @@ const struct file_operations cifs_file_nobrl_ops = {
 	.fsync = cifs_fsync,
 	.flush = cifs_flush,
 	.mmap  = cifs_file_mmap,
-	.splice_read = generic_file_splice_read,
+	.splice_read = cifs_splice_read,
 	.splice_write = iter_file_splice_write,
 	.llseek = cifs_llseek,
 	.unlocked_ioctl	= cifs_ioctl,
@@ -1434,7 +1434,7 @@ const struct file_operations cifs_file_strict_nobrl_ops = {
 	.fsync = cifs_strict_fsync,
 	.flush = cifs_flush,
 	.mmap = cifs_file_strict_mmap,
-	.splice_read = generic_file_splice_read,
+	.splice_read = cifs_splice_read,
 	.splice_write = iter_file_splice_write,
 	.llseek = cifs_llseek,
 	.unlocked_ioctl	= cifs_ioctl,
@@ -1452,7 +1452,7 @@ const struct file_operations cifs_file_direct_nobrl_ops = {
 	.fsync = cifs_fsync,
 	.flush = cifs_flush,
 	.mmap = cifs_file_mmap,
-	.splice_read = generic_file_splice_read,
+	.splice_read = cifs_splice_read,
 	.splice_write = iter_file_splice_write,
 	.unlocked_ioctl  = cifs_ioctl,
 	.copy_file_range = cifs_copy_file_range,
fs/cifs/cifsfs.h +3 -0
index 1705c76529d8..2e979d2f4e36 100644
@@ -100,6 +100,9 @@ extern ssize_t cifs_strict_readv(struct kiocb *iocb, struct iov_iter *to);
 extern ssize_t cifs_user_writev(struct kiocb *iocb, struct iov_iter *from);
 extern ssize_t cifs_direct_writev(struct kiocb *iocb, struct iov_iter *from);
 extern ssize_t cifs_strict_writev(struct kiocb *iocb, struct iov_iter *from);
+extern ssize_t cifs_splice_read(struct file *in, loff_t *ppos,
+				struct pipe_inode_info *pipe, size_t len,
+				unsigned int flags);
 extern int cifs_flock(struct file *pfile, int cmd, struct file_lock *plock);
 extern int cifs_lock(struct file *, int, struct file_lock *);
 extern int cifs_fsync(struct file *, loff_t, loff_t, int);
fs/cifs/file.c +92 -0

95 lines not on this page · raw message

fs/splice.c +1 -0
index 5969b7a1d353..95435b5cca2a 100644
@@ -330,6 +330,7 @@ const struct pipe_buf_operations default_pipe_buf_ops = {
 	.try_steal	= generic_pipe_buf_try_steal,
 	.get		= generic_pipe_buf_get,
 };
+EXPORT_SYMBOL(default_pipe_buf_ops);
 
 /* Pipe buffer operations for a socket and similar. */
 const struct pipe_buf_operations nosteal_pipe_buf_ops = {
#4 lore raw

Re: [PATCH 03/12] cifs: Implement splice_read to pass down ITER_BVEC not ITER_PIPE

Quoted 2 lines · David Howells

Provide cifs_splice_read() to use a bvec rather than an pipe iterator as the latter cannot so easily be split and advanced,

What exactly are the issues? If cifs is running into them others will as well, so we might better fix them in the infrastructure.

Quoted 3 lines

(1) Bulk-allocate a bunch of pages to carry as much of the requested amount of data as possible, but without overrunning the available slots in the pipe and add them to an ITER_BVEC.

And that basically makes the splice_read entirely pointless and you might as well simply not support it.

author maintainer
#5 lore raw

PATCH 4/12 cifs: Add a function to build an RDMA SGE list from an iterator

In reply to David Howells

Add a function to add elements onto an RDMA SGE list representing page fragments extracted from a BVEC-, KVEC- or XARRAY-type iterator and DMA mapped until the maximum number of elements is reached.

Nothing is done to make sure the pages remain present - that must be done by the caller.

---
 fs/cifs/smbdirect.c | 220 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 220 insertions(+)
fs/cifs/smbdirect.c +220 -0
index 8c816b25ce7c..44eff1cf4ed0 100644
@@ -44,6 +44,17 @@ static int smbd_post_send_page(struct smbd_connection *info,
 static void destroy_mr_list(struct smbd_connection *info);
 static int allocate_mr_list(struct smbd_connection *info);
 
+struct smb_extract_to_rdma {
+	struct ib_sge		*sge;
+	unsigned int		nr_sge;
+	unsigned int		max_sge;
+	struct ib_device	*device;
+	u32			local_dma_lkey;
+	enum dma_data_direction	direction;
+};
+static ssize_t smb_extract_iter_to_rdma(struct iov_iter *iter, size_t len,
+					struct smb_extract_to_rdma *rdma);
+
 /* SMBD version number */
 #define SMBD_V1	0x0100
 
@@ -2490,3 +2501,212 @@ int smbd_deregister_mr(struct smbd_mr *smbdirect_mr)
 
 	return rc;
 }
+
+static bool smb_set_sge(struct smb_extract_to_rdma *rdma,
+			struct page *lowest_page, size_t off, size_t len)
+{
+	struct ib_sge *sge = &rdma->sge[rdma->nr_sge];
+	u64 addr;
+
+	addr = ib_dma_map_page(rdma->device, lowest_page,
+			       off, len, rdma->direction);
+	if (ib_dma_mapping_error(rdma->device, addr))
+		return false;
+
+	sge->addr   = addr;
+	sge->length = len;
+	sge->lkey   = rdma->local_dma_lkey;
+	rdma->nr_sge++;
+	return true;
+}
+
+/*
+ * Extract page fragments from a BVEC-class iterator and add them to an RDMA
+ * element list.  The pages are not pinned.
+ */
+static ssize_t smb_extract_bvec_to_rdma(struct iov_iter *iter,
+					struct smb_extract_to_rdma *rdma,
+					ssize_t maxsize)
+{
+	const struct bio_vec *bv = iter->bvec;
+	unsigned long start = iter->iov_offset;
+	unsigned int i;
+	ssize_t ret = 0;
+
+	for (i = 0; i < iter->nr_segs; i++) {
+		size_t off, len;
+
+		len = bv[i].bv_len;
+		if (start >= len) {
+			start -= len;
+			continue;
+		}
+
+		len = min_t(size_t, maxsize, len - start);
+		off = bv[i].bv_offset + start;
+
+		if (!smb_set_sge(rdma, bv[i].bv_page, off, len))
+			return -EIO;
+
+		ret += len;
+		maxsize -= len;
+		if (rdma->nr_sge >= rdma->max_sge || maxsize <= 0)
+			break;
+		start = 0;
+	}
+
+	return ret;
+}
+
+/*
+ * Extract fragments from a KVEC-class iterator and add them to an RDMA list.
+ * This can deal with vmalloc'd buffers as well as kmalloc'd or static buffers.
+ * The pages are not pinned.
+ */
+static ssize_t smb_extract_kvec_to_rdma(struct iov_iter *iter,
+					struct smb_extract_to_rdma *rdma,
+					ssize_t maxsize)
+{
+	const struct kvec *kv = iter->kvec;
+	unsigned long start = iter->iov_offset;
+	unsigned int i;
+	ssize_t ret = 0;
+
+	for (i = 0; i < iter->nr_segs; i++) {
+		struct page *page;
+		unsigned long kaddr;
+		size_t off, len, seg;
+
+		len = kv[i].iov_len;
+		if (start >= len) {
+			start -= len;
+			continue;
+		}
+
+		kaddr = (unsigned long)kv[i].iov_base + start;
+		off = kaddr & ~PAGE_MASK;
+		len = min_t(size_t, maxsize, len - start);
+		kaddr &= PAGE_MASK;
+
+		maxsize -= len;
+		do {
+			seg = min_t(size_t, len, PAGE_SIZE - off);
+
+			if (is_vmalloc_or_module_addr((void *)kaddr))
+				page = vmalloc_to_page((void *)kaddr);
+			else
+				page = virt_to_page(kaddr);
+
+			if (!smb_set_sge(rdma, page, off, seg))
+				return -EIO;
+
+			ret += seg;
+			len -= seg;
+			kaddr += PAGE_SIZE;
+			off = 0;
+		} while (len > 0 && rdma->nr_sge < rdma->max_sge);
+
+		if (rdma->nr_sge >= rdma->max_sge || maxsize <= 0)
+			break;
+		start = 0;
+	}
+
+	return ret;
+}
+
+/*
+ * Extract folio fragments from an XARRAY-class iterator and add them to an
+ * RDMA list.  The folios are not pinned.
+ */
+static ssize_t smb_extract_xarray_to_rdma(struct iov_iter *iter,
+					  struct smb_extract_to_rdma *rdma,
+					  ssize_t maxsize)
+{
+	struct xarray *xa = iter->xarray;
+	struct folio *folio;
+	loff_t start = iter->xarray_start + iter->iov_offset;
+	pgoff_t index = start / PAGE_SIZE;
+	ssize_t ret = 0;
+	size_t off, len;
+	XA_STATE(xas, xa, index);
+
+	rcu_read_lock();
+
+	xas_for_each(&xas, folio, ULONG_MAX) {
+		if (xas_retry(&xas, folio))
+			continue;
+		if (WARN_ON(xa_is_value(folio)))
+			break;
+		if (WARN_ON(folio_test_hugetlb(folio)))
+			break;
+
+		off = offset_in_folio(folio, start);
+		len = min_t(size_t, maxsize, folio_size(folio) - off);
+
+		if (!smb_set_sge(rdma, folio_page(folio, 0), off, len)) {
+			rcu_read_lock();
+			return -EIO;
+		}
+
+		maxsize -= len;
+		ret += len;
+		if (rdma->nr_sge >= rdma->max_sge || maxsize <= 0)
+			break;
+	}
+
+	rcu_read_unlock();
+	return ret;
+}
+
+/*
+ * Extract page fragments from up to the given amount of the source iterator
+ * and build up an RDMA list that refers to all of those bits.  The RDMA list
+ * is appended to, up to the maximum number of elements set in the parameter
+ * block.
+ *
+ * The extracted page fragments are not pinned or ref'd in any way; if an
+ * IOVEC/UBUF-type iterator is to be used, it should be converted to a
+ * BVEC-type iterator and the pages pinned, ref'd or otherwise held in some
+ * way.
+ */
+static ssize_t smb_extract_iter_to_rdma(struct iov_iter *iter, size_t len,
+					struct smb_extract_to_rdma *rdma)
+{
+	ssize_t ret;
+	int before = rdma->nr_sge;
+
+	if (iov_iter_is_discard(iter) ||
+	    iov_iter_is_pipe(iter) ||
+	    user_backed_iter(iter)) {
+		WARN_ON_ONCE(1);
+		return -EIO;
+	}
+
+	switch (iov_iter_type(iter)) {
+	case ITER_BVEC:
+		ret = smb_extract_bvec_to_rdma(iter, rdma, len);
+		break;
+	case ITER_KVEC:
+		ret = smb_extract_kvec_to_rdma(iter, rdma, len);
+		break;
+	case ITER_XARRAY:
+		ret = smb_extract_xarray_to_rdma(iter, rdma, len);
+		break;
+	default:
+		BUG();
+	}
+
+	if (ret > 0) {
+		iov_iter_advance(iter, ret);
+	} else if (ret < 0) {
+		while (rdma->nr_sge > before) {
+			struct ib_sge *sge = &rdma->sge[rdma->nr_sge--];
+
+			ib_dma_unmap_single(rdma->device, sge->addr, sge->length,
+					    rdma->direction);
+			sge->addr = 0;
+		}
+	}
+
+	return ret;
+}
author maintainer
#6 lore raw

PATCH 5/12 cifs: Add a function to Hash the contents of an iterator

In reply to David Howells

Add a function to push the contents of a BVEC-, KVEC- or XARRAY-type iterator into a symmetric hash algorithm.

