From: Brian Foster <bfoster@redhat.com>
To: Christoph Hellwig <hch@lst.de>
Cc: linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
Dave Chinner <dchinner@redhat.com>
Subject: Re: [PATCH 12/21] xfs: make xfs_writepage_map extent map centric
Date: Mon, 4 Jun 2018 08:26:58 -0400 [thread overview]
Message-ID: <20180604122658.GE110455@bfoster.bfoster> (raw)
In-Reply-To: <20180531180759.21631-13-hch@lst.de>
On Thu, May 31, 2018 at 08:07:50PM +0200, Christoph Hellwig wrote:
> From: Dave Chinner <dchinner@redhat.com>
>
...
>
> Signed-Off-By: Dave Chinner <dchinner@redhat.com>
> [hch: forward port + slight refactoring]
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> fs/xfs/xfs_aops.c | 96 ++++++++++++++++++++++-------------------------
> 1 file changed, 44 insertions(+), 52 deletions(-)
>
> diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
> index 6f57a785f4d3..a0ecfd63b858 100644
> --- a/fs/xfs/xfs_aops.c
> +++ b/fs/xfs/xfs_aops.c
...
> @@ -822,61 +830,46 @@ xfs_writepage_map(
> {
> LIST_HEAD(submit_list);
> struct xfs_ioend *ioend, *next;
> - struct buffer_head *bh, *head;
> + struct buffer_head *bh;
> ssize_t len = i_blocksize(inode);
> uint64_t file_offset; /* file offset of page */
> + unsigned poffset; /* offset into page */
> int error = 0;
> int count = 0;
> - unsigned int new_type;
>
> - bh = head = page_buffers(page);
> + /*
> + * Walk the blocks on the page, and we we run off then end of the
> + * current map or find the current map invalid, grab a new one.
> + * We only use bufferheads here to check per-block state - they no
> + * longer control the iteration through the page. This allows us to
> + * replace the bufferhead with some other state tracking mechanism in
> + * future.
> + */
> file_offset = page_offset(page);
> - do {
> + bh = page_buffers(page);
> + for (poffset = 0;
> + poffset < PAGE_SIZE;
> + poffset += len, file_offset += len, bh = bh->b_this_page) {
> + /* past the range we are writing, so nothing more to write. */
> if (file_offset >= end_offset)
> break;
>
> - /*
> - * set_page_dirty dirties all buffers in a page, independent
> - * of their state. The dirty state however is entirely
> - * meaningless for holes (!mapped && uptodate), so skip
> - * buffers covering holes here.
> - */
> - if (!buffer_mapped(bh) && buffer_uptodate(bh))
> - continue;
> -
> - if (buffer_unwritten(bh))
> - new_type = XFS_IO_UNWRITTEN;
> - else if (buffer_delay(bh))
> - new_type = XFS_IO_DELALLOC;
> - else if (buffer_uptodate(bh))
> - new_type = XFS_IO_OVERWRITE;
> - else {
> + if (!buffer_uptodate(bh)) {
> if (PageUptodate(page))
> ASSERT(buffer_mapped(bh));
> - /*
> - * This buffer is not uptodate and will not be
> - * written to disk.
> - */
> continue;
> }
>
> - /*
> - * If we already have a valid COW mapping keep using it.
> - */
> - if (wpc->io_type == XFS_IO_COW &&
> - xfs_imap_valid(inode, &wpc->imap, file_offset)) {
> - wpc->imap_valid = true;
> - new_type = XFS_IO_COW;
> - }
> -
> - if (wpc->io_type != new_type) {
> - wpc->io_type = new_type;
> - wpc->imap_valid = false;
> - }
> -
We drop the previously lifted check but unless I'm missing something wrt
my previous comments on the imap_valid && io_type == COW case, we still
have that unconditional allocation behavior. I suspect if the previous
patch updated the logic below (and preserved it moving forward while
also perhaps adding to the comment to explain why we must call
xfs_map_blocks() in certain cases for reflink inodes), that would fix up
both of these patches.
Otherwise the rest of the changes here look fine to me.
Brian
> if (wpc->imap_valid)
> wpc->imap_valid = xfs_imap_valid(inode, &wpc->imap,
> file_offset);
> + /*
> + * If we don't have a valid map, now it's time to get a new one
> + * for this offset. This will convert delayed allocations
> + * (including COW ones) into real extents. If we return without
> + * a valid map, it means we landed in a hole and we skip the
> + * block.
