summaryrefslogtreecommitdiff
path: root/fs
diff options
context:
space:
mode:
Diffstat (limited to 'fs')
-rw-r--r--fs/Kconfig2
-rw-r--r--fs/afs/dir.c1
-rw-r--r--fs/afs/file.c1
-rw-r--r--fs/afs/fs_operation.c7
-rw-r--r--fs/afs/inode.c1
-rw-r--r--fs/afs/internal.h1
-rw-r--r--fs/afs/mntpt.c1
-rw-r--r--fs/afs/xattr.c31
-rw-r--r--fs/binfmt_misc.c29
-rw-r--r--fs/block_dev.c17
-rw-r--r--fs/btrfs/block-group.c33
-rw-r--r--fs/btrfs/block-group.h9
-rw-r--r--fs/btrfs/compression.c68
-rw-r--r--fs/btrfs/ctree.h5
-rw-r--r--fs/btrfs/delayed-inode.c5
-rw-r--r--fs/btrfs/extent_io.c23
-rw-r--r--fs/btrfs/file.c5
-rw-r--r--fs/btrfs/free-space-cache.c21
-rw-r--r--fs/btrfs/inode.c46
-rw-r--r--fs/btrfs/ioctl.c19
-rw-r--r--fs/btrfs/lzo.c4
-rw-r--r--fs/btrfs/qgroup.c8
-rw-r--r--fs/btrfs/qgroup.h2
-rw-r--r--fs/btrfs/raid56.c31
-rw-r--r--fs/btrfs/ref-verify.c4
-rw-r--r--fs/btrfs/reflink.c24
-rw-r--r--fs/btrfs/scrub.c11
-rw-r--r--fs/btrfs/send.c7
-rw-r--r--fs/btrfs/super.c4
-rw-r--r--fs/btrfs/tree-checker.c16
-rw-r--r--fs/btrfs/tree-log.c3
-rw-r--r--fs/btrfs/xattr.c31
-rw-r--r--fs/btrfs/zlib.c5
-rw-r--r--fs/btrfs/zoned.c4
-rw-r--r--fs/btrfs/zstd.c6
-rw-r--r--fs/cifs/cifs_debug.c2
-rw-r--r--fs/cifs/cifsfs.c2
-rw-r--r--fs/cifs/cifsglob.h19
-rw-r--r--fs/cifs/connect.c14
-rw-r--r--fs/cifs/sess.c1
-rw-r--r--fs/cifs/smb2inode.c1
-rw-r--r--fs/cifs/smb2misc.c10
-rw-r--r--fs/cifs/smb2ops.c10
-rw-r--r--fs/cifs/smb2pdu.c6
-rw-r--r--fs/cifs/smb2proto.h3
-rw-r--r--fs/cifs/transport.c4
-rw-r--r--fs/configfs/file.c6
-rw-r--r--fs/crypto/bio.c6
-rw-r--r--fs/erofs/data.c28
-rw-r--r--fs/erofs/zdata.c2
-rw-r--r--fs/ext4/page-io.c2
-rw-r--r--fs/f2fs/checkpoint.c2
-rw-r--r--fs/f2fs/data.c4
-rw-r--r--fs/f2fs/segment.c2
-rw-r--r--fs/f2fs/segment.h4
-rw-r--r--fs/f2fs/super.c4
-rw-r--r--fs/gfs2/log.c6
-rw-r--r--fs/gfs2/lops.c2
-rw-r--r--fs/gfs2/ops_fstype.c4
-rw-r--r--fs/gfs2/super.c10
-rw-r--r--fs/gfs2/super.h2
-rw-r--r--fs/gfs2/trans.c2
-rw-r--r--fs/gfs2/util.c17
-rw-r--r--fs/io-wq.c268
-rw-r--r--fs/io-wq.h3
-rw-r--r--fs/io_uring.c1149
-rw-r--r--fs/iomap/buffered-io.c4
-rw-r--r--fs/iomap/direct-io.c4
-rw-r--r--fs/locks.c3
-rw-r--r--fs/mpage.c2
-rw-r--r--fs/nfs/Kconfig2
-rw-r--r--fs/nfs/dir.c58
-rw-r--r--fs/nfs/inode.c7
-rw-r--r--fs/nfs/internal.h3
-rw-r--r--fs/nfs/nfs3xdr.c3
-rw-r--r--fs/nfs/nfs42proc.c12
-rw-r--r--fs/nfs/nfs4proc.c33
-rw-r--r--fs/nfs/unlink.c6
-rw-r--r--fs/nfs/write.c8
-rw-r--r--fs/nfsd/Kconfig1
-rw-r--r--fs/nfsd/filecache.c2
-rw-r--r--fs/nfsd/nfs4callback.c1
-rw-r--r--fs/nfsd/nfs4proc.c2
-rw-r--r--fs/nfsd/nfs4state.c55
-rw-r--r--fs/nilfs2/segbuf.c2
-rw-r--r--fs/pnode.h2
-rw-r--r--fs/proc/task_mmu.c2
-rw-r--r--fs/pstore/inode.c2
-rw-r--r--fs/pstore/ram_core.c2
-rw-r--r--fs/squashfs/block.c2
-rw-r--r--fs/zonefs/super.c2
91 files changed, 1195 insertions, 1105 deletions
diff --git a/fs/Kconfig b/fs/Kconfig
index 462253ae483a..a55bda4233bb 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -203,7 +203,7 @@ config TMPFS_XATTR
config TMPFS_INODE64
bool "Use 64-bit ino_t by default in tmpfs"
- depends on TMPFS && 64BIT && !(S390 || ALPHA)
+ depends on TMPFS && 64BIT
default n
help
tmpfs has historically used only inode numbers as wide as an unsigned
diff --git a/fs/afs/dir.c b/fs/afs/dir.c
index 714fcca9af99..17548c1faf02 100644
--- a/fs/afs/dir.c
+++ b/fs/afs/dir.c
@@ -70,7 +70,6 @@ const struct inode_operations afs_dir_inode_operations = {
.permission = afs_permission,
.getattr = afs_getattr,
.setattr = afs_setattr,
- .listxattr = afs_listxattr,
};
const struct address_space_operations afs_dir_aops = {
diff --git a/fs/afs/file.c b/fs/afs/file.c
index 85f5adf21aa0..960b64268623 100644
--- a/fs/afs/file.c
+++ b/fs/afs/file.c
@@ -43,7 +43,6 @@ const struct inode_operations afs_file_inode_operations = {
.getattr = afs_getattr,
.setattr = afs_setattr,
.permission = afs_permission,
- .listxattr = afs_listxattr,
};
const struct address_space_operations afs_fs_aops = {
diff --git a/fs/afs/fs_operation.c b/fs/afs/fs_operation.c
index 97cab12b0a6c..71c58723763d 100644
--- a/fs/afs/fs_operation.c
+++ b/fs/afs/fs_operation.c
@@ -181,10 +181,13 @@ void afs_wait_for_operation(struct afs_operation *op)
if (test_bit(AFS_SERVER_FL_IS_YFS, &op->server->flags) &&
op->ops->issue_yfs_rpc)
op->ops->issue_yfs_rpc(op);
- else
+ else if (op->ops->issue_afs_rpc)
op->ops->issue_afs_rpc(op);
+ else
+ op->ac.error = -ENOTSUPP;
- op->error = afs_wait_for_call_to_complete(op->call, &op->ac);
+ if (op->call)
+ op->error = afs_wait_for_call_to_complete(op->call, &op->ac);
}
switch (op->error) {
diff --git a/fs/afs/inode.c b/fs/afs/inode.c
index 1156b2df28d3..12be88716e4c 100644
--- a/fs/afs/inode.c
+++ b/fs/afs/inode.c
@@ -27,7 +27,6 @@
static const struct inode_operations afs_symlink_inode_operations = {
.get_link = page_get_link,
- .listxattr = afs_listxattr,
};
static noinline void dump_vnode(struct afs_vnode *vnode, struct afs_vnode *parent_vnode)
diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index b626e38e9ab5..1627b1872812 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -1509,7 +1509,6 @@ extern int afs_launder_page(struct page *);
* xattr.c
*/
extern const struct xattr_handler *afs_xattr_handlers[];
-extern ssize_t afs_listxattr(struct dentry *, char *, size_t);
/*
* yfsclient.c
diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c
index 052dab2f5c03..bbb2c210d139 100644
--- a/fs/afs/mntpt.c
+++ b/fs/afs/mntpt.c
@@ -32,7 +32,6 @@ const struct inode_operations afs_mntpt_inode_operations = {
.lookup = afs_mntpt_lookup,
.readlink = page_readlink,
.getattr = afs_getattr,
- .listxattr = afs_listxattr,
};
const struct inode_operations afs_autocell_inode_operations = {
diff --git a/fs/afs/xattr.c b/fs/afs/xattr.c
index c629caae5002..7751b0b3f81d 100644
--- a/fs/afs/xattr.c
+++ b/fs/afs/xattr.c
@@ -11,29 +11,6 @@
#include <linux/xattr.h>
#include "internal.h"
-static const char afs_xattr_list[] =
- "afs.acl\0"
- "afs.cell\0"
- "afs.fid\0"
- "afs.volume\0"
- "afs.yfs.acl\0"
- "afs.yfs.acl_inherited\0"
- "afs.yfs.acl_num_cleaned\0"
- "afs.yfs.vol_acl";
-
-/*
- * Retrieve a list of the supported xattrs.
- */
-ssize_t afs_listxattr(struct dentry *dentry, char *buffer, size_t size)
-{
- if (size == 0)
- return sizeof(afs_xattr_list);
- if (size < sizeof(afs_xattr_list))
- return -ERANGE;
- memcpy(buffer, afs_xattr_list, sizeof(afs_xattr_list));
- return sizeof(afs_xattr_list);
-}
-
/*
* Deal with the result of a successful fetch ACL operation.
*/
@@ -231,6 +208,8 @@ static int afs_xattr_get_yfs(const struct xattr_handler *handler,
else
ret = -ERANGE;
}
+ } else if (ret == -ENOTSUPP) {
+ ret = -ENODATA;
}
error_yacl:
@@ -256,6 +235,7 @@ static int afs_xattr_set_yfs(const struct xattr_handler *handler,
{
struct afs_operation *op;
struct afs_vnode *vnode = AFS_FS_I(inode);
+ int ret;
if (flags == XATTR_CREATE ||
strcmp(name, "acl") != 0)
@@ -270,7 +250,10 @@ static int afs_xattr_set_yfs(const struct xattr_handler *handler,
return afs_put_operation(op);
op->ops = &yfs_store_opaque_acl2_operation;
- return afs_do_sync_operation(op);
+ ret = afs_do_sync_operation(op);
+ if (ret == -ENOTSUPP)
+ ret = -ENODATA;
+ return ret;
}
static const struct xattr_handler afs_xattr_yfs_handler = {
diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
index c457334de43f..e1eae7ea823a 100644
--- a/fs/binfmt_misc.c
+++ b/fs/binfmt_misc.c
@@ -649,12 +649,24 @@ static ssize_t bm_register_write(struct file *file, const char __user *buffer,
struct super_block *sb = file_inode(file)->i_sb;
struct dentry *root = sb->s_root, *dentry;
int err = 0;
+ struct file *f = NULL;
e = create_entry(buffer, count);
if (IS_ERR(e))
return PTR_ERR(e);
+ if (e->flags & MISC_FMT_OPEN_FILE) {
+ f = open_exec(e->interpreter);
+ if (IS_ERR(f)) {
+ pr_notice("register: failed to install interpreter file %s\n",
+ e->interpreter);
+ kfree(e);
+ return PTR_ERR(f);
+ }
+ e->interp_file = f;
+ }
+
inode_lock(d_inode(root));
dentry = lookup_one_len(e->name, root, strlen(e->name));
err = PTR_ERR(dentry);
@@ -678,21 +690,6 @@ static ssize_t bm_register_write(struct file *file, const char __user *buffer,
goto out2;
}
- if (e->flags & MISC_FMT_OPEN_FILE) {
- struct file *f;
-
- f = open_exec(e->interpreter);
- if (IS_ERR(f)) {
- err = PTR_ERR(f);
- pr_notice("register: failed to install interpreter file %s\n", e->interpreter);
- simple_release_fs(&bm_mnt, &entry_count);
- iput(inode);
- inode = NULL;
- goto out2;
- }
- e->interp_file = f;
- }
-
e->dentry = dget(dentry);
inode->i_private = e;
inode->i_fop = &bm_entry_operations;
@@ -709,6 +706,8 @@ out:
inode_unlock(d_inode(root));
if (err) {
+ if (f)
+ filp_close(f, NULL);
kfree(e);
return err;
}
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 4aa1f88d5bf8..92ed7d5df677 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -118,13 +118,22 @@ int truncate_bdev_range(struct block_device *bdev, fmode_t mode,
if (!(mode & FMODE_EXCL)) {
int err = bd_prepare_to_claim(bdev, truncate_bdev_range);
if (err)
- return err;
+ goto invalidate;
}
truncate_inode_pages_range(bdev->bd_inode->i_mapping, lstart, lend);
if (!(mode & FMODE_EXCL))
bd_abort_claiming(bdev, truncate_bdev_range);
return 0;
+
+invalidate:
+ /*
+ * Someone else has handle exclusively open. Try invalidating instead.
+ * The 'end' argument is inclusive so the rounding is safe.
+ */
+ return invalidate_inode_pages2_range(bdev->bd_inode->i_mapping,
+ lstart >> PAGE_SHIFT,
+ lend >> PAGE_SHIFT);
}
static void set_init_blocksize(struct block_device *bdev)
@@ -423,7 +432,7 @@ static ssize_t __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter,
dio->size += bio->bi_iter.bi_size;
pos += bio->bi_iter.bi_size;
- nr_pages = bio_iov_vecs_to_alloc(iter, BIO_MAX_PAGES);
+ nr_pages = bio_iov_vecs_to_alloc(iter, BIO_MAX_VECS);
if (!nr_pages) {
bool polled = false;
@@ -491,8 +500,8 @@ blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
if (!iov_iter_count(iter))
return 0;
- nr_pages = bio_iov_vecs_to_alloc(iter, BIO_MAX_PAGES + 1);
- if (is_sync_kiocb(iocb) && nr_pages <= BIO_MAX_PAGES)
+ nr_pages = bio_iov_vecs_to_alloc(iter, BIO_MAX_VECS + 1);
+ if (is_sync_kiocb(iocb) && nr_pages <= BIO_MAX_VECS)
return __blkdev_direct_IO_simple(iocb, iter, nr_pages);
return __blkdev_direct_IO(iocb, iter, bio_max_segs(nr_pages));
diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c
index 5064be59dac5..744b99ddc28c 100644
--- a/fs/btrfs/block-group.c
+++ b/fs/btrfs/block-group.c
@@ -1162,6 +1162,11 @@ static int inc_block_group_ro(struct btrfs_block_group *cache, int force)
spin_lock(&sinfo->lock);
spin_lock(&cache->lock);
+ if (cache->swap_extents) {
+ ret = -ETXTBSY;
+ goto out;
+ }
+
if (cache->ro) {
cache->ro++;
ret = 0;
@@ -2307,7 +2312,7 @@ again:
}
ret = inc_block_group_ro(cache, 0);
- if (!do_chunk_alloc)
+ if (!do_chunk_alloc || ret == -ETXTBSY)
goto unlock_out;
if (!ret)
goto out;
@@ -2316,6 +2321,8 @@ again:
if (ret < 0)
goto out;
ret = inc_block_group_ro(cache, 0);
+ if (ret == -ETXTBSY)
+ goto unlock_out;
out:
if (cache->flags & BTRFS_BLOCK_GROUP_SYSTEM) {
alloc_flags = btrfs_get_alloc_profile(fs_info, cache->flags);
@@ -3406,6 +3413,7 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
ASSERT(list_empty(&block_group->io_list));
ASSERT(list_empty(&block_group->bg_list));
ASSERT(refcount_read(&block_group->refs) == 1);
+ ASSERT(block_group->swap_extents == 0);
btrfs_put_block_group(block_group);
spin_lock(&info->block_group_cache_lock);
@@ -3472,3 +3480,26 @@ void btrfs_unfreeze_block_group(struct btrfs_block_group *block_group)
__btrfs_remove_free_space_cache(block_group->free_space_ctl);
}
}
+
+bool btrfs_inc_block_group_swap_extents(struct btrfs_block_group *bg)
+{
+ bool ret = true;
+
+ spin_lock(&bg->lock);
+ if (bg->ro)
+ ret = false;
+ else
+ bg->swap_extents++;
+ spin_unlock(&bg->lock);
+
+ return ret;
+}
+
+void btrfs_dec_block_group_swap_extents(struct btrfs_block_group *bg, int amount)
+{
+ spin_lock(&bg->lock);
+ ASSERT(!bg->ro);
+ ASSERT(bg->swap_extents >= amount);
+ bg->swap_extents -= amount;
+ spin_unlock(&bg->lock);
+}
diff --git a/fs/btrfs/block-group.h b/fs/btrfs/block-group.h
index 29678426247d..3ecc3372a5ce 100644
--- a/fs/btrfs/block-group.h
+++ b/fs/btrfs/block-group.h
@@ -186,6 +186,12 @@ struct btrfs_block_group {
/* Flag indicating this block group is placed on a sequential zone */
bool seq_zone;
+ /*
+ * Number of extents in this block group used for swap files.
+ * All accesses protected by the spinlock 'lock'.
+ */
+ int swap_extents;
+
/* Record locked full stripes for RAID5/6 block group */
struct btrfs_full_stripe_locks_tree full_stripe_locks_root;
@@ -312,4 +318,7 @@ static inline int btrfs_block_group_done(struct btrfs_block_group *cache)
void btrfs_freeze_block_group(struct btrfs_block_group *cache);
void btrfs_unfreeze_block_group(struct btrfs_block_group *cache);
+bool btrfs_inc_block_group_swap_extents(struct btrfs_block_group *bg);
+void btrfs_dec_block_group_swap_extents(struct btrfs_block_group *bg, int amount);
+
#endif /* BTRFS_BLOCK_GROUP_H */
diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index 6d203acfdeb3..3f4c832abfed 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -141,6 +141,7 @@ static int check_compressed_csum(struct btrfs_inode *inode, struct bio *bio,
struct btrfs_fs_info *fs_info = inode->root->fs_info;
SHASH_DESC_ON_STACK(shash, fs_info->csum_shash);
const u32 csum_size = fs_info->csum_size;
+ const u32 sectorsize = fs_info->sectorsize;
struct page *page;
unsigned long i;
char *kaddr;
@@ -154,22 +155,34 @@ static int check_compressed_csum(struct btrfs_inode *inode, struct bio *bio,
shash->tfm = fs_info->csum_shash;
for (i = 0; i < cb->nr_pages; i++) {
+ u32 pg_offset;
+ u32 bytes_left = PAGE_SIZE;
page = cb->compressed_pages[i];
- kaddr = kmap_atomic(page);
- crypto_shash_digest(shash, kaddr, PAGE_SIZE, csum);
- kunmap_atomic(kaddr);
-
- if (memcmp(&csum, cb_sum, csum_size)) {
- btrfs_print_data_csum_error(inode, disk_start,
- csum, cb_sum, cb->mirror_num);
- if (btrfs_io_bio(bio)->device)
- btrfs_dev_stat_inc_and_print(
- btrfs_io_bio(bio)->device,
- BTRFS_DEV_STAT_CORRUPTION_ERRS);
- return -EIO;
+ /* Determine the remaining bytes inside the page first */
+ if (i == cb->nr_pages - 1)
+ bytes_left = cb->compressed_len - i * PAGE_SIZE;
+
+ /* Hash through the page sector by sector */
+ for (pg_offset = 0; pg_offset < bytes_left;
+ pg_offset += sectorsize) {
+ kaddr = kmap_atomic(page);
+ crypto_shash_digest(shash, kaddr + pg_offset,
+ sectorsize, csum);
+ kunmap_atomic(kaddr);
+
+ if (memcmp(&csum, cb_sum, csum_size) != 0) {
+ btrfs_print_data_csum_error(inode, disk_start,
+ csum, cb_sum, cb->mirror_num);
+ if (btrfs_io_bio(bio)->device)
+ btrfs_dev_stat_inc_and_print(
+ btrfs_io_bio(bio)->device,
+ BTRFS_DEV_STAT_CORRUPTION_ERRS);
+ return -EIO;
+ }
+ cb_sum += csum_size;
+ disk_start += sectorsize;
}
- cb_sum += csum_size;
}
return 0;
}
@@ -640,7 +653,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
read_lock(&em_tree->lock);
em = lookup_extent_mapping(em_tree,
page_offset(bio_first_page_all(bio)),
- PAGE_SIZE);
+ fs_info->sectorsize);
read_unlock(&em_tree->lock);
if (!em)
return BLK_STS_IOERR;
@@ -698,19 +711,30 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
refcount_set(&cb->pending_bios, 1);
for (pg_index = 0; pg_index < nr_pages; pg_index++) {
+ u32 pg_len = PAGE_SIZE;
int submit = 0;
+ /*
+ * To handle subpage case, we need to make sure the bio only
+ * covers the range we need.
