From: Christoph Hellwig <hch@lst.de>
To: Carlos Maiolino <cem@kernel.org>
Cc: "Cc : Hans Holmberg" <hans.holmberg@wdc.com>, linux-xfs@vger.kernel.org
Subject: [PATCH 1/2] xfs: add a separate bio_set for spliting GC writes
Date: Tue, 28 Jul 2026 10:06:44 +0200 [thread overview]
Message-ID: <20260728080650.1777545-2-hch@lst.de> (raw)
In-Reply-To: <20260728080650.1777545-1-hch@lst.de>
Allocating the new bio for a split from the same pool as the original
one can deadlock under memory pressure. Add a separate pool for
splitting GC write bios to avoid this.
Fixes: 080d01c41d44 ("xfs: implement zoned garbage collection")
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
fs/xfs/xfs_zone_gc.c | 12 +++++++++++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/xfs_zone_gc.c b/fs/xfs/xfs_zone_gc.c
index f76a09130852..e4f70e024632 100644
--- a/fs/xfs/xfs_zone_gc.c
+++ b/fs/xfs/xfs_zone_gc.c
@@ -130,6 +130,9 @@ struct xfs_zone_gc_data {
/* bioset used to allocate the gc_bios */
struct bio_set bio_set;
+ /* bioset used when writes need to be split to hardware limits */
+ struct bio_set split_bio_set;
+
/*
* Scratchpad to buffer GC data, organized as a ring buffer over
* discontiguous folios. scratch_head is where the buffer is filled,
@@ -221,6 +224,9 @@ xfs_zone_gc_data_alloc(
if (bioset_init(&data->bio_set, 16, offsetof(struct xfs_gc_bio, bio),
BIOSET_NEED_BVECS))
goto out_free_recs;
+ if (bioset_init(&data->split_bio_set, 16,
+ offsetof(struct xfs_gc_bio, bio), 0))
+ goto out_exit_bio_set;
for (i = 0; i < XFS_GC_NR_BUFS; i++) {
data->scratch_folios[i] =
folio_alloc(GFP_KERNEL, get_order(XFS_GC_BUF_SIZE));
@@ -238,6 +244,8 @@ xfs_zone_gc_data_alloc(
out_free_scratch:
while (--i >= 0)
folio_put(data->scratch_folios[i]);
+ bioset_exit(&data->split_bio_set);
+out_exit_bio_set:
bioset_exit(&data->bio_set);
out_free_recs:
kfree(data->iter.recs);
@@ -254,6 +262,7 @@ xfs_zone_gc_data_free(
for (i = 0; i < XFS_GC_NR_BUFS; i++)
folio_put(data->scratch_folios[i]);
+ bioset_exit(&data->split_bio_set);
bioset_exit(&data->bio_set);
kfree(data->iter.recs);
kfree(data);
@@ -810,7 +819,8 @@ xfs_zone_gc_split_write(
data->mp->m_sb.sb_blocksize) >> SECTOR_SHIFT;
split_len = split_sectors << SECTOR_SHIFT;
- split = bio_split(&chunk->bio, split_sectors, GFP_NOFS, &data->bio_set);
+ split = bio_split(&chunk->bio, split_sectors, GFP_NOFS,
+ &data->split_bio_set);
split_chunk = container_of(split, struct xfs_gc_bio, bio);
split_chunk->data = data;
ihold(VFS_I(chunk->ip));
--
2.53.0
next prev parent reply other threads:[~2026-07-28 8:06 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-28 8:06 zoned GC fix Christoph Hellwig
2026-07-28 8:06 ` Christoph Hellwig [this message]
2026-07-28 15:27 ` [PATCH 1/2] xfs: add a separate bio_set for spliting GC writes Darrick J. Wong
2026-07-29 4:07 ` Damien Le Moal
2026-07-28 8:06 ` [PATCH 2/2] xfs: add a comment to describe xfs_gc_bio.victim_rtg Christoph Hellwig
2026-07-28 15:27 ` Darrick J. Wong
2026-07-29 4:05 ` Damien Le Moal
-- strict thread matches above, loose matches on Subject: below --
2026-07-29 13:02 zoned GC fix v2 Christoph Hellwig
2026-07-29 13:02 ` [PATCH 1/2] xfs: add a separate bio_set for spliting GC writes 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=20260728080650.1777545-2-hch@lst.de \
--to=hch@lst.de \
--cc=cem@kernel.org \
--cc=hans.holmberg@wdc.com \
--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