block: Add BIO_PAGE_PINNED and associated infrastructure

Add BIO_PAGE_PINNED to indicate that the pages in a bio are pinned
(FOLL_PIN) and that the pin will need removing.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Jan Kara <jack@suse.cz>
cc: Matthew Wilcox <willy@infradead.org>
cc: Logan Gunthorpe <logang@deltatee.com>
cc: linux-block@vger.kernel.org
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230522205744.2825689-5-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Replace BIO_NO_PAGE_REF with BIO_PAGE_REFFED with inverted logic

Replace BIO_NO_PAGE_REF with a BIO_PAGE_REFFED flag that has the inverted
meaning is only set when a page reference has been acquired that needs to
be released by bio_release_pages().

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Jan Kara <jack@suse.cz>
cc: Matthew Wilcox <willy@infradead.org>
cc: Logan Gunthorpe <logang@deltatee.com>
cc: linux-block@vger.kernel.org
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230522205744.2825689-4-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Fix bio_flagged() so that gcc can better optimise it

Fix bio_flagged() so that multiple instances of it, such as:

if (bio_flagged(bio, BIO_PAGE_REFFED) ||
    bio_flagged(bio, BIO_PAGE_PINNED))

can be combined by the gcc optimiser into a single test in assembly
(arguably, this is a compiler optimisation issue[1]).

The missed optimisation stems from bio_flagged() comparing the result of
the bitwise-AND to zero.  This results in an out-of-line bio_release_page()
being compiled to something like:

   <+0>:     mov    0x14(%rdi),%eax
   <+3>:     test   $0x1,%al
   <+5>:     jne    0xffffffff816dac53 <bio_release_pages+11>
   <+7>:     test   $0x2,%al
   <+9>:     je     0xffffffff816dac5c <bio_release_pages+20>
   <+11>:    movzbl %sil,%esi
   <+15>:    jmp    0xffffffff816daba1 <__bio_release_pages>
   <+20>:    jmp    0xffffffff81d0b800 <__x86_return_thunk>

However, the test is superfluous as the return type is bool.  Removing it
results in:

   <+0>:     testb  $0x3,0x14(%rdi)
   <+4>:     je     0xffffffff816e4af4 <bio_release_pages+15>
   <+6>:     movzbl %sil,%esi
   <+10>:    jmp    0xffffffff816dab7c <__bio_release_pages>
   <+15>:    jmp    0xffffffff81d0b7c0 <__x86_return_thunk>

instead.

Also, the MOVZBL instruction looks unnecessary[2] - I think it's just
're-booling' the mark_dirty parameter.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
cc: Jens Axboe <axboe@kernel.dk>
cc: linux-block@vger.kernel.org
Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108370
Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108371
Link: https://lore.kernel.org/r/167391056756.2311931.356007731815807265.stgit@warthog.procyon.org.uk/
Reviewed-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230522205744.2825689-3-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

iomap: Don't get an reference on ZERO_PAGE for direct I/O block zeroing

ZERO_PAGE can't go away, no need to hold an extra reference.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: linux-fsdevel@vger.kernel.org
Reviewed-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230522205744.2825689-2-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Merge branch 'for-6.5/splice' into for-6.5/block

Merge splice bits as subsequent block cleanups and improvements for DIO
depend on them.

* for-6.5/splice: (31 commits)
  splice: kdoc for filemap_splice_read() and copy_splice_read()
  iov_iter: Kill ITER_PIPE
  splice: Remove generic_file_splice_read()
  splice: Use filemap_splice_read() instead of generic_file_splice_read()
  cifs: Use filemap_splice_read()
  trace: Convert trace/seq to use copy_splice_read()
  zonefs: Provide a splice-read wrapper
  xfs: Provide a splice-read wrapper
  orangefs: Provide a splice-read wrapper
  ocfs2: Provide a splice-read wrapper
  ntfs3: Provide a splice-read wrapper
  nfs: Provide a splice-read wrapper
  f2fs: Provide a splice-read wrapper
  ext4: Provide a splice-read wrapper
  ecryptfs: Provide a splice-read wrapper
  ceph: Provide a splice-read wrapper
  afs: Provide a splice-read wrapper
  9p: Add splice_read wrapper
  net: Make sock_splice_read() use copy_splice_read() by default
  tty, proc, kernfs, random: Use copy_splice_read()
  ...

splice: kdoc for filemap_splice_read() and copy_splice_read()

Provide kerneldoc comments for filemap_splice_read() and
copy_splice_read().

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christian Brauner <brauner@kernel.org>
cc: Christoph Hellwig <hch@lst.de>
cc: Jens Axboe <axboe@kernel.dk>
cc: Steve French <smfrench@gmail.com>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-32-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

iov_iter: Kill ITER_PIPE

The ITER_PIPE-type iterator was only used by generic_file_splice_read() and
that has been replaced and removed.  This leaves ITER_PIPE unused - so
remove it too.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-31-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Remove generic_file_splice_read()

Remove generic_file_splice_read() as it has been replaced with calls to
filemap_splice_read() and copy_splice_read().

With this, ITER_PIPE is no longer used.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Steve French <smfrench@gmail.com>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-30-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Use filemap_splice_read() instead of generic_file_splice_read()

Replace pointers to generic_file_splice_read() with calls to
filemap_splice_read().

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-29-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

cifs: Use filemap_splice_read()

Make cifs use filemap_splice_read() rather than doing its own version of
generic_file_splice_read().

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
cc: Jens Axboe <axboe@kernel.dk>
cc: Steve French <smfrench@gmail.com>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-28-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

trace: Convert trace/seq to use copy_splice_read()

For the splice from the trace seq buffer, just use copy_splice_read().

In the future, something better can probably be done by gifting pages from
seq->buf into the pipe, but that would require changing seq->buf into a
vmap over an array of pages.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Steven Rostedt <rostedt@goodmis.org>
cc: Masami Hiramatsu <mhiramat@kernel.org>
cc: linux-kernel@vger.kernel.org
cc: linux-trace-kernel@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-27-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

zonefs: Provide a splice-read wrapper

Provide a splice_read wrapper for zonefs.  This does some checks before
proceeding and locks the inode across the call to filemap_splice_read() and
a size check in case of truncation.  Splicing from direct I/O is handled by
the caller.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Darrick J. Wong <djwong@kernel.org>
cc: linux-xfs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Acked-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230522135018.2742245-26-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

xfs: Provide a splice-read wrapper

Provide a splice_read wrapper for XFS.  This does a stat count and a
shutdown check before proceeding, then emits a new trace line and locks the
inode across the call to filemap_splice_read() and adds to the stats
afterwards.  Splicing from direct I/O or DAX is handled by the caller.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Darrick J. Wong <djwong@kernel.org>
cc: linux-xfs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-25-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

orangefs: Provide a splice-read wrapper

Provide a splice_read wrapper for ocfs2.  This increments the read stats
and then locks the inode across the call to filemap_splice_read() and a
revalidation of the mapping.  Splicing from direct I/O is done by the
caller.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Mike Marshall <hubcap@omnibond.com>
cc: Martin Brandenburg <martin@omnibond.com>
cc: devel@lists.orangefs.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-24-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ocfs2: Provide a splice-read wrapper

Provide a splice_read wrapper for ocfs2.  This emits trace lines and does
an atime lock/update before calling filemap_splice_read().  Splicing from
direct I/O is handled by the caller.