+ *
+ * If we're at the last page, truncate the length to only cover
+ * the remaining part.
+ */
+ if (pg_index == nr_pages - 1)
+ pg_len = min_t(u32, PAGE_SIZE,
+ compressed_len - pg_index * PAGE_SIZE);
+
page = cb->compressed_pages[pg_index];
page->mapping = inode->i_mapping;
page->index = em_start >> PAGE_SHIFT;
if (comp_bio->bi_iter.bi_size)
- submit = btrfs_bio_fits_in_stripe(page, PAGE_SIZE,
+ submit = btrfs_bio_fits_in_stripe(page, pg_len,
comp_bio, 0);
page->mapping = NULL;
- if (submit || bio_add_page(comp_bio, page, PAGE_SIZE, 0) <
- PAGE_SIZE) {
+ if (submit || bio_add_page(comp_bio, page, pg_len, 0) < pg_len) {
unsigned int nr_sectors;
ret = btrfs_bio_wq_end_io(fs_info, comp_bio,
@@ -743,9 +767,9 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
comp_bio->bi_private = cb;
comp_bio->bi_end_io = end_compressed_bio_read;
- bio_add_page(comp_bio, page, PAGE_SIZE, 0);
+ bio_add_page(comp_bio, page, pg_len, 0);
}
- cur_disk_byte += PAGE_SIZE;
+ cur_disk_byte += pg_len;
}
ret = btrfs_bio_wq_end_io(fs_info, comp_bio, BTRFS_WQ_ENDIO_DATA);
@@ -1237,7 +1261,6 @@ int btrfs_decompress_buf2page(const char *buf, unsigned long buf_start,
unsigned long prev_start_byte;
unsigned long working_bytes = total_out - buf_start;
unsigned long bytes;
- char *kaddr;
struct bio_vec bvec = bio_iter_iovec(bio, bio->bi_iter);
/*
@@ -1268,9 +1291,8 @@ int btrfs_decompress_buf2page(const char *buf, unsigned long buf_start,
PAGE_SIZE - (buf_offset % PAGE_SIZE));
bytes = min(bytes, working_bytes);
- kaddr = kmap_atomic(bvec.bv_page);
- memcpy(kaddr + bvec.bv_offset, buf + buf_offset, bytes);
- kunmap_atomic(kaddr);
+ memcpy_to_page(bvec.bv_page, bvec.bv_offset, buf + buf_offset,
+ bytes);
flush_dcache_page(bvec.bv_page);
buf_offset += bytes;
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index bd659354d043..9ae776ab3967 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -524,6 +524,11 @@ struct btrfs_swapfile_pin {
* points to a struct btrfs_device.
*/
bool is_block_group;
+ /*
+ * Only used when 'is_block_group' is true and it is the number of
+ * extents used by a swapfile for this block group ('ptr' field).
+ */
+ int bg_extent_count;
};
bool btrfs_pinned_by_swapfile(struct btrfs_fs_info *fs_info, void *ptr);
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index ec0b50b8c5d6..bf25401c9768 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -627,7 +627,8 @@ static int btrfs_delayed_inode_reserve_metadata(
*/
if (!src_rsv || (!trans->bytes_reserved &&
src_rsv->type != BTRFS_BLOCK_RSV_DELALLOC)) {
- ret = btrfs_qgroup_reserve_meta_prealloc(root, num_bytes, true);
+ ret = btrfs_qgroup_reserve_meta(root, num_bytes,
+ BTRFS_QGROUP_RSV_META_PREALLOC, true);
if (ret < 0)
return ret;
ret = btrfs_block_rsv_add(root, dst_rsv, num_bytes,
@@ -649,7 +650,7 @@ static int btrfs_delayed_inode_reserve_metadata(
btrfs_ino(inode),
num_bytes, 1);
} else {
- btrfs_qgroup_free_meta_prealloc(root, fs_info->nodesize);
+ btrfs_qgroup_free_meta_prealloc(root, num_bytes);
}
return ret;
}
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 4dfb3ead1175..191e358f1322 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -3008,12 +3008,23 @@ readpage_ok:
if (likely(uptodate)) {
loff_t i_size = i_size_read(inode);
pgoff_t end_index = i_size >> PAGE_SHIFT;
- unsigned off;
- /* Zero out the end if this page straddles i_size */
- off = offset_in_page(i_size);
- if (page->index == end_index && off)
- zero_user_segment(page, off, PAGE_SIZE);
+ /*
+ * Zero out the remaining part if this range straddles
+ * i_size.
+ *
+ * Here we should only zero the range inside the bvec,
+ * not touch anything else.
+ *
+ * NOTE: i_size is exclusive while end is inclusive.
+ */
+ if (page->index == end_index && i_size <= end) {
+ u32 zero_start = max(offset_in_page(i_size),
+ offset_in_page(end));
+
+ zero_user_segment(page, zero_start,
+ offset_in_page(end) + 1);
+ }
}
ASSERT(bio_offset + len > bio_offset);
bio_offset += len;
@@ -3048,7 +3059,7 @@ struct bio *btrfs_bio_alloc(u64 first_byte)
{
struct bio *bio;
- bio = bio_alloc_bioset(GFP_NOFS, BIO_MAX_PAGES, &btrfs_bioset);
+ bio = bio_alloc_bioset(GFP_NOFS, BIO_MAX_VECS, &btrfs_bioset);
bio->bi_iter.bi_sector = first_byte >> 9;
btrfs_io_bio_init(btrfs_io_bio(bio));
return bio;
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index bf2c51a9607a..0e155f013839 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -3260,8 +3260,11 @@ reserve_space:
goto out;
ret = btrfs_qgroup_reserve_data(BTRFS_I(inode), &data_reserved,
alloc_start, bytes_to_reserve);
- if (ret)
+ if (ret) {
+ unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart,
+ lockend, &cached_state);
goto out;
+ }
ret = btrfs_prealloc_file_range(inode, mode, alloc_start,
alloc_end - alloc_start,
i_blocksize(inode),
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 5400294bd271..9988decd5717 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -2555,7 +2555,12 @@ static int __btrfs_add_free_space_zoned(struct btrfs_block_group *block_group,
to_unusable = size - to_free;
ctl->free_space += to_free;
- block_group->zone_unusable += to_unusable;
+ /*
+ * If the block group is read-only, we should account freed space into
+ * bytes_readonly.
+ */
+ if (!block_group->ro)
+ block_group->zone_unusable += to_unusable;
spin_unlock(&ctl->tree_lock);
if (!used) {
spin_lock(&block_group->lock);
@@ -2801,8 +2806,10 @@ static void __btrfs_return_cluster_to_free_space(
struct rb_node *node;
spin_lock(&cluster->lock);
- if (cluster->block_group != block_group)
- goto out;
+ if (cluster->block_group != block_group) {
+ spin_unlock(&cluster->lock);
+ return;
+ }
cluster->block_group = NULL;
cluster->window_start = 0;
@@ -2840,8 +2847,6 @@ static void __btrfs_return_cluster_to_free_space(
entry->offset, &entry->offset_index, bitmap);
}
cluster->root = RB_ROOT;
-
-out:
spin_unlock(&cluster->lock);
btrfs_put_block_group(block_group);
}
@@ -3125,8 +3130,6 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group *block_group,
entry->bytes -= bytes;
}
- if (entry->bytes == 0)
- rb_erase(&entry->offset_index, &cluster->root);
break;
}
out:
@@ -3143,7 +3146,10 @@ out:
ctl->free_space -= bytes;
if (!entry->bitmap && !btrfs_free_space_trimmed(entry))
ctl->discardable_bytes[BTRFS_STAT_CURR] -= bytes;
+
+ spin_lock(&cluster->lock);
if (entry->bytes == 0) {
+ rb_erase(&entry->offset_index, &cluster->root);
ctl->free_extents--;
if (entry->bitmap) {
kmem_cache_free(btrfs_free_space_bitmap_cachep,
@@ -3156,6 +3162,7 @@ out:
kmem_cache_free(btrfs_free_space_cachep, entry);
}
+ spin_unlock(&cluster->lock);
spin_unlock(&ctl->tree_lock);
return ret;
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 2e1c282c202d..35bfa0533f23 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1674,9 +1674,6 @@ next_slot:
*/
btrfs_release_path(path);
- /* If extent is RO, we must COW it */
- if (btrfs_extent_readonly(fs_info, disk_bytenr))
- goto out_check;
ret = btrfs_cross_ref_exist(root, ino,
found_key.offset -
extent_offset, disk_bytenr, false);
@@ -1723,6 +1720,7 @@ next_slot:
WARN_ON_ONCE(freespace_inode);
goto out_check;
}
+ /* If the extent's block group is RO, we must COW */
if (!btrfs_inc_nocow_writers(fs_info, disk_bytenr))
goto out_check;
nocow = true;
@@ -6085,7 +6083,7 @@ static int btrfs_dirty_inode(struct inode *inode)
return PTR_ERR(trans);
ret = btrfs_update_inode(trans, root, BTRFS_I(inode));
- if (ret && ret == -ENOSPC) {
+ if (ret && (ret == -ENOSPC || ret == -EDQUOT)) {
/* whoops, lets try again with the full transaction */
btrfs_end_transaction(trans);
trans = btrfs_start_transaction(root, 1);
@@ -10200,6 +10198,7 @@ static int btrfs_add_swapfile_pin(struct inode *inode, void *ptr,
sp->ptr = ptr;
sp->inode = inode;
sp->is_block_group = is_block_group;
+ sp->bg_extent_count = 1;
spin_lock(&fs_info->swapfile_pins_lock);
p = &fs_info->swapfile_pins.rb_node;
@@ -10213,6 +10212,8 @@ static int btrfs_add_swapfile_pin(struct inode *inode, void *ptr,
(sp->ptr == entry->ptr && sp->inode > entry->inode)) {
p = &(*p)->rb_right;
} else {
+ if (is_block_group)
+ entry->bg_extent_count++;
spin_unlock(&fs_info->swapfile_pins_lock);
kfree(sp);
return 1;
@@ -10238,8 +10239,11 @@ static void btrfs_free_swapfile_pins(struct inode *inode)
sp = rb_entry(node, struct btrfs_swapfile_pin, node);
if (sp->inode == inode) {
rb_erase(&sp->node, &fs_info->swapfile_pins);
- if (sp->is_block_group)
+ if (sp->is_block_group) {
+ btrfs_dec_block_group_swap_extents(sp->ptr,
+ sp->bg_extent_count);
btrfs_put_block_group(sp->ptr);
+ }
kfree(sp);
}
node = next;
@@ -10300,7 +10304,8 @@ static int btrfs_swap_activate(struct swap_info_struct *sis, struct file *file,
sector_t *span)
{
struct inode *inode = file_inode(file);
- struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ struct btrfs_fs_info *fs_info = root->fs_info;
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
struct extent_state *cached_state = NULL;
struct extent_map *em = NULL;
@@ -10351,13 +10356,27 @@ static int btrfs_swap_activate(struct swap_info_struct *sis, struct file *file,
"cannot activate swapfile while exclusive operation is running");
return -EBUSY;
}
+
+ /*
+ * Prevent snapshot creation while we are activating the swap file.
+ * We do not want to race with snapshot creation. If snapshot creation
+ * already started before we bumped nr_swapfiles from 0 to 1 and
+ * completes before the first write into the swap file after it is
+ * activated, than that write would fallback to COW.
+ */
+ if (!btrfs_drew_try_write_lock(&root->snapshot_lock)) {
+ btrfs_exclop_finish(fs_info);
+ btrfs_warn(fs_info,
+ "cannot activate swapfile because snapshot creation is in progress");
+ return -EINVAL;
+ }
/*
* Snapshots can create extents which require COW even if NODATACOW is
* set. We use this counter to prevent snapshots. We must increment it
* before walking the extents because we don't want a concurrent
* snapshot to run after we've already checked the extents.
*/
- atomic_inc(&BTRFS_I(inode)->root->nr_swapfiles);
+ atomic_inc(&root->nr_swapfiles);
isize = ALIGN_DOWN(inode->i_size, fs_info->sectorsize);
@@ -10454,6 +10473,17 @@ static int btrfs_swap_activate(struct swap_info_struct *sis, struct file *file,
goto out;
}
+ if (!btrfs_inc_block_group_swap_extents(bg)) {
+ btrfs_warn(fs_info,
+ "block group for swapfile at %llu is read-only%s",
+ bg->start,
+ atomic_read(&fs_info->scrubs_running) ?
+ " (scrub running)" : "");
+ btrfs_put_block_group(bg);
+ ret = -EINVAL;
+ goto out;
+ }
+
ret = btrfs_add_swapfile_pin(inode, bg, true);
if (ret) {
btrfs_put_block_group(bg);
@@ -10492,6 +10522,8 @@ out:
if (ret)
btrfs_swap_deactivate(file);
+ btrfs_drew_write_unlock(&root->snapshot_lock);
+
btrfs_exclop_finish(fs_info);
if (ret)
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 072e77726e94..e8d53fea4c61 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1936,7 +1936,10 @@ static noinline int btrfs_ioctl_snap_create_v2(struct file *file,
if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
readonly = true;
if (vol_args->flags & BTRFS_SUBVOL_QGROUP_INHERIT) {
- if (vol_args->size > PAGE_SIZE) {
+ u64 nums;
+
+ if (vol_args->size < sizeof(*inherit) ||
+ vol_args->size > PAGE_SIZE) {
ret = -EINVAL;
goto free_args;
}
@@ -1945,6 +1948,20 @@ static noinline int btrfs_ioctl_snap_create_v2(struct file *file,
ret = PTR_ERR(inherit);
goto free_args;
}
+
+ if (inherit->num_qgroups > PAGE_SIZE ||
+ inherit->num_ref_copies > PAGE_SIZE ||
+ inherit->num_excl_copies > PAGE_SIZE) {
+ ret = -EINVAL;
+ goto free_inherit;
+ }
+
+ nums = inherit->num_qgroups + 2 * inherit->num_ref_copies +
+ 2 * inherit->num_excl_copies;
+ if (vol_args->size != struct_size(inherit, qgroups, nums)) {
+ ret = -EINVAL;
+ goto free_inherit;
+ }
}
ret = __btrfs_ioctl_snap_create(file, vol_args->name, vol_args->fd,
diff --git a/fs/btrfs/lzo.c b/fs/btrfs/lzo.c
index aa9cd11f4b78..9084a950dc09 100644
--- a/fs/btrfs/lzo.c
+++ b/fs/btrfs/lzo.c
@@ -467,7 +467,7 @@ int lzo_decompress(struct list_head *ws, unsigned char *data_in,
destlen = min_t(unsigned long, destlen, PAGE_SIZE);
bytes = min_t(unsigned long, destlen, out_len - start_byte);
- kaddr = kmap_atomic(dest_page);
+ kaddr = kmap_local_page(dest_page);
memcpy(kaddr, workspace->buf + start_byte, bytes);
/*
@@ -477,7 +477,7 @@ int lzo_decompress(struct list_head *ws, unsigned char *data_in,
*/
if (bytes < destlen)
memset(kaddr+bytes, 0, destlen-bytes);
- kunmap_atomic(kaddr);
+ kunmap_local(kaddr);
out:
return ret;
}
diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index 808370ada888..14ff388fd3bd 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -3841,8 +3841,8 @@ static int sub_root_meta_rsv(struct btrfs_root *root, int num_bytes,
return num_bytes;
}
-static int qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
- enum btrfs_qgroup_rsv_type type, bool enforce)
+int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
+ enum btrfs_qgroup_rsv_type type, bool enforce)
{
struct btrfs_fs_info *fs_info = root->fs_info;
int ret;
@@ -3873,14 +3873,14 @@ int __btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
{
int ret;
- ret = qgroup_reserve_meta(root, num_bytes, type, enforce);
+ ret = btrfs_qgroup_reserve_meta(root, num_bytes, type, enforce);
if (ret <= 0 && ret != -EDQUOT)
return ret;
ret = try_flush_qgroup(root);
if (ret < 0)
return ret;
- return qgroup_reserve_meta(root, num_bytes, type, enforce);
+ return btrfs_qgroup_reserve_meta(root, num_bytes, type, enforce);
}
void btrfs_qgroup_free_meta_all_pertrans(struct btrfs_root *root)
diff --git a/fs/btrfs/qgroup.h b/fs/btrfs/qgroup.h
index 50dea9a2d8fb..7283e4f549af 100644
--- a/fs/btrfs/qgroup.h
+++ b/fs/btrfs/qgroup.h
@@ -361,6 +361,8 @@ int btrfs_qgroup_release_data(struct btrfs_inode *inode, u64 start, u64 len);
int btrfs_qgroup_free_data(struct btrfs_inode *inode,
struct extent_changeset *reserved, u64 start,
u64 len);
+int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
+ enum btrfs_qgroup_rsv_type type, bool enforce);
int __btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
enum btrfs_qgroup_rsv_type type, bool enforce);
/* Reserve metadata space for pertrans and prealloc type */
diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c
index 8ec34ecb6d68..8c31357f08ed 100644
--- a/fs/btrfs/raid56.c
+++ b/fs/btrfs/raid56.c
@@ -249,8 +249,6 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info)
static void cache_rbio_pages(struct btrfs_raid_bio *rbio)
{
int i;
- char *s;
- char *d;
int ret;
ret = alloc_rbio_pages(rbio);
@@ -261,13 +259,7 @@ static void cache_rbio_pages(struct btrfs_raid_bio *rbio)
if (!rbio->bio_pages[i])
continue;
- s = kmap(rbio->bio_pages[i]);
- d = kmap(rbio->stripe_pages[i]);
-
- copy_page(d, s);
-
- kunmap(rbio->bio_pages[i]);
- kunmap(rbio->stripe_pages[i]);
+ copy_highpage(rbio->stripe_pages[i], rbio->bio_pages[i]);
SetPageUptodate(rbio->stripe_pages[i]);
}
set_bit(RBIO_CACHE_READY_BIT, &rbio->flags);
@@ -2359,16 +2351,21 @@ static noinline void finish_parity_scrub(struct btrfs_raid_bio *rbio,
SetPageUptodate(p_page);
if (has_qstripe) {
+ /* RAID6, allocate and map temp space for the Q stripe */
q_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM);
if (!q_page) {
__free_page(p_page);
goto cleanup;
}
SetPageUptodate(q_page);
+ pointers[rbio->real_stripes - 1] = kmap(q_page);
}
atomic_set(&rbio->error, 0);
+ /* Map the parity stripe just once */
+ pointers[nr_data] = kmap(p_page);
+
for_each_set_bit(pagenr, rbio->dbitmap, rbio->stripe_npages) {
struct page *p;
void *parity;
@@ -2378,16 +2375,8 @@ static noinline void finish_parity_scrub(struct btrfs_raid_bio *rbio,
pointers[stripe] = kmap(p);
}
- /* then add the parity stripe */
- pointers[stripe++] = kmap(p_page);
-
if (has_qstripe) {
- /*
- * raid6, add the qstripe and call the
- * library function to fill in our p/q
- */
- pointers[stripe++] = kmap(q_page);
-
+ /* RAID6, call the library function to fill in our P/Q */
raid6_call.gen_syndrome(rbio->real_stripes, PAGE_SIZE,
pointers);
} else {
@@ -2408,12 +2397,14 @@ static noinline void finish_parity_scrub(struct btrfs_raid_bio *rbio,
for (stripe = 0; stripe < nr_data; stripe++)
kunmap(page_in_rbio(rbio, stripe, pagenr, 0));
- kunmap(p_page);
}
+ kunmap(p_page);
__free_page(p_page);
- if (q_page)
+ if (q_page) {
+ kunmap(q_page);
__free_page(q_page);
+ }
writeback:
/*
diff --git a/fs/btrfs/ref-verify.c b/fs/btrfs/ref-verify.c
index 2b490becbe67..8e026de74c44 100644
--- a/fs/btrfs/ref-verify.c
+++ b/fs/btrfs/ref-verify.c
@@ -218,11 +218,11 @@ static void __print_stack_trace(struct btrfs_fs_info *fs_info,
stack_trace_print(ra->trace, ra->trace_len, 2);
}
#else
-static void inline __save_stack_trace(struct ref_action *ra)
+static inline void __save_stack_trace(struct ref_action *ra)
{
}
-static void inline __print_stack_trace(struct btrfs_fs_info *fs_info,
+static inline void __print_stack_trace(struct btrfs_fs_info *fs_info,
struct ref_action *ra)
{
btrfs_err(fs_info, " ref-verify: no stacktrace support");
diff --git a/fs/btrfs/reflink.c b/fs/btrfs/reflink.c
index b24396cf2f99..762881b777b3 100644
--- a/fs/btrfs/reflink.c
+++ b/fs/btrfs/reflink.c
@@ -106,12 +106,8 @@ static int copy_inline_to_page(struct btrfs_inode *inode,
set_bit(BTRFS_INODE_NO_DELALLOC_FLUSH, &inode->runtime_flags);
if (comp_type == BTRFS_COMPRESS_NONE) {
- char *map;
-
- map = kmap(page);
- memcpy(map, data_start, datal);
+ memcpy_to_page(page, 0, data_start, datal);
flush_dcache_page(page);
- kunmap(page);
} else {
ret = btrfs_decompress(comp_type, data_start, page, 0,
inline_size, datal);
@@ -553,6 +549,24 @@ process_slot:
*/
btrfs_release_path(path);
+ /*
+ * When using NO_HOLES and we are cloning a range that covers
+ * only a hole (no extents) into a range beyond the current
+ * i_size, punching a hole in the target range will not create
+ * an extent map defining a hole, because the range starts at or
+ * beyond current i_size. If the file previously had an i_size
+ * greater than the new i_size set by this clone operation, we
+ * need to make sure the next fsync is a full fsync, so that it
+ * detects and logs a hole covering a range from the current
+ * i_size to the new i_size. If the clone range covers extents,
+ * besides a hole, then we know the full sync flag was already
+ * set by previous calls to btrfs_replace_file_extents() that
+ * replaced file extent items.