UBUF- and IOBUF-type iterators are not supported on the assumption that either we're doing buffered I/O, in which case we won't see them, or we're doing direct I/O, in which case the iterator will have been extracted into a BVEC-type iterator higher up.

---
 fs/cifs/cifsencrypt.c | 144 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 144 insertions(+)
fs/cifs/cifsencrypt.c +144 -0
index cbc18b4a9cb2..7be589aeb520 100644
@@ -24,6 +24,150 @@
 #include "../smbfs_common/arc4.h"
 #include <crypto/aead.h>
 
+/*
+ * Hash data from a BVEC-type iterator.
+ */
+static int cifs_shash_bvec(const struct iov_iter *iter, ssize_t maxsize,
+			   struct shash_desc *shash)
+{
+	const struct bio_vec *bv = iter->bvec;
+	unsigned long start = iter->iov_offset;
+	unsigned int i;
+	void *p;
+	int ret;
+
+	for (i = 0; i < iter->nr_segs; i++) {
+		size_t off, len;
+
+		len = bv[i].bv_len;
+		if (start >= len) {
+			start -= len;
+			continue;
+		}
+
+		len = min_t(size_t, maxsize, len - start);
+		off = bv[i].bv_offset + start;
+
+		p = kmap_local_page(bv[i].bv_page);
+		ret = crypto_shash_update(shash, p + off, len);
+		kunmap_local(p);
+		if (ret < 0)
+			return ret;
+
+		maxsize -= len;
+		if (maxsize <= 0)
+			break;
+		start = 0;
+	}
+
+	return 0;
+}
+
+/*
+ * Hash data from a KVEC-type iterator.
+ */
+static int cifs_shash_kvec(const struct iov_iter *iter, ssize_t maxsize,
+			   struct shash_desc *shash)
+{
+	const struct kvec *kv = iter->kvec;
+	unsigned long start = iter->iov_offset;
+	unsigned int i;
+	int ret;
+
+	for (i = 0; i < iter->nr_segs; i++) {
+		size_t len;
+
+		len = kv[i].iov_len;
+		if (start >= len) {
+			start -= len;
+			continue;
+		}
+
+		len = min_t(size_t, maxsize, len - start);
+		ret = crypto_shash_update(shash, kv[i].iov_base + start, len);
+		if (ret < 0)
+			return ret;
+		maxsize -= len;
+
+		if (maxsize <= 0)
+			break;
+		start = 0;
+	}
+
+	return 0;
+}
+
+/*
+ * Hash data from an XARRAY-type iterator.
+ */
+static ssize_t cifs_shash_xarray(const struct iov_iter *iter, ssize_t maxsize,
+				 struct shash_desc *shash)
+{
+	struct folio *folios[16], *folio;
+	unsigned int nr, i, j, npages;
+	loff_t start = iter->xarray_start + iter->iov_offset;
+	pgoff_t last, index = start / PAGE_SIZE;
+	ssize_t ret = 0;
+	size_t len, offset, foffset;
+	void *p;
+
+	if (maxsize == 0)
+		return 0;
+
+	last = (start + maxsize - 1) / PAGE_SIZE;
+	do {
+		nr = xa_extract(iter->xarray, (void **)folios, index, last,
+				ARRAY_SIZE(folios), XA_PRESENT);
+		if (nr == 0)
+			return -EIO;
+
+		for (i = 0; i < nr; i++) {
+			folio = folios[i];
+			npages = folio_nr_pages(folio);
+			foffset = start - folio_pos(folio);
+			offset = foffset % PAGE_SIZE;
+			for (j = foffset / PAGE_SIZE; j < npages; j++) {
+				len = min_t(size_t, maxsize, PAGE_SIZE - offset);
+				p = kmap_local_page(folio_page(folio, j));
+				ret = crypto_shash_update(shash, p, len);
+				kunmap_local(p);
+				if (ret < 0)
+					return ret;
+				maxsize -= len;
+				if (maxsize <= 0)
+					return 0;
+				start += len;
+				offset = 0;
+				index++;
+			}
+		}
+	} while (nr == ARRAY_SIZE(folios));
+	return 0;
+}
+
+/*
+ * Pass the data from an iterator into a hash.
+ */
+static int cifs_shash_iter(const struct iov_iter *iter, size_t maxsize,
+			   struct shash_desc *shash)
+{
+	if (maxsize == 0)
+		return 0;
+
+	switch (iov_iter_type(iter)) {
+	case ITER_BVEC:
+		return cifs_shash_bvec(iter, maxsize, shash);
+	case ITER_KVEC:
+		return cifs_shash_kvec(iter, maxsize, shash);
+	case ITER_XARRAY:
+		return cifs_shash_xarray(iter, maxsize, shash);
+	default:
+		pr_err("cifs_shash_iter(%u) unsupported\n", iov_iter_type(iter));
+		WARN_ON_ONCE(1);
+		return -EIO;
+	}
+}
+
 int __cifs_calc_signature(struct smb_rqst *rqst,
 			struct TCP_Server_Info *server, char *signature,
 			struct shash_desc *shash)
maintainer
#7 lore raw

Re: [PATCH 05/12] cifs: Add a function to Hash the contents of an iterator

Quoted 2 lines · David Howells

Add a function to push the contents of a BVEC-, KVEC- or XARRAY-type iterator into a symmetric hash algorithm.

There is no such thing as a symmetric hash. You're using shash which stands for synchronous hash.

In any case, as I said in the previous review this is abusing the shash API. Please use ahash instead.

Thanks,

Signature · 3 lines
Email: Herbert Xu <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
author maintainer
#8 lore raw

Re: [PATCH 05/12] cifs: Add a function to Hash the contents of an iterator

Quoted 5 lines · Herbert Xu
Quoted 2 lines

Add a function to push the contents of a BVEC-, KVEC- or XARRAY-type iterator into a symmetric hash algorithm.

There is no such thing as a symmetric hash. You're using shash which stands for synchronous hash.

Sorry, you're right.

Quoted 2 lines

In any case, as I said in the previous review this is abusing the shash API. Please use ahash instead.

It's already abusing the shash API, this doesn't change that, except where it gets the data from.

I can have a go at conversion to ahash, but that will have to be a separate patch set as it isn't trivial. The problem is that the current code just assumes it can push bits into the hash as it gets them - this is not possible with ahash.

David

maintainer
#9 lore raw

Re: [PATCH 05/12] cifs: Add a function to Hash the contents of an iterator

Quoted 3 lines · David Howells

It's already abusing the shash API, this doesn't change that, except where it gets the data from.

Oh OK. All I saw was this one patch which added shash code. Were you also deleting shash code in another patch that wasn't cc'ed to me?

Cheers,

Signature · 3 lines
Email: Herbert Xu <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
author maintainer
#10 lore raw

Re: [PATCH 05/12] cifs: Add a function to Hash the contents of an iterator

Quoted 2 lines · Herbert Xu

Were you also deleting shash code in another patch that wasn't cc'ed to me?

No, I don't want to change it this late into the dev cycle. I can look at it after the merge window.

David

author maintainer
#11 lore raw

PATCH 6/12 cifs: Add some helper functions

In reply to David Howells

Add some helper functions to manipulate the folio marks by iterating through a list of folios held in an xarray rather than using a page list.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Jeff Layton <[email protected]>
---
 fs/cifs/cifsfs.h |  3 ++
 fs/cifs/file.c   | 93 ++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 96 insertions(+)
fs/cifs/cifsfs.h +3 -0
index 2e979d2f4e36..845a0dda5864 100644
@@ -113,6 +113,9 @@ extern int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma);
 extern const struct file_operations cifs_dir_ops;
 extern int cifs_dir_open(struct inode *inode, struct file *file);
 extern int cifs_readdir(struct file *file, struct dir_context *ctx);
+extern void cifs_pages_written_back(struct inode *inode, loff_t start, unsigned int len);
+extern void cifs_pages_write_failed(struct inode *inode, loff_t start, unsigned int len);
+extern void cifs_pages_write_redirty(struct inode *inode, loff_t start, unsigned int len);
 
 /* Functions related to dir entries */
 extern const struct dentry_operations cifs_dentry_ops;
fs/cifs/file.c +93 -0

99 lines not on this page · raw message

author maintainer
#12 lore raw

PATCH 7/12 cifs: Add a function to read into an iter from a socket

In reply to David Howells

Add a helper function to read data from a socket into the given iterator.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Jeff Layton <[email protected]>
---
 fs/cifs/cifsproto.h |  3 +++
 fs/cifs/connect.c   | 14 ++++++++++++++
 2 files changed, 17 insertions(+)
fs/cifs/cifsproto.h +3 -0
index 1207b39686fb..cb7a3fe89278 100644
@@ -244,6 +244,9 @@ extern int cifs_read_page_from_socket(struct TCP_Server_Info *server,
 					struct page *page,
 					unsigned int page_offset,
 					unsigned int to_read);
+int cifs_read_iter_from_socket(struct TCP_Server_Info *server,
+			       struct iov_iter *iter,
+			       unsigned int to_read);
 extern int cifs_setup_cifs_sb(struct cifs_sb_info *cifs_sb);
 void cifs_mount_put_conns(struct cifs_mount_ctx *mnt_ctx);
 int cifs_mount_get_session(struct cifs_mount_ctx *mnt_ctx);
fs/cifs/connect.c +14 -0
index b2a04b4e89a5..152b457b849f 100644
@@ -765,6 +765,20 @@ cifs_read_page_from_socket(struct TCP_Server_Info *server, struct page *page,
 	return cifs_readv_from_socket(server, &smb_msg);
 }
 
+int
+cifs_read_iter_from_socket(struct TCP_Server_Info *server, struct iov_iter *iter,
+			   unsigned int to_read)
+{
+	struct msghdr smb_msg = { .msg_iter = *iter };
+	int ret;
+
+	iov_iter_truncate(&smb_msg.msg_iter, to_read);
+	ret = cifs_readv_from_socket(server, &smb_msg);
+	if (ret > 0)
+		iov_iter_advance(iter, ret);
+	return ret;
+}
+
 static bool
 is_smb_response(struct TCP_Server_Info *server, unsigned char type)
 {
author maintainer
#13 lore raw

PATCH 8/12 cifs: Change the I/O paths to use an iterator rather than a page list

In reply to David Howells

Currently, the cifs I/O paths hand lists of pages from the VM interface routines at the top all the way through the intervening layers to the socket interface at the bottom.