A couple of new tracepoints are added for this purpose.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Mark Fasheh <mark@fasheh.com>
cc: Joel Becker <jlbec@evilplan.org>
cc: ocfs2-devel@oss.oracle.com
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-23-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ntfs3: Provide a splice-read wrapper

Provide a splice_read wrapper for NTFS3 to perform various checks before
allowing the operation to proceed.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
cc: ntfs3@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-22-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nfs: Provide a splice-read wrapper

Provide a splice_read wrapper for NFS.  This locks the inode around
filemap_splice_read() and revalidates the mapping.  Splicing from direct
I/O is handled by the caller.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Trond Myklebust <trond.myklebust@hammerspace.com>
cc: Anna Schumaker <anna@kernel.org>
cc: linux-nfs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-21-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

f2fs: Provide a splice-read wrapper

Provide a splice_read wrapper for f2fs.  This does some checks and tracing
before calling filemap_splice_read() and will update the iostats
afterwards.  Direct I/O is handled by the caller.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Jaegeuk Kim <jaegeuk@kernel.org>
cc: Chao Yu <chao@kernel.org>
cc: linux-f2fs-devel@lists.sourceforge.net
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-20-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ext4: Provide a splice-read wrapper

Provide a splice_read wrapper for Ext4.  This does the inode shutdown check
before proceeding.  Splicing from DAX files and O_DIRECT fds is handled by
the caller.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Theodore Ts'o <tytso@mit.edu>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Andreas Dilger <adilger.kernel@dilger.ca>
cc: linux-ext4@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-19-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ecryptfs: Provide a splice-read wrapper

Provide a splice_read wrapper for ecryptfs to update the access time on the
lower file after the operation.  Splicing from a direct I/O fd will update
the access time when ->read_iter() is called.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Tyler Hicks <code@tyhicks.com>
cc: ecryptfs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-18-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ceph: Provide a splice-read wrapper

Provide a splice_read wrapper for Ceph.  This does the inode shutdown check
before proceeding and jumps to copy_splice_read() if the file has inline
data or is a synchronous file.

We try and get FILE_RD and either FILE_CACHE and/or FILE_LAZYIO caps and
hold them across filemap_splice_read().  If we fail to get FILE_CACHE or
FILE_LAZYIO capabilities, we use copy_splice_read() instead.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Ilya Dryomov <idryomov@gmail.com>
cc: Jeff Layton <jlayton@kernel.org>
cc: ceph-devel@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-17-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

afs: Provide a splice-read wrapper

Provide a splice_read wrapper for AFS to call afs_validate() before going
into generic_file_splice_read() so that we're likely to have a callback
promise from the server.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-16-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

9p: Add splice_read wrapper

Add a splice_read wrapper for 9p.  We should use copy_splice_read() if
9PL_DIRECT is set and filemap_splice_read() otherwise.  Note that this
doesn't seem to be particularly related to O_DIRECT.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: Dominique Martinet <asmadeus@codewreck.org>
cc: Eric Van Hensbergen <ericvh@kernel.org>
cc: Latchesar Ionkov <lucho@ionkov.net>
cc: Christian Schoenebeck <linux_oss@crudebyte.com>
cc: v9fs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-15-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

net: Make sock_splice_read() use copy_splice_read() by default

Make sock_splice_read() use copy_splice_read() by default as
file_splice_read() will return immediately with 0 as a socket has no
pagecache and is a zero-size file.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: "David S. Miller" <davem@davemloft.net>
cc: Eric Dumazet <edumazet@google.com>
cc: Jakub Kicinski <kuba@kernel.org>
cc: Paolo Abeni <pabeni@redhat.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: netdev@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-14-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

tty, proc, kernfs, random: Use copy_splice_read()

Use copy_splice_read() for tty, procfs, kernfs and random files rather
than going through generic_file_splice_read() as they just copy the file
into the output buffer and don't splice pages.  This avoids the need for
them to have a ->read_folio() to satisfy filemap_splice_read().

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
cc: Christoph Hellwig <hch@lst.de>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: John Hubbard <jhubbard@nvidia.com>
cc: David Hildenbrand <david@redhat.com>
cc: Matthew Wilcox <willy@infradead.org>
cc: Miklos Szeredi <miklos@szeredi.hu>
cc: Arnd Bergmann <arnd@arndb.de>
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-13-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

coda: Implement splice-read

Implement splice-read for coda by passing the request down a layer rather
than going through generic_file_splice_read() which is going to be changed
to assume that ->read_folio() is present on buffered files.

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Jan Harkes <jaharkes@cs.cmu.edu>
cc: Christoph Hellwig <hch@lst.de>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: John Hubbard <jhubbard@nvidia.com>
cc: David Hildenbrand <david@redhat.com>
cc: Matthew Wilcox <willy@infradead.org>
cc: coda@cs.cmu.edu
cc: codalist@coda.cs.cmu.edu
cc: linux-unionfs@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-12-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

overlayfs: Implement splice-read

Implement splice-read for overlayfs by passing the request down a layer
rather than going through generic_file_splice_read() which is going to be
changed to assume that ->read_folio() is present on buffered files.

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Christian Brauner <brauner@kernel.org>
cc: Christoph Hellwig <hch@lst.de>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: John Hubbard <jhubbard@nvidia.com>
cc: David Hildenbrand <david@redhat.com>
cc: Matthew Wilcox <willy@infradead.org>
cc: Miklos Szeredi <miklos@szeredi.hu>
cc: Amir Goldstein <amir73il@gmail.com>
cc: linux-unionfs@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-11-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

shmem: Implement splice-read

The new filemap_splice_read() has an implicit expectation via
filemap_get_pages() that ->read_folio() exists if ->readahead() doesn't
fully populate the pagecache of the file it is reading from[1], potentially
leading to a jump to NULL if this doesn't exist.  shmem, however, (and by
extension, tmpfs, ramfs and rootfs), doesn't have ->read_folio(),

Work around this by equipping shmem with its own splice-read
implementation, based on filemap_splice_read(), but able to paste in
zero_page when there's a page missing.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Daniel Golle <daniel@makrotopia.org>
cc: Guenter Roeck <groeck7@gmail.com>
cc: Christoph Hellwig <hch@lst.de>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: John Hubbard <jhubbard@nvidia.com>
cc: David Hildenbrand <david@redhat.com>
cc: Matthew Wilcox <willy@infradead.org>
cc: Hugh Dickins <hughd@google.com>
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/Y+pdHFFTk1TTEBsO@makrotopia.org/
Link: https://lore.kernel.org/r/20230522135018.2742245-10-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Make splice from a DAX file use copy_splice_read()

Make a read splice from a DAX file go directly to copy_splice_read() to do
the reading as filemap_splice_read() is unlikely to find any pagecache to
splice.