+ */
+ if (last_dest_end >= i_size_read(inode))
+ set_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
+ &BTRFS_I(inode)->runtime_flags);
+
ret = btrfs_replace_file_extents(inode, path, last_dest_end,
destoff + len - 1, NULL, &trans);
if (ret)
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index 582df11d298a..3d9088eab2fc 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -1428,7 +1428,7 @@ static void scrub_recheck_block_on_raid56(struct btrfs_fs_info *fs_info,
if (!first_page->dev->bdev)
goto out;
- bio = btrfs_io_bio_alloc(BIO_MAX_PAGES);
+ bio = btrfs_io_bio_alloc(BIO_MAX_VECS);
bio_set_dev(bio, first_page->dev->bdev);
for (page_num = 0; page_num < sblock->page_count; page_num++) {
@@ -3767,6 +3767,13 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
* commit_transactions.
*/
ro_set = 0;
+ } else if (ret == -ETXTBSY) {
+ btrfs_warn(fs_info,
+ "skipping scrub of block group %llu due to active swapfile",
+ cache->start);
+ scrub_pause_off(fs_info);
+ ret = 0;
+ goto skip_unfreeze;
} else {
btrfs_warn(fs_info,
"failed setting block group ro: %d", ret);
@@ -3862,7 +3869,7 @@ done:
} else {
spin_unlock(&cache->lock);
}
-
+skip_unfreeze:
btrfs_unfreeze_block_group(cache);
btrfs_put_block_group(cache);
if (ret)
diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index f87878274e9f..8f323859156b 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -4932,7 +4932,6 @@ static int put_file_data(struct send_ctx *sctx, u64 offset, u32 len)
struct btrfs_fs_info *fs_info = root->fs_info;
struct inode *inode;
struct page *page;
- char *addr;
pgoff_t index = offset >> PAGE_SHIFT;
pgoff_t last_index;
unsigned pg_offset = offset_in_page(offset);
@@ -4985,10 +4984,8 @@ static int put_file_data(struct send_ctx *sctx, u64 offset, u32 len)
}
}
- addr = kmap(page);
- memcpy(sctx->send_buf + sctx->send_size, addr + pg_offset,
- cur_len);
- kunmap(page);
+ memcpy_from_page(sctx->send_buf + sctx->send_size, page,
+ pg_offset, cur_len);
unlock_page(page);
put_page(page);
index++;
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index f8435641b912..f7a4ad86adee 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -1918,8 +1918,8 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
btrfs_resize_thread_pool(fs_info,
fs_info->thread_pool_size, old_thread_pool_size);
- if (btrfs_test_opt(fs_info, FREE_SPACE_TREE) !=
- btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE) &&
+ if ((bool)btrfs_test_opt(fs_info, FREE_SPACE_TREE) !=
+ (bool)btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE) &&
(!sb_rdonly(sb) || (*flags & SB_RDONLY))) {
btrfs_warn(fs_info,
"remount supports changing free space tree only from ro to rw");
diff --git a/fs/btrfs/tree-checker.c b/fs/btrfs/tree-checker.c
index 582061c7b547..f4ade821307d 100644
--- a/fs/btrfs/tree-checker.c
+++ b/fs/btrfs/tree-checker.c
@@ -1453,22 +1453,14 @@ static int check_extent_data_ref(struct extent_buffer *leaf,
return -EUCLEAN;
}
for (; ptr < end; ptr += sizeof(*dref)) {
- u64 root_objectid;
- u64 owner;
u64 offset;
- u64 hash;
+ /*
+ * We cannot check the extent_data_ref hash due to possible
+ * overflow from the leaf due to hash collisions.
+ */
dref = (struct btrfs_extent_data_ref *)ptr;
- root_objectid = btrfs_extent_data_ref_root(leaf, dref);
- owner = btrfs_extent_data_ref_objectid(leaf, dref);
offset = btrfs_extent_data_ref_offset(leaf, dref);
- hash = hash_extent_data_ref(root_objectid, owner, offset);
- if (unlikely(hash != key->offset)) {
- extent_err(leaf, slot,
- "invalid extent data ref hash, item has 0x%016llx key has 0x%016llx",
- hash, key->offset);
- return -EUCLEAN;
- }
if (unlikely(!IS_ALIGNED(offset, leaf->fs_info->sectorsize))) {
extent_err(leaf, slot,
"invalid extent data backref offset, have %llu expect aligned to %u",
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index d90695c1ab6c..2f1acc9aea9e 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3174,16 +3174,13 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
root_log_ctx.log_transid = log_root_tree->log_transid;
if (btrfs_is_zoned(fs_info)) {
- mutex_lock(&fs_info->tree_root->log_mutex);
if (!log_root_tree->node) {
ret = btrfs_alloc_log_tree_node(trans, log_root_tree);
if (ret) {
- mutex_unlock(&fs_info->tree_log_mutex);
mutex_unlock(&log_root_tree->log_mutex);
goto out;
}
}
- mutex_unlock(&fs_info->tree_root->log_mutex);
}
/*
diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c
index b025102e435f..8a4514283a4b 100644
--- a/fs/btrfs/xattr.c
+++ b/fs/btrfs/xattr.c
@@ -229,11 +229,33 @@ int btrfs_setxattr_trans(struct inode *inode, const char *name,
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_trans_handle *trans;
+ const bool start_trans = (current->journal_info == NULL);
int ret;
- trans = btrfs_start_transaction(root, 2);
- if (IS_ERR(trans))
- return PTR_ERR(trans);
+ if (start_trans) {
+ /*
+ * 1 unit for inserting/updating/deleting the xattr
+ * 1 unit for the inode item update
+ */
+ trans = btrfs_start_transaction(root, 2);
+ if (IS_ERR(trans))
+ return PTR_ERR(trans);
+ } else {
+ /*
+ * This can happen when smack is enabled and a directory is being
+ * created. It happens through d_instantiate_new(), which calls
+ * smack_d_instantiate(), which in turn calls __vfs_setxattr() to
+ * set the transmute xattr (XATTR_NAME_SMACKTRANSMUTE) on the
+ * inode. We have already reserved space for the xattr and inode
+ * update at btrfs_mkdir(), so just use the transaction handle.
+ * We don't join or start a transaction, as that will reset the
+ * block_rsv of the handle and trigger a warning for the start
+ * case.
+ */
+ ASSERT(strncmp(name, XATTR_SECURITY_PREFIX,
+ XATTR_SECURITY_PREFIX_LEN) == 0);
+ trans = current->journal_info;
+ }
ret = btrfs_setxattr(trans, inode, name, value, size, flags);
if (ret)
@@ -244,7 +266,8 @@ int btrfs_setxattr_trans(struct inode *inode, const char *name,
ret = btrfs_update_inode(trans, root, BTRFS_I(inode));
BUG_ON(ret);
out:
- btrfs_end_transaction(trans);
+ if (start_trans)
+ btrfs_end_transaction(trans);
return ret;
}
diff --git a/fs/btrfs/zlib.c b/fs/btrfs/zlib.c
index 05615a1099db..d524acf7b3e5 100644
--- a/fs/btrfs/zlib.c
+++ b/fs/btrfs/zlib.c
@@ -432,9 +432,8 @@ int zlib_decompress(struct list_head *ws, unsigned char *data_in,
PAGE_SIZE - (buf_offset % PAGE_SIZE));
bytes = min(bytes, bytes_left);
- kaddr = kmap_atomic(dest_page);
- memcpy(kaddr + pg_offset, workspace->buf + buf_offset, bytes);
- kunmap_atomic(kaddr);
+ memcpy_to_page(dest_page, pg_offset,
+ workspace->buf + buf_offset, bytes);
pg_offset += bytes;
bytes_left -= bytes;
diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
index d0eb0c8d6269..1f972b75a9ab 100644
--- a/fs/btrfs/zoned.c
+++ b/fs/btrfs/zoned.c
@@ -269,7 +269,7 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
sector_t sector = 0;
struct blk_zone *zones = NULL;
unsigned int i, nreported = 0, nr_zones;
- unsigned int zone_sectors;
+ sector_t zone_sectors;
char *model, *emulated;
int ret;
@@ -658,7 +658,7 @@ int btrfs_sb_log_location_bdev(struct block_device *bdev, int mirror, int rw,
u64 *bytenr_ret)
{
struct blk_zone zones[BTRFS_NR_SB_LOG_ZONES];
- unsigned int zone_sectors;
+ sector_t zone_sectors;
u32 sb_zone;
int ret;
u8 zone_sectors_shift;
diff --git a/fs/btrfs/zstd.c b/fs/btrfs/zstd.c
index 9a4871636c6c..8e9626d63976 100644
--- a/fs/btrfs/zstd.c
+++ b/fs/btrfs/zstd.c
@@ -688,10 +688,8 @@ int zstd_decompress(struct list_head *ws, unsigned char *data_in,
bytes = min_t(unsigned long, destlen - pg_offset,
workspace->out_buf.size - buf_offset);
- kaddr = kmap_atomic(dest_page);
- memcpy(kaddr + pg_offset, workspace->out_buf.dst + buf_offset,
- bytes);
- kunmap_atomic(kaddr);
+ memcpy_to_page(dest_page, pg_offset,
+ workspace->out_buf.dst + buf_offset, bytes);
pg_offset += bytes;
}
diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c
index 3aedc484e440..88a7958170ee 100644
--- a/fs/cifs/cifs_debug.c
+++ b/fs/cifs/cifs_debug.c
@@ -207,7 +207,7 @@ static int cifs_debug_files_proc_show(struct seq_file *m, void *v)
from_kuid(&init_user_ns, cfile->uid),
cfile->dentry);
#ifdef CONFIG_CIFS_DEBUG2
- seq_printf(m, " 0x%llx\n", cfile->fid.mid);
+ seq_printf(m, " %llu\n", cfile->fid.mid);
#else
seq_printf(m, "\n");
#endif /* CIFS_DEBUG2 */
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index d43e935d2df4..099ad9f3660b 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -290,7 +290,7 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf)
rc = server->ops->queryfs(xid, tcon, cifs_sb, buf);
free_xid(xid);
- return 0;
+ return rc;
}
static long cifs_fallocate(struct file *file, int mode, loff_t off, loff_t len)
diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h
index 3de3c5908a72..31fc8695abd6 100644
--- a/fs/cifs/cifsglob.h
+++ b/fs/cifs/cifsglob.h
@@ -257,7 +257,7 @@ struct smb_version_operations {
/* verify the message */
int (*check_message)(char *, unsigned int, struct TCP_Server_Info *);
bool (*is_oplock_break)(char *, struct TCP_Server_Info *);
- int (*handle_cancelled_mid)(char *, struct TCP_Server_Info *);
+ int (*handle_cancelled_mid)(struct mid_q_entry *, struct TCP_Server_Info *);
void (*downgrade_oplock)(struct TCP_Server_Info *server,
struct cifsInodeInfo *cinode, __u32 oplock,
unsigned int epoch, bool *purge_cache);
@@ -1705,16 +1705,17 @@ static inline bool is_retryable_error(int error)
#define CIFS_NO_RSP_BUF 0x040 /* no response buffer required */
/* Type of request operation */
-#define CIFS_ECHO_OP 0x080 /* echo request */
-#define CIFS_OBREAK_OP 0x0100 /* oplock break request */
-#define CIFS_NEG_OP 0x0200 /* negotiate request */
+#define CIFS_ECHO_OP 0x080 /* echo request */
+#define CIFS_OBREAK_OP 0x0100 /* oplock break request */
+#define CIFS_NEG_OP 0x0200 /* negotiate request */
+#define CIFS_CP_CREATE_CLOSE_OP 0x0400 /* compound create+close request */
/* Lower bitmask values are reserved by others below. */
-#define CIFS_SESS_OP 0x2000 /* session setup request */
-#define CIFS_OP_MASK 0x2380 /* mask request type */
+#define CIFS_SESS_OP 0x2000 /* session setup request */
+#define CIFS_OP_MASK 0x2780 /* mask request type */
-#define CIFS_HAS_CREDITS 0x0400 /* already has credits */
-#define CIFS_TRANSFORM_REQ 0x0800 /* transform request before sending */
-#define CIFS_NO_SRV_RSP 0x1000 /* there is no server response */
+#define CIFS_HAS_CREDITS 0x0400 /* already has credits */
+#define CIFS_TRANSFORM_REQ 0x0800 /* transform request before sending */
+#define CIFS_NO_SRV_RSP 0x1000 /* there is no server response */
/* Security Flags: indicate type of session setup needed */
#define CIFSSEC_MAY_SIGN 0x00001
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 112692300fb6..eec8a2052da2 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -741,7 +741,7 @@ static void clean_demultiplex_info(struct TCP_Server_Info *server)
spin_lock(&GlobalMid_Lock);
list_for_each_safe(tmp, tmp2, &server->pending_mid_q) {
mid_entry = list_entry(tmp, struct mid_q_entry, qhead);
- cifs_dbg(FYI, "Clearing mid 0x%llx\n", mid_entry->mid);
+ cifs_dbg(FYI, "Clearing mid %llu\n", mid_entry->mid);
kref_get(&mid_entry->refcount);
mid_entry->mid_state = MID_SHUTDOWN;
list_move(&mid_entry->qhead, &dispose_list);
@@ -752,7 +752,7 @@ static void clean_demultiplex_info(struct TCP_Server_Info *server)
/* now walk dispose list and issue callbacks */
list_for_each_safe(tmp, tmp2, &dispose_list) {
mid_entry = list_entry(tmp, struct mid_q_entry, qhead);
- cifs_dbg(FYI, "Callback mid 0x%llx\n", mid_entry->mid);
+ cifs_dbg(FYI, "Callback mid %llu\n", mid_entry->mid);
list_del_init(&mid_entry->qhead);
mid_entry->callback(mid_entry);
cifs_mid_q_entry_release(mid_entry);
@@ -1429,6 +1429,11 @@ smbd_connected:
tcp_ses->min_offload = ctx->min_offload;
tcp_ses->tcpStatus = CifsNeedNegotiate;
+ if ((ctx->max_credits < 20) || (ctx->max_credits > 60000))
+ tcp_ses->max_credits = SMB2_MAX_CREDITS_AVAILABLE;
+ else
+ tcp_ses->max_credits = ctx->max_credits;
+
tcp_ses->nr_targets = 1;
tcp_ses->ignore_signature = ctx->ignore_signature;
/* thread spawned, put it on the list */
@@ -2832,11 +2837,6 @@ static int mount_get_conns(struct smb3_fs_context *ctx, struct cifs_sb_info *cif
*nserver = server;
- if ((ctx->max_credits < 20) || (ctx->max_credits > 60000))
- server->max_credits = SMB2_MAX_CREDITS_AVAILABLE;
- else
- server->max_credits = ctx->max_credits;
-
/* get a reference to a SMB session */
ses = cifs_get_smb_ses(server, ctx);
if (IS_ERR(ses)) {
diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c
index 183a3a868d7b..63d517b9f2ff 100644
--- a/fs/cifs/sess.c
+++ b/fs/cifs/sess.c
@@ -230,6 +230,7 @@ cifs_ses_add_channel(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses,
ctx.noautotune = ses->server->noautotune;
ctx.sockopt_tcp_nodelay = ses->server->tcp_nodelay;
ctx.echo_interval = ses->server->echo_interval / HZ;
+ ctx.max_credits = ses->server->max_credits;
/*
* This will be used for encoding/decoding user/domain/pw
diff --git a/fs/cifs/smb2inode.c b/fs/cifs/smb2inode.c
index 1f900b81c34a..a718dc77e604 100644
--- a/fs/cifs/smb2inode.c
+++ b/fs/cifs/smb2inode.c
@@ -358,6 +358,7 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon,
if (cfile)
goto after_close;
/* Close */
+ flags |= CIFS_CP_CREATE_CLOSE_OP;
rqst[num_rqst].rq_iov = &vars->close_iov[0];
rqst[num_rqst].rq_nvec = 1;
rc = SMB2_close_init(tcon, server,
diff --git a/fs/cifs/smb2misc.c b/fs/cifs/smb2misc.c
index 60d4bd1eae2b..b50164e2c88d 100644
--- a/fs/cifs/smb2misc.c
+++ b/fs/cifs/smb2misc.c
@@ -767,7 +767,7 @@ smb2_cancelled_close_fid(struct work_struct *work)
int rc;
if (cancelled->mid)
- cifs_tcon_dbg(VFS, "Close unmatched open for MID:%llx\n",
+ cifs_tcon_dbg(VFS, "Close unmatched open for MID:%llu\n",
cancelled->mid);
else
cifs_tcon_dbg(VFS, "Close interrupted close\n");
@@ -844,14 +844,14 @@ smb2_handle_cancelled_close(struct cifs_tcon *tcon, __u64 persistent_fid,
}
int
-smb2_handle_cancelled_mid(char *buffer, struct TCP_Server_Info *server)
+smb2_handle_cancelled_mid(struct mid_q_entry *mid, struct TCP_Server_Info *server)
{
- struct smb2_sync_hdr *sync_hdr = (struct smb2_sync_hdr *)buffer;
- struct smb2_create_rsp *rsp = (struct smb2_create_rsp *)buffer;
+ struct smb2_sync_hdr *sync_hdr = mid->resp_buf;
+ struct smb2_create_rsp *rsp = mid->resp_buf;
struct cifs_tcon *tcon;
int rc;
- if (sync_hdr->Command != SMB2_CREATE ||
+ if ((mid->optype & CIFS_CP_CREATE_CLOSE_OP) || sync_hdr->Command != SMB2_CREATE ||
sync_hdr->Status != STATUS_SUCCESS)
return 0;
diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c
index f5087295424c..9bae7e8deb09 100644
--- a/fs/cifs/smb2ops.c
+++ b/fs/cifs/smb2ops.c
@@ -1195,7 +1195,7 @@ smb2_set_ea(const unsigned int xid, struct cifs_tcon *tcon,
struct TCP_Server_Info *server = cifs_pick_channel(ses);
__le16 *utf16_path = NULL;
int ea_name_len = strlen(ea_name);
- int flags = 0;
+ int flags = CIFS_CP_CREATE_CLOSE_OP;
int len;
struct smb_rqst rqst[3];
int resp_buftype[3];
@@ -1573,7 +1573,7 @@ smb2_ioctl_query_info(const unsigned int xid,
struct smb_query_info qi;
struct smb_query_info __user *pqi;
int rc = 0;
- int flags = 0;
+ int flags = CIFS_CP_CREATE_CLOSE_OP;
struct smb2_query_info_rsp *qi_rsp = NULL;
struct smb2_ioctl_rsp *io_rsp = NULL;
void *buffer = NULL;
@@ -2577,7 +2577,7 @@ smb2_query_info_compound(const unsigned int xid, struct cifs_tcon *tcon,
{
struct cifs_ses *ses = tcon->ses;
struct TCP_Server_Info *server = cifs_pick_channel(ses);
- int flags = 0;
+ int flags = CIFS_CP_CREATE_CLOSE_OP;
struct smb_rqst rqst[3];
int resp_buftype[3];
struct kvec rsp_iov[3];
@@ -2975,7 +2975,7 @@ smb2_query_symlink(const unsigned int xid, struct cifs_tcon *tcon,
unsigned int sub_offset;
unsigned int print_len;
unsigned int print_offset;
- int flags = 0;
+ int flags = CIFS_CP_CREATE_CLOSE_OP;
struct smb_rqst rqst[3];
int resp_buftype[3];
struct kvec rsp_iov[3];
@@ -3157,7 +3157,7 @@ smb2_query_reparse_tag(const unsigned int xid, struct cifs_tcon *tcon,
struct cifs_open_parms oparms;
struct cifs_fid fid;
struct TCP_Server_Info *server = cifs_pick_channel(tcon->ses);
- int flags = 0;
+ int flags = CIFS_CP_CREATE_CLOSE_OP;
struct smb_rqst rqst[3];
int resp_buftype[3];
struct kvec rsp_iov[3];
diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c
index 4bbb6126b14d..2199a9bfae8f 100644
--- a/fs/cifs/smb2pdu.c
+++ b/fs/cifs/smb2pdu.c
@@ -4041,8 +4041,7 @@ smb2_async_readv(struct cifs_readdata *rdata)
if (rdata->credits.value > 0) {
shdr->CreditCharge = cpu_to_le16(DIV_ROUND_UP(rdata->bytes,
SMB2_MAX_BUFFER_SIZE));
- shdr->CreditRequest =
- cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 1);
+ shdr->CreditRequest = cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 8);
