linux.git/fs/netfs, branch v6.3

netfs: Fix netfs_extract_iter_to_sg() for ITER_UBUF/IOVEC

2023-04-12T16:26:36+00:00

Fix netfs_extract_iter_to_sg() for ITER_UBUF and ITER_IOVEC to set the
size of the page to the part of the page extracted, not the remaining
amount of data in the extracted page array at that point.

This doesn't yet affect anything as cifs, the only current user, only
passes in non-user-backed iterators.

Fixes: 018584697533 ("netfs: Add a function to extract an iterator into a scatterlist")
Signed-off-by: David Howells 
Reviewed-by: Jeff Layton 
Cc: Steve French 
Cc: Shyam Prasad N 
Cc: Rohith Surabattula 
Signed-off-by: Linus Torvalds

iov: Fix netfs_extract_user_to_sg()

2023-03-02T00:18:25+00:00

Fix the loop check in netfs_extract_user_to_sg() for extraction from
user-backed iterators to do the body if npages > 0, not if npages < 0
(which it can never be).

This isn't currently used by cifs, which only ever extracts data from BVEC,
KVEC and XARRAY iterators at this level, user-backed iterators having being
decanted into BVEC iterators at a higher level to accommodate the work
being done in a kernel thread.

Found by smatch:
	fs/netfs/iterator.c:139 netfs_extract_user_to_sg() warn: unsigned 'npages' is never less than zero.

Fixes: 018584697533 ("netfs: Add a function to extract an iterator into a scatterlist")
Reported-by: kernel test robot 
Link: https://lore.kernel.org/oe-kbuild-all/202302261115.P3TQi1ZO-lkp@intel.com/
Reported-by: Dan Carpenter 
Link: https://lore.kernel.org/r/Y/yYnAhoAYDBKixX@kili
Reviewed-by: Paulo Alcantara (SUSE) 
Signed-off-by: David Howells 
cc: Jeff Layton 
cc: linux-cifs@vger.kernel.org
cc: linux-cachefs@redhat.com
Signed-off-by: Steve French

Merge tag '6.3-rc-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6

2023-02-23T01:12:44+00:00

Pull cifs client updates from Steve French:
 "The largest subset of this is from David Howells et al: making the
  cifs/smb3 driver pass iov_iters down to the lowest layers, directly to
  the network transport rather than passing lists of pages around,
  helping multiple areas:

   - Pin user pages, thereby fixing the race between concurrent DIO read
     and fork, where 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.

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

   - Allowing cifs to use the new splice function

  The remainder are:

   - fixes for stable, including various fixes for uninitialized memory,
     wrong length field causing mount issue to very old servers,
     important directory lease fixes and reconnect fixes

   - cleanups (unused code removal, change one element array usage, and
     a change form strtobool to kstrtobool, and Kconfig cleanups)

   - SMBDIRECT (RDMA) fixes including iov_iter integration and UAF fixes

   - reconnect fixes

   - multichannel fixes, including improving channel allocation (to
     least used channel)

   - remove the last use of lock_page_killable by moving to
     folio_lock_killable"

* tag '6.3-rc-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6: (46 commits)
  update internal module version number for cifs.ko
  cifs: update ip_addr for ses only for primary chan setup
  cifs: use tcon allocation functions even for dummy tcon
  cifs: use the least loaded channel for sending requests
  cifs: DIO to/from KVEC-type iterators should now work
  cifs: Remove unused code
  cifs: Build the RDMA SGE list directly from an iterator
  cifs: Change the I/O paths to use an iterator rather than a page list
  cifs: Add a function to read into an iter from a socket
  cifs: Add some helper functions
  cifs: Add a function to Hash the contents of an iterator
  cifs: Add a function to build an RDMA SGE list from an iterator
  netfs: Add a function to extract an iterator into a scatterlist
  netfs: Add a function to extract a UBUF or IOVEC into a BVEC iterator
  cifs: Implement splice_read to pass down ITER_BVEC not ITER_PIPE
  splice: Export filemap/direct_splice_read()
  iov_iter: Add a function to extract a page list from an iterator
  iov_iter: Define flags to qualify page extraction.
  splice: Add a func to do a splice from an O_DIRECT file without ITER_PIPE
  splice: Add a func to do a splice from a buffered file without ITER_PIPE
  ...

netfs: Add a function to extract an iterator into a scatterlist

2023-02-20T23:25:43+00:00

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.

Signed-off-by: David Howells 
cc: Jeff Layton 
cc: Steve French 
cc: Shyam Prasad N 
cc: Rohith Surabattula 
cc: linux-cachefs@redhat.com
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French

netfs: Add a function to extract a UBUF or IOVEC into a BVEC iterator

2023-02-20T23:25:43+00:00

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).

Signed-off-by: David Howells 
cc: Jeff Layton 
cc: Steve French 
cc: Shyam Prasad N 
cc: Rohith Surabattula 
cc: linux-cachefs@redhat.com
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French

use less confusing names for iov_iter direction initializers

2022-11-25T18:01:55+00:00

READ/WRITE proved to be actively confusing - the meanings are
"data destination, as used with read(2)" and "data source, as
used with write(2)", but people keep interpreting those as
"we read data from it" and "we write data to it", i.e. exactly
the wrong way.