I think this affects only erofs, Ext2, Ext4, fuse and XFS.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: linux-erofs@lists.ozlabs.org
cc: linux-ext4@vger.kernel.org
cc: linux-xfs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-9-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Make splice from an O_DIRECT fd use copy_splice_read()

Make a read splice from a file descriptor that's open O_DIRECT use
copy_splice_read() to do the reading as filemap_splice_read() is unlikely
to find any pagecache to splice.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: Jens Axboe <axboe@kernel.dk>
cc: linux-fsdevel@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-8-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Check for zero count in vfs_splice_read()

Make vfs_splice_read() return immediately if the length is 0.

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-7-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Make do_splice_to() generic and export it

Rename do_splice_to() to vfs_splice_read() and export it so that it can be
used as a helper when calling down to a lower layer filesystem as it
performs all the necessary checks[1].

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Miklos Szeredi <miklos@szeredi.hu>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: John Hubbard <jhubbard@nvidia.com>
cc: David Hildenbrand <david@redhat.com>
cc: Matthew Wilcox <willy@infradead.org>
cc: linux-unionfs@vger.kernel.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/CAJfpeguGksS3sCigmRi9hJdUec8qtM9f+_9jC1rJhsXT+dV01w@mail.gmail.com/
Link: https://lore.kernel.org/r/20230522135018.2742245-6-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Clean up copy_splice_read() a bit

Do a couple of cleanups to copy_splice_read():

 (1) Cast to struct page **, not void *.

 (2) Simplify the calculation of the number of pages to keep/reclaim in
     copy_splice_read().

Suggested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-5-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Rename direct_splice_read() to copy_splice_read()

Rename direct_splice_read() to copy_splice_read() to better reflect as to
what it does.

Suggested-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Steve French <sfrench@samba.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: linux-cifs@vger.kernel.org
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-4-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Make filemap_splice_read() check s_maxbytes

Make filemap_splice_read() check s_maxbytes analogously to filemap_read().

Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Steve French <stfrench@microsoft.com>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-3-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

splice: Fix filemap_splice_read() to use the correct inode

Fix filemap_splice_read() to use file->f_mapping->host, not file->f_inode,
as the source of the file size because in the case of a block device,
file->f_inode points to the block-special file (which is typically 0
length) and not the backing store.

Fixes: 07073eb01c5f ("splice: Add a func to do a splice from a buffered file without ITER_PIPE")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
cc: Steve French <stfrench@microsoft.com>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: David Hildenbrand <david@redhat.com>
cc: John Hubbard <jhubbard@nvidia.com>
cc: linux-mm@kvack.org
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Link: https://lore.kernel.org/r/20230522135018.2742245-2-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: introduce block_io_start/block_io_done tracepoints

Currently, several BCC ([0]) tools (biosnoop/biostacks/biotop) use
kprobes to blk_account_io_start/blk_account_io_done to implement
their functionalities. This is fragile because the target kernel
functions may be renamed ([1]) or inlined ([2]). So introduce two
new tracepoints for such use cases.

  [0]: https://github.com/iovisor/bcc
  [1]: https://github.com/iovisor/bcc/issues/3954
  [2]: https://github.com/iovisor/bcc/issues/4261

Tested-by: Francis Laniel <flaniel@linux.microsoft.com>
Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com>
Tested-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/r/20230520084057.1467003-1-hengqi.chen@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block/rq_qos: protect rq_qos apis with a new lock

commit 50e34d78815e ("block: disable the elevator int del_gendisk")
move rq_qos_exit() from disk_release() to del_gendisk(), this will
introduce some problems:

1) If rq_qos_add() is triggered by enabling iocost/iolatency through
   cgroupfs, then it can concurrent with del_gendisk(), it's not safe to
   write 'q->rq_qos' concurrently.

2) Activate cgroup policy that is relied on rq_qos will call
   rq_qos_add() and blkcg_activate_policy(), and if rq_qos_exit() is
   called in the middle, null-ptr-dereference will be triggered in
   blkcg_activate_policy().

3) blkg_conf_open_bdev() can call blkdev_get_no_open() first to find the
   disk, then if rq_qos_exit() from del_gendisk() is done before
   rq_qos_add(), then memory will be leaked.

This patch add a new disk level mutex 'rq_qos_mutex':

1) The lock will protect rq_qos_exit() directly.

2) For wbt that doesn't relied on blk-cgroup, rq_qos_add() can only be
   called from disk initialization for now because wbt can't be
   destructed until rq_qos_exit(), so it's safe not to protect wbt for
   now. Hoever, in case that rq_qos dynamically destruction is supported
   in the furture, this patch also protect rq_qos_add() from wbt_init()
   directly, this is enough because blk-sysfs already synchronize
   writers with disk removal.

3) For iocost and iolatency, in order to synchronize disk removal and
   cgroup configuration, the lock is held after blkdev_get_no_open()
   from blkg_conf_open_bdev(), and is released in blkg_conf_exit().
   In order to fix the above memory leak, disk_live() is checked after
   holding the new lock.

Fixes: 50e34d78815e ("block: disable the elevator int del_gendisk")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230414084008.2085155-1-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: remove redundant req_op in blk_rq_is_passthrough

op &= REQ_OP_MASK in blk_op_is_passthrough() is exactly what req_op() do.
Therefore, it is redundant to call req_op() for blk_op_is_passthrough().

Signed-off-by: Li Nan <linan122@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20230522085355.1740772-1-linan666@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: fix build warning on iov_iter_get_pages2

Return type of iov_iter_get_pages2() is ssize_t instead of size_t, so
fix it.

Fixes: 981f95a571e3 ("ublk: cleanup ublk_copy_user_pages")
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Julia Lawall <julia.lawall@inria.fr>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230520151134.459679-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: don't plug in blkdev_write_iter

For direct I/O writes that issues more than a single bio, the plugging
is already done in __blkdev_direct_IO.
For synchronous buffered writes the plugging is done deep down in
writeback_inodes_wb / wb_writeback.