> + */
> if (!wpc->imap_valid || xfs_is_reflink_inode(XFS_I(inode))) {
> error = xfs_map_blocks(wpc, inode, file_offset);
> if (error)
> @@ -889,11 +882,10 @@ xfs_writepage_map(
> continue;
>
> lock_buffer(bh);
> - if (wpc->io_type != XFS_IO_OVERWRITE)
> - xfs_map_at_offset(inode, bh, &wpc->imap, file_offset);
> + xfs_map_at_offset(inode, bh, &wpc->imap, file_offset);
> xfs_add_to_ioend(inode, bh, file_offset, wpc, wbc, &submit_list);
> count++;
> - } while (file_offset += len, ((bh = bh->b_this_page) != head));
> + }
>
> ASSERT(wpc->ioend || list_empty(&submit_list));
>
> --
> 2.17.0
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2018-06-04 12:27 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-05-31 18:07 buffered writes without buffer heads in xfs and iomap v5 Christoph Hellwig
2018-05-31 18:07 ` [PATCH 01/21] fs: factor out a __generic_write_end helper Christoph Hellwig
2018-05-31 18:07 ` [PATCH 02/21] iomap: add initial support for writes without buffer heads Christoph Hellwig
2018-05-31 18:07 ` [PATCH 03/21] xfs: simplify xfs_bmap_punch_delalloc_range Christoph Hellwig
2018-05-31 18:07 ` [PATCH 04/21] xfs: simplify xfs_aops_discard_page Christoph Hellwig
2018-05-31 18:07 ` [PATCH 05/21] xfs: move locking into xfs_bmap_punch_delalloc_range Christoph Hellwig
2018-05-31 18:07 ` [PATCH 06/21] xfs: do not set the page uptodate in xfs_writepage_map Christoph Hellwig
2018-05-31 18:07 ` [PATCH 07/21] xfs: don't clear imap_valid for a non-uptodate buffers Christoph Hellwig
2018-05-31 18:07 ` [PATCH 08/21] xfs: don't use XFS_BMAPI_IGSTATE in xfs_map_blocks Christoph Hellwig
2018-06-04 12:25 ` Brian Foster
2018-05-31 18:07 ` [PATCH 09/21] xfs: remove xfs_reflink_trim_irec_to_next_cow Christoph Hellwig
2018-06-04 12:25 ` Brian Foster
2018-05-31 18:07 ` [PATCH 10/21] xfs: remove xfs_map_cow Christoph Hellwig
2018-06-04 12:25 ` Brian Foster
2018-06-13 15:30 ` Christoph Hellwig
2018-05-31 18:07 ` [PATCH 11/21] xfs: rename the offset variable in xfs_writepage_map Christoph Hellwig
2018-06-04 12:25 ` Brian Foster
2018-05-31 18:07 ` [PATCH 12/21] xfs: make xfs_writepage_map extent map centric Christoph Hellwig
2018-06-04 12:26 ` Brian Foster [this message]
2018-05-31 18:07 ` [PATCH 13/21] xfs: remove the now unused XFS_BMAPI_IGSTATE flag Christoph Hellwig
2018-06-04 12:27 ` Brian Foster
2018-05-31 18:07 ` [PATCH 14/21] xfs: remove xfs_reflink_find_cow_mapping Christoph Hellwig
2018-05-31 18:07 ` [PATCH 15/21] xfs: simplify xfs_map_blocks by using xfs_iext_lookup_extent directly Christoph Hellwig
2018-05-31 18:07 ` [PATCH 16/21] xfs: remove the imap_valid flag Christoph Hellwig
2018-05-31 18:07 ` [PATCH 17/21] xfs: don't look at buffer heads in xfs_add_to_ioend Christoph Hellwig
2018-05-31 18:07 ` [PATCH 18/21] xfs: move all writeback buffer_head manipulation into xfs_map_at_offset Christoph Hellwig
2018-05-31 18:07 ` [PATCH 19/21] xfs: remove xfs_start_page_writeback Christoph Hellwig
2018-05-31 18:07 ` [PATCH 20/21] xfs: refactor the tail of xfs_writepage_map Christoph Hellwig
2018-05-31 18:07 ` [PATCH 21/21] xfs: allow writeback on pages without buffer heads Christoph Hellwig
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=20180604122658.GE110455@bfoster.bfoster \
--to=bfoster@redhat.com \
--cc=dchinner@redhat.com \
--cc=hch@lst.de \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).