This is a problem, however, for interfacing with netfslib which passes an iterator through to the ->issue_read() method (and will pass an iterator through to the ->issue_write() method in future). Netfslib takes over bounce buffering for direct I/O, async I/O and encrypted content, so cifs doesn't need to do that. Netfslib also converts IOVEC-type iterators into BVEC-type iterators if necessary.

Further, cifs needs foliating - and folios may come in a variety of sizes, so a page list pointing to an array of heterogeneous pages may cause problems in places such as where crypto is done.

Change the cifs I/O paths to hand iov_iter iterators all the way through instead.

Notes:

(1) Some old routines are #if'd out to be removed in a follow up patch so as to avoid confusing diff, thereby making the diff output easier to follow. I've removed functions that don't overlap with anything added.

(2) struct smb_rqst loses rq_pages, rq_offset, rq_npages, rq_pagesz and rq_tailsz which describe the pages forming the buffer; instead there's an rq_iter describing the source buffer and an rq_buffer which is used to hold the buffer for encryption.

(3) struct cifs_readdata and cifs_writedata are similarly modified to smb_rqst. The ->read_into_pages() and ->copy_into_pages() are then replaced with passing the iterator directly to the socket.

The iterators are stored in these structs so that they are persistent and don't get deallocated when the function returns (unlike if they were stack variables).

(4) Buffered writeback is overhauled, borrowing the code from the afs filesystem to gather up contiguous runs of folios. The XARRAY-type iterator is then used to refer directly to the pagecache and can be passed to the socket to transmit data directly from there.

This includes:

cifs_extend_writeback()
cifs_write_back_from_locked_folio()
cifs_writepages_region()
cifs_writepages()

(5) Pages are converted to folios.

(6) Direct I/O uses netfs_extract_user_iter() to create a BVEC-type iterator from an IOBUF/UBUF-type source iterator.

(7) smb2_get_aead_req() uses netfs_extract_iter_to_sg() to extract page fragments from the iterator into the scatterlists that the crypto layer prefers.

(8) smb2_init_transform_rq() attached pages to smb_rqst::rq_buffer, an xarray, to use as a bounce buffer for encryption. An XARRAY-type iterator can then be used to pass the bounce buffer to lower layers.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Paulo Alcantara <[email protected]>
cc Jeff Layton <[email protected]>
---
 fs/cifs/Kconfig       |    1 +
 fs/cifs/cifsencrypt.c |   28 +-
 fs/cifs/cifsglob.h    |   66 +--
 fs/cifs/cifsproto.h   |    8 +-
 fs/cifs/cifssmb.c     |   15 +-
 fs/cifs/file.c        | 1199 ++++++++++++++++++++++++++---------------
 fs/cifs/fscache.c     |   22 +-
 fs/cifs/fscache.h     |   10 +-
 fs/cifs/misc.c        |  128 +----
 fs/cifs/smb2ops.c     |  362 ++++++-------
 fs/cifs/smb2pdu.c     |   53 +-
 fs/cifs/smbdirect.c   |  262 ++++-----
 fs/cifs/smbdirect.h   |    4 +-
 fs/cifs/transport.c   |   54 +-
 14 files changed, 1123 insertions(+), 1089 deletions(-)
fs/cifs/Kconfig +1 -0
index bbf63a9eb927..4c0d53bf931a 100644
@@ -18,6 +18,7 @@ config CIFS
 	select DNS_RESOLVER
 	select ASN1
 	select OID_REGISTRY
+	select NETFS_SUPPORT
 	help
 	  This is the client VFS module for the SMB3 family of network file
 	  protocols (including the most recent, most secure dialect SMB3.1.1).
fs/cifs/cifsencrypt.c +6 -22
index 7be589aeb520..357bd27a7fd1 100644
@@ -169,11 +169,11 @@ static int cifs_shash_iter(const struct iov_iter *iter, size_t maxsize,
 }
 
 int __cifs_calc_signature(struct smb_rqst *rqst,
-			struct TCP_Server_Info *server, char *signature,
-			struct shash_desc *shash)
+			  struct TCP_Server_Info *server, char *signature,
+			  struct shash_desc *shash)
 {
 	int i;
-	int rc;
+	ssize_t rc;
 	struct kvec *iov = rqst->rq_iov;
 	int n_vec = rqst->rq_nvec;
 
@@ -205,25 +205,9 @@ int __cifs_calc_signature(struct smb_rqst *rqst,
 		}
 	}
 
-	/* now hash over the rq_pages array */
-	for (i = 0; i < rqst->rq_npages; i++) {
-		void *kaddr;
-		unsigned int len, offset;
-
-		rqst_page_get_length(rqst, i, &len, &offset);
-
-		kaddr = (char *) kmap(rqst->rq_pages[i]) + offset;
-
-		rc = crypto_shash_update(shash, kaddr, len);
-		if (rc) {
-			cifs_dbg(VFS, "%s: Could not update with payload\n",
-				 __func__);
-			kunmap(rqst->rq_pages[i]);
-			return rc;
-		}
-
-		kunmap(rqst->rq_pages[i]);
-	}
+	rc = cifs_shash_iter(&rqst->rq_iter, iov_iter_count(&rqst->rq_iter), shash);
+	if (rc < 0)
+		return rc;
 
 	rc = crypto_shash_final(shash, signature);
 	if (rc)
fs/cifs/cifsglob.h +33 -33

147 lines not on this page · raw message

fs/cifs/cifsproto.h +1 -7
index cb7a3fe89278..2873f68a051c 100644
@@ -584,10 +584,7 @@ int cifs_readv_receive(struct TCP_Server_Info *server, struct mid_q_entry *mid);
 int cifs_async_writev(struct cifs_writedata *wdata,
 		      void (*release)(struct kref *kref));
 void cifs_writev_complete(struct work_struct *work);
-struct cifs_writedata *cifs_writedata_alloc(unsigned int nr_pages,
-						work_func_t complete);
-struct cifs_writedata *cifs_writedata_direct_alloc(struct page **pages,
-						work_func_t complete);
+struct cifs_writedata *cifs_writedata_alloc(work_func_t complete);
 void cifs_writedata_release(struct kref *refcount);
 int cifs_query_mf_symlink(unsigned int xid, struct cifs_tcon *tcon,
 			  struct cifs_sb_info *cifs_sb,
@@ -604,13 +601,10 @@ enum securityEnum cifs_select_sectype(struct TCP_Server_Info *,
 					enum securityEnum);
 struct cifs_aio_ctx *cifs_aio_ctx_alloc(void);
 void cifs_aio_ctx_release(struct kref *refcount);
-int setup_aio_ctx_iter(struct cifs_aio_ctx *ctx, struct iov_iter *iter, int rw);
 
 int cifs_alloc_hash(const char *name, struct shash_desc **sdesc);
 void cifs_free_hash(struct shash_desc **sdesc);
 
-void rqst_page_get_length(const struct smb_rqst *rqst, unsigned int page,
-			  unsigned int *len, unsigned int *offset);
 struct cifs_chan *
 cifs_ses_find_chan(struct cifs_ses *ses, struct TCP_Server_Info *server);
 int cifs_try_adding_channels(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses);
fs/cifs/cifssmb.c +5 -10
index 23f10e0d6e7e..e6c44a742bf3 100644
@@ -24,6 +24,7 @@
 #include <linux/task_io_accounting_ops.h>
 #include <linux/uaccess.h>
 #include "cifspdu.h"
+#include "cifsfs.h"
 #include "cifsglob.h"
 #include "cifsacl.h"
 #include "cifsproto.h"
@@ -1294,11 +1295,8 @@ cifs_readv_callback(struct mid_q_entry *mid)
 	struct TCP_Server_Info *server = tcon->ses->server;
 	struct smb_rqst rqst = { .rq_iov = rdata->iov,
 				 .rq_nvec = 2,
-				 .rq_pages = rdata->pages,
-				 .rq_offset = rdata->page_offset,
-				 .rq_npages = rdata->nr_pages,
-				 .rq_pagesz = rdata->pagesz,
-				 .rq_tailsz = rdata->tailsz };
+				 .rq_iter_size = iov_iter_count(&rdata->iter),
+				 .rq_iter = rdata->iter };
 	struct cifs_credits credits = { .value = 1, .instance = 0 };
 
 	cifs_dbg(FYI, "%s: mid=%llu state=%d result=%d bytes=%u\n",
@@ -1737,11 +1735,8 @@ cifs_async_writev(struct cifs_writedata *wdata,
 
 	rqst.rq_iov = iov;
 	rqst.rq_nvec = 2;
-	rqst.rq_pages = wdata->pages;
-	rqst.rq_offset = wdata->page_offset;
-	rqst.rq_npages = wdata->nr_pages;
-	rqst.rq_pagesz = wdata->pagesz;
-	rqst.rq_tailsz = wdata->tailsz;
+	rqst.rq_iter = wdata->iter;
+	rqst.rq_iter_size = iov_iter_count(&wdata->iter);
 
 	cifs_dbg(FYI, "async write at %llu %u bytes\n",
 		 wdata->offset, wdata->bytes);
fs/cifs/file.c +750 -449