For the other cases there is no point in plugging as as single bio or no
bio at all is submitted.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230520044503.334444-1-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: BFQ: Move an invariant check

Check bfqq->dispatched for each BFQ queue instead of checking it for an
invalid bfqq pointer.

Fixes: 3e49c1e4a615 ("block: BFQ: Add several invariant checks")
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230519220347.3643295-1-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: support user copy

Currently copy between io request buffer(pages) and userspace buffer is
done inside ublk_map_io() or ublk_unmap_io(). This way performs very
well in case of pre-allocated userspace io buffer.

For dynamically allocated or external userspace backend io buffer,
UBLK_F_NEED_GET_DATA is added for ublk server to provide buffer by one
extra command communication for WRITE request. For READ, userspace
simply provides buffer, but can't know when the buffer is done[1].

Add UBLK_F_USER_COPY by moving io data copy out of kernel by providing
read()/write() on /dev/ublkcN, and simply let ublk server do the io
data copy. This way makes both side cleaner, the cost is that one extra
syscall for copy io data between request and backend buffer.

With UBLK_F_USER_COPY, it actually becomes possible to run per-io zero
copy now, such as, only do zero copy for big size IO, so it can be
thought as one prep patch for supporting zero copy. Meantime zero copy
still needs to expose read()/write() buffer for some corner case, such
as passthrough IO.

[1] READ buffer in UBLK_F_NEED_GET_DATA
https://lore.kernel.org/linux-block/116d8a56-0881-56d3-9bcc-78ff3e1dc4e5@linux.alibaba.com/T/#m23bd4b8634c0a054e6797063167b469949a247bb

ublksrv loop usercopy code:

https://github.com/ming1/ubdsrv/commits/usercopy

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-8-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: add read()/write() support for ublk char device

Support pread()/pwrite() on ublk char device for reading/writing request
io buffer, so data copy between io request buffer and userspace buffer
can be moved to ublk server from ublk driver. Then UBLK_F_NEED_GET_DATA
becomes not necessary, so ublk server can allocate buffer without one
extra round uring command communication for userspace to provide buffer.

IO buffer can be located by iocb->ki_pos which encodes buffer offset, io
tag and queue id info, and type of iocb->ki_pos is u64, so it is big
enough for holding reasonable queue depth, nr_queues and max io buffer
size.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-7-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: support to copy any part of request pages

Add 'offset' to 'struct ublk_map_data', so that ublk_copy_user_pages()
can be used to copy any sub-buffer(linear mapped) of the request.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-6-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: grab request reference when the request is handled by userspace

Add one reference counter into request pdu data, and hold this reference
in the request's lifetime.

Prepare for supporting to move request data copy into userspace, which
needs to copy request data by read()/write() on /dev/ublkcN, so we have
to guarantee that read()/write() is done on one valid/active request,
and that will be enhanced by holding the io request reference in
read()/write().

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-5-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: cleanup ublk_copy_user_pages

Clean up ublk_copy_user_pages() by using iov_iter_get_pages2, and code
gets simplified a lot and becomes much more readable than before.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-4-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: cleanup io cmd code path by adding ublk_fill_io_cmd()

Add one small helper to cleanup io command hanlding code path.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-3-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ublk: kill queuing request by task_work_add

task_work_add() is used from early ublk development stage for handling
request in batch. However, since commit 7d4a93176e01 ("ublk_drv: don't
forward io commands in reserve order"), we can get similar batch
processing with io_uring_cmd_complete_in_task(), and similar performance
data is observed between task_work_add() and
io_uring_cmd_complete_in_task().

Meantime we can kill one fast code path, which is actually seldom used
given it is common to build ublk driver as module.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230519065030.351216-2-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: don't use the requeue list to queue flush commands

Currently both requeues of commands that were already sent to the driver
and flush commands submitted from the flush state machine share the same
requeue_list struct request_queue, despite requeues doing head
insertions and flushes not.  Switch to using two separate lists instead.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230519044050.107790-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: do not do head insertions post-pre-flush commands

blk_flush_complete_seq currently queues requests that write data after
a pre-flush from the flush state machine at the head of the queue.
This doesn't really make sense, as the original request bypassed all
queue lists by directly diverting to blk_insert_flush from
blk_mq_submit_bio.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230519044050.107790-7-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: defer to the normal submission path for post-flush requests

Requests with the FUA bit on hardware without FUA support need a post
flush before returning to the caller, but they can still be sent using
the normal I/O path after initializing the flush-related fields and
end I/O handler.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230519044050.107790-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: use the I/O scheduler for writes from the flush state machine

Send write requests issued by the flush state machine through the normal
I/O submission path including the I/O scheduler (if present) so that I/O
scheduler policies are applied to writes with the FUA flag set.

Separate the I/O scheduler members from the flush members in struct
request since now a request may pass through both an I/O scheduler
and the flush machinery.

Note that the actual flush requests, which have no bio attached to the
request still bypass the I/O schedulers.

Signed-off-by: Bart Van Assche <bvanassche@acm.org>
[hch: rebased]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230519044050.107790-5-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: defer to the normal submission path for non-flush flush commands

If blk_insert_flush decides that a command does not need to use the
flush state machine, return false and let blk_mq_submit_bio handle
it the normal way (including using an I/O scheduler) instead of doing
a bypass insert.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230519044050.107790-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: reflow blk_insert_flush

Use a switch statement to decide on the disposition of a flush request
instead of multiple if statements, out of which one does checks that are
more complex than required.  Also warn on a malformed request early
on instead of doing a BUG_ON later.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230519044050.107790-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: factor out a blk_rq_init_flush helper

Factor out a helper from blk_insert_flush that initializes the flush
machine related fields in struct request, and don't bother with the
full memset as there's just a few fields to initialize, and all but
one already have explicit initializers.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230519044050.107790-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

fs: remove the special !CONFIG_BLOCK def_blk_fops

def_blk_fops always returns -ENODEV, which dosn't match the return value
of a non-existing block device with CONFIG_BLOCK, which is -ENXIO.
Just remove the extra implementation and fall back to the default
no_open_fops that always returns -ENXIO.

Fixes: 9361401eb761 ("[PATCH] BLOCK: Make it possible to disable the block layer [try #6]")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230508144405.41792-1-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: BFQ: Add several invariant checks

If anything goes wrong with the counters that track the number of
requests, I/O locks up. Make such scenarios easier to debug by adding
invariant checks for the request counters. Additionally, check that
BFQ queues are empty before these are freed.

