LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Andreas Gruenbacher <agruenba@redhat.com>
To: Gao Xiang <hsiangkao@linux.alibaba.com>
Cc: linux-erofs@lists.ozlabs.org,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Huang Jianan <huangjianan@oppo.com>,
	Joseph Qi <joseph.qi@linux.alibaba.com>,
	"Darrick J . Wong" <djwong@kernel.org>,
	Christoph Hellwig <hch@lst.de>,
	Matthew Wilcox <willy@infradead.org>
Subject: Re: [PATCH v9] iomap: Support file tail packing
Date: Sun, 1 Aug 2021 14:03:33 +0200	[thread overview]
Message-ID: <CAHc6FU5x3XOTyu8vooReSZ-hacfTdo3cu7wFJRcQrfTH8NkVeg@mail.gmail.com> (raw)
In-Reply-To: <20210727025956.80684-1-hsiangkao@linux.alibaba.com>

On Tue, Jul 27, 2021 at 5:00 AM Gao Xiang <hsiangkao@linux.alibaba.com> wrote:
> The existing inline data support only works for cases where the entire
> file is stored as inline data.  For larger files, EROFS stores the
> initial blocks separately and then can pack a small tail adjacent to the
> inode.  Generalise inline data to allow for tail packing.  Tails may not
> cross a page boundary in memory.
>
> We currently have no filesystems that support tail packing and writing,
> so that case is currently disabled (see iomap_write_begin_inline).
>
> Cc: Darrick J. Wong <djwong@kernel.org>
> Reviewed-by: Christoph Hellwig <hch@lst.de>
> Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
> Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
> ---
> v8: https://lore.kernel.org/r/20210726145734.214295-1-hsiangkao@linux.alibaba.com
> changes since v8:
>  - update the subject to 'iomap: Support file tail packing' as there
>    are clearly a number of ways to make the inline data support more
>    flexible (Matthew);
>
>  - add one extra safety check (Darrick):
>         if (WARN_ON_ONCE(size > iomap->length))
>                 return -EIO;
>
>  fs/iomap/buffered-io.c | 42 ++++++++++++++++++++++++++++++------------
>  fs/iomap/direct-io.c   | 10 ++++++----
>  include/linux/iomap.h  | 18 ++++++++++++++++++
>  3 files changed, 54 insertions(+), 16 deletions(-)
>
> diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
> index 87ccb3438bec..f429b9d87dbe 100644
> --- a/fs/iomap/buffered-io.c
> +++ b/fs/iomap/buffered-io.c
> @@ -205,25 +205,32 @@ struct iomap_readpage_ctx {
>         struct readahead_control *rac;
>  };
>
> -static void
> -iomap_read_inline_data(struct inode *inode, struct page *page,
> +static int iomap_read_inline_data(struct inode *inode, struct page *page,
>                 struct iomap *iomap)
>  {
> -       size_t size = i_size_read(inode);
> +       size_t size = i_size_read(inode) - iomap->offset;
>         void *addr;
>
>         if (PageUptodate(page))
> -               return;
> +               return 0;
>
> -       BUG_ON(page_has_private(page));
> -       BUG_ON(page->index);
> -       BUG_ON(size > PAGE_SIZE - offset_in_page(iomap->inline_data));
> +       /* inline data must start page aligned in the file */
> +       if (WARN_ON_ONCE(offset_in_page(iomap->offset)))
> +               return -EIO;
> +       if (WARN_ON_ONCE(size > PAGE_SIZE -
> +                        offset_in_page(iomap->inline_data)))
> +               return -EIO;
> +       if (WARN_ON_ONCE(size > iomap->length))
> +               return -EIO;
> +       if (WARN_ON_ONCE(page_has_private(page)))
> +               return -EIO;
>
>         addr = kmap_atomic(page);
>         memcpy(addr, iomap->inline_data, size);
>         memset(addr + size, 0, PAGE_SIZE - size);
>         kunmap_atomic(addr);
>         SetPageUptodate(page);
> +       return 0;
>  }
>
>  static inline bool iomap_block_needs_zeroing(struct inode *inode,
> @@ -247,8 +254,10 @@ iomap_readpage_actor(struct inode *inode, loff_t pos, loff_t length, void *data,
>         sector_t sector;
>
>         if (iomap->type == IOMAP_INLINE) {
> -               WARN_ON_ONCE(pos);
> -               iomap_read_inline_data(inode, page, iomap);
> +               int ret = iomap_read_inline_data(inode, page, iomap);
> +
> +               if (ret)
> +                       return ret;
>                 return PAGE_SIZE;
>         }
>
> @@ -589,6 +598,15 @@ __iomap_write_begin(struct inode *inode, loff_t pos, unsigned len, int flags,
>         return 0;
>  }
>
> +static int iomap_write_begin_inline(struct inode *inode,
> +               struct page *page, struct iomap *srcmap)
> +{
> +       /* needs more work for the tailpacking case, disable for now */

Nit: the comma should be a semicolon or period here.