1,647 lines not on this page · raw message

fs/cifs/fscache.c +8 -14
index f6f3a6b75601..47c9f36c11fb 100644
@@ -165,22 +165,16 @@ static int fscache_fallback_read_page(struct inode *inode, struct page *page)
 /*
  * Fallback page writing interface.
  */
-static int fscache_fallback_write_page(struct inode *inode, struct page *page,
-				       bool no_space_allocated_yet)
+static int fscache_fallback_write_pages(struct inode *inode, loff_t start, size_t len,
+					bool no_space_allocated_yet)
 {
 	struct netfs_cache_resources cres;
 	struct fscache_cookie *cookie = cifs_inode_cookie(inode);
 	struct iov_iter iter;
-	struct bio_vec bvec[1];
-	loff_t start = page_offset(page);
-	size_t len = PAGE_SIZE;
 	int ret;
 
 	memset(&cres, 0, sizeof(cres));
-	bvec[0].bv_page		= page;
-	bvec[0].bv_offset	= 0;
-	bvec[0].bv_len		= PAGE_SIZE;
-	iov_iter_bvec(&iter, ITER_SOURCE, bvec, ARRAY_SIZE(bvec), PAGE_SIZE);
+	iov_iter_xarray(&iter, ITER_SOURCE, &inode->i_mapping->i_pages, start, len);
 
 	ret = fscache_begin_write_operation(&cres, cookie);
 	if (ret < 0)
@@ -189,7 +183,7 @@ static int fscache_fallback_write_page(struct inode *inode, struct page *page,
 	ret = cres.ops->prepare_write(&cres, &start, &len, i_size_read(inode),
 				      no_space_allocated_yet);
 	if (ret == 0)
-		ret = fscache_write(&cres, page_offset(page), &iter, NULL, NULL);
+		ret = fscache_write(&cres, start, &iter, NULL, NULL);
 	fscache_end_operation(&cres);
 	return ret;
 }
@@ -213,12 +207,12 @@ int __cifs_readpage_from_fscache(struct inode *inode, struct page *page)
 	return 0;
 }
 
-void __cifs_readpage_to_fscache(struct inode *inode, struct page *page)
+void __cifs_readahead_to_fscache(struct inode *inode, loff_t pos, size_t len)
 {
-	cifs_dbg(FYI, "%s: (fsc: %p, p: %p, i: %p)\n",
-		 __func__, cifs_inode_cookie(inode), page, inode);
+	cifs_dbg(FYI, "%s: (fsc: %p, p: %llx, l: %zx, i: %p)\n",
+		 __func__, cifs_inode_cookie(inode), pos, len, inode);
 
-	fscache_fallback_write_page(inode, page, true);
+	fscache_fallback_write_pages(inode, pos, len, true);
 }
 
 /*
fs/cifs/fscache.h +5 -5
index 67b601041f0a..173999610997 100644
@@ -90,7 +90,7 @@ static inline int cifs_fscache_query_occupancy(struct inode *inode,
 }
 
 extern int __cifs_readpage_from_fscache(struct inode *pinode, struct page *ppage);
-extern void __cifs_readpage_to_fscache(struct inode *pinode, struct page *ppage);
+extern void __cifs_readahead_to_fscache(struct inode *pinode, loff_t pos, size_t len);
 
 
 static inline int cifs_readpage_from_fscache(struct inode *inode,
@@ -101,11 +101,11 @@ static inline int cifs_readpage_from_fscache(struct inode *inode,
 	return -ENOBUFS;
 }
 
-static inline void cifs_readpage_to_fscache(struct inode *inode,
-					    struct page *page)
+static inline void cifs_readahead_to_fscache(struct inode *inode,
+					     loff_t pos, size_t len)
 {
 	if (cifs_inode_cookie(inode))
-		__cifs_readpage_to_fscache(inode, page);
+		__cifs_readahead_to_fscache(inode, pos, len);
 }
 
 #else /* CONFIG_CIFS_FSCACHE */
@@ -141,7 +141,7 @@ cifs_readpage_from_fscache(struct inode *inode, struct page *page)
 }
 
 static inline
-void cifs_readpage_to_fscache(struct inode *inode, struct page *page) {}
+void cifs_readahead_to_fscache(struct inode *inode, loff_t pos, size_t len) {}
 
 #endif /* CONFIG_CIFS_FSCACHE */
 
fs/cifs/misc.c +13 -115

149 lines not on this page · raw message

fs/cifs/smb2ops.c +169 -193

620 lines not on this page · raw message

fs/cifs/smb2pdu.c +17 -36

100 lines not on this page · raw message

fs/cifs/smbdirect.c +97 -165

152 lines not on this page · raw message

fs/cifs/smbdirect.h +2 -2
index 207ef979cd51..be2cf18b7fec 100644

Lines of this file are not shown. View the raw message.

fs/cifs/transport.c +16 -38

54 lines not on this page · raw message

The rest of this diff is not shown. View the raw message.

Part of the diff is not shown here. View the raw message.

author maintainer
#14 lore raw

PATCH 9/12 cifs: Build the RDMA SGE list directly from an iterator

In reply to David Howells

In the depths of the cifs RDMA code, extract part of an iov iterator directly into an SGE list without going through an intermediate scatterlist.

Note that this doesn't support extraction from an IOBUF- or UBUF-type iterator (ie. user-supplied buffer). The assumption is that the higher layers will extract those to a BVEC-type iterator first and do whatever is required to stop the pages from going away.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Tom Talpey <[email protected]>
cc Jeff Layton <[email protected]>
---
 fs/cifs/smbdirect.c | 163 +++++++++++++++++---------------------------
 fs/cifs/smbdirect.h |   3 +-
 2 files changed, 65 insertions(+), 101 deletions(-)
fs/cifs/smbdirect.c +64 -99
index be7dfa2f7fb0..31c4dc8212c3 100644
@@ -828,16 +828,16 @@ static int smbd_post_send(struct smbd_connection *info,
 	return rc;
 }
 
-static int smbd_post_send_sgl(struct smbd_connection *info,
-	struct scatterlist *sgl, int data_length, int remaining_data_length)
+static int smbd_post_send_iter(struct smbd_connection *info,
+			       struct iov_iter *iter,
+			       int *_remaining_data_length)
 {
-	int num_sgs;
 	int i, rc;
 	int header_length;
+	int data_length;
 	struct smbd_request *request;
 	struct smbd_data_transfer *packet;
 	int new_credits;
-	struct scatterlist *sg;
 
 wait_credit:
 	/* Wait for send credits. A SMBD packet needs one credit */
@@ -881,6 +881,30 @@ static int smbd_post_send_sgl(struct smbd_connection *info,
 	}
 
 	request->info = info;
+	memset(request->sge, 0, sizeof(request->sge));
+
+	/* Fill in the data payload to find out how much data we can add */
+	if (iter) {
+		struct smb_extract_to_rdma extract = {
+			.nr_sge		= 1,
+			.max_sge	= SMBDIRECT_MAX_SEND_SGE,
+			.sge		= request->sge,
+			.device		= info->id->device,
+			.local_dma_lkey	= info->pd->local_dma_lkey,
+			.direction	= DMA_TO_DEVICE,
+		};
+
+		rc = smb_extract_iter_to_rdma(iter, *_remaining_data_length,
+					      &extract);
+		if (rc < 0)
+			goto err_dma;
+		data_length = rc;
+		request->num_sge = extract.nr_sge;
+		*_remaining_data_length -= data_length;
+	} else {
+		data_length = 0;
+		request->num_sge = 1;
+	}
 
 	/* Fill in the packet header */
 	packet = smbd_request_payload(request);
@@ -902,7 +926,7 @@ static int smbd_post_send_sgl(struct smbd_connection *info,
 	else
 		packet->data_offset = cpu_to_le32(24);
 	packet->data_length = cpu_to_le32(data_length);
-	packet->remaining_data_length = cpu_to_le32(remaining_data_length);
+	packet->remaining_data_length = cpu_to_le32(*_remaining_data_length);
 	packet->padding = 0;
 
 	log_outgoing(INFO, "credits_requested=%d credits_granted=%d data_offset=%d data_length=%d remaining_data_length=%d\n",
@@ -918,7 +942,6 @@ static int smbd_post_send_sgl(struct smbd_connection *info,
 	if (!data_length)
 		header_length = offsetof(struct smbd_data_transfer, padding);
 
-	request->num_sge = 1;
 	request->sge[0].addr = ib_dma_map_single(info->id->device,
 						 (void *)packet,
 						 header_length,
@@ -932,23 +955,6 @@ static int smbd_post_send_sgl(struct smbd_connection *info,
 	request->sge[0].length = header_length;
 	request->sge[0].lkey = info->pd->local_dma_lkey;
 
-	/* Fill in the packet data payload */
-	num_sgs = sgl ? sg_nents(sgl) : 0;
-	for_each_sg(sgl, sg, num_sgs, i) {
-		request->sge[i+1].addr =
-			ib_dma_map_page(info->id->device, sg_page(sg),
-			       sg->offset, sg->length, DMA_TO_DEVICE);
-		if (ib_dma_mapping_error(
-				info->id->device, request->sge[i+1].addr)) {
-			rc = -EIO;
-			request->sge[i+1].addr = 0;
-			goto err_dma;
-		}
-		request->sge[i+1].length = sg->length;
-		request->sge[i+1].lkey = info->pd->local_dma_lkey;
-		request->num_sge++;
-	}
-
 	rc = smbd_post_send(info, request);
 	if (!rc)
 		return 0;
@@ -987,8 +993,10 @@ static int smbd_post_send_sgl(struct smbd_connection *info,
  */
 static int smbd_post_send_empty(struct smbd_connection *info)
 {
+	int remaining_data_length = 0;
+
 	info->count_send_empty++;
-	return smbd_post_send_sgl(info, NULL, 0, 0);
+	return smbd_post_send_iter(info, NULL, &remaining_data_length);
 }
 
 /*
@@ -1933,42 +1941,6 @@ int smbd_recv(struct smbd_connection *info, struct msghdr *msg)
 	return rc;
 }
 