Cc: Jan Kara <jack@suse.cz>
Cc: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230516223853.1385255-1-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Fix handling of at-head zoned writes

Before dispatching a zoned write from the FIFO list, check whether there
are any zoned writes in the RB-tree with a lower LBA for the same zone.
This patch ensures that zoned writes happen in order even if at_head is
set for some writes for a zone and not for others.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230517174230.897144-12-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Handle requeued requests correctly

Start dispatching from the start of a zone instead of from the starting
position of the most recently dispatched request.

If a zoned write is requeued with an LBA that is lower than already
inserted zoned writes, make sure that it is submitted first.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cc: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230517174230.897144-11-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Track the dispatch position

Track the position (sector_t) of the most recently dispatched request
instead of tracking a pointer to the next request to dispatch. This
patch is the basis for patch "Handle requeued requests correctly".
Without this patch it would be significantly more complicated to make
sure that zoned writes are dispatched in LBA order per zone.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230517174230.897144-10-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Reduce lock contention

blk_mq_free_requests() calls dd_finish_request() indirectly. Prevent
nested locking of dd->lock and dd->zone_lock by moving the code for
freeing requests.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230517174230.897144-9-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Simplify deadline_skip_seq_writes()

Make the deadline_skip_seq_writes() code shorter without changing its
functionality.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230517174230.897144-8-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Clean up deadline_check_fifo()

Change the return type of deadline_check_fifo() from 'int' into 'bool'.
Use time_is_before_eq_jiffies() instead of time_after_eq(). No
functionality has been changed.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cc: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230517174230.897144-7-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Introduce blk_rq_is_seq_zoned_write()

Introduce the function blk_rq_is_seq_zoned_write(). This function will
be used in later patches to preserve the order of zoned writes that
require write serialization.

This patch includes an optimization: instead of using
rq->q->disk->part0->bd_queue to check whether or not the queue is
associated with a zoned block device, use rq->q->disk->queue.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230517174230.897144-6-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Introduce op_needs_zoned_write_locking()

Introduce a helper function for checking whether write serialization is
required if the operation will be sent to a zoned device. A second caller
for op_needs_zoned_write_locking() will be introduced in the next patch
in this series.

Suggested-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230517174230.897144-5-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Fix the type of the second bdev_op_is_zoned_write() argument

Change the type of the second argument of bdev_op_is_zoned_write() from
blk_opf_t into enum req_op because this function expects an operation
without flags as second argument.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Pankaj Raghav <p.raghav@samsung.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Fixes: 8cafdb5ab94c ("block: adapt blk_mq_plug() to not plug for writes that require a zone lock")
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230517174230.897144-4-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Simplify blk_req_needs_zone_write_lock()

Remove the blk_rq_is_passthrough() check because it is redundant:
blk_req_needs_zone_write_lock() also calls bdev_op_is_zoned_write()
and the latter function returns false for pass-through requests.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230517174230.897144-3-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: mq-deadline: Add a word in a source code comment

Add the missing word "and".

Cc: Damien Le Moal <dlemoal@kernel.org>
Suggested-by: Damien Le Moal <dlemoal@kernel.org>
Fixes: 945ffb60c11d ("mq-deadline: add blk-mq adaptation of the deadline IO scheduler")
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Tested-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230517174230.897144-2-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: make sure elevator callbacks aren't called for passthrough request

In case of q->elevator, passthrough request can still be marked as
RQF_ELV, so some elevator callbacks will be called for them.

Fix this by splitting RQF_SCHED_TAGS, which is set for all requests that
are issued on a queue that uses an I/O scheduler, and RQF_USE_SCHED for
non-flush, non-passthrough requests on such a queue.

Roughly based on two different patches from
Ming Lei <ming.lei@redhat.com>.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230518053101.760632-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: remove RQF_ELVPRIV

RQF_ELVPRIV is set for all non-flush requests that have RQF_ELV set.
Expand this condition in the two users of the flag and remove it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230518053101.760632-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: don't queue plugged passthrough requests into scheduler

Passthrough requests should never be queued to the I/O scheduler,
as scheduling these opaque requests doesn't make sense, and I/O
schedulers might require req->bio to be always valid.

We never let passthrough requests insert into the scheduler before
commit 1c2d2fff6dc0 ("block: wire-up support for passthrough plugging"),
restore this behavior even for passthrough requests issued under a plug.

[hch: use blk_mq_insert_requests for passthrough requests,
      fix up the commit message and comments]

Reported-by: Guangwu Zhang <guazhang@redhat.com>
Closes: https://lore.kernel.org/linux-block/CAGS2=YosaYaUTEMU3uaf+y=8MqSrhL7sYsJn8EwbaM=76p_4Qg@mail.gmail.com/
Investigated-by: Yu Kuai <yukuai1@huaweicloud.com>
Fixes: 1c2d2fff6dc0 ("block: wire-up support for passthrough plugging")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230518053101.760632-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Decode all flag names in the debugfs output

See also:
* Commit 4d337cebcb1c ("blk-mq: avoid to touch q->elevator without any protection").
* Commit 414dd48e882c ("blk-mq: add tagset quiesce interface").

Cc: Christoph Hellwig <hch@lst.de>
Cc: Damien Le Moal <dlemoal@kernel.org>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20230518222708.1190867-1-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

brd: use XArray instead of radix-tree to index backing pages

XArray was introduced to hold large array of pointers with a simple API.
XArray API also provides array semantics which simplifies the way we store
and access the backing pages, and the code becomes significantly easier
to understand.

No performance difference was noticed between the two implementation
using fio with direct=1 [1].

[1] Performance in KIOPS:

          |  radix-tree |    XArray  |   Diff
          |             |            |
write     |    315      |     313    |   -0.6%
randwrite |    286      |     290    |   +1.3%
read      |    330      |     335    |   +1.5%
randread  |    309      |     312    |   +0.9%

Signed-off-by: Pankaj Raghav <p.raghav@samsung.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20230511121544.111648-1-p.raghav@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Linux 6.4-rc2

Merge tag 'cxl-fixes-6.4-rc2' of git://git./linux/kernel/git/cxl/cxl

Pull compute express link fixes from Dan Williams:

 - Fix a compilation issue with DEFINE_STATIC_SRCU() in the unit tests

 - Fix leaking kernel memory to a root-only sysfs attribute

* tag 'cxl-fixes-6.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl:
  cxl: Add missing return to cdat read error path
  tools/testing/cxl: Use DEFINE_STATIC_SRCU()

Merge tag 'parisc-for-6.4-2' of git://git./linux/kernel/git/deller/parisc-linux

Pull parisc architecture fixes from Helge Deller:

