From: Hao Xu <[email protected]>
To: [email protected], Jens Axboe <[email protected]>
Cc: Dominique Martinet <[email protected]>,
Pavel Begunkov <[email protected]>,
Christian Brauner <[email protected]>,
Alexander Viro <[email protected]>,
Stefan Roesch <[email protected]>, Clay Harris <[email protected]>,
Dave Chinner <[email protected]>,
"Darrick J . Wong" <[email protected]>,
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
[email protected], [email protected],
Wanpeng Li <[email protected]>
Subject: [PATCH 13/29] xfs: make xfs_trans_alloc() support nowait semantics
Date: Fri, 25 Aug 2023 21:54:15 +0800 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
From: Hao Xu <[email protected]>
There are locks in xfs_trans_alloc(), spot them and apply trylock logic.
Make them return -EAGAIN when it would block. To achieve this, add
nowait parameter for those functions in the path. Besides, add a generic
transaction flag XFS_TRANS_NOWAIT to deliver nowait info.
Signed-off-by: Hao Xu <[email protected]>
---
fs/xfs/libxfs/xfs_shared.h | 2 ++
fs/xfs/xfs_iops.c | 3 ++-
fs/xfs/xfs_trans.c | 21 ++++++++++++++++++---
3 files changed, 22 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_shared.h b/fs/xfs/libxfs/xfs_shared.h
index c4381388c0c1..0ba3d6f53405 100644
--- a/fs/xfs/libxfs/xfs_shared.h
+++ b/fs/xfs/libxfs/xfs_shared.h
@@ -83,6 +83,8 @@ void xfs_log_get_max_trans_res(struct xfs_mount *mp,
* made then this algorithm will eventually find all the space it needs.
*/
#define XFS_TRANS_LOWMODE 0x100 /* allocate in low space mode */
+/* Transaction should follow nowait semantics */
+#define XFS_TRANS_NOWAIT (1u << 9)
/*
* Field values for xfs_trans_mod_sb.
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index 5fa391083de9..47b4fd5f8f5c 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -1054,7 +1054,8 @@ xfs_vn_update_time(
if (nowait)
old_pflags = memalloc_noio_save();
- error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0, &tp);
+ error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0,
+ nowait ? XFS_TRANS_NOWAIT : 0, &tp);
if (error)
goto out;
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 8c0bfc9a33b1..dbec685f4f4a 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -251,6 +251,9 @@ xfs_trans_alloc(
struct xfs_trans *tp;
bool want_retry = true;
int error;
+ bool nowait = flags & XFS_TRANS_NOWAIT;
+ gfp_t gfp_flags = GFP_KERNEL |
+ (nowait ? 0 : __GFP_NOFAIL);
/*
* Allocate the handle before we do our freeze accounting and setting up
@@ -258,9 +261,21 @@ xfs_trans_alloc(
* by doing GFP_KERNEL allocations inside sb_start_intwrite().