-/*
- * Send the contents of an iterator
- * @iter: The iterator to send
- * @_remaining_data_length: remaining data to send in this payload
- */
-static int smbd_post_send_iter(struct smbd_connection *info,
-			       struct iov_iter *iter,
-			       int *_remaining_data_length)
-{
-	struct scatterlist sgl[SMBDIRECT_MAX_SEND_SGE - 1];
-	unsigned int max_payload = info->max_send_size - sizeof(struct smbd_data_transfer);
-	ssize_t rc;
-
-	/* We're not expecting a user-backed iter */
-	WARN_ON(iov_iter_extract_will_pin(iter));
-
-	do {
-		struct sg_table sgtable = { .sgl = sgl };
-		size_t maxlen = min_t(size_t, *_remaining_data_length, max_payload);
-
-		sg_init_table(sgtable.sgl, ARRAY_SIZE(sgl));
-		rc = netfs_extract_iter_to_sg(iter, maxlen,
-					      &sgtable, ARRAY_SIZE(sgl), 0);
-		if (rc < 0)
-			break;
-		if (WARN_ON_ONCE(sgtable.nents == 0))
-			return -EIO;
-
-		sg_mark_end(&sgl[sgtable.nents - 1]);
-		*_remaining_data_length -= rc;
-		rc = smbd_post_send_sgl(info, sgl, rc, *_remaining_data_length);
-	} while (rc == 0 && iov_iter_count(iter) > 0);
-
-	return rc;
-}
-
 /*
  * Send data to transport
  * Each rqst is transported as a SMBDirect payload
@@ -2129,10 +2101,10 @@ static void destroy_mr_list(struct smbd_connection *info)
 	cancel_work_sync(&info->mr_recovery_work);
 	list_for_each_entry_safe(mr, tmp, &info->mr_list, list) {
 		if (mr->state == MR_INVALIDATED)
-			ib_dma_unmap_sg(info->id->device, mr->sgl,
-				mr->sgl_count, mr->dir);
+			ib_dma_unmap_sg(info->id->device, mr->sgt.sgl,
+				mr->sgt.nents, mr->dir);
 		ib_dereg_mr(mr->mr);
-		kfree(mr->sgl);
+		kfree(mr->sgt.sgl);
 		kfree(mr);
 	}
 }
@@ -2167,11 +2139,10 @@ static int allocate_mr_list(struct smbd_connection *info)
 				    info->mr_type, info->max_frmr_depth);
 			goto out;
 		}
-		smbdirect_mr->sgl = kcalloc(
-					info->max_frmr_depth,
-					sizeof(struct scatterlist),
-					GFP_KERNEL);
-		if (!smbdirect_mr->sgl) {
+		smbdirect_mr->sgt.sgl = kcalloc(info->max_frmr_depth,
+						sizeof(struct scatterlist),
+						GFP_KERNEL);
+		if (!smbdirect_mr->sgt.sgl) {
 			log_rdma_mr(ERR, "failed to allocate sgl\n");
 			ib_dereg_mr(smbdirect_mr->mr);
 			goto out;
@@ -2190,7 +2161,7 @@ static int allocate_mr_list(struct smbd_connection *info)
 
 	list_for_each_entry_safe(smbdirect_mr, tmp, &info->mr_list, list) {
 		ib_dereg_mr(smbdirect_mr->mr);
-		kfree(smbdirect_mr->sgl);
+		kfree(smbdirect_mr->sgt.sgl);
 		kfree(smbdirect_mr);
 	}
 	return -ENOMEM;
@@ -2244,22 +2215,20 @@ static struct smbd_mr *get_mr(struct smbd_connection *info)
 
 /*
  * Transcribe the pages from an iterator into an MR scatterlist.
- * @iter: The iterator to transcribe
- * @_remaining_data_length: remaining data to send in this payload
  */
 static int smbd_iter_to_mr(struct smbd_connection *info,
 			   struct iov_iter *iter,
-			   struct scatterlist *sgl,
-			   unsigned int num_pages)
+			   struct sg_table *sgt,
+			   unsigned int max_sg)
 {
-	struct sg_table sgtable = { .sgl = sgl };
 	int ret;
 
-	sg_init_table(sgl, num_pages);
+	memset(sgt->sgl, 0, max_sg * sizeof(struct scatterlist));
 
-	ret = netfs_extract_iter_to_sg(iter, iov_iter_count(iter),
-				       &sgtable, num_pages, 0);
+	ret = netfs_extract_iter_to_sg(iter, iov_iter_count(iter), sgt, max_sg, 0);
 	WARN_ON(ret < 0);
+	if (sgt->nents > 0)
+		sg_mark_end(&sgt->sgl[sgt->nents - 1]);
 	return ret;
 }
 
@@ -2296,25 +2265,27 @@ struct smbd_mr *smbd_register_mr(struct smbd_connection *info,
 	dir = writing ? DMA_FROM_DEVICE : DMA_TO_DEVICE;
 	smbdirect_mr->dir = dir;
 	smbdirect_mr->need_invalidate = need_invalidate;
-	smbdirect_mr->sgl_count = num_pages;
+	smbdirect_mr->sgt.nents = 0;
+	smbdirect_mr->sgt.orig_nents = 0;
 
-	log_rdma_mr(INFO, "num_pages=0x%x count=0x%zx\n",
-		    num_pages, iov_iter_count(iter));
-	smbd_iter_to_mr(info, iter, smbdirect_mr->sgl, num_pages);
+	log_rdma_mr(INFO, "num_pages=0x%x count=0x%zx depth=%u\n",
+		    num_pages, iov_iter_count(iter), info->max_frmr_depth);
+	smbd_iter_to_mr(info, iter, &smbdirect_mr->sgt, info->max_frmr_depth);
 
-	rc = ib_dma_map_sg(info->id->device, smbdirect_mr->sgl, num_pages, dir);
+	rc = ib_dma_map_sg(info->id->device, smbdirect_mr->sgt.sgl,
+			   smbdirect_mr->sgt.nents, dir);
 	if (!rc) {
 		log_rdma_mr(ERR, "ib_dma_map_sg num_pages=%x dir=%x rc=%x\n",
 			num_pages, dir, rc);
 		goto dma_map_error;
 	}
 
-	rc = ib_map_mr_sg(smbdirect_mr->mr, smbdirect_mr->sgl, num_pages,
-		NULL, PAGE_SIZE);
-	if (rc != num_pages) {
+	rc = ib_map_mr_sg(smbdirect_mr->mr, smbdirect_mr->sgt.sgl,
+			  smbdirect_mr->sgt.nents, NULL, PAGE_SIZE);
+	if (rc != smbdirect_mr->sgt.nents) {
 		log_rdma_mr(ERR,
-			"ib_map_mr_sg failed rc = %d num_pages = %x\n",
-			rc, num_pages);
+			"ib_map_mr_sg failed rc = %d nents = %x\n",
+			rc, smbdirect_mr->sgt.nents);
 		goto map_mr_error;
 	}
 
@@ -2346,8 +2317,8 @@ struct smbd_mr *smbd_register_mr(struct smbd_connection *info,
 
 	/* If all failed, attempt to recover this MR by setting it MR_ERROR*/
 map_mr_error:
-	ib_dma_unmap_sg(info->id->device, smbdirect_mr->sgl,
-		smbdirect_mr->sgl_count, smbdirect_mr->dir);
+	ib_dma_unmap_sg(info->id->device, smbdirect_mr->sgt.sgl,
+			smbdirect_mr->sgt.nents, smbdirect_mr->dir);
 
 dma_map_error:
 	smbdirect_mr->state = MR_ERROR;
@@ -2414,8 +2385,8 @@ int smbd_deregister_mr(struct smbd_mr *smbdirect_mr)
 
 	if (smbdirect_mr->state == MR_INVALIDATED) {
 		ib_dma_unmap_sg(
-			info->id->device, smbdirect_mr->sgl,
-			smbdirect_mr->sgl_count,
+			info->id->device, smbdirect_mr->sgt.sgl,
+			smbdirect_mr->sgt.nents,
 			smbdirect_mr->dir);
 		smbdirect_mr->state = MR_READY;
 		if (atomic_inc_return(&info->mr_ready_count) == 1)
@@ -2607,13 +2578,6 @@ static ssize_t smb_extract_iter_to_rdma(struct iov_iter *iter, size_t len,
 	ssize_t ret;
 	int before = rdma->nr_sge;
 
-	if (iov_iter_is_discard(iter) ||
-	    iov_iter_is_pipe(iter) ||
-	    user_backed_iter(iter)) {
-		WARN_ON_ONCE(1);
-		return -EIO;
-	}
-
 	switch (iov_iter_type(iter)) {
 	case ITER_BVEC:
 		ret = smb_extract_bvec_to_rdma(iter, rdma, len);
@@ -2625,7 +2589,8 @@ static ssize_t smb_extract_iter_to_rdma(struct iov_iter *iter, size_t len,
 		ret = smb_extract_xarray_to_rdma(iter, rdma, len);
 		break;
 	default:
-		BUG();
+		WARN_ON_ONCE(1);
+		return -EIO;
 	}
 
 	if (ret > 0) {
fs/cifs/smbdirect.h +1 -2
index be2cf18b7fec..83f239f376f0 100644
@@ -288,8 +288,7 @@ struct smbd_mr {
 	struct list_head	list;
 	enum mr_state		state;
 	struct ib_mr		*mr;
-	struct scatterlist	*sgl;
-	int			sgl_count;
+	struct sg_table		sgt;
 	enum dma_data_direction	dir;
 	union {
 		struct ib_reg_wr	wr;
author maintainer
#15 lore raw

PATCH 10/12 cifs: Remove unused code

In reply to David Howells

Remove a bunch of functions that are no longer used and are commented out after the conversion to use iterators throughout the I/O path.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Jeff Layton <[email protected]>
---
 fs/cifs/file.c | 606 -------------------------------------------------
 1 file changed, 606 deletions(-)
fs/cifs/file.c +0 -606
index a3d87feb4476..ed84104b669d 100644
@@ -2603,314 +2603,6 @@ static int cifs_partialpagewrite(struct page *page, unsigned from, unsigned to)
 	return rc;
 }
 