 - Fix encoding of swp_entry due to added SWP_EXCLUSIVE flag

 - Include reboot.h to avoid gcc-12 compiler warning

* tag 'parisc-for-6.4-2' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
  parisc: Fix encoding of swp_entry due to added SWP_EXCLUSIVE flag
  parisc: kexec: include reboot.h

Merge tag 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm

Pull ARM fixes from Russell King:

 - fix unwinder for uleb128 case

 - fix kernel-doc warnings for HP Jornada 7xx

 - fix unbalanced stack on vfp success path

* tag 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm:
  ARM: 9297/1: vfp: avoid unbalanced stack on 'success' return path
  ARM: 9296/1: HP Jornada 7XX: fix kernel-doc warnings
  ARM: 9295/1: unwind:fix unwind abort for uleb128 case

Merge tag 'locking_urgent_for_v6.4_rc2' of git://git./linux/kernel/git/tip/tip

Pull locking fix from Borislav Petkov:

 - Make sure __down_read_common() is always inlined so that the callers'
   names land in traceevents output and thus the blocked function can be
   identified

* tag 'locking_urgent_for_v6.4_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  locking/rwsem: Add __always_inline annotation to __down_read_common() and inlined callers

Merge tag 'perf_urgent_for_v6.4_rc2' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Borislav Petkov:

 - Make sure the PEBS buffer is flushed before reprogramming the
   hardware so that the correct record sizes are used

 - Update the sample size for AMD BRS events

 - Fix a confusion with using the same on-stack struct with different
   events in the event processing path

* tag 'perf_urgent_for_v6.4_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/intel/ds: Flush PEBS DS when changing PEBS_DATA_CFG
  perf/x86: Fix missing sample size update on AMD BRS
  perf/core: Fix perf_sample_data not properly initialized for different swevents in perf_tp_event()

Merge tag 'sched_urgent_for_v6.4_rc2' of git://git./linux/kernel/git/tip/tip

Pull scheduler fix from Borislav Petkov:

 - Fix a couple of kernel-doc warnings

* tag 'sched_urgent_for_v6.4_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: fix cid_lock kernel-doc warnings

Merge tag 'x86_urgent_for_v6.4_rc2' of git://git./linux/kernel/git/tip/tip

Pull x86 fix from Borislav Petkov:

 - Add the required PCI IDs so that the generic SMN accesses provided by
   amd_nb.c work for drivers which switch to them. Add a PCI device ID
   to k10temp's table so that latter is loaded on such systems too

* tag 'x86_urgent_for_v6.4_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  hwmon: (k10temp) Add PCI ID for family 19, model 78h
  x86/amd_nb: Add PCI ID for family 19h model 78h

Merge tag 'timers_urgent_for_v6.4_rc2' of git://git./linux/kernel/git/tip/tip

Pull timer fix from Borislav Petkov:

 - Prevent CPU state corruption when an active clockevent broadcast
   device is replaced while the system is already in oneshot mode

* tag 'timers_urgent_for_v6.4_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  tick/broadcast: Make broadcast device replacement work correctly

Merge tag 'ext4_for_linus_stable' of git://git./linux/kernel/git/tytso/ext4

Pull ext4 fixes from Ted Ts'o:
 "Some ext4 bug fixes (mostly to address Syzbot reports)"

* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: bail out of ext4_xattr_ibody_get() fails for any reason
  ext4: add bounds checking in get_max_inline_xattr_value_size()
  ext4: add indication of ro vs r/w mounts in the mount message
  ext4: fix deadlock when converting an inline directory in nojournal mode
  ext4: improve error recovery code paths in __ext4_remount()
  ext4: improve error handling from ext4_dirhash()
  ext4: don't clear SB_RDONLY when remounting r/w until quota is re-enabled
  ext4: check iomap type only if ext4_iomap_begin() does not fail
  ext4: avoid a potential slab-out-of-bounds in ext4_group_desc_csum
  ext4: fix data races when using cached status extents
  ext4: avoid deadlock in fs reclaim with page writeback
  ext4: fix invalid free tracking in ext4_xattr_move_to_block()
  ext4: remove a BUG_ON in ext4_mb_release_group_pa()
  ext4: allow ext4_get_group_info() to fail
  ext4: fix lockdep warning when enabling MMP
  ext4: fix WARNING in mb_find_extent

Merge tag 'fbdev-for-6.4-rc2' of git://git./linux/kernel/git/deller/linux-fbdev

Pull fbdev fixes from Helge Deller:

 - use after free fix in imsttfb (Zheng Wang)

 - fix error handling in arcfb (Zongjie Li)

 - lots of whitespace cleanups (Thomas Zimmermann)

 - add 1920x1080 modedb entry (me)

* tag 'fbdev-for-6.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/linux-fbdev:
  fbdev: stifb: Fix info entry in sti_struct on error path
  fbdev: modedb: Add 1920x1080 at 60 Hz video mode
  fbdev: imsttfb: Fix use after free bug in imsttfb_probe
  fbdev: vfb: Remove trailing whitespaces
  fbdev: valkyriefb: Remove trailing whitespaces
  fbdev: stifb: Remove trailing whitespaces
  fbdev: sa1100fb: Remove trailing whitespaces
  fbdev: platinumfb: Remove trailing whitespaces
  fbdev: p9100: Remove trailing whitespaces
  fbdev: maxinefb: Remove trailing whitespaces
  fbdev: macfb: Remove trailing whitespaces
  fbdev: hpfb: Remove trailing whitespaces
  fbdev: hgafb: Remove trailing whitespaces
  fbdev: g364fb: Remove trailing whitespaces
  fbdev: controlfb: Remove trailing whitespaces
  fbdev: cg14: Remove trailing whitespaces
  fbdev: atmel_lcdfb: Remove trailing whitespaces
  fbdev: 68328fb: Remove trailing whitespaces
  fbdev: arcfb: Fix error handling in arcfb_probe()

Merge tag 'scsi-fixes' of git://git./linux/kernel/git/jejb/scsi

Pull SCSI fix from James Bottomley:
 "A single small fix for the UFS driver to fix a power management
  failure"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: ufs: core: Fix I/O hang that occurs when BKOPS fails in W-LUN suspend

parisc: Fix encoding of swp_entry due to added SWP_EXCLUSIVE flag

Fix the __swp_offset() and __swp_entry() macros due to commit 6d239fc78c0b
("parisc/mm: support __HAVE_ARCH_PTE_SWP_EXCLUSIVE") which introduced the
SWP_EXCLUSIVE flag by reusing the _PAGE_ACCESSED flag.