rc = adjust_credits(server, &rdata->credits, rdata->bytes);
if (rc)
@@ -4348,8 +4347,7 @@ smb2_async_writev(struct cifs_writedata *wdata,
if (wdata->credits.value > 0) {
shdr->CreditCharge = cpu_to_le16(DIV_ROUND_UP(wdata->bytes,
SMB2_MAX_BUFFER_SIZE));
- shdr->CreditRequest =
- cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 1);
+ shdr->CreditRequest = cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 8);
rc = adjust_credits(server, &wdata->credits, wdata->bytes);
if (rc)
diff --git a/fs/cifs/smb2proto.h b/fs/cifs/smb2proto.h
index 9565e27681a5..a2eb34a8d9c9 100644
--- a/fs/cifs/smb2proto.h
+++ b/fs/cifs/smb2proto.h
@@ -246,8 +246,7 @@ extern int SMB2_oplock_break(const unsigned int xid, struct cifs_tcon *tcon,
extern int smb2_handle_cancelled_close(struct cifs_tcon *tcon,
__u64 persistent_fid,
__u64 volatile_fid);
-extern int smb2_handle_cancelled_mid(char *buffer,
- struct TCP_Server_Info *server);
+extern int smb2_handle_cancelled_mid(struct mid_q_entry *mid, struct TCP_Server_Info *server);
void smb2_cancelled_close_fid(struct work_struct *work);
extern int SMB2_QFS_info(const unsigned int xid, struct cifs_tcon *tcon,
u64 persistent_file_id, u64 volatile_file_id,
diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c
index e90a1d1380b0..007d99437c77 100644
--- a/fs/cifs/transport.c
+++ b/fs/cifs/transport.c
@@ -101,7 +101,7 @@ static void _cifs_mid_q_entry_release(struct kref *refcount)
if (midEntry->resp_buf && (midEntry->mid_flags & MID_WAIT_CANCELLED) &&
midEntry->mid_state == MID_RESPONSE_RECEIVED &&
server->ops->handle_cancelled_mid)
- server->ops->handle_cancelled_mid(midEntry->resp_buf, server);
+ server->ops->handle_cancelled_mid(midEntry, server);
midEntry->mid_state = MID_FREE;
atomic_dec(&midCount);
@@ -1207,7 +1207,7 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses,
}
if (rc != 0) {
for (; i < num_rqst; i++) {
- cifs_server_dbg(VFS, "Cancelling wait for mid %llu cmd: %d\n",
+ cifs_server_dbg(FYI, "Cancelling wait for mid %llu cmd: %d\n",
midQ[i]->mid, le16_to_cpu(midQ[i]->command));
send_cancel(server, &rqst[i], midQ[i]);
spin_lock(&GlobalMid_Lock);
diff --git a/fs/configfs/file.c b/fs/configfs/file.c
index 1f0270229d7b..da8351d1e455 100644
--- a/fs/configfs/file.c
+++ b/fs/configfs/file.c
@@ -378,7 +378,7 @@ static int __configfs_open_file(struct inode *inode, struct file *file, int type
attr = to_attr(dentry);
if (!attr)
- goto out_put_item;
+ goto out_free_buffer;
if (type & CONFIGFS_ITEM_BIN_ATTR) {
buffer->bin_attr = to_bin_attr(dentry);
@@ -391,7 +391,7 @@ static int __configfs_open_file(struct inode *inode, struct file *file, int type
/* Grab the module reference for this attribute if we have one */
error = -ENODEV;
if (!try_module_get(buffer->owner))
- goto out_put_item;
+ goto out_free_buffer;
error = -EACCES;
if (!buffer->item->ci_type)
@@ -435,8 +435,6 @@ static int __configfs_open_file(struct inode *inode, struct file *file, int type
out_put_module:
module_put(buffer->owner);
-out_put_item:
- config_item_put(buffer->item);
out_free_buffer:
up_read(&frag->frag_sem);
kfree(buffer);
diff --git a/fs/crypto/bio.c b/fs/crypto/bio.c
index b048a0e38516..68a2de6b5a9b 100644
--- a/fs/crypto/bio.c
+++ b/fs/crypto/bio.c
@@ -52,7 +52,7 @@ static int fscrypt_zeroout_range_inline_crypt(const struct inode *inode,
int num_pages = 0;
/* This always succeeds since __GFP_DIRECT_RECLAIM is set. */
- bio = bio_alloc(GFP_NOFS, BIO_MAX_PAGES);
+ bio = bio_alloc(GFP_NOFS, BIO_MAX_VECS);
while (len) {
unsigned int blocks_this_page = min(len, blocks_per_page);
@@ -74,7 +74,7 @@ static int fscrypt_zeroout_range_inline_crypt(const struct inode *inode,
len -= blocks_this_page;
lblk += blocks_this_page;
pblk += blocks_this_page;
- if (num_pages == BIO_MAX_PAGES || !len ||
+ if (num_pages == BIO_MAX_VECS || !len ||
!fscrypt_mergeable_bio(bio, inode, lblk)) {
err = submit_bio_wait(bio);
if (err)
@@ -126,7 +126,7 @@ int fscrypt_zeroout_range(const struct inode *inode, pgoff_t lblk,
return fscrypt_zeroout_range_inline_crypt(inode, lblk, pblk,
len);
- BUILD_BUG_ON(ARRAY_SIZE(pages) > BIO_MAX_PAGES);
+ BUILD_BUG_ON(ARRAY_SIZE(pages) > BIO_MAX_VECS);
nr_pages = min_t(unsigned int, ARRAY_SIZE(pages),
(len + blocks_per_page - 1) >> blocks_per_page_bits);
diff --git a/fs/erofs/data.c b/fs/erofs/data.c
index f88851c5c250..1249e74b3bf0 100644
--- a/fs/erofs/data.c
+++ b/fs/erofs/data.c
@@ -129,6 +129,7 @@ static inline struct bio *erofs_read_raw_page(struct bio *bio,
struct page *page,
erofs_off_t *last_block,
unsigned int nblocks,
+ unsigned int *eblks,
bool ra)
{
struct inode *const inode = mapping->host;
@@ -145,8 +146,7 @@ static inline struct bio *erofs_read_raw_page(struct bio *bio,
/* note that for readpage case, bio also equals to NULL */
if (bio &&
- /* not continuous */
- *last_block + 1 != current_block) {
+ (*last_block + 1 != current_block || !*eblks)) {
submit_bio_retry:
submit_bio(bio);
bio = NULL;
@@ -216,7 +216,8 @@ submit_bio_retry:
if (nblocks > DIV_ROUND_UP(map.m_plen, PAGE_SIZE))
nblocks = DIV_ROUND_UP(map.m_plen, PAGE_SIZE);
- bio = bio_alloc(GFP_NOIO, bio_max_segs(nblocks));
+ *eblks = bio_max_segs(nblocks);
+ bio = bio_alloc(GFP_NOIO, *eblks);
bio->bi_end_io = erofs_readendio;
bio_set_dev(bio, sb->s_bdev);
@@ -229,16 +230,8 @@ submit_bio_retry:
/* out of the extent or bio is full */
if (err < PAGE_SIZE)
goto submit_bio_retry;
-
+ --*eblks;
*last_block = current_block;
-
- /* shift in advance in case of it followed by too many gaps */
- if (bio->bi_iter.bi_size >= bio->bi_max_vecs * PAGE_SIZE) {
- /* err should reassign to 0 after submitting */
- err = 0;
- goto submit_bio_out;
- }
-
return bio;
err_out:
@@ -252,7 +245,6 @@ has_updated:
/* if updated manually, continuous pages has a gap */
if (bio)
-submit_bio_out:
submit_bio(bio);
return err ? ERR_PTR(err) : NULL;
}
@@ -264,23 +256,26 @@ submit_bio_out:
static int erofs_raw_access_readpage(struct file *file, struct page *page)
{
erofs_off_t last_block;
+ unsigned int eblks;
struct bio *bio;
trace_erofs_readpage(page, true);
bio = erofs_read_raw_page(NULL, page->mapping,
- page, &last_block, 1, false);
+ page, &last_block, 1, &eblks, false);
if (IS_ERR(bio))
return PTR_ERR(bio);
- DBG_BUGON(bio); /* since we have only one bio -- must be NULL */
+ if (bio)
+ submit_bio(bio);
return 0;
}
static void erofs_raw_access_readahead(struct readahead_control *rac)
{
erofs_off_t last_block;
+ unsigned int eblks;
struct bio *bio = NULL;
struct page *page;
@@ -291,7 +286,7 @@ static void erofs_raw_access_readahead(struct readahead_control *rac)
prefetchw(&page->flags);
bio = erofs_read_raw_page(bio, rac->mapping, page, &last_block,
- readahead_count(rac), true);
+ readahead_count(rac), &eblks, true);
/* all the page errors are ignored when readahead */
if (IS_ERR(bio)) {
@@ -305,7 +300,6 @@ static void erofs_raw_access_readahead(struct readahead_control *rac)
put_page(page);
}
- /* the rare case (end in gaps) */
if (bio)
submit_bio(bio);
}
diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 6cb356c4217b..3851e1a64f73 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -1235,7 +1235,7 @@ submit_bio_retry:
}
if (!bio) {
- bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES);
+ bio = bio_alloc(GFP_NOIO, BIO_MAX_VECS);
bio->bi_end_io = z_erofs_decompressqueue_endio;
bio_set_dev(bio, sb->s_bdev);
diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c
index 03a44a0de86a..f038d578d8d8 100644
--- a/fs/ext4/page-io.c
+++ b/fs/ext4/page-io.c
@@ -398,7 +398,7 @@ static void io_submit_init_bio(struct ext4_io_submit *io,
* bio_alloc will _always_ be able to allocate a bio if
* __GFP_DIRECT_RECLAIM is set, see comments for bio_alloc_bioset().
*/
- bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES);
+ bio = bio_alloc(GFP_NOIO, BIO_MAX_VECS);
fscrypt_set_bio_crypt_ctx_bh(bio, bh, GFP_NOIO);
bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9);
bio_set_dev(bio, bh->b_bdev);
diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 174a0819ad96..be5415a0dbbc 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -292,7 +292,7 @@ void f2fs_ra_meta_pages_cond(struct f2fs_sb_info *sbi, pgoff_t index)
f2fs_put_page(page, 0);
if (readahead)
- f2fs_ra_meta_pages(sbi, index, BIO_MAX_PAGES, META_POR, true);
+ f2fs_ra_meta_pages(sbi, index, BIO_MAX_VECS, META_POR, true);
}
static int __f2fs_write_meta_page(struct page *page,
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 7c95818639a6..4e5257c763d0 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -857,7 +857,7 @@ int f2fs_merge_page_bio(struct f2fs_io_info *fio)
f2fs_submit_merged_ipu_write(fio->sbi, &bio, NULL);
alloc_new:
if (!bio) {
- bio = __bio_alloc(fio, BIO_MAX_PAGES);
+ bio = __bio_alloc(fio, BIO_MAX_VECS);
__attach_io_flag(fio);
f2fs_set_bio_crypt_ctx(bio, fio->page->mapping->host,
fio->page->index, fio, GFP_NOIO);
@@ -932,7 +932,7 @@ alloc_new:
fio->retry = true;
goto skip;
}
- io->bio = __bio_alloc(fio, BIO_MAX_PAGES);
+ io->bio = __bio_alloc(fio, BIO_MAX_VECS);
f2fs_set_bio_crypt_ctx(io->bio, fio->page->mapping->host,
bio_page->index, fio, GFP_NOIO);
io->fio = *fio;
diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 993004f06a77..c2866561263e 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -4381,7 +4381,7 @@ static int build_sit_entries(struct f2fs_sb_info *sbi)
block_t total_node_blocks = 0;
do {
- readed = f2fs_ra_meta_pages(sbi, start_blk, BIO_MAX_PAGES,
+ readed = f2fs_ra_meta_pages(sbi, start_blk, BIO_MAX_VECS,
META_SIT, true);
start = start_blk * sit_i->sents_per_block;
diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h
index 229814b4f4a6..e9a7a637d688 100644
--- a/fs/f2fs/segment.h
+++ b/fs/f2fs/segment.h
@@ -851,7 +851,7 @@ static inline int nr_pages_to_skip(struct f2fs_sb_info *sbi, int type)
else if (type == NODE)
return 8 * sbi->blocks_per_seg;
else if (type == META)
- return 8 * BIO_MAX_PAGES;
+ return 8 * BIO_MAX_VECS;
else
return 0;
}
@@ -868,7 +868,7 @@ static inline long nr_pages_to_write(struct f2fs_sb_info *sbi, int type,
return 0;
nr_to_write = wbc->nr_to_write;
- desired = BIO_MAX_PAGES;
+ desired = BIO_MAX_VECS;
if (type == NODE)
desired <<= 1;
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 7069793752f1..82592b19b4e0 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -753,9 +753,9 @@ static int parse_options(struct super_block *sb, char *options, bool is_remount)
case Opt_io_size_bits:
if (args->from && match_int(args, &arg))
return -EINVAL;
- if (arg <= 0 || arg > __ilog2_u32(BIO_MAX_PAGES)) {
+ if (arg <= 0 || arg > __ilog2_u32(BIO_MAX_VECS)) {
f2fs_warn(sbi, "Not support %d, larger than %d",
- 1 << arg, BIO_MAX_PAGES);
+ 1 << arg, BIO_MAX_VECS);
return -EINVAL;
}
F2FS_OPTION(sbi).write_io_size_bits = arg;
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 16937ebb2a3e..6410281546f9 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -998,12 +998,16 @@ static void trans_drain(struct gfs2_trans *tr)
while (!list_empty(head)) {
bd = list_first_entry(head, struct gfs2_bufdata, bd_list);
list_del_init(&bd->bd_list);
+ if (!list_empty(&bd->bd_ail_st_list))
+ gfs2_remove_from_ail(bd);
kmem_cache_free(gfs2_bufdata_cachep, bd);
}
head = &tr->tr_databuf;
while (!list_empty(head)) {
bd = list_first_entry(head, struct gfs2_bufdata, bd_list);
list_del_init(&bd->bd_list);
+ if (!list_empty(&bd->bd_ail_st_list))
+ gfs2_remove_from_ail(bd);
kmem_cache_free(gfs2_bufdata_cachep, bd);
}
}
@@ -1032,7 +1036,7 @@ repeat:
* Do this check while holding the log_flush_lock to prevent new
* buffers from being added to the ail via gfs2_pin()
*/
- if (gfs2_withdrawn(sdp))
+ if (gfs2_withdrawn(sdp) || !test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
goto out;
/* Log might have been flushed while we waited for the flush lock */
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index dc1b93a877c6..a82f4747aa8d 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -267,7 +267,7 @@ static struct bio *gfs2_log_alloc_bio(struct gfs2_sbd *sdp, u64 blkno,
bio_end_io_t *end_io)
{
struct super_block *sb = sdp->sd_vfs;
- struct bio *bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES);
+ struct bio *bio = bio_alloc(GFP_NOIO, BIO_MAX_VECS);
bio->bi_iter.bi_sector = blkno << sdp->sd_fsb2bb_shift;
bio_set_dev(bio, sb->s_bdev);
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 74c7d01723b9..aa4136055a83 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -1539,9 +1539,7 @@ static int gfs2_reconfigure(struct fs_context *fc)
return -EINVAL;
if (fc->sb_flags & SB_RDONLY) {
- error = gfs2_make_fs_ro(sdp);
- if (error)
- errorfc(fc, "unable to remount read-only");
+ gfs2_make_fs_ro(sdp);
} else {
error = gfs2_make_fs_rw(sdp);
if (error)
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index 861ed5fe02a5..97076d3f562f 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -587,9 +587,8 @@ out:
* Returns: errno
*/
-int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
+void gfs2_make_fs_ro(struct gfs2_sbd *sdp)
{
- int error = 0;
int log_write_allowed = test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
gfs2_flush_delete_work(sdp);
@@ -624,8 +623,6 @@ int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
if (!log_write_allowed)
sdp->sd_vfs->s_flags |= SB_RDONLY;
-
- return error;
}
/**
@@ -637,7 +634,6 @@ int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
static void gfs2_put_super(struct super_block *sb)
{
struct gfs2_sbd *sdp = sb->s_fs_info;
- int error;
struct gfs2_jdesc *jd;
/* No more recovery requests */
@@ -658,9 +654,7 @@ restart:
spin_unlock(&sdp->sd_jindex_spin);
if (!sb_rdonly(sb)) {
- error = gfs2_make_fs_ro(sdp);
- if (error)
- gfs2_io_error(sdp);
+ gfs2_make_fs_ro(sdp);
}
WARN_ON(gfs2_withdrawing(sdp));
diff --git a/fs/gfs2/super.h b/fs/gfs2/super.h
index 08e502dec7ec..ec4affb33ed5 100644
--- a/fs/gfs2/super.h
+++ b/fs/gfs2/super.h
@@ -34,7 +34,7 @@ extern int gfs2_lookup_in_master_dir(struct gfs2_sbd *sdp, char *filename,
struct gfs2_inode **ipp);
extern int gfs2_make_fs_rw(struct gfs2_sbd *sdp);
-extern int gfs2_make_fs_ro(struct gfs2_sbd *sdp);
+extern void gfs2_make_fs_ro(struct gfs2_sbd *sdp);
extern void gfs2_online_uevent(struct gfs2_sbd *sdp);
extern int gfs2_statfs_init(struct gfs2_sbd *sdp);
extern void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free,
diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index ab96cf0bf26b..63fec11ef2ce 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -169,6 +169,8 @@ static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl,
bd->bd_bh = bh;
bd->bd_gl = gl;
INIT_LIST_HEAD(&bd->bd_list);
+ INIT_LIST_HEAD(&bd->bd_ail_st_list);
+ INIT_LIST_HEAD(&bd->bd_ail_gl_list);
bh->b_private = bd;
return bd;
}
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 8d3c670c990f..4f034b87b427 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -119,17 +119,22 @@ void gfs2_freeze_unlock(struct gfs2_holder *freeze_gh)
static void signal_our_withdraw(struct gfs2_sbd *sdp)
{
struct gfs2_glock *live_gl = sdp->sd_live_gh.gh_gl;
- struct inode *inode = sdp->sd_jdesc->jd_inode;
- struct gfs2_inode *ip = GFS2_I(inode);
- struct gfs2_glock *i_gl = ip->i_gl;
- u64 no_formal_ino = ip->i_no_formal_ino;
+ struct inode *inode;
+ struct gfs2_inode *ip;
+ struct gfs2_glock *i_gl;
+ u64 no_formal_ino;
int log_write_allowed = test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
int ret = 0;
int tries;
- if (test_bit(SDF_NORECOVERY, &sdp->sd_flags))
+ if (test_bit(SDF_NORECOVERY, &sdp->sd_flags) || !sdp->sd_jdesc)
return;
+ inode = sdp->sd_jdesc->jd_inode;
+ ip = GFS2_I(inode);
+ i_gl = ip->i_gl;
+ no_formal_ino = ip->i_no_formal_ino;
+
/* Prevent any glock dq until withdraw recovery is complete */
set_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
/*
@@ -156,7 +161,7 @@ static void signal_our_withdraw(struct gfs2_sbd *sdp)
ret = 0;
}
if (!ret)
- ret = gfs2_make_fs_ro(sdp);
+ gfs2_make_fs_ro(sdp);
gfs2_freeze_unlock(&freeze_gh);
}
diff --git a/fs/io-wq.c b/fs/io-wq.c
index 44e20248805a..0ae9ecadf295 100644
--- a/fs/io-wq.c
+++ b/fs/io-wq.c
@@ -16,6 +16,7 @@
#include <linux/rculist_nulls.h>
#include <linux/cpu.h>
#include <linux/tracehook.h>
+#include <linux/freezer.h>
#include "../kernel/sched/sched.h"
#include "io-wq.h"
@@ -52,9 +53,6 @@ struct io_worker {
struct io_wq_work *cur_work;
spinlock_t lock;
- const struct cred *cur_creds;
- const struct cred *saved_creds;
-
struct completion ref_done;
struct rcu_head rcu;
@@ -112,12 +110,14 @@ struct io_wq {
io_wq_work_fn *do_work;
struct task_struct *manager;
- struct user_struct *user;
struct io_wq_hash *hash;
refcount_t refs;
- struct completion done;
+ struct completion exited;
+
