linux-stable/fs
Xu Yang 9ee7a77c15 iomap: fault in smaller chunks for non-large folio mappings
commit 4e527d5841 upstream.

Since commit (5d8edfb900 "iomap: Copy larger chunks from userspace"),
iomap will try to copy in larger chunks than PAGE_SIZE. However, if the
mapping doesn't support large folio, only one page of maximum 4KB will
be created and 4KB data will be writen to pagecache each time. Then,
next 4KB will be handled in next iteration. This will cause potential
write performance problem.

If chunk is 2MB, total 512 pages need to be handled finally. During this
period, fault_in_iov_iter_readable() is called to check iov_iter readable
validity. Since only 4KB will be handled each time, below address space
will be checked over and over again:

start         	end
-
buf,    	buf+2MB
buf+4KB, 	buf+2MB
buf+8KB, 	buf+2MB
...
buf+2044KB 	buf+2MB

Obviously the checking size is wrong since only 4KB will be handled each
time. So this will get a correct chunk to let iomap work well in non-large
folio case.

With this change, the write speed will be stable. Tested on ARM64 device.

Before:

 - dd if=/dev/zero of=/dev/sda bs=400K  count=10485  (334 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=800K  count=5242   (278 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=1600K count=2621   (204 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=2200K count=1906   (170 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=3000K count=1398   (150 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=4500K count=932    (139 MB/s)

After:

 - dd if=/dev/zero of=/dev/sda bs=400K  count=10485  (339 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=800K  count=5242   (330 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=1600K count=2621   (332 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=2200K count=1906   (333 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=3000K count=1398   (333 MB/s)
 - dd if=/dev/zero of=/dev/sda bs=4500K count=932    (333 MB/s)