Reported-by: Christoph Biedl <linux-kernel.bfrz@manchmal.in-ulm.de>
Tested-by: Christoph Biedl <linux-kernel.bfrz@manchmal.in-ulm.de>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Helge Deller <deller@gmx.de>
Fixes: 6d239fc78c0b ("parisc/mm: support __HAVE_ARCH_PTE_SWP_EXCLUSIVE")
Cc: <stable@vger.kernel.org> # v6.3+

ext4: bail out of ext4_xattr_ibody_get() fails for any reason

In ext4_update_inline_data(), if ext4_xattr_ibody_get() fails for any
reason, it's best if we just fail as opposed to stumbling on,
especially if the failure is EFSCORRUPTED.

Cc: stable@kernel.org
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: add bounds checking in get_max_inline_xattr_value_size()

Normally the extended attributes in the inode body would have been
checked when the inode is first opened, but if someone is writing to
the block device while the file system is mounted, it's possible for
the inode table to get corrupted.  Add bounds checking to avoid
reading beyond the end of allocated memory if this happens.

Reported-by: syzbot+1966db24521e5f6e23f7@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?extid=1966db24521e5f6e23f7
Cc: stable@kernel.org
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: add indication of ro vs r/w mounts in the mount message

Whether the file system is mounted read-only or read/write is more
important than the quota mode, which we are already printing.  Add the
ro vs r/w indication since this can be helpful in debugging problems
from the console log.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: fix deadlock when converting an inline directory in nojournal mode

In no journal mode, ext4_finish_convert_inline_dir() can self-deadlock
by calling ext4_handle_dirty_dirblock() when it already has taken the
directory lock.  There is a similar self-deadlock in
ext4_incvert_inline_data_nolock() for data files which we'll fix at
the same time.

A simple reproducer demonstrating the problem:

    mke2fs -Fq -t ext2 -O inline_data -b 4k /dev/vdc 64
    mount -t ext4 -o dirsync /dev/vdc /vdc
    cd /vdc
    mkdir file0
    cd file0
    touch file0
    touch file1
    attr -s BurnSpaceInEA -V abcde .
    touch supercalifragilisticexpialidocious

Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20230507021608.1290720-1-tytso@mit.edu
Reported-by: syzbot+91dccab7c64e2850a4e5@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?id=ba84cc80a9491d65416bc7877e1650c87530fe8a
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: improve error recovery code paths in __ext4_remount()

If there are failures while changing the mount options in
__ext4_remount(), we need to restore the old mount options.

This commit fixes two problem.  The first is there is a chance that we
will free the old quota file names before a potential failure leading
to a use-after-free.  The second problem addressed in this commit is
if there is a failed read/write to read-only transition, if the quota
has already been suspended, we need to renable quota handling.

Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20230506142419.984260-2-tytso@mit.edu
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: improve error handling from ext4_dirhash()

The ext4_dirhash() will *almost* never fail, especially when the hash
tree feature was first introduced.  However, with the addition of
support of encrypted, casefolded file names, that function can most
certainly fail today.

So make sure the callers of ext4_dirhash() properly check for
failures, and reflect the errors back up to their callers.

Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20230506142419.984260-1-tytso@mit.edu
Reported-by: syzbot+394aa8a792cb99dbc837@syzkaller.appspotmail.com
Reported-by: syzbot+344aaa8697ebd232bfc8@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?id=db56459ea4ac4a676ae4b4678f633e55da005a9b
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: don't clear SB_RDONLY when remounting r/w until quota is re-enabled

When a file system currently mounted read/only is remounted
read/write, if we clear the SB_RDONLY flag too early, before the quota
is initialized, and there is another process/thread constantly
attempting to create a directory, it's possible to trigger the

WARN_ON_ONCE(dquot_initialize_needed(inode));

in ext4_xattr_block_set(), with the following stack trace:

   WARNING: CPU: 0 PID: 5338 at fs/ext4/xattr.c:2141 ext4_xattr_block_set+0x2ef2/0x3680
   RIP: 0010:ext4_xattr_block_set+0x2ef2/0x3680 fs/ext4/xattr.c:2141
   Call Trace:
    ext4_xattr_set_handle+0xcd4/0x15c0 fs/ext4/xattr.c:2458
    ext4_initxattrs+0xa3/0x110 fs/ext4/xattr_security.c:44
    security_inode_init_security+0x2df/0x3f0 security/security.c:1147
    __ext4_new_inode+0x347e/0x43d0 fs/ext4/ialloc.c:1324
    ext4_mkdir+0x425/0xce0 fs/ext4/namei.c:2992
    vfs_mkdir+0x29d/0x450 fs/namei.c:4038
    do_mkdirat+0x264/0x520 fs/namei.c:4061
    __do_sys_mkdirat fs/namei.c:4076 [inline]
    __se_sys_mkdirat fs/namei.c:4074 [inline]
    __x64_sys_mkdirat+0x89/0xa0 fs/namei.c:4074

Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20230506142419.984260-1-tytso@mit.edu
Reported-by: syzbot+6385d7d3065524c5ca6d@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?id=6513f6cb5cd6b5fc9f37e3bb70d273b94be9c34c
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: check iomap type only if ext4_iomap_begin() does not fail

When ext4_iomap_overwrite_begin() calls ext4_iomap_begin() map blocks may
fail for some reason (e.g. memory allocation failure, bare disk write), and
later because "iomap->type ! = IOMAP_MAPPED" triggers WARN_ON(). When ext4
iomap_begin() returns an error, it is normal that the type of iomap->type
may not match the expectation. Therefore, we only determine if iomap->type
is as expected when ext4_iomap_begin() is executed successfully.

Cc: stable@kernel.org
Reported-by: syzbot+08106c4b7d60702dbc14@syzkaller.appspotmail.com
Link: https://lore.kernel.org/all/00000000000015760b05f9b4eee9@google.com
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230505132429.714648-1-libaokun1@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: avoid a potential slab-out-of-bounds in ext4_group_desc_csum

When modifying the block device while it is mounted by the filesystem,
syzbot reported the following:

BUG: KASAN: slab-out-of-bounds in crc16+0x206/0x280 lib/crc16.c:58
Read of size 1 at addr ffff888075f5c0a8 by task syz-executor.2/15586