+ atomic_t worker_refs;
+ struct completion worker_done;
struct hlist_node cpuhp_node;
@@ -126,6 +126,17 @@ struct io_wq {
static enum cpuhp_state io_wq_online;
+struct io_cb_cancel_data {
+ work_cancel_fn *fn;
+ void *data;
+ int nr_running;
+ int nr_pending;
+ bool cancel_all;
+};
+
+static void io_wqe_cancel_pending_work(struct io_wqe *wqe,
+ struct io_cb_cancel_data *match);
+
static bool io_worker_get(struct io_worker *worker)
{
return refcount_inc_not_zero(&worker->ref);
@@ -175,11 +186,6 @@ static void io_worker_exit(struct io_worker *worker)
worker->flags = 0;
preempt_enable();
- if (worker->saved_creds) {
- revert_creds(worker->saved_creds);
- worker->cur_creds = worker->saved_creds = NULL;
- }
-
raw_spin_lock_irq(&wqe->lock);
if (flags & IO_WORKER_F_FREE)
hlist_nulls_del_rcu(&worker->nulls_node);
@@ -188,7 +194,9 @@ static void io_worker_exit(struct io_worker *worker)
raw_spin_unlock_irq(&wqe->lock);
kfree_rcu(worker, rcu);
- io_wq_put(wqe->wq);
+ if (atomic_dec_and_test(&wqe->wq->worker_refs))
+ complete(&wqe->wq->worker_done);
+ do_exit(0);
}
static inline bool io_wqe_run_queue(struct io_wqe *wqe)
@@ -263,12 +271,6 @@ static void io_wqe_dec_running(struct io_worker *worker)
io_wqe_wake_worker(wqe, acct);
}
-static void io_worker_start(struct io_worker *worker)
-{
- worker->flags |= (IO_WORKER_F_UP | IO_WORKER_F_RUNNING);
- io_wqe_inc_running(worker);
-}
-
/*
* Worker will start processing some work. Move it to the busy list, if
* it's currently on the freelist
@@ -319,10 +321,6 @@ static void __io_worker_idle(struct io_wqe *wqe, struct io_worker *worker)
worker->flags |= IO_WORKER_F_FREE;
hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list);
}
- if (worker->saved_creds) {
- revert_creds(worker->saved_creds);
- worker->cur_creds = worker->saved_creds = NULL;
- }
}
static inline unsigned int io_get_work_hash(struct io_wq_work *work)
@@ -397,18 +395,6 @@ static void io_flush_signals(void)
}
}
-static void io_wq_switch_creds(struct io_worker *worker,
- struct io_wq_work *work)
-{
- const struct cred *old_creds = override_creds(work->creds);
-
- worker->cur_creds = work->creds;
- if (worker->saved_creds)
- put_cred(old_creds); /* creds set by previous switch */
- else
- worker->saved_creds = old_creds;
-}
-
static void io_assign_current_work(struct io_worker *worker,
struct io_wq_work *work)
{
@@ -458,8 +444,6 @@ get_next:
unsigned int hash = io_get_work_hash(work);
next_hashed = wq_next_work(work);
- if (work->creds && worker->cur_creds != work->creds)
- io_wq_switch_creds(worker, work);
wq->do_work(work);
io_assign_current_work(worker, NULL);
@@ -495,8 +479,13 @@ static int io_wqe_worker(void *data)
struct io_worker *worker = data;
struct io_wqe *wqe = worker->wqe;
struct io_wq *wq = wqe->wq;
+ char buf[TASK_COMM_LEN];
- io_worker_start(worker);
+ worker->flags |= (IO_WORKER_F_UP | IO_WORKER_F_RUNNING);
+ io_wqe_inc_running(worker);
+
+ sprintf(buf, "iou-wrk-%d", wq->task_pid);
+ set_task_comm(current, buf);
while (!test_bit(IO_WQ_BIT_EXIT, &wq->state)) {
set_current_state(TASK_INTERRUPTIBLE);
@@ -571,67 +560,11 @@ void io_wq_worker_sleeping(struct task_struct *tsk)
raw_spin_unlock_irq(&worker->wqe->lock);
}
-static int task_thread(void *data, int index)
-{
- struct io_worker *worker = data;
- struct io_wqe *wqe = worker->wqe;
- struct io_wqe_acct *acct = &wqe->acct[index];
- struct io_wq *wq = wqe->wq;
- char buf[TASK_COMM_LEN];
-
- sprintf(buf, "iou-wrk-%d", wq->task_pid);
- set_task_comm(current, buf);
-
- current->pf_io_worker = worker;
- worker->task = current;
-
- set_cpus_allowed_ptr(current, cpumask_of_node(wqe->node));
- current->flags |= PF_NO_SETAFFINITY;
-
- raw_spin_lock_irq(&wqe->lock);
- hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list);
- list_add_tail_rcu(&worker->all_list, &wqe->all_list);
- worker->flags |= IO_WORKER_F_FREE;
- if (index == IO_WQ_ACCT_BOUND)
- worker->flags |= IO_WORKER_F_BOUND;
- if (!acct->nr_workers && (worker->flags & IO_WORKER_F_BOUND))
- worker->flags |= IO_WORKER_F_FIXED;
- acct->nr_workers++;
- raw_spin_unlock_irq(&wqe->lock);
-
- io_wqe_worker(data);
- do_exit(0);
-}
-
-static int task_thread_bound(void *data)
-{
- return task_thread(data, IO_WQ_ACCT_BOUND);
-}
-
-static int task_thread_unbound(void *data)
-{
- return task_thread(data, IO_WQ_ACCT_UNBOUND);
-}
-
-pid_t io_wq_fork_thread(int (*fn)(void *), void *arg)
-{
- unsigned long flags = CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|
- CLONE_IO|SIGCHLD;
- struct kernel_clone_args args = {
- .flags = ((lower_32_bits(flags) | CLONE_VM |
- CLONE_UNTRACED) & ~CSIGNAL),
- .exit_signal = (lower_32_bits(flags) & CSIGNAL),
- .stack = (unsigned long)fn,
- .stack_size = (unsigned long)arg,
- };
-
- return kernel_clone(&args);
-}
-
static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)
{
+ struct io_wqe_acct *acct = &wqe->acct[index];
struct io_worker *worker;
- pid_t pid;
+ struct task_struct *tsk;
__set_current_state(TASK_RUNNING);
@@ -645,17 +578,32 @@ static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)
spin_lock_init(&worker->lock);
init_completion(&worker->ref_done);
- refcount_inc(&wq->refs);
+ atomic_inc(&wq->worker_refs);
- if (index == IO_WQ_ACCT_BOUND)
- pid = io_wq_fork_thread(task_thread_bound, worker);
- else
- pid = io_wq_fork_thread(task_thread_unbound, worker);
- if (pid < 0) {
- io_wq_put(wq);
+ tsk = create_io_thread(io_wqe_worker, worker, wqe->node);
+ if (IS_ERR(tsk)) {
+ if (atomic_dec_and_test(&wq->worker_refs))
+ complete(&wq->worker_done);
kfree(worker);
return false;
}
+
+ tsk->pf_io_worker = worker;
+ worker->task = tsk;
+ set_cpus_allowed_ptr(tsk, cpumask_of_node(wqe->node));
+ tsk->flags |= PF_NO_SETAFFINITY;
+
+ raw_spin_lock_irq(&wqe->lock);
+ hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list);
+ list_add_tail_rcu(&worker->all_list, &wqe->all_list);
+ worker->flags |= IO_WORKER_F_FREE;
+ if (index == IO_WQ_ACCT_BOUND)
+ worker->flags |= IO_WORKER_F_BOUND;
+ if (!acct->nr_workers && (worker->flags & IO_WORKER_F_BOUND))
+ worker->flags |= IO_WORKER_F_FIXED;
+ acct->nr_workers++;
+ raw_spin_unlock_irq(&wqe->lock);
+ wake_up_new_task(tsk);
return true;
}
@@ -664,6 +612,8 @@ static inline bool io_wqe_need_worker(struct io_wqe *wqe, int index)
{
struct io_wqe_acct *acct = &wqe->acct[index];
+ if (acct->nr_workers && test_bit(IO_WQ_BIT_EXIT, &wqe->wq->state))
+ return false;
/* if we have available workers or no work, no need */
if (!hlist_nulls_empty(&wqe->free_list) || !io_wqe_run_queue(wqe))
return false;
@@ -697,6 +647,7 @@ static bool io_wq_for_each_worker(struct io_wqe *wqe,
static bool io_wq_worker_wake(struct io_worker *worker, void *data)
{
+ set_notify_signal(worker->task);
wake_up_process(worker->task);
return false;
}
@@ -725,6 +676,23 @@ static void io_wq_check_workers(struct io_wq *wq)
}
}
+static bool io_wq_work_match_all(struct io_wq_work *work, void *data)
+{
+ return true;
+}
+
+static void io_wq_cancel_pending(struct io_wq *wq)
+{
+ struct io_cb_cancel_data match = {
+ .fn = io_wq_work_match_all,
+ .cancel_all = true,
+ };
+ int node;
+
+ for_each_node(node)
+ io_wqe_cancel_pending_work(wq->wqes[node], &match);
+}
+
/*
* Manager thread. Tasked with creating new workers, if we need them.
*/
@@ -732,13 +700,10 @@ static int io_wq_manager(void *data)
{
struct io_wq *wq = data;
char buf[TASK_COMM_LEN];
+ int node;
sprintf(buf, "iou-mgr-%d", wq->task_pid);
set_task_comm(current, buf);
- current->flags |= PF_IO_WORKER;
- wq->manager = current;
-
- complete(&wq->done);
do {
set_current_state(TASK_INTERRUPTIBLE);
@@ -749,8 +714,23 @@ static int io_wq_manager(void *data)
} while (!test_bit(IO_WQ_BIT_EXIT, &wq->state));
io_wq_check_workers(wq);
- wq->manager = NULL;
- io_wq_put(wq);
+
+ rcu_read_lock();
+ for_each_node(node)
+ io_wq_for_each_worker(wq->wqes[node], io_wq_worker_wake, NULL);
+ rcu_read_unlock();
+
+ if (atomic_dec_and_test(&wq->worker_refs))
+ complete(&wq->worker_done);
+ wait_for_completion(&wq->worker_done);
+
+ spin_lock_irq(&wq->hash->wait.lock);
+ for_each_node(node)
+ list_del_init(&wq->wqes[node]->wait.entry);
+ spin_unlock_irq(&wq->hash->wait.lock);
+
+ io_wq_cancel_pending(wq);
+ complete(&wq->exited);
do_exit(0);
}
@@ -787,23 +767,26 @@ append:
static int io_wq_fork_manager(struct io_wq *wq)
{
- int ret;
+ struct task_struct *tsk;
if (wq->manager)
return 0;
- clear_bit(IO_WQ_BIT_EXIT, &wq->state);
- refcount_inc(&wq->refs);
- current->flags |= PF_IO_WORKER;
- ret = io_wq_fork_thread(io_wq_manager, wq);
- current->flags &= ~PF_IO_WORKER;
- if (ret >= 0) {
- wait_for_completion(&wq->done);
+ WARN_ON_ONCE(test_bit(IO_WQ_BIT_EXIT, &wq->state));
+
+ init_completion(&wq->worker_done);
+ atomic_set(&wq->worker_refs, 1);
+ tsk = create_io_thread(io_wq_manager, wq, NUMA_NO_NODE);
+ if (!IS_ERR(tsk)) {
+ wq->manager = get_task_struct(tsk);
+ wake_up_new_task(tsk);
return 0;
}
- io_wq_put(wq);
- return ret;
+ if (atomic_dec_and_test(&wq->worker_refs))
+ complete(&wq->worker_done);
+
+ return PTR_ERR(tsk);
}
static void io_wqe_enqueue(struct io_wqe *wqe, struct io_wq_work *work)
@@ -813,9 +796,9 @@ static void io_wqe_enqueue(struct io_wqe *wqe, struct io_wq_work *work)
unsigned long flags;
/* Can only happen if manager creation fails after exec */
- if (unlikely(io_wq_fork_manager(wqe->wq))) {
- work->flags |= IO_WQ_WORK_CANCEL;
- wqe->wq->do_work(work);
+ if (io_wq_fork_manager(wqe->wq) ||
+ test_bit(IO_WQ_BIT_EXIT, &wqe->wq->state)) {
+ io_run_cancel(work, wqe);
return;
}
@@ -849,14 +832,6 @@ void io_wq_hash_work(struct io_wq_work *work, void *val)
work->flags |= (IO_WQ_WORK_HASHED | (bit << IO_WQ_HASH_SHIFT));
}
-struct io_cb_cancel_data {
- work_cancel_fn *fn;
- void *data;
- int nr_running;
- int nr_pending;
- bool cancel_all;
-};
-
static bool io_wq_worker_cancel(struct io_worker *worker, void *data)
{
struct io_cb_cancel_data *match = data;
@@ -1043,16 +1018,14 @@ struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data)
}
wq->task_pid = current->pid;
- init_completion(&wq->done);
+ init_completion(&wq->exited);
refcount_set(&wq->refs, 1);
ret = io_wq_fork_manager(wq);
if (!ret)
return wq;
-
- io_wq_put(wq);
- io_wq_put_hash(data->hash);
err:
+ io_wq_put_hash(data->hash);
cpuhp_state_remove_instance_nocalls(io_wq_online, &wq->cpuhp_node);
for_each_node(node)
kfree(wq->wqes[node]);
@@ -1063,6 +1036,16 @@ err_wq:
return ERR_PTR(ret);
}
+static void io_wq_destroy_manager(struct io_wq *wq)
+{
+ if (wq->manager) {
+ wake_up_process(wq->manager);
+ wait_for_completion(&wq->exited);
+ put_task_struct(wq->manager);
+ wq->manager = NULL;
+ }
+}
+
static void io_wq_destroy(struct io_wq *wq)
{
int node;
@@ -1070,26 +1053,16 @@ static void io_wq_destroy(struct io_wq *wq)
cpuhp_state_remove_instance_nocalls(io_wq_online, &wq->cpuhp_node);
set_bit(IO_WQ_BIT_EXIT, &wq->state);
- if (wq->manager)
- wake_up_process(wq->manager);
-
- rcu_read_lock();
- for_each_node(node)
- io_wq_for_each_worker(wq->wqes[node], io_wq_worker_wake, NULL);
- rcu_read_unlock();
+ io_wq_destroy_manager(wq);
- spin_lock_irq(&wq->hash->wait.lock);
for_each_node(node) {
struct io_wqe *wqe = wq->wqes[node];
-
- list_del_init(&wqe->wait.entry);
+ WARN_ON_ONCE(!wq_list_empty(&wqe->work_list));
kfree(wqe);
}
- spin_unlock_irq(&wq->hash->wait.lock);
io_wq_put_hash(wq->hash);
kfree(wq->wqes);
kfree(wq);
-
}
void io_wq_put(struct io_wq *wq)
@@ -1098,6 +1071,13 @@ void io_wq_put(struct io_wq *wq)
io_wq_destroy(wq);
}
+void io_wq_put_and_exit(struct io_wq *wq)
+{
+ set_bit(IO_WQ_BIT_EXIT, &wq->state);
+ io_wq_destroy_manager(wq);
+ io_wq_put(wq);
+}
+
static bool io_wq_worker_affinity(struct io_worker *worker, void *data)
{
struct task_struct *task = worker->task;
diff --git a/fs/io-wq.h b/fs/io-wq.h
index b6ca12b60c35..1ac2f3248088 100644
--- a/fs/io-wq.h
+++ b/fs/io-wq.h
@@ -114,12 +114,11 @@ struct io_wq_data {
struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data);
void io_wq_put(struct io_wq *wq);
+void io_wq_put_and_exit(struct io_wq *wq);
void io_wq_enqueue(struct io_wq *wq, struct io_wq_work *work);
void io_wq_hash_work(struct io_wq_work *work, void *val);
-pid_t io_wq_fork_thread(int (*fn)(void *), void *arg);
-
static inline bool io_wq_is_hashed(struct io_wq_work *work)
{
return work->flags & IO_WQ_WORK_HASHED;
diff --git a/fs/io_uring.c b/fs/io_uring.c
index 4a088581b0f2..a4bce17af506 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -74,13 +74,11 @@
#include <linux/fsnotify.h>
#include <linux/fadvise.h>
#include <linux/eventpoll.h>
-#include <linux/fs_struct.h>
#include <linux/splice.h>
#include <linux/task_work.h>
#include <linux/pagemap.h>
#include <linux/io_uring.h>
-#include <linux/blk-cgroup.h>
-#include <linux/audit.h>
+#include <linux/freezer.h>
#define CREATE_TRACE_POINTS
#include <trace/events/io_uring.h>
@@ -260,12 +258,10 @@ enum {
struct io_sq_data {
refcount_t refs;
- struct mutex lock;
+ struct rw_semaphore rw_lock;
/* ctx's that are using this sqd */
struct list_head ctx_list;
- struct list_head ctx_new_list;
- struct mutex ctx_lock;
struct task_struct *thread;
struct wait_queue_head wait;
@@ -273,10 +269,9 @@ struct io_sq_data {
unsigned sq_thread_idle;
int sq_cpu;
pid_t task_pid;
+ pid_t task_tgid;
unsigned long state;
- struct completion startup;
- struct completion completion;
struct completion exited;
};
@@ -338,8 +333,6 @@ struct io_ring_ctx {
unsigned int drain_next: 1;
unsigned int eventfd_async: 1;
unsigned int restricted: 1;
- unsigned int sqo_dead: 1;
- unsigned int sqo_exec: 1;
/*
* Ring buffer of indices into array of io_uring_sqe, which is
@@ -380,14 +373,10 @@ struct io_ring_ctx {
struct io_rings *rings;
- /*
- * For SQPOLL usage
- */
- struct task_struct *sqo_task;
-
/* Only used for accounting purposes */
struct mm_struct *mm_account;
+ const struct cred *sq_creds; /* cred used for __io_sq_thread() */
struct io_sq_data *sq_data; /* if using sq thread polling */
struct wait_queue_head sqo_sq_wait;
@@ -408,7 +397,6 @@ struct io_ring_ctx {
struct user_struct *user;
struct completion ref_comp;
- struct completion sq_thread_comp;
#if defined(CONFIG_UNIX)
struct socket *ring_sock;
@@ -416,7 +404,8 @@ struct io_ring_ctx {
struct idr io_buffer_idr;
- struct idr personality_idr;
+ struct xarray personalities;
+ u32 pers_next;
struct {
unsigned cached_cq_tail;
@@ -462,6 +451,7 @@ struct io_ring_ctx {
/* Keep this last, we don't need it for the fast path */
struct work_struct exit_work;
+ struct list_head tctx_list;
};
/*
@@ -688,7 +678,6 @@ enum {
REQ_F_POLLED_BIT,
REQ_F_BUFFER_SELECTED_BIT,
REQ_F_NO_FILE_TABLE_BIT,
- REQ_F_WORK_INITIALIZED_BIT,
REQ_F_LTIMEOUT_ACTIVE_BIT,
REQ_F_COMPLETE_INLINE_BIT,
@@ -712,7 +701,7 @@ enum {
/* fail rest of links */
REQ_F_FAIL_LINK = BIT(REQ_F_FAIL_LINK_BIT),
- /* on inflight list */
+ /* on inflight list, should be cancelled and waited on exit reliably */
REQ_F_INFLIGHT = BIT(REQ_F_INFLIGHT_BIT),
/* read/write uses file position */
REQ_F_CUR_POS = BIT(REQ_F_CUR_POS_BIT),
@@ -730,8 +719,6 @@ enum {
REQ_F_BUFFER_SELECTED = BIT(REQ_F_BUFFER_SELECTED_BIT),
/* doesn't need file table for this request */
REQ_F_NO_FILE_TABLE = BIT(REQ_F_NO_FILE_TABLE_BIT),
- /* io_wq_work is initialized */
- REQ_F_WORK_INITIALIZED = BIT(REQ_F_WORK_INITIALIZED_BIT),
/* linked timeout is active, i.e. prepared by link's head */
REQ_F_LTIMEOUT_ACTIVE = BIT(REQ_F_LTIMEOUT_ACTIVE_BIT),
/* completion is deferred through io_comp_state */
@@ -816,6 +803,12 @@ struct io_kiocb {
struct io_wq_work work;
};
+struct io_tctx_node {
+ struct list_head ctx_node;
+ struct task_struct *task;
+ struct io_ring_ctx *ctx;
+};
+
struct io_defer_entry {
struct list_head list;
struct io_kiocb *req;
@@ -990,6 +983,8 @@ static const struct io_op_def io_op_defs[] = {
[IORING_OP_UNLINKAT] = {},
};
+static bool io_disarm_next(struct io_kiocb *req);
+static void io_uring_del_task_file(unsigned long index);
static void io_uring_try_cancel_requests(struct io_ring_ctx *ctx,
struct task_struct *task,
struct files_struct *files);
@@ -1080,9 +1075,7 @@ static bool io_match_task(struct io_kiocb *head,
return true;
io_for_each_link(req, head) {
- if (!(req->flags & REQ_F_WORK_INITIALIZED))
- continue;
- if (req->file && req->file->f_op == &io_uring_fops)
+ if (req->flags & REQ_F_INFLIGHT)
return true;
if (req->task->files == files)
return true;
@@ -1096,24 +1089,6 @@ static inline void req_set_fail_links(struct io_kiocb *req)
req->flags |= REQ_F_FAIL_LINK;
}
-static inline void __io_req_init_async(struct io_kiocb *req)
-{
- memset(&req->work, 0, sizeof(req->work));
- req->flags |= REQ_F_WORK_INITIALIZED;
-}
-
-/*
- * Note: must call io_req_init_async() for the first time you
- * touch any members of io_wq_work.