*/
retry:
- tp = kmem_cache_zalloc(xfs_trans_cache, GFP_KERNEL | __GFP_NOFAIL);
- if (!(flags & XFS_TRANS_NO_WRITECOUNT))
- sb_start_intwrite(mp->m_super);
+ tp = kmem_cache_zalloc(xfs_trans_cache, gfp_flags);
+ if (!tp)
+ return -EAGAIN;
+ if (!(flags & XFS_TRANS_NO_WRITECOUNT)) {
+ if (nowait) {
+ bool locked = sb_start_intwrite_trylock(mp->m_super);
+
+ if (!locked) {
+ xfs_trans_cancel(tp);
+ return -EAGAIN;
+ }
+ } else {
+ sb_start_intwrite(mp->m_super);
+ }
+ }
xfs_trans_set_context(tp);
/*
--
2.25.1
next prev parent reply other threads:[~2023-08-25 14:02 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-25 13:54 [PATCH RFC v5 00/29] io_uring getdents Hao Xu
2023-08-25 13:54 ` [PATCH 01/29] fs: split off vfs_getdents function of getdents64 syscall Hao Xu
2023-08-25 13:54 ` [PATCH 02/29] xfs: rename XBF_TRYLOCK to XBF_NOWAIT Hao Xu
2023-08-25 21:39 ` Dave Chinner
2023-08-25 13:54 ` [PATCH 03/29] xfs: add NOWAIT semantics for readdir Hao Xu
2023-08-25 13:54 ` [PATCH 04/29] vfs: add nowait flag for struct dir_context Hao Xu
2023-08-25 13:54 ` [PATCH 05/29] vfs: add a vfs helper for io_uring file pos lock Hao Xu
2023-08-25 13:54 ` [PATCH 06/29] vfs: add file_pos_unlock() for io_uring usage Hao Xu
2023-08-25 13:54 ` [PATCH 07/29] vfs: add a nowait parameter for touch_atime() Hao Xu
2023-08-25 13:54 ` [PATCH 08/29] vfs: add nowait parameter for file_accessed() Hao Xu
2023-08-25 13:54 ` [PATCH 09/29] vfs: move file_accessed() to the beginning of iterate_dir() Hao Xu
2023-08-25 13:54 ` [PATCH 10/29] vfs: add S_NOWAIT for nowait time update Hao Xu
2023-08-25 13:54 ` [PATCH 11/29] vfs: trylock inode->i_rwsem in iterate_dir() to support nowait Hao Xu
2023-08-25 13:54 ` [PATCH 12/29] xfs: enforce GFP_NOIO implicitly during nowait time update Hao Xu
2023-08-25 14:20 ` Matthew Wilcox
2023-08-25 13:54 ` Hao Xu [this message]
2023-08-25 13:54 ` [PATCH 14/29] xfs: support nowait for xfs_log_reserve() Hao Xu
2023-08-25 13:54 ` [PATCH 15/29] xfs: don't wait for free space in xlog_grant_head_check() in nowait case Hao Xu
2023-08-25 13:54 ` [PATCH 16/29] xfs: add nowait parameter for xfs_inode_item_init() Hao Xu
2023-08-25 13:54 ` [PATCH 17/29] xfs: make xfs_trans_ijoin() error out -EAGAIN Hao Xu
2023-08-25 13:54 ` [PATCH 18/29] xfs: set XBF_NOWAIT for xfs_buf_read_map if necessary Hao Xu
2023-08-25 13:54 ` [PATCH 19/29] xfs: support nowait memory allocation in _xfs_buf_alloc() Hao Xu
2023-08-25 13:54 ` [PATCH 20/29] xfs: distinguish error type of memory allocation failure for nowait case Hao Xu
2023-08-25 13:54 ` [PATCH 21/29] xfs: return -EAGAIN when bulk memory allocation fails in " Hao Xu
2023-08-25 13:54 ` [PATCH 22/29] xfs: comment page allocation for nowait case in xfs_buf_find_insert() Hao Xu
2023-08-25 14:09 ` Matthew Wilcox
2023-08-25 13:54 ` [PATCH 23/29] xfs: don't print warn info for -EAGAIN error in xfs_buf_get_map() Hao Xu
2023-08-25 13:54 ` [PATCH 24/29] xfs: support nowait for xfs_buf_read_map() Hao Xu
2023-08-25 21:53 ` Dave Chinner
2023-08-25 13:54 ` [PATCH 25/29] xfs: support nowait for xfs_buf_item_init() Hao Xu
2023-08-25 22:16 ` Dave Chinner
2023-08-25 13:54 ` [PATCH 26/29] xfs: return -EAGAIN when nowait meets sync in transaction commit Hao Xu
2023-08-25 21:58 ` Dave Chinner
2023-08-25 13:54 ` [PATCH 27/29] xfs: add a comment for xlog_kvmalloc() Hao Xu
2023-08-25 13:54 ` [PATCH 28/29] xfs: support nowait semantics for xc_ctx_lock in xlog_cil_commit() Hao Xu
2023-08-25 21:59 ` Dave Chinner
2023-08-25 13:54 ` [PATCH 29/29] io_uring: add support for getdents Hao Xu
2023-08-25 15:11 ` [PATCH RFC v5 00/29] io_uring getdents Darrick J. Wong
2023-08-25 22:53 ` Dave Chinner
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 \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/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