CPU: 1 PID: 15586 Comm: syz-executor.2 Not tainted 6.2.0-rc5-syzkaller-00205-gc96618275234 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/12/2023
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:88 [inline]
 dump_stack_lvl+0x1b1/0x290 lib/dump_stack.c:106
 print_address_description+0x74/0x340 mm/kasan/report.c:306
 print_report+0x107/0x1f0 mm/kasan/report.c:417
 kasan_report+0xcd/0x100 mm/kasan/report.c:517
 crc16+0x206/0x280 lib/crc16.c:58
 ext4_group_desc_csum+0x81b/0xb20 fs/ext4/super.c:3187
 ext4_group_desc_csum_set+0x195/0x230 fs/ext4/super.c:3210
 ext4_mb_clear_bb fs/ext4/mballoc.c:6027 [inline]
 ext4_free_blocks+0x191a/0x2810 fs/ext4/mballoc.c:6173
 ext4_remove_blocks fs/ext4/extents.c:2527 [inline]
 ext4_ext_rm_leaf fs/ext4/extents.c:2710 [inline]
 ext4_ext_remove_space+0x24ef/0x46a0 fs/ext4/extents.c:2958
 ext4_ext_truncate+0x177/0x220 fs/ext4/extents.c:4416
 ext4_truncate+0xa6a/0xea0 fs/ext4/inode.c:4342
 ext4_setattr+0x10c8/0x1930 fs/ext4/inode.c:5622
 notify_change+0xe50/0x1100 fs/attr.c:482
 do_truncate+0x200/0x2f0 fs/open.c:65
 handle_truncate fs/namei.c:3216 [inline]
 do_open fs/namei.c:3561 [inline]
 path_openat+0x272b/0x2dd0 fs/namei.c:3714
 do_filp_open+0x264/0x4f0 fs/namei.c:3741
 do_sys_openat2+0x124/0x4e0 fs/open.c:1310
 do_sys_open fs/open.c:1326 [inline]
 __do_sys_creat fs/open.c:1402 [inline]
 __se_sys_creat fs/open.c:1396 [inline]
 __x64_sys_creat+0x11f/0x160 fs/open.c:1396
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x3d/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7f72f8a8c0c9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 f1 19 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f72f97e3168 EFLAGS: 00000246 ORIG_RAX: 0000000000000055
RAX: ffffffffffffffda RBX: 00007f72f8bac050 RCX: 00007f72f8a8c0c9
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000020000280
RBP: 00007f72f8ae7ae9 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007ffd165348bf R14: 00007f72f97e3300 R15: 0000000000022000

Replace
le16_to_cpu(sbi->s_es->s_desc_size)
with
sbi->s_desc_size

It reduces ext4's compiled text size, and makes the code more efficient
(we remove an extra indirect reference and a potential byte
swap on big endian systems), and there is no downside. It also avoids the
potential KASAN / syzkaller failure, as a bonus.

Reported-by: syzbot+fc51227e7100c9294894@syzkaller.appspotmail.com
Reported-by: syzbot+8785e41224a3afd04321@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?id=70d28d11ab14bd7938f3e088365252aa923cff42
Link: https://syzkaller.appspot.com/bug?id=b85721b38583ecc6b5e72ff524c67302abbc30f3
Link: https://lore.kernel.org/all/000000000000ece18705f3b20934@google.com/
Fixes: 717d50e4971b ("Ext4: Uninitialized Block Groups")
Cc: stable@vger.kernel.org
Signed-off-by: Tudor Ambarus <tudor.ambarus@linaro.org>
Link: https://lore.kernel.org/r/20230504121525.3275886-1-tudor.ambarus@linaro.org
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: fix data races when using cached status extents

When using cached extent stored in extent status tree in tree->cache_es
another process holding ei->i_es_lock for reading can be racing with us
setting new value of tree->cache_es. If the compiler would decide to
refetch tree->cache_es at an unfortunate moment, it could result in a
bogus in_range() check. Fix the possible race by using READ_ONCE() when
using tree->cache_es only under ei->i_es_lock for reading.

Cc: stable@kernel.org
Reported-by: syzbot+4a03518df1e31b537066@syzkaller.appspotmail.com
Link: https://lore.kernel.org/all/000000000000d3b33905fa0fd4a6@google.com
Suggested-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230504125524.10802-1-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: avoid deadlock in fs reclaim with page writeback

Ext4 has a filesystem wide lock protecting ext4_writepages() calls to
avoid races with switching of journalled data flag or inode format. This
lock can however cause a deadlock like:

CPU0                            CPU1

ext4_writepages()
  percpu_down_read(sbi->s_writepages_rwsem);
                                ext4_change_inode_journal_flag()
                                  percpu_down_write(sbi->s_writepages_rwsem);
                                    - blocks, all readers block from now on
  ext4_do_writepages()
    ext4_init_io_end()
      kmem_cache_zalloc(io_end_cachep, GFP_KERNEL)
        fs_reclaim frees dentry...
          dentry_unlink_inode()
            iput() - last ref =>
              iput_final() - inode dirty =>
                write_inode_now()...
                  ext4_writepages() tries to acquire sbi->s_writepages_rwsem
                    and blocks forever

Make sure we cannot recurse into filesystem reclaim from writeback code
to avoid the deadlock.

Reported-by: syzbot+6898da502aef574c5f8a@syzkaller.appspotmail.com
Link: https://lore.kernel.org/all/0000000000004c66b405fa108e27@google.com
Fixes: c8585c6fcaf2 ("ext4: fix races between changing inode journal mode and ext4_writepages")
CC: stable@vger.kernel.org
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230504124723.20205-1-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: fix invalid free tracking in ext4_xattr_move_to_block()

In ext4_xattr_move_to_block(), the value of the extended attribute
which we need to move to an external block may be allocated by
kvmalloc() if the value is stored in an external inode.  So at the end
of the function the code tried to check if this was the case by
testing entry->e_value_inum.

However, at this point, the pointer to the xattr entry is no longer
valid, because it was removed from the original location where it had
been stored.  So we could end up calling kvfree() on a pointer which
was not allocated by kvmalloc(); or we could also potentially leak
memory by not freeing the buffer when it should be freed.  Fix this by
storing whether it should be freed in a separate variable.

Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20230430160426.581366-1-tytso@mit.edu
Link: https://syzkaller.appspot.com/bug?id=5c2aee8256e30b55ccf57312c16d88417adbd5e1
Link: https://syzkaller.appspot.com/bug?id=41a6b5d4917c0412eb3b3c3c604965bed7d7420b
Reported-by: syzbot+64b645917ce07d89bde5@syzkaller.appspotmail.com
Reported-by: syzbot+0d042627c4f2ad332195@syzkaller.appspotmail.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

ext4: remove a BUG_ON in ext4_mb_release_group_pa()

If a malicious fuzzer overwrites the ext4 superblock while it is
mounted such that the s_first_data_block is set to a very large
number, the calculation of the block group can underflow, and trigger
a BUG_ON check.  Change this to be an ext4_warning so that we don't
crash the kernel.

Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20230430154311.579720-3-tytso@mit.edu
Reported-by: syzbot+e2efa3efc15a1c9e95c3@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?id=69b28112e098b070f639efb356393af3ffec4220
Signed-off-by: Theodore Ts'o <tytso@mit.edu>