-#if 0 // TODO: Remove for iov_iter support
-static struct cifs_writedata *
-wdata_alloc_and_fillpages(pgoff_t tofind, struct address_space *mapping,
-			  pgoff_t end, pgoff_t *index,
-			  unsigned int *found_pages)
-{
-	struct cifs_writedata *wdata;
-
-	wdata = cifs_writedata_alloc((unsigned int)tofind,
-				     cifs_writev_complete);
-	if (!wdata)
-		return NULL;
-
-	*found_pages = find_get_pages_range_tag(mapping, index, end,
-				PAGECACHE_TAG_DIRTY, tofind, wdata->pages);
-	return wdata;
-}
-
-static unsigned int
-wdata_prepare_pages(struct cifs_writedata *wdata, unsigned int found_pages,
-		    struct address_space *mapping,
-		    struct writeback_control *wbc,
-		    pgoff_t end, pgoff_t *index, pgoff_t *next, bool *done)
-{
-	unsigned int nr_pages = 0, i;
-	struct page *page;
-
-	for (i = 0; i < found_pages; i++) {
-		page = wdata->pages[i];
-		/*
-		 * At this point we hold neither the i_pages lock nor the
-		 * page lock: the page may be truncated or invalidated
-		 * (changing page->mapping to NULL), or even swizzled
-		 * back from swapper_space to tmpfs file mapping
-		 */
-
-		if (nr_pages == 0)
-			lock_page(page);
-		else if (!trylock_page(page))
-			break;
-
-		if (unlikely(page->mapping != mapping)) {
-			unlock_page(page);
-			break;
-		}
-
-		if (!wbc->range_cyclic && page->index > end) {
-			*done = true;
-			unlock_page(page);
-			break;
-		}
-
-		if (*next && (page->index != *next)) {
-			/* Not next consecutive page */
-			unlock_page(page);
-			break;
-		}
-
-		if (wbc->sync_mode != WB_SYNC_NONE)
-			wait_on_page_writeback(page);
-
-		if (PageWriteback(page) ||
-				!clear_page_dirty_for_io(page)) {
-			unlock_page(page);
-			break;
-		}
-
-		/*
-		 * This actually clears the dirty bit in the radix tree.
-		 * See cifs_writepage() for more commentary.
-		 */
-		set_page_writeback(page);
-		if (page_offset(page) >= i_size_read(mapping->host)) {
-			*done = true;
-			unlock_page(page);
-			end_page_writeback(page);
-			break;
-		}
-
-		wdata->pages[i] = page;
-		*next = page->index + 1;
-		++nr_pages;
-	}
-
-	/* reset index to refind any pages skipped */
-	if (nr_pages == 0)
-		*index = wdata->pages[0]->index + 1;
-
-	/* put any pages we aren't going to use */
-	for (i = nr_pages; i < found_pages; i++) {
-		put_page(wdata->pages[i]);
-		wdata->pages[i] = NULL;
-	}
-
-	return nr_pages;
-}
-
-static int
-wdata_send_pages(struct cifs_writedata *wdata, unsigned int nr_pages,
-		 struct address_space *mapping, struct writeback_control *wbc)
-{
-	int rc;
-
-	wdata->sync_mode = wbc->sync_mode;
-	wdata->nr_pages = nr_pages;
-	wdata->offset = page_offset(wdata->pages[0]);
-	wdata->pagesz = PAGE_SIZE;
-	wdata->tailsz = min(i_size_read(mapping->host) -
-			page_offset(wdata->pages[nr_pages - 1]),
-			(loff_t)PAGE_SIZE);
-	wdata->bytes = ((nr_pages - 1) * PAGE_SIZE) + wdata->tailsz;
-	wdata->pid = wdata->cfile->pid;
-
-	rc = adjust_credits(wdata->server, &wdata->credits, wdata->bytes);
-	if (rc)
-		return rc;
-
-	if (wdata->cfile->invalidHandle)
-		rc = -EAGAIN;
-	else
-		rc = wdata->server->ops->async_writev(wdata,
-						      cifs_writedata_release);
-
-	return rc;
-}
-
-static int
-cifs_writepage_locked(struct page *page, struct writeback_control *wbc);
-
-static int cifs_write_one_page(struct page *page, struct writeback_control *wbc,
-		void *data)
-{
-	struct address_space *mapping = data;
-	int ret;
-
-	ret = cifs_writepage_locked(page, wbc);
-	unlock_page(page);
-	mapping_set_error(mapping, ret);
-	return ret;
-}
-
-static int cifs_writepages(struct address_space *mapping,
-			   struct writeback_control *wbc)
-{
-	struct inode *inode = mapping->host;
-	struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb);
-	struct TCP_Server_Info *server;
-	bool done = false, scanned = false, range_whole = false;
-	pgoff_t end, index;
-	struct cifs_writedata *wdata;
-	struct cifsFileInfo *cfile = NULL;
-	int rc = 0;
-	int saved_rc = 0;
-	unsigned int xid;
-
-	/*
-	 * If wsize is smaller than the page cache size, default to writing
-	 * one page at a time.
-	 */
-	if (cifs_sb->ctx->wsize < PAGE_SIZE)
-		return write_cache_pages(mapping, wbc, cifs_write_one_page,
-				mapping);
-
-	xid = get_xid();
-	if (wbc->range_cyclic) {
-		index = mapping->writeback_index; /* Start from prev offset */
-		end = -1;
-	} else {
-		index = wbc->range_start >> PAGE_SHIFT;
-		end = wbc->range_end >> PAGE_SHIFT;
-		if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX)
-			range_whole = true;
-		scanned = true;
-	}
-	server = cifs_pick_channel(cifs_sb_master_tcon(cifs_sb)->ses);
-
-retry:
-	while (!done && index <= end) {
-		unsigned int i, nr_pages, found_pages, wsize;
-		pgoff_t next = 0, tofind, saved_index = index;
-		struct cifs_credits credits_on_stack;
-		struct cifs_credits *credits = &credits_on_stack;
-		int get_file_rc = 0;
-
-		if (cfile)
-			cifsFileInfo_put(cfile);
-
-		rc = cifs_get_writable_file(CIFS_I(inode), FIND_WR_ANY, &cfile);
-
-		/* in case of an error store it to return later */
-		if (rc)
-			get_file_rc = rc;
-
-		rc = server->ops->wait_mtu_credits(server, cifs_sb->ctx->wsize,
-						   &wsize, credits);
-		if (rc != 0) {
-			done = true;
-			break;
-		}
-
-		tofind = min((wsize / PAGE_SIZE) - 1, end - index) + 1;
-
-		wdata = wdata_alloc_and_fillpages(tofind, mapping, end, &index,
-						  &found_pages);
-		if (!wdata) {
-			rc = -ENOMEM;
-			done = true;
-			add_credits_and_wake_if(server, credits, 0);
-			break;
-		}
-
-		if (found_pages == 0) {
-			kref_put(&wdata->refcount, cifs_writedata_release);
-			add_credits_and_wake_if(server, credits, 0);
-			break;
-		}
-
-		nr_pages = wdata_prepare_pages(wdata, found_pages, mapping, wbc,
-					       end, &index, &next, &done);
-
-		/* nothing to write? */
-		if (nr_pages == 0) {
-			kref_put(&wdata->refcount, cifs_writedata_release);
-			add_credits_and_wake_if(server, credits, 0);
-			continue;
-		}
-
-		wdata->credits = credits_on_stack;
-		wdata->cfile = cfile;
-		wdata->server = server;
-		cfile = NULL;
-
-		if (!wdata->cfile) {
-			cifs_dbg(VFS, "No writable handle in writepages rc=%d\n",
-				 get_file_rc);
-			if (is_retryable_error(get_file_rc))
-				rc = get_file_rc;
-			else
-				rc = -EBADF;
-		} else
-			rc = wdata_send_pages(wdata, nr_pages, mapping, wbc);
-
-		for (i = 0; i < nr_pages; ++i)
-			unlock_page(wdata->pages[i]);
-
-		/* send failure -- clean up the mess */
-		if (rc != 0) {
-			add_credits_and_wake_if(server, &wdata->credits, 0);
-			for (i = 0; i < nr_pages; ++i) {
-				if (is_retryable_error(rc))
-					redirty_page_for_writepage(wbc,
-							   wdata->pages[i]);
-				else
-					SetPageError(wdata->pages[i]);
-				end_page_writeback(wdata->pages[i]);
-				put_page(wdata->pages[i]);
-			}
-			if (!is_retryable_error(rc))
-				mapping_set_error(mapping, rc);
-		}
-		kref_put(&wdata->refcount, cifs_writedata_release);
-
-		if (wbc->sync_mode == WB_SYNC_ALL && rc == -EAGAIN) {
-			index = saved_index;
-			continue;
-		}
-
-		/* Return immediately if we received a signal during writing */
-		if (is_interrupt_error(rc)) {
-			done = true;
-			break;
-		}
-
-		if (rc != 0 && saved_rc == 0)
-			saved_rc = rc;
-
-		wbc->nr_to_write -= nr_pages;
-		if (wbc->nr_to_write <= 0)
-			done = true;
-
-		index = next;
-	}
-
-	if (!scanned && !done) {
-		/*
-		 * We hit the last page and there is more work to be done: wrap
-		 * back to the start of the file
-		 */
-		scanned = true;
-		index = 0;
-		goto retry;
-	}
-
-	if (saved_rc != 0)
-		rc = saved_rc;
-
-	if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0))
-		mapping->writeback_index = index;
-
-	if (cfile)
-		cifsFileInfo_put(cfile);
-	free_xid(xid);
-	/* Indication to update ctime and mtime as close is deferred */
-	set_bit(CIFS_INO_MODIFIED_ATTR, &CIFS_I(inode)->flags);
-	return rc;
-}
-#endif
-
 /*
  * Extend the region to be written back to include subsequent contiguously
  * dirty pages if possible, but don't sleep while doing so.
@@ -3506,49 +3198,6 @@ int cifs_flush(struct file *file, fl_owner_t id)
 	return rc;
 }
 
-#if 0 // TODO: Remove for iov_iter support
-static int
-cifs_write_allocate_pages(struct page **pages, unsigned long num_pages)
-{
-	int rc = 0;
-	unsigned long i;
-
-	for (i = 0; i < num_pages; i++) {
-		pages[i] = alloc_page(GFP_KERNEL|__GFP_HIGHMEM);
-		if (!pages[i]) {
-			/*
-			 * save number of pages we have already allocated and
-			 * return with ENOMEM error
-			 */
-			num_pages = i;
-			rc = -ENOMEM;
-			break;
-		}
-	}
-
-	if (rc) {
-		for (i = 0; i < num_pages; i++)
-			put_page(pages[i]);
-	}
-	return rc;
-}
-
-static inline
-size_t get_numpages(const size_t wsize, const size_t len, size_t *cur_len)
-{
-	size_t num_pages;
-	size_t clen;
-
-	clen = min_t(const size_t, len, wsize);
-	num_pages = DIV_ROUND_UP(clen, PAGE_SIZE);
-
-	if (cur_len)
-		*cur_len = clen;
-
-	return num_pages;
-}
-#endif
-
 static void
 cifs_uncached_writedata_release(struct kref *refcount)
 {
@@ -3581,50 +3230,6 @@ cifs_uncached_writev_complete(struct work_struct *work)
 	kref_put(&wdata->refcount, cifs_uncached_writedata_release);
 }
 