Fixes: 5d8edfb900 ("iomap: Copy larger chunks from userspace")
Cc: stable@vger.kernel.org
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Xu Yang <xu.yang_2@nxp.com>
Link: https://lore.kernel.org/r/20240521114939.2541461-2-xu.yang_2@nxp.com
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2024-06-16 13:47:40 +02:00
..
9p 9p: add missing locking around taking dentry fid list 2024-06-16 13:47:37 +02:00
adfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
affs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
afs afs: Don't cross .backup mountpoint from backup volume 2024-06-16 13:47:30 +02:00
autofs v6.6-vfs.autofs 2023-08-28 11:39:14 -07:00
befs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
bfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
btrfs btrfs: do not wait for short bulk allocation 2024-05-17 12:02:39 +02:00
cachefiles cachefiles: fix memory leak in cachefiles_add_cache() 2024-03-01 13:35:00 +01:00
ceph ceph: redirty page before returning AOP_WRITEPAGE_ACTIVATE 2024-04-27 17:11:29 +02:00
coda v6.6-vfs.ctime 2023-08-28 09:31:32 -07:00
configfs configfs: convert to ctime accessor functions 2023-07-13 10:28:05 +02:00
cramfs v6.6-vfs.super 2023-08-28 11:04:18 -07:00
crypto
debugfs debugfs: fix automount d_fsdata usage 2024-01-20 11:51:37 +01:00
devpts v6.6-vfs.misc 2023-08-28 10:17:14 -07:00
dlm dlm: fix user space lock decision to copy lvb 2024-06-12 11:11:38 +02:00
ecryptfs ecryptfs: Fix buffer size for tag 66 packet 2024-06-12 11:11:31 +02:00
efivarfs efivarfs: Request at most 512 bytes for variable names 2024-03-06 14:48:41 +00:00
efs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
erofs erofs: avoid allocating DEFLATE streams before mounting 2024-06-16 13:47:31 +02:00
exfat exfat: support handle zero-size directory 2023-11-28 17:19:44 +00:00
exportfs exportfs: remove kernel-doc warnings in exportfs 2023-08-29 17:45:22 -04:00
ext2 quota: Properly annotate i_dquot arrays with __rcu 2024-03-26 18:19:46 -04:00
ext4 ext4: remove the redundant folio_wait_stable() 2024-06-12 11:12:10 +02:00
f2fs f2fs: fix to do sanity check on i_xattr_nid in sanity_check_inode() 2024-06-16 13:47:32 +02:00
fat fat: fix uninitialized field in nostale filehandles 2024-04-03 15:28:20 +02:00
freevxfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
fscache netfs, fscache: Prevent Oops in fscache_put_cache() 2024-01-31 16:19:01 -08:00
fuse fuse: fix leaked ENOSYS error on first statx call 2024-04-27 17:11:42 +02:00
gfs2 kthread: add kthread_stop_put 2024-06-12 11:12:52 +02:00
hfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
hfsplus for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
hostfs hostfs: convert to ctime accessor functions 2023-07-24 10:30:00 +02:00
hpfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
hugetlbfs mm: hugetlb pages should not be reserved by shmat() if SHM_NORESERVE 2024-02-23 09:25:16 +01:00
iomap iomap: fault in smaller chunks for non-large folio mappings 2024-06-16 13:47:40 +02:00
isofs isofs: handle CDs with bad root inode but good Joliet root directory 2024-04-13 13:07:34 +02:00
jbd2 jbd2: fix soft lockup in journal_finish_inode_data_buffers() 2024-01-20 11:51:43 +01:00
jffs2 jffs2: prevent xattr node from overflowing the eraseblock 2024-06-12 11:11:33 +02:00
jfs quota: Properly annotate i_dquot arrays with __rcu 2024-03-26 18:19:46 -04:00
kernfs kernfs: RCU protect kernfs_nodes and avoid kernfs_idr_lock in kernfs_find_and_get_node_by_id() 2024-04-13 13:07:38 +02:00
lockd SUNRPC: Add enum svc_auth_status 2023-08-29 17:45:22 -04:00
minix for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
netfs netfs: Only call folio_start_fscache() one time for each folio 2023-09-18 12:03:46 -07:00
nfs pNFS/filelayout: fixup pNfs allocation modes 2024-06-12 11:12:49 +02:00
nfs_common
nfsd nfsd: hold a lighter-weight client reference over CB_RECALL_ANY 2024-04-10 16:36:01 +02:00
nilfs2 nilfs2: fix out-of-range warning 2024-06-12 11:11:31 +02:00
nls nls: Hide new NLS_UCS2_UTILS 2023-08-31 12:07:34 -05:00
notify fanotify: limit reporting of event with non-decodeable file handles 2023-10-19 16:19:20 +02:00
ntfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
ntfs3 fs/ntfs3: Use variable length array instead of fixed size 2024-06-12 11:12:39 +02:00
ocfs2 quota: Properly annotate i_dquot arrays with __rcu 2024-03-26 18:19:46 -04:00
omfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
openpromfs openpromfs: finish conversion to the new mount API 2024-06-12 11:11:30 +02:00
orangefs Julia Lawall reported this null pointer dereference, this should fix it. 2024-04-13 13:07:35 +02:00
overlayfs ovl: remove upper umask handling from ovl_create_upper() 2024-06-12 11:12:24 +02:00
proc proc: Move fdinfo PTRACE_MODE_READ check into the inode .permission operation 2024-06-16 13:47:33 +02:00
pstore pstore/zone: Add a null pointer check to the psz_kmsg_read 2024-04-13 13:07:31 +02:00
qnx4 for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
qnx6 for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
quota quota: Properly annotate i_dquot arrays with __rcu 2024-03-26 18:19:46 -04:00
ramfs ramfs: convert to ctime accessor functions 2023-07-24 10:30:04 +02:00