> +       if (WARN_ON_ONCE(srcmap->offset != 0))
> +               return -EIO;
> +       return iomap_read_inline_data(inode, page, srcmap);
> +}
> +
>  static int
>  iomap_write_begin(struct inode *inode, loff_t pos, unsigned len, unsigned flags,
>                 struct page **pagep, struct iomap *iomap, struct iomap *srcmap)
> @@ -618,7 +636,7 @@ iomap_write_begin(struct inode *inode, loff_t pos, unsigned len, unsigned flags,
>         }
>
>         if (srcmap->type == IOMAP_INLINE)
> -               iomap_read_inline_data(inode, page, srcmap);
> +               status = iomap_write_begin_inline(inode, page, srcmap);
>         else if (iomap->flags & IOMAP_F_BUFFER_HEAD)
>                 status = __block_write_begin_int(page, pos, len, NULL, srcmap);
>         else
> @@ -671,11 +689,11 @@ static size_t iomap_write_end_inline(struct inode *inode, struct page *page,
>         void *addr;
>
>         WARN_ON_ONCE(!PageUptodate(page));
> -       BUG_ON(pos + copied > PAGE_SIZE - offset_in_page(iomap->inline_data));
> +       BUG_ON(!iomap_inline_data_valid(iomap));
>
>         flush_dcache_page(page);
>         addr = kmap_atomic(page);
> -       memcpy(iomap->inline_data + pos, addr + pos, copied);
> +       memcpy(iomap_inline_data(iomap, pos), addr + pos, copied);
>         kunmap_atomic(addr);
>
>         mark_inode_dirty(inode);
> diff --git a/fs/iomap/direct-io.c b/fs/iomap/direct-io.c
> index 9398b8c31323..41ccbfc9dc82 100644
> --- a/fs/iomap/direct-io.c
> +++ b/fs/iomap/direct-io.c
> @@ -378,23 +378,25 @@ iomap_dio_inline_actor(struct inode *inode, loff_t pos, loff_t length,
>                 struct iomap_dio *dio, struct iomap *iomap)
>  {
>         struct iov_iter *iter = dio->submit.iter;
> +       void *inline_data = iomap_inline_data(iomap, pos);
>         size_t copied;
>
> -       BUG_ON(pos + length > PAGE_SIZE - offset_in_page(iomap->inline_data));
> +       if (WARN_ON_ONCE(!iomap_inline_data_valid(iomap)))
> +               return -EIO;
>
>         if (dio->flags & IOMAP_DIO_WRITE) {
>                 loff_t size = inode->i_size;
>
>                 if (pos > size)
> -                       memset(iomap->inline_data + size, 0, pos - size);
> -               copied = copy_from_iter(iomap->inline_data + pos, length, iter);
> +                       memset(iomap_inline_data(iomap, size), 0, pos - size);
> +               copied = copy_from_iter(inline_data, length, iter);
>                 if (copied) {
>                         if (pos + copied > size)
>                                 i_size_write(inode, pos + copied);
>                         mark_inode_dirty(inode);
>                 }
>         } else {
> -               copied = copy_to_iter(iomap->inline_data + pos, length, iter);
> +               copied = copy_to_iter(inline_data, length, iter);
>         }
>         dio->size += copied;
>         return copied;
> diff --git a/include/linux/iomap.h b/include/linux/iomap.h
> index 479c1da3e221..b8ec145b2975 100644
> --- a/include/linux/iomap.h
> +++ b/include/linux/iomap.h
> @@ -97,6 +97,24 @@ iomap_sector(struct iomap *iomap, loff_t pos)
>         return (iomap->addr + pos - iomap->offset) >> SECTOR_SHIFT;
>  }
>
> +/*
> + * Returns the inline data pointer for logical offset @pos.
> + */
> +static inline void *iomap_inline_data(struct iomap *iomap, loff_t pos)
> +{
> +       return iomap->inline_data + pos - iomap->offset;
> +}
> +
> +/*
> + * Check if the mapping's length is within the valid range for inline data.
> + * This is used to guard against accessing data beyond the page inline_data
> + * points at.
> + */
> +static inline bool iomap_inline_data_valid(struct iomap *iomap)
> +{
> +       return iomap->length <= PAGE_SIZE - offset_in_page(iomap->inline_data);
> +}
> +
>  /*
>   * When a filesystem sets page_ops in an iomap mapping it returns, page_prepare
>   * and page_done will be called for each page written to.  This only applies to
> --
> 2.24.4
>

Andreas


  parent reply	other threads:[~2021-08-01 12:03 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-27  2:59 Gao Xiang
2021-07-27 15:10 ` Darrick J. Wong
2021-07-27 15:23   ` Andreas Grünbacher
2021-07-27 15:30   ` Re: Gao Xiang
2021-08-01 12:03 ` Andreas Gruenbacher [this message]
2021-08-02 12:15   ` [PATCH v9] iomap: Support file tail packing Gao Xiang
2021-08-02 21:58     ` Darrick J. Wong
2021-08-03  0:17     ` [PATCH v10] iomap: support reading inline data from non-zero pos Gao Xiang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAHc6FU5x3XOTyu8vooReSZ-hacfTdo3cu7wFJRcQrfTH8NkVeg@mail.gmail.com \
    --to=agruenba@redhat.com \
    --cc=djwong@kernel.org \
    --cc=hch@lst.de \
    --cc=hsiangkao@linux.alibaba.com \
    --cc=huangjianan@oppo.com \
    --cc=joseph.qi@linux.alibaba.com \
    --cc=linux-erofs@lists.ozlabs.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=willy@infradead.org \
    --subject='Re: [PATCH v9] iomap: Support file tail packing' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).