-#if 0 // TODO: Remove for iov_iter support
-static int
-wdata_fill_from_iovec(struct cifs_writedata *wdata, struct iov_iter *from,
-		      size_t *len, unsigned long *num_pages)
-{
-	size_t save_len, copied, bytes, cur_len = *len;
-	unsigned long i, nr_pages = *num_pages;
-
-	save_len = cur_len;
-	for (i = 0; i < nr_pages; i++) {
-		bytes = min_t(const size_t, cur_len, PAGE_SIZE);
-		copied = copy_page_from_iter(wdata->pages[i], 0, bytes, from);
-		cur_len -= copied;
-		/*
-		 * If we didn't copy as much as we expected, then that
-		 * may mean we trod into an unmapped area. Stop copying
-		 * at that point. On the next pass through the big
-		 * loop, we'll likely end up getting a zero-length
-		 * write and bailing out of it.
-		 */
-		if (copied < bytes)
-			break;
-	}
-	cur_len = save_len - cur_len;
-	*len = cur_len;
-
-	/*
-	 * If we have no data to send, then that probably means that
-	 * the copy above failed altogether. That's most likely because
-	 * the address in the iovec was bogus. Return -EFAULT and let
-	 * the caller free anything we allocated and bail out.
-	 */
-	if (!cur_len)
-		return -EFAULT;
-
-	/*
-	 * i + 1 now represents the number of pages we actually used in
-	 * the copy phase above.
-	 */
-	*num_pages = i + 1;
-	return 0;
-}
-#endif
-
 static int
 cifs_resend_wdata(struct cifs_writedata *wdata, struct list_head *wdata_list,
 	struct cifs_aio_ctx *ctx)
@@ -4211,83 +3816,6 @@ cifs_uncached_readv_complete(struct work_struct *work)
 	kref_put(&rdata->refcount, cifs_readdata_release);
 }
 
-#if 0 // TODO: Remove for iov_iter support
-
-static int
-uncached_fill_pages(struct TCP_Server_Info *server,
-		    struct cifs_readdata *rdata, struct iov_iter *iter,
-		    unsigned int len)
-{
-	int result = 0;
-	unsigned int i;
-	unsigned int nr_pages = rdata->nr_pages;
-	unsigned int page_offset = rdata->page_offset;
-
-	rdata->got_bytes = 0;
-	rdata->tailsz = PAGE_SIZE;
-	for (i = 0; i < nr_pages; i++) {
-		struct page *page = rdata->pages[i];
-		size_t n;
-		unsigned int segment_size = rdata->pagesz;
-
-		if (i == 0)
-			segment_size -= page_offset;
-		else
-			page_offset = 0;
-
-
-		if (len <= 0) {
-			/* no need to hold page hostage */
-			rdata->pages[i] = NULL;
-			rdata->nr_pages--;
-			put_page(page);
-			continue;
-		}
-
-		n = len;
-		if (len >= segment_size)
-			/* enough data to fill the page */
-			n = segment_size;
-		else
-			rdata->tailsz = len;
-		len -= n;
-
-		if (iter)
-			result = copy_page_from_iter(
-					page, page_offset, n, iter);
-#ifdef CONFIG_CIFS_SMB_DIRECT
-		else if (rdata->mr)
-			result = n;
-#endif
-		else
-			result = cifs_read_page_from_socket(
-					server, page, page_offset, n);
-		if (result < 0)
-			break;
-
-		rdata->got_bytes += result;
-	}
-
-	return rdata->got_bytes > 0 && result != -ECONNABORTED ?
-						rdata->got_bytes : result;
-}
-
-static int
-cifs_uncached_read_into_pages(struct TCP_Server_Info *server,
-			      struct cifs_readdata *rdata, unsigned int len)
-{
-	return uncached_fill_pages(server, rdata, NULL, len);
-}
-
-static int
-cifs_uncached_copy_into_pages(struct TCP_Server_Info *server,
-			      struct cifs_readdata *rdata,
-			      struct iov_iter *iter)
-{
-	return uncached_fill_pages(server, rdata, iter, iter->count);
-}
-#endif
-
 static int cifs_resend_rdata(struct cifs_readdata *rdata,
 			struct list_head *rdata_list,
 			struct cifs_aio_ctx *ctx)
@@ -4898,140 +4426,6 @@ int cifs_file_mmap(struct file *file, struct vm_area_struct *vma)
 	return rc;
 }
 
-#if 0 // TODO: Remove for iov_iter support
-
-static void
-cifs_readv_complete(struct work_struct *work)
-{
-	unsigned int i, got_bytes;
-	struct cifs_readdata *rdata = container_of(work,
-						struct cifs_readdata, work);
-
-	got_bytes = rdata->got_bytes;
-	for (i = 0; i < rdata->nr_pages; i++) {
-		struct page *page = rdata->pages[i];
-
-		if (rdata->result == 0 ||
-		    (rdata->result == -EAGAIN && got_bytes)) {
-			flush_dcache_page(page);
-			SetPageUptodate(page);
-		} else
-			SetPageError(page);
-
-		if (rdata->result == 0 ||
-		    (rdata->result == -EAGAIN && got_bytes))
-			cifs_readpage_to_fscache(rdata->mapping->host, page);
-
-		unlock_page(page);
-
-		got_bytes -= min_t(unsigned int, PAGE_SIZE, got_bytes);
-
-		put_page(page);
-		rdata->pages[i] = NULL;
-	}
-	kref_put(&rdata->refcount, cifs_readdata_release);
-}
-
-static int
-readpages_fill_pages(struct TCP_Server_Info *server,
-		     struct cifs_readdata *rdata, struct iov_iter *iter,
-		     unsigned int len)
-{
-	int result = 0;
-	unsigned int i;
-	u64 eof;
-	pgoff_t eof_index;
-	unsigned int nr_pages = rdata->nr_pages;
-	unsigned int page_offset = rdata->page_offset;
-
-	/* determine the eof that the server (probably) has */
-	eof = CIFS_I(rdata->mapping->host)->server_eof;
-	eof_index = eof ? (eof - 1) >> PAGE_SHIFT : 0;
-	cifs_dbg(FYI, "eof=%llu eof_index=%lu\n", eof, eof_index);
-
-	rdata->got_bytes = 0;
-	rdata->tailsz = PAGE_SIZE;
-	for (i = 0; i < nr_pages; i++) {
-		struct page *page = rdata->pages[i];
-		unsigned int to_read = rdata->pagesz;
-		size_t n;
-
-		if (i == 0)
-			to_read -= page_offset;
-		else
-			page_offset = 0;
-
-		n = to_read;
-
-		if (len >= to_read) {
-			len -= to_read;
-		} else if (len > 0) {
-			/* enough for partial page, fill and zero the rest */
-			zero_user(page, len + page_offset, to_read - len);
-			n = rdata->tailsz = len;
-			len = 0;
-		} else if (page->index > eof_index) {
-			/*
-			 * The VFS will not try to do readahead past the
-			 * i_size, but it's possible that we have outstanding
-			 * writes with gaps in the middle and the i_size hasn't
-			 * caught up yet. Populate those with zeroed out pages
-			 * to prevent the VFS from repeatedly attempting to
-			 * fill them until the writes are flushed.
-			 */
-			zero_user(page, 0, PAGE_SIZE);
-			flush_dcache_page(page);
-			SetPageUptodate(page);
-			unlock_page(page);
-			put_page(page);
-			rdata->pages[i] = NULL;
-			rdata->nr_pages--;
-			continue;
-		} else {
-			/* no need to hold page hostage */
-			unlock_page(page);
-			put_page(page);
-			rdata->pages[i] = NULL;
-			rdata->nr_pages--;
-			continue;
-		}
-
-		if (iter)
-			result = copy_page_from_iter(
-					page, page_offset, n, iter);
-#ifdef CONFIG_CIFS_SMB_DIRECT
-		else if (rdata->mr)
-			result = n;
-#endif
-		else
-			result = cifs_read_page_from_socket(
-					server, page, page_offset, n);
-		if (result < 0)
-			break;
-
-		rdata->got_bytes += result;
-	}
-
-	return rdata->got_bytes > 0 && result != -ECONNABORTED ?
-						rdata->got_bytes : result;
-}
-
-static int
-cifs_readpages_read_into_pages(struct TCP_Server_Info *server,
-			       struct cifs_readdata *rdata, unsigned int len)
-{
-	return readpages_fill_pages(server, rdata, NULL, len);
-}
-
-static int
-cifs_readpages_copy_into_pages(struct TCP_Server_Info *server,
-			       struct cifs_readdata *rdata,
-			       struct iov_iter *iter)
-{
-	return readpages_fill_pages(server, rdata, iter, iter->count);
-}
-#endif
-
 /*
  * Unlock a bunch of folios in the pagecache.
  */
author maintainer
#16 lore raw

PATCH 11/12 cifs: Fix problem with encrypted RDMA data read

In reply to David Howells

When the cifs client is talking to the ksmbd server by RDMA and the ksmbd server has "smb3 encryption = yes" in its config file, the normal PDU stream is encrypted, but the directly-delivered data isn't in the stream (and isn't encrypted), but is rather delivered by DDP/RDMA packets (at least with IWarp).

Currently, the direct delivery fails with:

buf can not contain only a part of read data

   WARNING: CPU: 0 PID: 4619 at fs/cifs/smb2ops.c:4731 handle_read_data+0x393/0x405
   ...
   RIP: 0010:handle_read_data+0x393/0x405
   ...
    smb3_handle_read_data+0x30/0x37
    receive_encrypted_standard+0x141/0x224
    cifs_demultiplex_thread+0x21a/0x63b
    kthread+0xe7/0xef
    ret_from_fork+0x22/0x30

The problem apparently stemming from the fact that it's trying to manage the decryption, but the data isn't in the smallbuf, the bigbuf or the page array).

This can be fixed simply by inserting an extra case into handle_read_data() that checks to see if use_rdma_mr is true, and if it is, just setting rdata->got_bytes to the length of data delivered and allowing normal continuation.

This can be seen in an IWarp packet trace. With the upstream code, it does a DDP/RDMA packet, which produces the warning above and then retries, retrieving the data inline, spread across several SMBDirect messages that get glued together into a single PDU. With the patch applied, only the DDP/RDMA packet is seen.