- */
-static inline void io_req_init_async(struct io_kiocb *req)
-{
- if (req->flags & REQ_F_WORK_INITIALIZED)
- return;
-
- __io_req_init_async(req);
-}
-
static void io_ring_ctx_ref_free(struct percpu_ref *ref)
{
struct io_ring_ctx *ctx = container_of(ref, struct io_ring_ctx, refs);
@@ -1160,9 +1135,8 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
init_waitqueue_head(&ctx->cq_wait);
INIT_LIST_HEAD(&ctx->cq_overflow_list);
init_completion(&ctx->ref_comp);
- init_completion(&ctx->sq_thread_comp);
idr_init(&ctx->io_buffer_idr);
- idr_init(&ctx->personality_idr);
+ xa_init_flags(&ctx->personalities, XA_FLAGS_ALLOC1);
mutex_init(&ctx->uring_lock);
init_waitqueue_head(&ctx->wait);
spin_lock_init(&ctx->completion_lock);
@@ -1175,6 +1149,7 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
INIT_LIST_HEAD(&ctx->rsrc_ref_list);
INIT_DELAYED_WORK(&ctx->rsrc_put_work, io_rsrc_put_work);
init_llist_head(&ctx->rsrc_put_llist);
+ INIT_LIST_HEAD(&ctx->tctx_list);
INIT_LIST_HEAD(&ctx->submit_state.comp.free_list);
INIT_LIST_HEAD(&ctx->submit_state.comp.locked_free_list);
return ctx;
@@ -1196,37 +1171,11 @@ static bool req_need_defer(struct io_kiocb *req, u32 seq)
return false;
}
-static void io_req_clean_work(struct io_kiocb *req)
-{
- if (!(req->flags & REQ_F_WORK_INITIALIZED))
- return;
-
- if (req->work.creds) {
- put_cred(req->work.creds);
- req->work.creds = NULL;
- }
- if (req->flags & REQ_F_INFLIGHT) {
- struct io_ring_ctx *ctx = req->ctx;
- struct io_uring_task *tctx = req->task->io_uring;
- unsigned long flags;
-
- spin_lock_irqsave(&ctx->inflight_lock, flags);
- list_del(&req->inflight_entry);
- spin_unlock_irqrestore(&ctx->inflight_lock, flags);
- req->flags &= ~REQ_F_INFLIGHT;
- if (atomic_read(&tctx->in_idle))
- wake_up(&tctx->wait);
- }
-
- req->flags &= ~REQ_F_WORK_INITIALIZED;
-}
-
static void io_req_track_inflight(struct io_kiocb *req)
{
struct io_ring_ctx *ctx = req->ctx;
if (!(req->flags & REQ_F_INFLIGHT)) {
- io_req_init_async(req);
req->flags |= REQ_F_INFLIGHT;
spin_lock_irq(&ctx->inflight_lock);
@@ -1240,7 +1189,8 @@ static void io_prep_async_work(struct io_kiocb *req)
const struct io_op_def *def = &io_op_defs[req->opcode];
struct io_ring_ctx *ctx = req->ctx;
- io_req_init_async(req);
+ if (!req->work.creds)
+ req->work.creds = get_current_cred();
if (req->flags & REQ_F_FORCE_ASYNC)
req->work.flags |= IO_WQ_WORK_CONCURRENT;
@@ -1252,8 +1202,6 @@ static void io_prep_async_work(struct io_kiocb *req)
if (def->unbound_nonreg_file)
req->work.flags |= IO_WQ_WORK_UNBOUND;
}
- if (!req->work.creds)
- req->work.creds = get_current_cred();
}
static void io_prep_async_link(struct io_kiocb *req)
@@ -1264,7 +1212,7 @@ static void io_prep_async_link(struct io_kiocb *req)
io_prep_async_work(cur);
}
-static struct io_kiocb *__io_queue_async_work(struct io_kiocb *req)
+static void io_queue_async_work(struct io_kiocb *req)
{
struct io_ring_ctx *ctx = req->ctx;
struct io_kiocb *link = io_prep_linked_timeout(req);
@@ -1275,18 +1223,9 @@ static struct io_kiocb *__io_queue_async_work(struct io_kiocb *req)
trace_io_uring_queue_async_work(ctx, io_wq_is_hashed(&req->work), req,
&req->work, req->flags);
- io_wq_enqueue(tctx->io_wq, &req->work);
- return link;
-}
-
-static void io_queue_async_work(struct io_kiocb *req)
-{
- struct io_kiocb *link;
-
/* init ->work of the whole link before punting */
io_prep_async_link(req);
- link = __io_queue_async_work(req);
-
+ io_wq_enqueue(tctx->io_wq, &req->work);
if (link)
io_queue_linked_timeout(link);
}
@@ -1521,18 +1460,22 @@ static bool __io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
return all_flushed;
}
-static void io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
+static bool io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
struct task_struct *tsk,
struct files_struct *files)
{
+ bool ret = true;
+
if (test_bit(0, &ctx->cq_check_overflow)) {
/* iopoll syncs against uring_lock, not completion_lock */
if (ctx->flags & IORING_SETUP_IOPOLL)
mutex_lock(&ctx->uring_lock);
- __io_cqring_overflow_flush(ctx, force, tsk, files);
+ ret = __io_cqring_overflow_flush(ctx, force, tsk, files);
if (ctx->flags & IORING_SETUP_IOPOLL)
mutex_unlock(&ctx->uring_lock);
}
+
+ return ret;
}
static void __io_cqring_fill_event(struct io_kiocb *req, long res, long cflags)
@@ -1580,15 +1523,14 @@ static void io_cqring_fill_event(struct io_kiocb *req, long res)
__io_cqring_fill_event(req, res, 0);
}
-static inline void io_req_complete_post(struct io_kiocb *req, long res,
- unsigned int cflags)
+static void io_req_complete_post(struct io_kiocb *req, long res,
+ unsigned int cflags)
{
struct io_ring_ctx *ctx = req->ctx;
unsigned long flags;
spin_lock_irqsave(&ctx->completion_lock, flags);
__io_cqring_fill_event(req, res, cflags);
- io_commit_cqring(ctx);
/*
* If we're the last reference to this request, add to our locked
* free_list cache.
@@ -1596,19 +1538,26 @@ static inline void io_req_complete_post(struct io_kiocb *req, long res,
if (refcount_dec_and_test(&req->refs)) {
struct io_comp_state *cs = &ctx->submit_state.comp;
+ if (req->flags & (REQ_F_LINK | REQ_F_HARDLINK)) {
+ if (req->flags & (REQ_F_LINK_TIMEOUT | REQ_F_FAIL_LINK))
+ io_disarm_next(req);
+ if (req->link) {
+ io_req_task_queue(req->link);
+ req->link = NULL;
+ }
+ }
io_dismantle_req(req);
io_put_task(req->task, 1);
list_add(&req->compl.list, &cs->locked_free_list);
cs->locked_free_nr++;
} else
req = NULL;
+ io_commit_cqring(ctx);
spin_unlock_irqrestore(&ctx->completion_lock, flags);
-
io_cqring_ev_posted(ctx);
- if (req) {
- io_queue_next(req);
+
+ if (req)
percpu_ref_put(&ctx->refs);
- }
}
static void io_req_complete_state(struct io_kiocb *req, long res,
@@ -1714,9 +1663,23 @@ static void io_dismantle_req(struct io_kiocb *req)
io_put_file(req, req->file, (req->flags & REQ_F_FIXED_FILE));
if (req->fixed_rsrc_refs)
percpu_ref_put(req->fixed_rsrc_refs);
- io_req_clean_work(req);
+ if (req->work.creds) {
+ put_cred(req->work.creds);
+ req->work.creds = NULL;
+ }
+
+ if (req->flags & REQ_F_INFLIGHT) {
+ struct io_ring_ctx *ctx = req->ctx;
+ unsigned long flags;
+
+ spin_lock_irqsave(&ctx->inflight_lock, flags);
+ list_del(&req->inflight_entry);
+ spin_unlock_irqrestore(&ctx->inflight_lock, flags);
+ req->flags &= ~REQ_F_INFLIGHT;
+ }
}
+/* must to be called somewhat shortly after putting a request */
static inline void io_put_task(struct task_struct *task, int nr)
{
struct io_uring_task *tctx = task->io_uring;
@@ -1746,15 +1709,11 @@ static inline void io_remove_next_linked(struct io_kiocb *req)
nxt->link = NULL;
}
-static void io_kill_linked_timeout(struct io_kiocb *req)
+static bool io_kill_linked_timeout(struct io_kiocb *req)
+ __must_hold(&req->ctx->completion_lock)
{
- struct io_ring_ctx *ctx = req->ctx;
- struct io_kiocb *link;
+ struct io_kiocb *link = req->link;
bool cancelled = false;
- unsigned long flags;
-
- spin_lock_irqsave(&ctx->completion_lock, flags);
- link = req->link;
/*
* Can happen if a linked timeout fired and link had been like
@@ -1769,58 +1728,48 @@ static void io_kill_linked_timeout(struct io_kiocb *req)
ret = hrtimer_try_to_cancel(&io->timer);
if (ret != -1) {
io_cqring_fill_event(link, -ECANCELED);
- io_commit_cqring(ctx);
+ io_put_req_deferred(link, 1);
cancelled = true;
}
}
req->flags &= ~REQ_F_LINK_TIMEOUT;
- spin_unlock_irqrestore(&ctx->completion_lock, flags);
-
- if (cancelled) {
- io_cqring_ev_posted(ctx);
- io_put_req(link);
- }
+ return cancelled;
}
-
static void io_fail_links(struct io_kiocb *req)
+ __must_hold(&req->ctx->completion_lock)
{
- struct io_kiocb *link, *nxt;
- struct io_ring_ctx *ctx = req->ctx;
- unsigned long flags;
+ struct io_kiocb *nxt, *link = req->link;
- spin_lock_irqsave(&ctx->completion_lock, flags);
- link = req->link;
req->link = NULL;
-
while (link) {
nxt = link->link;
link->link = NULL;
trace_io_uring_fail_link(req, link);
io_cqring_fill_event(link, -ECANCELED);
-
- /*
- * It's ok to free under spinlock as they're not linked anymore,
- * but avoid REQ_F_WORK_INITIALIZED because it may deadlock on
- * work.fs->lock.
- */
- if (link->flags & REQ_F_WORK_INITIALIZED)
- io_put_req_deferred(link, 2);
- else
- io_double_put_req(link);
+ io_put_req_deferred(link, 2);
link = nxt;
}
- io_commit_cqring(ctx);
- spin_unlock_irqrestore(&ctx->completion_lock, flags);
+}
- io_cqring_ev_posted(ctx);
+static bool io_disarm_next(struct io_kiocb *req)
+ __must_hold(&req->ctx->completion_lock)
+{
+ bool posted = false;
+
+ if (likely(req->flags & REQ_F_LINK_TIMEOUT))
+ posted = io_kill_linked_timeout(req);
+ if (unlikely(req->flags & REQ_F_FAIL_LINK)) {
+ posted |= (req->link != NULL);
+ io_fail_links(req);
+ }
+ return posted;
}
static struct io_kiocb *__io_req_find_next(struct io_kiocb *req)
{
- if (req->flags & REQ_F_LINK_TIMEOUT)
- io_kill_linked_timeout(req);
+ struct io_kiocb *nxt;
/*
* If LINK is set, we have dependent requests in this chain. If we
@@ -1828,14 +1777,22 @@ static struct io_kiocb *__io_req_find_next(struct io_kiocb *req)
* dependencies to the next request. In case of failure, fail the rest
* of the chain.
*/
- if (likely(!(req->flags & REQ_F_FAIL_LINK))) {
- struct io_kiocb *nxt = req->link;
+ if (req->flags & (REQ_F_LINK_TIMEOUT | REQ_F_FAIL_LINK)) {
+ struct io_ring_ctx *ctx = req->ctx;
+ unsigned long flags;
+ bool posted;
- req->link = NULL;
- return nxt;
+ spin_lock_irqsave(&ctx->completion_lock, flags);
+ posted = io_disarm_next(req);
+ if (posted)
+ io_commit_cqring(req->ctx);
+ spin_unlock_irqrestore(&ctx->completion_lock, flags);
+ if (posted)
+ io_cqring_ev_posted(ctx);
}
- io_fail_links(req);
- return NULL;
+ nxt = req->link;
+ req->link = NULL;
+ return nxt;
}
static inline struct io_kiocb *io_req_find_next(struct io_kiocb *req)
@@ -1845,6 +1802,18 @@ static inline struct io_kiocb *io_req_find_next(struct io_kiocb *req)
return __io_req_find_next(req);
}
+static void ctx_flush_and_put(struct io_ring_ctx *ctx)
+{
+ if (!ctx)
+ return;
+ if (ctx->submit_state.comp.nr) {
+ mutex_lock(&ctx->uring_lock);
+ io_submit_flush_completions(&ctx->submit_state.comp, ctx);
+ mutex_unlock(&ctx->uring_lock);
+ }
+ percpu_ref_put(&ctx->refs);
+}
+
static bool __tctx_task_work(struct io_uring_task *tctx)
{
struct io_ring_ctx *ctx = NULL;
@@ -1862,30 +1831,20 @@ static bool __tctx_task_work(struct io_uring_task *tctx)
node = list.first;
while (node) {
struct io_wq_work_node *next = node->next;
- struct io_ring_ctx *this_ctx;
struct io_kiocb *req;
req = container_of(node, struct io_kiocb, io_task_work.node);
- this_ctx = req->ctx;
- req->task_work.func(&req->task_work);
- node = next;
-
- if (!ctx) {
- ctx = this_ctx;
- } else if (ctx != this_ctx) {
- mutex_lock(&ctx->uring_lock);
- io_submit_flush_completions(&ctx->submit_state.comp, ctx);
- mutex_unlock(&ctx->uring_lock);
- ctx = this_ctx;
+ if (req->ctx != ctx) {
+ ctx_flush_and_put(ctx);
+ ctx = req->ctx;
+ percpu_ref_get(&ctx->refs);
}
- }
- if (ctx && ctx->submit_state.comp.nr) {
- mutex_lock(&ctx->uring_lock);
- io_submit_flush_completions(&ctx->submit_state.comp, ctx);
- mutex_unlock(&ctx->uring_lock);
+ req->task_work.func(&req->task_work);
+ node = next;
}
+ ctx_flush_and_put(ctx);
return list.first != NULL;
}
@@ -1893,10 +1852,10 @@ static void tctx_task_work(struct callback_head *cb)
{
struct io_uring_task *tctx = container_of(cb, struct io_uring_task, task_work);
+ clear_bit(0, &tctx->task_state);
+
while (__tctx_task_work(tctx))
cond_resched();
-
- clear_bit(0, &tctx->task_state);
}
static int io_task_work_add(struct task_struct *tsk, struct io_kiocb *req,
@@ -2010,7 +1969,7 @@ static void __io_req_task_submit(struct io_kiocb *req)
/* ctx stays valid until unlock, even if we drop all ours ctx->refs */
mutex_lock(&ctx->uring_lock);
- if (!ctx->sqo_dead && !(current->flags & PF_EXITING) && !current->in_execve)
+ if (!(current->flags & PF_EXITING) && !current->in_execve)
__io_queue_sqe(req);
else
__io_req_task_cancel(req, -EFAULT);
@@ -2472,23 +2431,32 @@ static bool io_resubmit_prep(struct io_kiocb *req)
return false;
return !io_setup_async_rw(req, iovec, inline_vecs, &iter, false);
}
-#endif
-static bool io_rw_reissue(struct io_kiocb *req)
+static bool io_rw_should_reissue(struct io_kiocb *req)
{
-#ifdef CONFIG_BLOCK
umode_t mode = file_inode(req->file)->i_mode;
+ struct io_ring_ctx *ctx = req->ctx;
if (!S_ISBLK(mode) && !S_ISREG(mode))
return false;
- if ((req->flags & REQ_F_NOWAIT) || io_wq_current_is_worker())
+ if ((req->flags & REQ_F_NOWAIT) || (io_wq_current_is_worker() &&
+ !(ctx->flags & IORING_SETUP_IOPOLL)))
return false;
/*
* If ref is dying, we might be running poll reap from the exit work.
* Don't attempt to reissue from that path, just let it fail with
* -EAGAIN.