reiserfs quota: Properly annotate i_dquot arrays with __rcu 2024-03-26 18:19:46 -04:00
romfs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
smb ksmbd: fix uninitialized symbol 'share' in smb2_tree_connect() 2024-06-12 11:11:30 +02:00
squashfs Squashfs: check the inode number is not the invalid value of zero 2024-05-02 16:32:41 +02:00
sysfs fs: sysfs: Fix reference leak in sysfs_break_active_protection() 2024-04-27 17:11:41 +02:00
sysv sysv: don't call sb_bread() with pointers_lock held 2024-04-13 13:07:34 +02:00
tracefs eventfs: Have "events" directory get permissions from its parent 2024-06-12 11:12:35 +02:00
ubifs ubifs: Set page uptodate in the correct place 2024-04-03 15:28:20 +02:00
udf udf: Convert udf_expand_file_adinicb() to use a folio 2024-06-12 11:12:23 +02:00
ufs for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
unicode
vboxsf vboxsf: explicitly deny setlease attempts 2024-05-17 12:02:13 +02:00
verity fsverity: use register_sysctl_init() to avoid kmemleak warning 2024-06-16 13:47:33 +02:00
xfs xfs: remove conditional building of rt geometry validator functions 2024-04-03 15:28:49 +02:00
zonefs zonefs: Improve error handling 2024-02-23 09:25:13 +01:00
aio.c fs/aio: Check IOCB_AIO_RW before the struct aio_kiocb conversion 2024-04-03 15:28:44 +02:00
anon_inodes.c
attr.c v6.6-vfs.misc 2023-08-28 10:17:14 -07:00
bad_inode.c fs: drop the timespec64 argument from update_time 2023-08-11 09:04:57 +02:00
binfmt_elf.c Merge branch 'expand-stack' 2023-06-28 20:35:21 -07:00
binfmt_elf_fdpic.c fs: binfmt_elf_efpic: fix personality for ELF-FDPIC 2023-09-29 17:20:45 -07:00
binfmt_elf_test.c
binfmt_flat.c
binfmt_misc.c fs: convert to ctime accessor functions 2023-07-13 10:28:04 +02:00
binfmt_script.c
buffer.c iomap: add a workaround for racy i_size updates on block devices 2023-09-25 08:55:00 -07:00
char_dev.c
compat_binfmt_elf.c
coredump.c
d_path.c
dax.c mm: convert DAX lock/unlock page to lock/unlock folio 2024-01-10 17:16:53 +01:00
dcache.c fast_dput(): handle underflows gracefully 2024-02-05 20:14:26 +00:00
direct-io.c - Yosry Ahmed brought back some cgroup v1 stats in OOM logs. 2023-06-28 10:28:11 -07:00
drop_caches.c fs: drop_caches: draining pages before dropping caches 2023-08-18 10:12:11 -07:00
eventfd.c eventfd: prevent underflow for eventfd semaphores 2023-07-11 11:41:34 +02:00
eventpoll.c epoll: be better about file lifetimes 2024-06-12 11:11:30 +02:00
exec.c exec: Fix NOMMU linux_binprm::exec in transfer_args_to_stack() 2024-04-03 15:28:55 +02:00
fcntl.c fs: Fix rw_hint validation 2024-03-26 18:19:17 -04:00
fhandle.c do_sys_name_to_handle(): use kzalloc() to fix kernel-infoleak 2024-03-26 18:19:15 -04:00
file.c v6.6-vfs.misc 2023-08-28 10:17:14 -07:00
file_table.c fs: use __fput_sync in close(2) 2023-08-08 19:36:51 +02:00
filesystems.c
fs-writeback.c writeback, cgroup: switch inodes with dirty timestamps to release dying cgwbs 2023-11-20 11:58:52 +01:00
fs_context.c fs: factor out vfs_parse_monolithic_sep() helper 2023-10-12 18:53:36 +03:00
fs_parser.c
fs_pin.c
fs_struct.c kill do_each_thread() 2023-08-21 13:46:25 -07:00
fs_types.c
fsopen.c fs: add FSCONFIG_CMD_CREATE_EXCL 2023-08-14 18:48:02 +02:00
init.c
inode.c filemap: add a per-mapping stable writes flag 2023-12-03 07:33:03 +01:00
internal.h for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
ioctl.c lsm: new security_file_ioctl_compat() hook 2024-01-31 16:18:54 -08:00
Kconfig for-6.6/block-2023-08-28 2023-08-29 20:21:42 -07:00
Kconfig.binfmt riscv: support the elf-fdpic binfmt loader 2023-08-23 14:17:43 -07:00
kernel_read_file.c fs: Fix kernel-doc warnings 2023-08-19 12:12:12 +02:00
libfs.c fs: new accessor methods for atime and mtime 2024-01-05 15:19:40 +01:00
locks.c NFSD 6.6 Release Notes 2023-08-31 15:32:18 -07:00
Makefile fs: add CONFIG_BUFFER_HEAD 2023-08-02 09:13:09 -06:00
mbcache.c
mnt_idmapping.c
mount.h
mpage.c
namei.c rename(): fix the locking of subdirectories 2024-01-31 16:18:57 -08:00
namespace.c fs: relax mount_setattr() permission checks 2024-02-23 09:25:15 +01:00
nsfs.c fs: convert to ctime accessor functions 2023-07-13 10:28:04 +02:00
open.c cred: get rid of CONFIG_DEBUG_CREDENTIALS 2023-12-20 17:01:51 +01:00
pipe.c fs/pipe: Fix lockdep false-positive in watchqueue pipe_write() 2024-04-10 16:35:57 +02:00
pnode.c
pnode.h
posix_acl.c fs: convert to ctime accessor functions 2023-07-13 10:28:04 +02:00
proc_namespace.c
read_write.c fs: Fix one kernel-doc comment 2023-08-15 08:32:45 +02:00
readdir.c vfs: get rid of old '->iterate' directory operation 2023-08-06 15:08:35 +02:00
remap_range.c
select.c fs/select: rework stack allocation hack for clang 2024-03-26 18:19:17 -04:00
seq_file.c
signalfd.c
splice.c - Some swap cleanups from Ma Wupeng ("fix WARN_ON in add_to_avail_list") 2023-08-29 14:25:26 -07:00
stack.c fs: convert to ctime accessor functions 2023-07-13 10:28:04 +02:00
stat.c fs: Pass AT_GETATTR_NOSEC flag to getattr interface function 2023-12-03 07:33:03 +01:00
statfs.c
super.c fs: export sget_dev() 2023-08-31 12:47:15 +02:00
sync.c
sysctls.c
timerfd.c
userfaultfd.c mm/userfaultfd: reset ptes when close() for wr-protected ones 2024-05-17 12:02:36 +02:00
utimes.c
xattr.c tmpfs,xattr: GFP_KERNEL_ACCOUNT for simple xattrs 2023-08-22 10:57:46 +02:00