Note that this doesn't happen if the server isn't told to encrypt stuff and it does also happen with softRoCE.

cc Steve French <[email protected]>
cc Tom Talpey <[email protected]>
cc Long Li <[email protected]>
cc Namjae Jeon <[email protected]>
cc Stefan Metzmacher <[email protected]>
---
 fs/cifs/smb2ops.c | 3 +++
 1 file changed, 3 insertions(+)
fs/cifs/smb2ops.c +3 -0
index cea578a45ed8..73b66ac86abf 100644
@@ -4733,6 +4733,9 @@ handle_read_data(struct TCP_Server_Info *server, struct mid_q_entry *mid,
 		if (length < 0)
 			return length;
 		rdata->got_bytes = data_len;
+	} else if (use_rdma_mr) {
+		/* The data was delivered directly by RDMA. */
+		rdata->got_bytes = data_len;
 	} else {
 		/* read response payload cannot be in both buf and pages */
 		WARN_ONCE(1, "buf can not contain only a part of read data");
#17 lore raw

Re: [PATCH 11/12] cifs: Fix problem with encrypted RDMA data read

Quoted 5 lines · David Howells

When the cifs client is talking to the ksmbd server by RDMA and the ksmbd server has "smb3 encryption = yes" in its config file, the normal PDU stream is encrypted, but the directly-delivered data isn't in the stream (and isn't encrypted), but is rather delivered by DDP/RDMA packets (at least with IWarp).

This really needs to be split into a separate backportable fix series. And it seems like Stefan has just send such a series.

#18 lore raw

Re: [PATCH 11/12] cifs: Fix problem with encrypted RDMA data read

Am 01.02.23 um 14:36 schrieb Christoph Hellwig:

Quoted 9 lines · David Howells
Quoted 5 lines

When the cifs client is talking to the ksmbd server by RDMA and the ksmbd server has "smb3 encryption = yes" in its config file, the normal PDU stream is encrypted, but the directly-delivered data isn't in the stream (and isn't encrypted), but is rather delivered by DDP/RDMA packets (at least with IWarp).

This really needs to be split into a separate backportable fix series. And it seems like Stefan has just send such a series.

My changes are triggered by this commit...

Now I'm looking more closely as I didn't understand it...

#19 lore raw

Re: [PATCH 11/12] cifs: Fix problem with encrypted RDMA data read

In reply to David Howells

Am 31.01.23 um 19:28 schrieb David Howells:

Quoted 55 lines

When the cifs client is talking to the ksmbd server by RDMA and the ksmbd server has "smb3 encryption = yes" in its config file, the normal PDU stream is encrypted, but the directly-delivered data isn't in the stream (and isn't encrypted), but is rather delivered by DDP/RDMA packets (at least with IWarp).

Currently, the direct delivery fails with:

    buf can not contain only a part of read data
    WARNING: CPU: 0 PID: 4619 at fs/cifs/smb2ops.c:4731 handle_read_data+0x393/0x405
    ...
    RIP: 0010:handle_read_data+0x393/0x405
    ...
     smb3_handle_read_data+0x30/0x37
     receive_encrypted_standard+0x141/0x224
     cifs_demultiplex_thread+0x21a/0x63b
     kthread+0xe7/0xef
     ret_from_fork+0x22/0x30

The problem apparently stemming from the fact that it's trying to manage the decryption, but the data isn't in the smallbuf, the bigbuf or the page array).

This can be fixed simply by inserting an extra case into handle_read_data() that checks to see if use_rdma_mr is true, and if it is, just setting rdata->got_bytes to the length of data delivered and allowing normal continuation.

This can be seen in an IWarp packet trace. With the upstream code, it does a DDP/RDMA packet, which produces the warning above and then retries, retrieving the data inline, spread across several SMBDirect messages that get glued together into a single PDU. With the patch applied, only the DDP/RDMA packet is seen.

Note that this doesn't happen if the server isn't told to encrypt stuff and it does also happen with softRoCE.

cc Steve French <[email protected]>
cc Tom Talpey <[email protected]>
cc Long Li <[email protected]>
cc Namjae Jeon <[email protected]>
cc Stefan Metzmacher <[email protected]>
---
  fs/cifs/smb2ops.c | 3 +++
  1 file changed, 3 insertions(+)
fs/cifs/smb2ops.c +0 -0
index cea578a45ed8..73b66ac86abf 100644

fs/cifs/smb2ops.c handle_read_data(struct TCP_Server_Info *server, struct mid_q_entry *mid, · line 4738

Quoted 6 lines
fs/cifs/smb2ops.c +3 -0
@@ -4733,6 +4733,9 @@ handle_read_data(struct TCP_Server_Info *server, struct mid_q_entry *mid,
  		if (length < 0)
  			return length;
  		rdata->got_bytes = data_len;
+	} else if (use_rdma_mr) {
+		/* The data was delivered directly by RDMA. */
+		rdata->got_bytes = data_len;

I actually don't understand why this would only be a problem with encryption.

I guess there's much more needed and data_offset should most likely be ignored completely in the rdma offload case. So I'd guess its just luck that we don't trigger the below warning/error more often.

Quoted 4 lines
fs/cifs/smb2ops.c +0 -0
  	} else {
  		/* read response payload cannot be in both buf and pages */
  		WARN_ONCE(1, "buf can not contain only a part of read data");
#20 lore raw

Re: [PATCH 11/12] cifs: Fix problem with encrypted RDMA data read

Am 01.02.23 um 15:05 schrieb Stefan Metzmacher:

Quoted 65 lines

Am 31.01.23 um 19:28 schrieb David Howells:

Quoted 55 lines

When the cifs client is talking to the ksmbd server by RDMA and the ksmbd server has "smb3 encryption = yes" in its config file, the normal PDU stream is encrypted, but the directly-delivered data isn't in the stream (and isn't encrypted), but is rather delivered by DDP/RDMA packets (at least with IWarp).

Currently, the direct delivery fails with:

    buf can not contain only a part of read data

    WARNING: CPU: 0 PID: 4619 at fs/cifs/smb2ops.c:4731 handle_read_data+0x393/0x405
    ...
    RIP: 0010:handle_read_data+0x393/0x405
    ...
     smb3_handle_read_data+0x30/0x37
     receive_encrypted_standard+0x141/0x224
     cifs_demultiplex_thread+0x21a/0x63b
     kthread+0xe7/0xef
     ret_from_fork+0x22/0x30

The problem apparently stemming from the fact that it's trying to manage the decryption, but the data isn't in the smallbuf, the bigbuf or the page array).

This can be fixed simply by inserting an extra case into handle_read_data() that checks to see if use_rdma_mr is true, and if it is, just setting rdata->got_bytes to the length of data delivered and allowing normal continuation.

This can be seen in an IWarp packet trace.  With the upstream code, it does a DDP/RDMA packet, which produces the warning above and then retries, retrieving the data inline, spread across several SMBDirect messages that get glued together into a single PDU.  With the patch applied, only the DDP/RDMA packet is seen.

Note that this doesn't happen if the server isn't told to encrypt stuff and it does also happen with softRoCE.

cc Steve French <[email protected]>
cc Tom Talpey <[email protected]>
cc Long Li <[email protected]>
cc Namjae Jeon <[email protected]>
cc Stefan Metzmacher <[email protected]>

---
  fs/cifs/smb2ops.c | 3 +++

  1 file changed, 3 insertions(+)
fs/cifs/smb2ops.c +0 -0
index cea578a45ed8..73b66ac86abf 100644

fs/cifs/smb2ops.c handle_read_data(struct TCP_Server_Info *server, struct mid_q_entry *mid, · line 4738

Quoted 6 lines
fs/cifs/smb2ops.c +3 -0
@@ -4733,6 +4733,9 @@ handle_read_data(struct TCP_Server_Info *server, struct mid_q_entry *mid,
           if (length < 0)
               return length;
           rdata->got_bytes = data_len;
+    } else if (use_rdma_mr) {
+        /* The data was delivered directly by RDMA. */
+        rdata->got_bytes = data_len;

I actually don't understand why this would only be a problem with encryption.

Quoted 3 lines

I guess there's much more needed and data_offset should most likely be ignored completely in the rdma offload case. So I'd guess its just luck that we don't trigger the below warning/error more often.

I guess it might be related to smb3_handle_read_data passing server->pdu_size to handle_read_data(), while server->pdu_size is the outer size of the transform and not the size of the decrypted pdu.

Maybe receive_encrypted_standard() needs to reset server->pdu_size during this:

        if (mid_entry && mid_entry->handle)
                ret = mid_entry->handle(server, mid_entry);
        else
                ret = cifs_handle_standard(server, mid_entry);

But this code is so overly complex, that's way to hard to understand...

author maintainer
#21 lore raw

PATCH 12/12 cifs: DIO to/from KVEC-type iterators should now work

In reply to David Howells

DIO to/from KVEC-type iterators should now work as the iterator is passed down to the socket in non-RDMA/non-crypto mode and in RDMA or crypto mode care is taken to handle vmap/vmalloc correctly and not take page refs when building a scatterlist.

cc Steve French <[email protected]>
cc Shyam Prasad N <[email protected]>
cc Rohith Surabattula <[email protected]>
cc Tom Talpey <[email protected]>
cc Jeff Layton <[email protected]>
---
 fs/cifs/file.c | 20 --------------------
 1 file changed, 20 deletions(-)
fs/cifs/file.c +0 -20
index ed84104b669d..11875224a917 100644
@@ -3546,16 +3546,6 @@ static ssize_t __cifs_writev(
 	struct cifs_aio_ctx *ctx;
 	int rc;
 
-	/*
-	 * iov_iter_get_pages_alloc doesn't work with ITER_KVEC.
-	 * In this case, fall back to non-direct write function.
-	 * this could be improved by getting pages directly in ITER_KVEC
-	 */
-	if (direct && iov_iter_is_kvec(from)) {
-		cifs_dbg(FYI, "use non-direct cifs_writev for kvec I/O\n");
-		direct = false;
-	}
-
 	rc = generic_write_checks(iocb, from);
 	if (rc <= 0)
 		return rc;
@@ -4089,16 +4079,6 @@ static ssize_t __cifs_readv(
 	loff_t offset = iocb->ki_pos;
 	struct cifs_aio_ctx *ctx;
 
-	/*
-	 * iov_iter_get_pages_alloc() doesn't work with ITER_KVEC,
-	 * fall back to data copy read path
-	 * this could be improved by getting pages directly in ITER_KVEC
-	 */
-	if (direct && iov_iter_is_kvec(to)) {
-		cifs_dbg(FYI, "use non-direct cifs_user_readv for kvec I/O\n");
-		direct = false;
-	}
-
 	len = iov_iter_count(to);
 	if (!len)
 		return 0;