*/
- if (percpu_ref_is_dying(&req->ctx->refs))
+ if (percpu_ref_is_dying(&ctx->refs))
+ return false;
+ return true;
+}
+#endif
+
+static bool io_rw_reissue(struct io_kiocb *req)
+{
+#ifdef CONFIG_BLOCK
+ if (!io_rw_should_reissue(req))
return false;
lockdep_assert_held(&req->ctx->uring_lock);
@@ -2531,6 +2499,19 @@ static void io_complete_rw_iopoll(struct kiocb *kiocb, long res, long res2)
{
struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw.kiocb);
+#ifdef CONFIG_BLOCK
+ /* Rewind iter, if we have one. iopoll path resubmits as usual */
+ if (res == -EAGAIN && io_rw_should_reissue(req)) {
+ struct io_async_rw *rw = req->async_data;
+
+ if (rw)
+ iov_iter_revert(&rw->iter,
+ req->result - iov_iter_count(&rw->iter));
+ else if (!io_resubmit_prep(req))
+ res = -EIO;
+ }
+#endif
+
if (kiocb->ki_flags & IOCB_WRITE)
kiocb_end_write(req);
@@ -3279,6 +3260,8 @@ static int io_read(struct io_kiocb *req, unsigned int issue_flags)
ret = io_iter_do_read(req, iter);
if (ret == -EIOCBQUEUED) {
+ if (req->async_data)
+ iov_iter_revert(iter, io_size - iov_iter_count(iter));
goto out_free;
} else if (ret == -EAGAIN) {
/* IOPOLL retry should happen for io-wq threads */
@@ -3324,6 +3307,7 @@ static int io_read(struct io_kiocb *req, unsigned int issue_flags)
if (ret == -EIOCBQUEUED)
return 0;
/* we got some bytes, but not all. retry. */
+ kiocb->ki_flags &= ~IOCB_WAITQ;
} while (ret > 0 && ret < io_size);
done:
kiocb_done(kiocb, ret, issue_flags);
@@ -3410,6 +3394,8 @@ static int io_write(struct io_kiocb *req, unsigned int issue_flags)
/* no retry on NONBLOCK nor RWF_NOWAIT */
if (ret2 == -EAGAIN && (req->flags & REQ_F_NOWAIT))
goto done;
+ if (ret2 == -EIOCBQUEUED && req->async_data)
+ iov_iter_revert(iter, io_size - iov_iter_count(iter));
if (!force_nonblock || ret2 != -EAGAIN) {
/* IOPOLL retry should happen for io-wq threads */
if ((req->ctx->flags & IORING_SETUP_IOPOLL) && ret2 == -EAGAIN)
@@ -3588,7 +3574,6 @@ static int __io_splice_prep(struct io_kiocb *req,
* Splice operation will be punted aync, and here need to
* modify io_wq_work.flags, so initialize io_wq_work firstly.
*/
- io_req_init_async(req);
req->work.flags |= IO_WQ_WORK_UNBOUND;
}
@@ -3864,7 +3849,7 @@ err:
static int io_openat(struct io_kiocb *req, unsigned int issue_flags)
{
- return io_openat2(req, issue_flags & IO_URING_F_NONBLOCK);
+ return io_openat2(req, issue_flags);
}
static int io_remove_buffers_prep(struct io_kiocb *req,
@@ -5003,6 +4988,9 @@ static void __io_queue_proc(struct io_poll_iocb *poll, struct io_poll_table *pt,
pt->error = -EINVAL;
return;
}
+ /* double add on the same waitqueue head, ignore */
+ if (poll->head == head)
+ return;
poll = kmalloc(sizeof(*poll), GFP_ATOMIC);
if (!poll) {
pt->error = -ENOMEM;
@@ -5538,6 +5526,7 @@ static int io_timeout_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe,
data->mode = io_translate_timeout_mode(flags);
hrtimer_init(&data->timer, CLOCK_MONOTONIC, data->mode);
+ io_req_track_inflight(req);
return 0;
}
@@ -5591,22 +5580,30 @@ add:
return 0;
}
+struct io_cancel_data {
+ struct io_ring_ctx *ctx;
+ u64 user_data;
+};
+
static bool io_cancel_cb(struct io_wq_work *work, void *data)
{
struct io_kiocb *req = container_of(work, struct io_kiocb, work);
+ struct io_cancel_data *cd = data;
- return req->user_data == (unsigned long) data;
+ return req->ctx == cd->ctx && req->user_data == cd->user_data;
}
-static int io_async_cancel_one(struct io_uring_task *tctx, void *sqe_addr)
+static int io_async_cancel_one(struct io_uring_task *tctx, u64 user_data,
+ struct io_ring_ctx *ctx)
{
+ struct io_cancel_data data = { .ctx = ctx, .user_data = user_data, };
enum io_wq_cancel cancel_ret;
int ret = 0;
- if (!tctx->io_wq)
+ if (!tctx || !tctx->io_wq)
return -ENOENT;
- cancel_ret = io_wq_cancel_cb(tctx->io_wq, io_cancel_cb, sqe_addr, false);
+ cancel_ret = io_wq_cancel_cb(tctx->io_wq, io_cancel_cb, &data, false);
switch (cancel_ret) {
case IO_WQ_CANCEL_OK:
ret = 0;
@@ -5629,8 +5626,7 @@ static void io_async_find_and_cancel(struct io_ring_ctx *ctx,
unsigned long flags;
int ret;
- ret = io_async_cancel_one(req->task->io_uring,
- (void *) (unsigned long) sqe_addr);
+ ret = io_async_cancel_one(req->task->io_uring, sqe_addr, ctx);
if (ret != -ENOENT) {
spin_lock_irqsave(&ctx->completion_lock, flags);
goto done;
@@ -5671,8 +5667,47 @@ static int io_async_cancel_prep(struct io_kiocb *req,
static int io_async_cancel(struct io_kiocb *req, unsigned int issue_flags)
{
struct io_ring_ctx *ctx = req->ctx;
+ u64 sqe_addr = req->cancel.addr;
+ struct io_tctx_node *node;
+ int ret;
- io_async_find_and_cancel(ctx, req, req->cancel.addr, 0);
+ /* tasks should wait for their io-wq threads, so safe w/o sync */
+ ret = io_async_cancel_one(req->task->io_uring, sqe_addr, ctx);
+ spin_lock_irq(&ctx->completion_lock);
+ if (ret != -ENOENT)
+ goto done;
+ ret = io_timeout_cancel(ctx, sqe_addr);
+ if (ret != -ENOENT)
+ goto done;
+ ret = io_poll_cancel(ctx, sqe_addr);
+ if (ret != -ENOENT)
+ goto done;
+ spin_unlock_irq(&ctx->completion_lock);
+
+ /* slow path, try all io-wq's */
+ io_ring_submit_lock(ctx, !(issue_flags & IO_URING_F_NONBLOCK));
+ ret = -ENOENT;
+ list_for_each_entry(node, &ctx->tctx_list, ctx_node) {
+ struct io_uring_task *tctx = node->task->io_uring;
+
+ if (!tctx || !tctx->io_wq)
+ continue;
+ ret = io_async_cancel_one(tctx, req->cancel.addr, ctx);
+ if (ret != -ENOENT)
+ break;
+ }
+ io_ring_submit_unlock(ctx, !(issue_flags & IO_URING_F_NONBLOCK));
+
+ spin_lock_irq(&ctx->completion_lock);
+done:
+ io_cqring_fill_event(req, ret);
+ io_commit_cqring(ctx);
+ spin_unlock_irq(&ctx->completion_lock);
+ io_cqring_ev_posted(ctx);
+
+ if (ret < 0)
+ req_set_fail_links(req);
+ io_put_req(req);
return 0;
}
@@ -5945,8 +5980,12 @@ static void __io_clean_op(struct io_kiocb *req)
static int io_issue_sqe(struct io_kiocb *req, unsigned int issue_flags)
{
struct io_ring_ctx *ctx = req->ctx;
+ const struct cred *creds = NULL;
int ret;
+ if (req->work.creds && req->work.creds != current_cred())
+ creds = override_creds(req->work.creds);
+
switch (req->opcode) {
case IORING_OP_NOP:
ret = io_nop(req, issue_flags);
@@ -6053,6 +6092,9 @@ static int io_issue_sqe(struct io_kiocb *req, unsigned int issue_flags)
break;
}
+ if (creds)
+ revert_creds(creds);
+
if (ret)
return ret;
@@ -6216,18 +6258,10 @@ static struct io_kiocb *io_prep_linked_timeout(struct io_kiocb *req)
static void __io_queue_sqe(struct io_kiocb *req)
{
struct io_kiocb *linked_timeout = io_prep_linked_timeout(req);
- const struct cred *old_creds = NULL;
int ret;
- if ((req->flags & REQ_F_WORK_INITIALIZED) && req->work.creds &&
- req->work.creds != current_cred())
- old_creds = override_creds(req->work.creds);
-
ret = io_issue_sqe(req, IO_URING_F_NONBLOCK|IO_URING_F_COMPLETE_DEFER);
- if (old_creds)
- revert_creds(old_creds);
-
/*
* We async punt it if the file wasn't marked NOWAIT, or if the file
* doesn't support non-blocking read/write attempts
@@ -6314,7 +6348,7 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req,
{
struct io_submit_state *state;
unsigned int sqe_flags;
- int id, ret = 0;
+ int personality, ret = 0;
req->opcode = READ_ONCE(sqe->opcode);
/* same numerical values with corresponding REQ_F_*, safe to copy */
@@ -6329,6 +6363,9 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req,
refcount_set(&req->refs, 2);
req->task = current;
req->result = 0;
+ req->work.list.next = NULL;
+ req->work.creds = NULL;
+ req->work.flags = 0;
/* enforce forwards compatibility on users */
if (unlikely(sqe_flags & ~SQE_VALID_FLAGS)) {
@@ -6346,15 +6383,13 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req,
!io_op_defs[req->opcode].buffer_select)
return -EOPNOTSUPP;
- id = READ_ONCE(sqe->personality);
- if (id) {
- __io_req_init_async(req);
- req->work.creds = idr_find(&ctx->personality_idr, id);
- if (unlikely(!req->work.creds))
+ personality = READ_ONCE(sqe->personality);
+ if (personality) {
+ req->work.creds = xa_load(&ctx->personalities, personality);
+ if (!req->work.creds)
return -EINVAL;
get_cred(req->work.creds);
}
-
state = &ctx->submit_state;
/*
@@ -6616,8 +6651,8 @@ static int __io_sq_thread(struct io_ring_ctx *ctx, bool cap_entries)
if (!list_empty(&ctx->iopoll_list))
io_do_iopoll(ctx, &nr_events, 0);
- if (to_submit && !ctx->sqo_dead &&
- likely(!percpu_ref_is_dying(&ctx->refs)))
+ if (to_submit && likely(!percpu_ref_is_dying(&ctx->refs)) &&
+ !(ctx->flags & IORING_SETUP_R_DISABLED))
ret = io_submit_sqes(ctx, to_submit);
mutex_unlock(&ctx->uring_lock);
}
@@ -6641,58 +6676,6 @@ static void io_sqd_update_thread_idle(struct io_sq_data *sqd)
sqd->sq_thread_idle = sq_thread_idle;
}
-static void io_sqd_init_new(struct io_sq_data *sqd)
-{
- struct io_ring_ctx *ctx;
-
- while (!list_empty(&sqd->ctx_new_list)) {
- ctx = list_first_entry(&sqd->ctx_new_list, struct io_ring_ctx, sqd_list);
- list_move_tail(&ctx->sqd_list, &sqd->ctx_list);
- complete(&ctx->sq_thread_comp);
- }
-
- io_sqd_update_thread_idle(sqd);
-}
-
-static bool io_sq_thread_should_stop(struct io_sq_data *sqd)
-{
- return test_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state);
-}
-
-static bool io_sq_thread_should_park(struct io_sq_data *sqd)
-{
- return test_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state);
-}
-
-static void io_sq_thread_parkme(struct io_sq_data *sqd)
-{
- for (;;) {
- /*
- * TASK_PARKED is a special state; we must serialize against
- * possible pending wakeups to avoid store-store collisions on
- * task->state.
- *
- * Such a collision might possibly result in the task state
- * changin from TASK_PARKED and us failing the
- * wait_task_inactive() in kthread_park().
- */
- set_special_state(TASK_PARKED);
- if (!test_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state))
- break;
-
- /*
- * Thread is going to call schedule(), do not preempt it,
- * or the caller of kthread_park() may spend more time in
- * wait_task_inactive().
- */
- preempt_disable();
- complete(&sqd->completion);
- schedule_preempt_disabled();
- preempt_enable();
- }
- __set_current_state(TASK_RUNNING);
-}
-
static int io_sq_thread(void *data)
{
struct io_sq_data *sqd = data;
@@ -6703,7 +6686,6 @@ static int io_sq_thread(void *data)
sprintf(buf, "iou-sqp-%d", sqd->task_pid);
set_task_comm(current, buf);
- sqd->thread = current;
current->pf_io_worker = NULL;
if (sqd->sq_cpu != -1)
@@ -6712,33 +6694,32 @@ static int io_sq_thread(void *data)
set_cpus_allowed_ptr(current, cpu_online_mask);
current->flags |= PF_NO_SETAFFINITY;
- complete(&sqd->completion);
+ down_read(&sqd->rw_lock);
- wait_for_completion(&sqd->startup);
-
- while (!io_sq_thread_should_stop(sqd)) {
+ while (!test_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state)) {
int ret;
bool cap_entries, sqt_spin, needs_sched;
- /*
- * Any changes to the sqd lists are synchronized through the
- * thread parking. This synchronizes the thread vs users,
- * the users are synchronized on the sqd->ctx_lock.
- */
- if (io_sq_thread_should_park(sqd)) {
- io_sq_thread_parkme(sqd);
- continue;
- }
- if (unlikely(!list_empty(&sqd->ctx_new_list))) {
- io_sqd_init_new(sqd);
+ if (test_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state)) {
+ up_read(&sqd->rw_lock);
+ cond_resched();
+ down_read(&sqd->rw_lock);
+ io_run_task_work();
timeout = jiffies + sqd->sq_thread_idle;
+ continue;
}
if (fatal_signal_pending(current))
break;
sqt_spin = false;
cap_entries = !list_is_singular(&sqd->ctx_list);
list_for_each_entry(ctx, &sqd->ctx_list, sqd_list) {
+ const struct cred *creds = NULL;
+
+ if (ctx->sq_creds != current_cred())
+ creds = override_creds(ctx->sq_creds);
ret = __io_sq_thread(ctx, cap_entries);
+ if (creds)
+ revert_creds(creds);
if (!sqt_spin && (ret > 0 || !list_empty(&ctx->iopoll_list)))
sqt_spin = true;
}
@@ -6765,11 +6746,13 @@ static int io_sq_thread(void *data)
}
}
- if (needs_sched && !io_sq_thread_should_park(sqd)) {
+ if (needs_sched && !test_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state)) {
list_for_each_entry(ctx, &sqd->ctx_list, sqd_list)
io_ring_set_wakeup_flag(ctx);
+ up_read(&sqd->rw_lock);
schedule();
+ down_read(&sqd->rw_lock);
list_for_each_entry(ctx, &sqd->ctx_list, sqd_list)
io_ring_clear_wakeup_flag(ctx);
}
@@ -6777,24 +6760,22 @@ static int io_sq_thread(void *data)
finish_wait(&sqd->wait, &wait);
timeout = jiffies + sqd->sq_thread_idle;
}
+ up_read(&sqd->rw_lock);
+ down_write(&sqd->rw_lock);
+ /*
+ * someone may have parked and added a cancellation task_work, run
+ * it first because we don't want it in io_uring_cancel_sqpoll()
+ */
+ io_run_task_work();
list_for_each_entry(ctx, &sqd->ctx_list, sqd_list)
io_uring_cancel_sqpoll(ctx);
-
- io_run_task_work();
-
- /*
- * Clear thread under lock so that concurrent parks work correctly
- */
- complete_all(&sqd->completion);
- mutex_lock(&sqd->lock);
sqd->thread = NULL;
- list_for_each_entry(ctx, &sqd->ctx_list, sqd_list) {
- ctx->sqo_exec = 1;
+ list_for_each_entry(ctx, &sqd->ctx_list, sqd_list)
io_ring_set_wakeup_flag(ctx);
- }
- mutex_unlock(&sqd->lock);
+ up_write(&sqd->rw_lock);
+ io_run_task_work();
complete(&sqd->exited);
do_exit(0);
}
@@ -6917,11 +6898,16 @@ static int io_cqring_wait(struct io_ring_ctx *ctx, int min_events,
iowq.nr_timeouts = atomic_read(&ctx->cq_timeouts);
trace_io_uring_cqring_wait(ctx, min_events);
do {
- io_cqring_overflow_flush(ctx, false, NULL, NULL);
+ /* if we can't even flush overflow, don't wait for more */
+ if (!io_cqring_overflow_flush(ctx, false, NULL, NULL)) {
+ ret = -EBUSY;
+ break;
+ }
prepare_to_wait_exclusive(&ctx->wait, &iowq.wq,
TASK_INTERRUPTIBLE);
ret = io_cqring_wait_schedule(ctx, &iowq, &timeout);
finish_wait(&ctx->wait, &iowq.wq);
+ cond_resched();
} while (ret > 0);
restore_saved_sigmask_unless(ret == -EINTR);
@@ -7089,41 +7075,36 @@ static int io_sqe_files_unregister(struct io_ring_ctx *ctx)
}
static void io_sq_thread_unpark(struct io_sq_data *sqd)
- __releases(&sqd->lock)
+ __releases(&sqd->rw_lock)
{
- if (!sqd->thread)
- return;
- if (sqd->thread == current)
- return;
+ WARN_ON_ONCE(sqd->thread == current);
+
clear_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state);
- wake_up_state(sqd->thread, TASK_PARKED);
- mutex_unlock(&sqd->lock);
+ up_write(&sqd->rw_lock);
}
-static bool io_sq_thread_park(struct io_sq_data *sqd)
- __acquires(&sqd->lock)
+static void io_sq_thread_park(struct io_sq_data *sqd)
+ __acquires(&sqd->rw_lock)
{
- if (sqd->thread == current)
- return true;
- mutex_lock(&sqd->lock);
- if (!sqd->thread) {
- mutex_unlock(&sqd->lock);
- return false;
- }
+ WARN_ON_ONCE(sqd->thread == current);
+
set_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state);
- wake_up_process(sqd->thread);
- wait_for_completion(&sqd->completion);
- return true;
+ down_write(&sqd->rw_lock);
+ /* set again for consistency, in case concurrent parks are happening */
+ set_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state);
+ if (sqd->thread)
+ wake_up_process(sqd->thread);
}
static void io_sq_thread_stop(struct io_sq_data *sqd)
{
- if (!sqd->thread)
- return;
+ WARN_ON_ONCE(sqd->thread == current);
+ down_write(&sqd->rw_lock);
set_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state);
- WARN_ON_ONCE(test_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state));
- wake_up_process(sqd->thread);
+ if (sqd->thread)
+ wake_up_process(sqd->thread);
+ up_write(&sqd->rw_lock);
wait_for_completion(&sqd->exited);
}
@@ -7140,22 +7121,15 @@ static void io_sq_thread_finish(struct io_ring_ctx *ctx)
struct io_sq_data *sqd = ctx->sq_data;
if (sqd) {
- complete(&sqd->startup);
- if (sqd->thread) {
- wait_for_completion(&ctx->sq_thread_comp);
- io_sq_thread_park(sqd);
- }
-
- mutex_lock(&sqd->ctx_lock);
- list_del(&ctx->sqd_list);
+ io_sq_thread_park(sqd);
+ list_del_init(&ctx->sqd_list);
io_sqd_update_thread_idle(sqd);
- mutex_unlock(&sqd->ctx_lock);
-
- if (sqd->thread)
- io_sq_thread_unpark(sqd);
+ io_sq_thread_unpark(sqd);
io_put_sq_data(sqd);
ctx->sq_data = NULL;
+ if (ctx->sq_creds)
+ put_cred(ctx->sq_creds);
}
}
@@ -7179,18 +7153,32 @@ static struct io_sq_data *io_attach_sq_data(struct io_uring_params *p)
fdput(f);
return ERR_PTR(-EINVAL);
}
+ if (sqd->task_tgid != current->tgid) {
+ fdput(f);
+ return ERR_PTR(-EPERM);
+ }
refcount_inc(&sqd->refs);
fdput(f);