Call them ITER_DEST and ITER_SOURCE - at least that is harder
to misinterpret...

Signed-off-by: Al Viro

netfs: Fix dodgy maths

2022-11-15T16:56:07+00:00

Fix the dodgy maths in netfs_rreq_unlock_folios().  start_page could be
inside the folio, in which case the calculation of pgpos will be come up
with a negative number (though for the moment rreq->start is rounded down
earlier and folios would have to get merged whilst locked)

Alter how this works to just frame the tracking in terms of absolute file
positions, rather than offsets from the start of the I/O request.  This
simplifies the maths and makes it easier to follow.

Fix the issue by using folio_pos() and folio_size() to calculate the end
position of the page.

Fixes: 3d3c95046742 ("netfs: Provide readahead and readpage netfs helpers")
Reported-by: Matthew Wilcox 
Signed-off-by: David Howells 
Reviewed-by: Jeff Layton 
Reviewed-by: Jingbo Xu 
cc: linux-cachefs@redhat.com
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/Y2SJw7w1IsIik3nb@casper.infradead.org/
Link: https://lore.kernel.org/r/166757988611.950645.7626959069846893164.stgit@warthog.procyon.org.uk/ # v2

netfs: Fix missing xas_retry() calls in xarray iteration

2022-11-15T16:49:06+00:00

netfslib has a number of places in which it performs iteration of an xarray
whilst being under the RCU read lock.  It *should* call xas_retry() as the
first thing inside of the loop and do "continue" if it returns true in case
the xarray walker passed out a special value indicating that the walk needs
to be redone from the root[*].

Fix this by adding the missing retry checks.

[*] I wonder if this should be done inside xas_find(), xas_next_node() and
    suchlike, but I'm told that's not an simple change to effect.

This can cause an oops like that below.  Note the faulting address - this
is an internal value (|0x2) returned from xarray.

BUG: kernel NULL pointer dereference, address: 0000000000000402
...
RIP: 0010:netfs_rreq_unlock+0xef/0x380 [netfs]
...
Call Trace:
 netfs_rreq_assess+0xa6/0x240 [netfs]
 netfs_readpage+0x173/0x3b0 [netfs]
 ? init_wait_var_entry+0x50/0x50
 filemap_read_page+0x33/0xf0
 filemap_get_pages+0x2f2/0x3f0
 filemap_read+0xaa/0x320
 ? do_filp_open+0xb2/0x150
 ? rmqueue+0x3be/0xe10
 ceph_read_iter+0x1fe/0x680 [ceph]
 ? new_sync_read+0x115/0x1a0
 new_sync_read+0x115/0x1a0
 vfs_read+0xf3/0x180
 ksys_read+0x5f/0xe0
 do_syscall_64+0x38/0x90
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Changes:
========
ver #2)
 - Changed an unsigned int to a size_t to reduce the likelihood of an
   overflow as per Willy's suggestion.
 - Added an additional patch to fix the maths.

Fixes: 3d3c95046742 ("netfs: Provide readahead and readpage netfs helpers")
Reported-by: George Law 
Signed-off-by: David Howells 
Reviewed-by: Jeff Layton 
Reviewed-by: Jingbo Xu 
cc: Matthew Wilcox 
cc: linux-cachefs@redhat.com
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/166749229733.107206.17482609105741691452.stgit@warthog.procyon.org.uk/ # v1
Link: https://lore.kernel.org/r/166757987929.950645.12595273010425381286.stgit@warthog.procyon.org.uk/ # v2

netfs: do not unlock and put the folio twice

2022-07-14T08:10:12+00:00

check_write_begin() will unlock and put the folio when return
non-zero.  So we should avoid unlocking and putting it twice in
netfs layer.

Change the way ->check_write_begin() works in the following two ways:

 (1) Pass it a pointer to the folio pointer, allowing it to unlock and put
     the folio prior to doing the stuff it wants to do, provided it clears
     the folio pointer.

 (2) Change the return values such that 0 with folio pointer set means
     continue, 0 with folio pointer cleared means re-get and all error
     codes indicating an error (no special treatment for -EAGAIN).

[ bagasdotme: use Sphinx code text syntax for *foliop pointer ]

Cc: stable@vger.kernel.org
Link: https://tracker.ceph.com/issues/56423
Link: https://lore.kernel.org/r/cf169f43-8ee7-8697-25da-0204d1b4343e@redhat.com
Co-developed-by: David Howells 
Signed-off-by: Xiubo Li 
Signed-off-by: David Howells 
Signed-off-by: Bagas Sanjaya 
Signed-off-by: Ilya Dryomov

netfs: Rename the netfs_io_request cleanup op and give it an op pointer

2022-06-10T19:55:21+00:00

The netfs_io_request cleanup op is now always in a position to be given a
pointer to a netfs_io_request struct, so this can be passed in instead of
the mapping and private data arguments (both of which are included in the
struct).

So rename the ->cleanup op to ->free_request (to match ->init_request) and
pass in the I/O pointer.

Signed-off-by: David Howells 
Reviewed-by: Jeff Layton 
cc: linux-cachefs@redhat.com