From: "Darrick J. Wong" <djwong@kernel.org>
To: sandeen@sandeen.net, djwong@kernel.org
Cc: Christoph Hellwig <hch@lst.de>, linux-xfs@vger.kernel.org
Subject: [PATCH 3/6] mkfs: don't let internal logs bump the root dir inode chunk to AG 1
Date: Thu, 05 May 2022 09:05:25 -0700 [thread overview]
Message-ID: <165176672573.248587.11315519052535516182.stgit@magnolia> (raw)
In-Reply-To: <165176670883.248587.2509972137741301804.stgit@magnolia>
From: Darrick J. Wong <djwong@kernel.org>
Currently, we don't let an internal log consume every last block in an
AG. According to the comment, we're doing this to avoid tripping AGF
verifiers if freeblks==0, but on a modern filesystem this isn't
sufficient to avoid problems because we need to have enough space in the
AG to allocate an aligned root inode chunk, if it should be the case
that the log also ends up in AG 0:
$ truncate -s 6366g /tmp/a ; mkfs.xfs -f /tmp/a -d agcount=3200 -l agnum=0
meta-data=/tmp/a isize=512 agcount=3200, agsize=521503 blks
= sectsz=512 attr=2, projid32bit=1
= crc=1 finobt=1, sparse=1, rmapbt=0
= reflink=1 bigtime=0 inobtcount=0
data = bsize=4096 blocks=1668808704, imaxpct=5
= sunit=0 swidth=0 blks
naming =version 2 bsize=4096 ascii-ci=0, ftype=1
log =internal log bsize=4096 blocks=521492, version=2
= sectsz=512 sunit=0 blks, lazy-count=1
realtime =none extsz=4096 blocks=0, rtextents=0
mkfs.xfs: root inode created in AG 1, not AG 0
Therefore, modify the maximum internal log size calculation to constrain
the maximum internal log size so that the aligned inode chunk allocation
will always succeed.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
mkfs/xfs_mkfs.c | 47 +++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 47 insertions(+)
diff --git a/mkfs/xfs_mkfs.c b/mkfs/xfs_mkfs.c
index eb4d7fa9..0b1fb746 100644
--- a/mkfs/xfs_mkfs.c
+++ b/mkfs/xfs_mkfs.c
@@ -3270,6 +3270,49 @@ validate_log_size(uint64_t logblocks, int blocklog, int min_logblocks)
}
}
+static void
+adjust_ag0_internal_logblocks(
+ struct mkfs_params *cfg,
+ struct xfs_mount *mp,
+ int min_logblocks,
+ int *max_logblocks)
+{
+ int backoff = 0;
+ int ichunk_blocks;
+
+ /*
+ * mkfs will trip over the write verifiers if the log is allocated in
+ * AG 0 and consumes enough space that we cannot allocate a non-sparse
+ * inode chunk for the root directory. The inode allocator requires
+ * that the AG have enough free space for the chunk itself plus enough
+ * to fix up the freelist with aligned blocks if we need to fill the
+ * allocation from the AGFL.
+ */
+ ichunk_blocks = XFS_INODES_PER_CHUNK * cfg->inodesize >> cfg->blocklog;
+ backoff = ichunk_blocks * 4;
+
+ /*
+ * We try to align inode allocations to the data device stripe unit,
+ * so ensure there's enough space to perform an aligned allocation.
+ * The inode geometry structure isn't set up yet, so compute this by
+ * hand.
+ */
+ backoff = max(backoff, cfg->dsunit * 2);
+
+ *max_logblocks -= backoff;
+
+ /* If the specified log size is too big, complain. */
+ if (cli_opt_set(&lopts, L_SIZE) && cfg->logblocks > *max_logblocks) {
+ fprintf(stderr,
+_("internal log size %lld too large, must be less than %d\n"),
+ (long long)cfg->logblocks,
+ *max_logblocks);
+ usage();
+ }
+
+ cfg->logblocks = min(cfg->logblocks, *max_logblocks);
+}
+
static void
calculate_log_size(
struct mkfs_params *cfg,
@@ -3382,6 +3425,10 @@ _("log ag number %lld too large, must be less than %lld\n"),
} else
cfg->logagno = (xfs_agnumber_t)(sbp->sb_agcount / 2);
+ if (cfg->logagno == 0)
+ adjust_ag0_internal_logblocks(cfg, mp, min_logblocks,
+ &max_logblocks);
+
cfg->logstart = XFS_AGB_TO_FSB(mp, cfg->logagno,
libxfs_prealloc_blocks(mp));
next prev parent reply other threads:[~2022-05-05 16:05 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-05 16:05 [PATCHSET 0/6] mkfs: various bug fixes Darrick J. Wong
2022-05-05 16:05 ` [PATCH 1/6] mkfs: fix missing validation of -l size against maximum internal log size Darrick J. Wong
2022-05-05 16:05 ` [PATCH 2/6] mkfs: reduce internal log size when log stripe units are in play Darrick J. Wong
2022-05-05 16:05 ` Darrick J. Wong [this message]
2022-05-05 16:05 ` [PATCH 4/6] mkfs: improve log extent validation Darrick J. Wong
2022-05-05 16:05 ` [PATCH 5/6] mkfs: round log size down if rounding log start up causes overflow Darrick J. Wong
2022-05-10 12:55 ` Christoph Hellwig
2022-05-05 16:05 ` [PATCH 6/6] mkfs: don't trample the gid set in the protofile Darrick J. Wong
2022-05-10 12:55 ` 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=165176672573.248587.11315519052535516182.stgit@magnolia \
--to=djwong@kernel.org \
--cc=hch@lst.de \
--cc=linux-xfs@vger.kernel.org \
--cc=sandeen@sandeen.net \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.