return sqd;
}
-static struct io_sq_data *io_get_sq_data(struct io_uring_params *p)
+static struct io_sq_data *io_get_sq_data(struct io_uring_params *p,
+ bool *attached)
{
struct io_sq_data *sqd;
- if (p->flags & IORING_SETUP_ATTACH_WQ)
- return io_attach_sq_data(p);
+ *attached = false;
+ if (p->flags & IORING_SETUP_ATTACH_WQ) {
+ sqd = io_attach_sq_data(p);
+ if (!IS_ERR(sqd)) {
+ *attached = true;
+ return sqd;
+ }
+ /* fall through for EPERM case, setup new sqd/task */
+ if (PTR_ERR(sqd) != -EPERM)
+ return sqd;
+ }
sqd = kzalloc(sizeof(*sqd), GFP_KERNEL);
if (!sqd)
@@ -7198,12 +7186,8 @@ static struct io_sq_data *io_get_sq_data(struct io_uring_params *p)
refcount_set(&sqd->refs, 1);
INIT_LIST_HEAD(&sqd->ctx_list);
- INIT_LIST_HEAD(&sqd->ctx_new_list);
- mutex_init(&sqd->ctx_lock);
- mutex_init(&sqd->lock);
+ init_rwsem(&sqd->rw_lock);
init_waitqueue_head(&sqd->wait);
- init_completion(&sqd->startup);
- init_completion(&sqd->completion);
init_completion(&sqd->exited);
return sqd;
}
@@ -7820,7 +7804,6 @@ static int io_uring_alloc_task_context(struct task_struct *task,
init_waitqueue_head(&tctx->wait);
tctx->last = NULL;
atomic_set(&tctx->in_idle, 0);
- tctx->sqpoll = false;
task->io_uring = tctx;
spin_lock_init(&tctx->task_lock);
INIT_WQ_LIST(&tctx->task_list);
@@ -7834,30 +7817,13 @@ void __io_uring_free(struct task_struct *tsk)
struct io_uring_task *tctx = tsk->io_uring;
WARN_ON_ONCE(!xa_empty(&tctx->xa));
+ WARN_ON_ONCE(tctx->io_wq);
+
percpu_counter_destroy(&tctx->inflight);
kfree(tctx);
tsk->io_uring = NULL;
}
-static int io_sq_thread_fork(struct io_sq_data *sqd, struct io_ring_ctx *ctx)
-{
- int ret;
-
- clear_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state);
- reinit_completion(&sqd->completion);
- ctx->sqo_dead = ctx->sqo_exec = 0;
- sqd->task_pid = current->pid;
- current->flags |= PF_IO_WORKER;
- ret = io_wq_fork_thread(io_sq_thread, sqd);
- current->flags &= ~PF_IO_WORKER;
- if (ret < 0) {
- sqd->thread = NULL;
- return ret;
- }
- wait_for_completion(&sqd->completion);
- return io_uring_alloc_task_context(sqd->thread, ctx);
-}
-
static int io_sq_offload_create(struct io_ring_ctx *ctx,
struct io_uring_params *p)
{
@@ -7878,40 +7844,53 @@ static int io_sq_offload_create(struct io_ring_ctx *ctx,
fdput(f);
}
if (ctx->flags & IORING_SETUP_SQPOLL) {
+ struct task_struct *tsk;
struct io_sq_data *sqd;
+ bool attached;
ret = -EPERM;
if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_NICE))
goto err;
- sqd = io_get_sq_data(p);
+ sqd = io_get_sq_data(p, &attached);
if (IS_ERR(sqd)) {
ret = PTR_ERR(sqd);
goto err;
}
+ ctx->sq_creds = get_current_cred();
ctx->sq_data = sqd;
- io_sq_thread_park(sqd);
- mutex_lock(&sqd->ctx_lock);
- list_add(&ctx->sqd_list, &sqd->ctx_new_list);
- mutex_unlock(&sqd->ctx_lock);
- io_sq_thread_unpark(sqd);
-
ctx->sq_thread_idle = msecs_to_jiffies(p->sq_thread_idle);
if (!ctx->sq_thread_idle)
ctx->sq_thread_idle = HZ;
- if (sqd->thread)
+ ret = 0;
+ io_sq_thread_park(sqd);
+ /* don't attach to a dying SQPOLL thread, would be racy */
+ if (attached && !sqd->thread) {
+ ret = -ENXIO;
+ } else {
+ list_add(&ctx->sqd_list, &sqd->ctx_list);
+ io_sqd_update_thread_idle(sqd);
+ }
+ io_sq_thread_unpark(sqd);
+
+ if (ret < 0) {
+ io_put_sq_data(sqd);
+ ctx->sq_data = NULL;
+ return ret;
+ } else if (attached) {
return 0;
+ }
if (p->flags & IORING_SETUP_SQ_AFF) {
int cpu = p->sq_thread_cpu;
ret = -EINVAL;
if (cpu >= nr_cpu_ids)
- goto err;
+ goto err_sqpoll;
if (!cpu_online(cpu))
- goto err;
+ goto err_sqpoll;
sqd->sq_cpu = cpu;
} else {
@@ -7919,15 +7898,16 @@ static int io_sq_offload_create(struct io_ring_ctx *ctx,
}
sqd->task_pid = current->pid;
- current->flags |= PF_IO_WORKER;
- ret = io_wq_fork_thread(io_sq_thread, sqd);
- current->flags &= ~PF_IO_WORKER;
- if (ret < 0) {
- sqd->thread = NULL;
- goto err;
+ sqd->task_tgid = current->tgid;
+ tsk = create_io_thread(io_sq_thread, sqd, NUMA_NO_NODE);
+ if (IS_ERR(tsk)) {
+ ret = PTR_ERR(tsk);
+ goto err_sqpoll;
}
- wait_for_completion(&sqd->completion);
- ret = io_uring_alloc_task_context(sqd->thread, ctx);
+
+ sqd->thread = tsk;
+ ret = io_uring_alloc_task_context(tsk, ctx);
+ wake_up_new_task(tsk);
if (ret)
goto err;
} else if (p->flags & IORING_SETUP_SQ_AFF) {
@@ -7940,14 +7920,9 @@ static int io_sq_offload_create(struct io_ring_ctx *ctx,
err:
io_sq_thread_finish(ctx);
return ret;
-}
-
-static void io_sq_offload_start(struct io_ring_ctx *ctx)
-{
- struct io_sq_data *sqd = ctx->sq_data;
-
- if (ctx->flags & IORING_SETUP_SQPOLL)
- complete(&sqd->startup);
+err_sqpoll:
+ complete(&ctx->sq_data->exited);
+ goto err;
}
static inline void __io_unaccount_mem(struct user_struct *user,
@@ -8384,7 +8359,7 @@ static void io_req_cache_free(struct list_head *list, struct task_struct *tsk)
}
}
-static void io_req_caches_free(struct io_ring_ctx *ctx, struct task_struct *tsk)
+static void io_req_caches_free(struct io_ring_ctx *ctx)
{
struct io_submit_state *submit_state = &ctx->submit_state;
struct io_comp_state *cs = &ctx->submit_state.comp;
@@ -8430,7 +8405,6 @@ static void io_ring_ctx_free(struct io_ring_ctx *ctx)
mutex_unlock(&ctx->uring_lock);
io_eventfd_unregister(ctx);
io_destroy_buffers(ctx);
- idr_destroy(&ctx->personality_idr);
#if defined(CONFIG_UNIX)
if (ctx->ring_sock) {
@@ -8444,7 +8418,7 @@ static void io_ring_ctx_free(struct io_ring_ctx *ctx)
percpu_ref_exit(&ctx->refs);
free_uid(ctx->user);
- io_req_caches_free(ctx, NULL);
+ io_req_caches_free(ctx);
if (ctx->hash_map)
io_wq_put_hash(ctx->hash_map);
kfree(ctx->cancel_hash);
@@ -8495,7 +8469,7 @@ static int io_unregister_personality(struct io_ring_ctx *ctx, unsigned id)
{
const struct cred *creds;
- creds = idr_remove(&ctx->personality_idr, id);
+ creds = xa_erase(&ctx->personalities, id);
if (creds) {
put_cred(creds);
return 0;
@@ -8504,24 +8478,13 @@ static int io_unregister_personality(struct io_ring_ctx *ctx, unsigned id)
return -EINVAL;
}
-static int io_remove_personalities(int id, void *p, void *data)
+static bool io_run_ctx_fallback(struct io_ring_ctx *ctx)
{
- struct io_ring_ctx *ctx = data;
-
- io_unregister_personality(ctx, id);
- return 0;
-}
-
-static void io_run_ctx_fallback(struct io_ring_ctx *ctx)
-{
- struct callback_head *work, *head, *next;
+ struct callback_head *work, *next;
+ bool executed = false;
do {
- do {
- head = NULL;
- work = READ_ONCE(ctx->exit_task_work);
- } while (cmpxchg(&ctx->exit_task_work, work, head) != work);
-
+ work = xchg(&ctx->exit_task_work, NULL);
if (!work)
break;
@@ -8531,13 +8494,40 @@ static void io_run_ctx_fallback(struct io_ring_ctx *ctx)
work = next;
cond_resched();
} while (work);
+ executed = true;
} while (1);
+
+ return executed;
+}
+
+struct io_tctx_exit {
+ struct callback_head task_work;
+ struct completion completion;
+ struct io_ring_ctx *ctx;
+};
+
+static void io_tctx_exit_cb(struct callback_head *cb)
+{
+ struct io_uring_task *tctx = current->io_uring;
+ struct io_tctx_exit *work;
+
+ work = container_of(cb, struct io_tctx_exit, task_work);
+ /*
+ * When @in_idle, we're in cancellation and it's racy to remove the
+ * node. It'll be removed by the end of cancellation, just ignore it.
+ */
+ if (!atomic_read(&tctx->in_idle))
+ io_uring_del_task_file((unsigned long)work->ctx);
+ complete(&work->completion);
}
static void io_ring_exit_work(struct work_struct *work)
{
- struct io_ring_ctx *ctx = container_of(work, struct io_ring_ctx,
- exit_work);
+ struct io_ring_ctx *ctx = container_of(work, struct io_ring_ctx, exit_work);
+ unsigned long timeout = jiffies + HZ * 60 * 5;
+ struct io_tctx_exit exit;
+ struct io_tctx_node *node;
+ int ret;
/*
* If we're doing polled IO and end up having requests being
@@ -8547,24 +8537,47 @@ static void io_ring_exit_work(struct work_struct *work)
*/
do {
io_uring_try_cancel_requests(ctx, NULL, NULL);
- io_run_ctx_fallback(ctx);
+
+ WARN_ON_ONCE(time_after(jiffies, timeout));
} while (!wait_for_completion_timeout(&ctx->ref_comp, HZ/20));
+
+ mutex_lock(&ctx->uring_lock);
+ while (!list_empty(&ctx->tctx_list)) {
+ WARN_ON_ONCE(time_after(jiffies, timeout));
+
+ node = list_first_entry(&ctx->tctx_list, struct io_tctx_node,
+ ctx_node);
+ exit.ctx = ctx;
+ init_completion(&exit.completion);
+ init_task_work(&exit.task_work, io_tctx_exit_cb);
+ ret = task_work_add(node->task, &exit.task_work, TWA_SIGNAL);
+ if (WARN_ON_ONCE(ret))
+ continue;
+ wake_up_process(node->task);
+
+ mutex_unlock(&ctx->uring_lock);
+ wait_for_completion(&exit.completion);
+ cond_resched();
+ mutex_lock(&ctx->uring_lock);
+ }
+ mutex_unlock(&ctx->uring_lock);
+
io_ring_ctx_free(ctx);
}
static void io_ring_ctx_wait_and_kill(struct io_ring_ctx *ctx)
{
+ unsigned long index;
+ struct creds *creds;
+
mutex_lock(&ctx->uring_lock);
percpu_ref_kill(&ctx->refs);
-
- if (WARN_ON_ONCE((ctx->flags & IORING_SETUP_SQPOLL) && !ctx->sqo_dead))
- ctx->sqo_dead = 1;
-
/* if force is set, the ring is going away. always drop after that */
ctx->cq_overflow_flushed = 1;
if (ctx->rings)
__io_cqring_overflow_flush(ctx, true, NULL, NULL);
- idr_for_each(&ctx->personality_idr, io_remove_personalities, ctx);
+ xa_for_each(&ctx->personalities, index, creds)
+ io_unregister_personality(ctx, index);
mutex_unlock(&ctx->uring_lock);
io_kill_timeouts(ctx, NULL, NULL);
@@ -8617,11 +8630,11 @@ static bool io_cancel_task_cb(struct io_wq_work *work, void *data)
return ret;
}
-static void io_cancel_defer_files(struct io_ring_ctx *ctx,
+static bool io_cancel_defer_files(struct io_ring_ctx *ctx,
struct task_struct *task,
struct files_struct *files)
{
- struct io_defer_entry *de = NULL;
+ struct io_defer_entry *de;
LIST_HEAD(list);
spin_lock_irq(&ctx->completion_lock);
@@ -8632,6 +8645,8 @@ static void io_cancel_defer_files(struct io_ring_ctx *ctx,
}
}
spin_unlock_irq(&ctx->completion_lock);
+ if (list_empty(&list))
+ return false;
while (!list_empty(&list)) {
de = list_first_entry(&list, struct io_defer_entry, list);
@@ -8641,6 +8656,38 @@ static void io_cancel_defer_files(struct io_ring_ctx *ctx,
io_req_complete(de->req, -ECANCELED);
kfree(de);
}
+ return true;
+}
+
+static bool io_cancel_ctx_cb(struct io_wq_work *work, void *data)
+{
+ struct io_kiocb *req = container_of(work, struct io_kiocb, work);
+
+ return req->ctx == data;
+}
+
+static bool io_uring_try_cancel_iowq(struct io_ring_ctx *ctx)
+{
+ struct io_tctx_node *node;
+ enum io_wq_cancel cret;
+ bool ret = false;
+
+ mutex_lock(&ctx->uring_lock);
+ list_for_each_entry(node, &ctx->tctx_list, ctx_node) {
+ struct io_uring_task *tctx = node->task->io_uring;
+
+ /*
+ * io_wq will stay alive while we hold uring_lock, because it's
+ * killed after ctx nodes, which requires to take the lock.
+ */
+ if (!tctx || !tctx->io_wq)
+ continue;
+ cret = io_wq_cancel_cb(tctx->io_wq, io_cancel_ctx_cb, ctx, true);
+ ret |= (cret != IO_WQ_CANCEL_NOTFOUND);
+ }
+ mutex_unlock(&ctx->uring_lock);
+
+ return ret;
}
static void io_uring_try_cancel_requests(struct io_ring_ctx *ctx,
@@ -8648,29 +8695,38 @@ static void io_uring_try_cancel_requests(struct io_ring_ctx *ctx,
struct files_struct *files)
{
struct io_task_cancel cancel = { .task = task, .files = files, };
- struct io_uring_task *tctx = current->io_uring;
+ struct io_uring_task *tctx = task ? task->io_uring : NULL;
while (1) {
enum io_wq_cancel cret;
bool ret = false;
- if (tctx && tctx->io_wq) {
+ if (!task) {
+ ret |= io_uring_try_cancel_iowq(ctx);
+ } else if (tctx && tctx->io_wq) {
+ /*
+ * Cancels requests of all rings, not only @ctx, but
+ * it's fine as the task is in exit/exec.
+ */
cret = io_wq_cancel_cb(tctx->io_wq, io_cancel_task_cb,
&cancel, true);
ret |= (cret != IO_WQ_CANCEL_NOTFOUND);
}
/* SQPOLL thread does its own polling */
- if (!(ctx->flags & IORING_SETUP_SQPOLL) && !files) {
+ if ((!(ctx->flags & IORING_SETUP_SQPOLL) && !files) ||
+ (ctx->sq_data && ctx->sq_data->thread == current)) {
while (!list_empty_careful(&ctx->iopoll_list)) {
io_iopoll_try_reap_events(ctx);
ret = true;
}
}
+ ret |= io_cancel_defer_files(ctx, task, files);
ret |= io_poll_remove_all(ctx, task, files);
ret |= io_kill_timeouts(ctx, task, files);
ret |= io_run_task_work();
+ ret |= io_run_ctx_fallback(ctx);
io_cqring_overflow_flush(ctx, true, task, files);
if (!ret)
break;
@@ -8706,67 +8762,21 @@ static void io_uring_cancel_files(struct io_ring_ctx *ctx,
io_uring_try_cancel_requests(ctx, task, files);
- if (ctx->sq_data)
- io_sq_thread_unpark(ctx->sq_data);
prepare_to_wait(&task->io_uring->wait, &wait,
TASK_UNINTERRUPTIBLE);
if (inflight == io_uring_count_inflight(ctx, task, files))
schedule();
finish_wait(&task->io_uring->wait, &wait);
- if (ctx->sq_data)
- io_sq_thread_park(ctx->sq_data);
- }
-}
-
-static void io_disable_sqo_submit(struct io_ring_ctx *ctx)
-{
- mutex_lock(&ctx->uring_lock);
- ctx->sqo_dead = 1;
- mutex_unlock(&ctx->uring_lock);
-
- /* make sure callers enter the ring to get error */
- if (ctx->rings)
- io_ring_set_wakeup_flag(ctx);
-}
-
-/*
- * We need to iteratively cancel requests, in case a request has dependent
- * hard links. These persist even for failure of cancelations, hence keep
- * looping until none are found.
- */
-static void io_uring_cancel_task_requests(struct io_ring_ctx *ctx,
- struct files_struct *files)
-{
- struct task_struct *task = current;
- bool did_park = false;
-
- if ((ctx->flags & IORING_SETUP_SQPOLL) && ctx->sq_data) {
- io_disable_sqo_submit(ctx);
- did_park = io_sq_thread_park(ctx->sq_data);
- if (did_park) {
- task = ctx->sq_data->thread;
- atomic_inc(&task->io_uring->in_idle);
- }
- }
-
- io_cancel_defer_files(ctx, task, files);
-
- io_uring_cancel_files(ctx, task, files);
- if (!files)
- io_uring_try_cancel_requests(ctx, task, NULL);
-
- if (did_park) {
- atomic_dec(&task->io_uring->in_idle);
- io_sq_thread_unpark(ctx->sq_data);
}
}
/*
* Note that this task has used io_uring. We use it for cancelation purposes.
*/
-static int io_uring_add_task_file(struct io_ring_ctx *ctx, struct file *file)
+static int io_uring_add_task_file(struct io_ring_ctx *ctx)
{
struct io_uring_task *tctx = current->io_uring;
+ struct io_tctx_node *node;
int ret;
if (unlikely(!tctx)) {
@@ -8775,98 +8785,143 @@ static int io_uring_add_task_file(struct io_ring_ctx *ctx, struct file *file)
return ret;
tctx = current->io_uring;
}
- if (tctx->last != file) {
- void *old = xa_load(&tctx->xa, (unsigned long)file);
+ if (tctx->last != ctx) {
+ void *old = xa_load(&tctx->xa, (unsigned long)ctx);
if (!old) {
- get_file(file);
- ret = xa_err(xa_store(&tctx->xa, (unsigned long)file,
- file, GFP_KERNEL));
+ node = kmalloc(sizeof(*node), GFP_KERNEL);
+ if (!node)
+ return -ENOMEM;
+ node->ctx = ctx;
+ node->task = current;
+
+ ret = xa_err(xa_store(&tctx->xa, (unsigned long)ctx,
+ node, GFP_KERNEL));
if (ret) {
- fput(file);
+ kfree(node);
return ret;
}
- /* one and only SQPOLL file note, held by sqo_task */
- WARN_ON_ONCE((ctx->flags & IORING_SETUP_SQPOLL) &&
- current != ctx->sqo_task);
+ mutex_lock(&ctx->uring_lock);
+ list_add(&node->ctx_node, &ctx->tctx_list);
+ mutex_unlock(&ctx->uring_lock);
}
- tctx->last = file;
+ tctx->last = ctx;
}
-
- /*
- * This is race safe in that the task itself is doing this, hence it
- * cannot be going through the exit/cancel paths at the same time.
- * This cannot be modified while exit/cancel is running.
- */
- if (!tctx->sqpoll && (ctx->flags & IORING_SETUP_SQPOLL))
- tctx->sqpoll = true;
-
return 0;
}
/*
* Remove this io_uring_file -> task mapping.
*/
-static void io_uring_del_task_file(struct file *file)
+static void io_uring_del_task_file(unsigned long index)
{
struct io_uring_task *tctx = current->io_uring;
+ struct io_tctx_node *node;
- if (tctx->last == file)
+ if (!tctx)
+ return;
+ node = xa_erase(&tctx->xa, index);
+ if (!node)
+ return;
+
+ WARN_ON_ONCE(current != node->task);
+ WARN_ON_ONCE(list_empty(&node->ctx_node));
+
+ mutex_lock(&node->ctx->uring_lock);
+ list_del(&node->ctx_node);
+ mutex_unlock(&node->ctx->uring_lock);
+
+ if (tctx->last == node->ctx)
tctx->last = NULL;
- file = xa_erase(&tctx->xa, (unsigned long)file);
- if (file)
- fput(file);
+ kfree(node);
}
-static void io_uring_remove_task_files(struct io_uring_task *tctx)
+static void io_uring_clean_tctx(struct io_uring_task *tctx)
{
- struct file *file;
+ struct io_tctx_node *node;
unsigned long index;
- xa_for_each(&tctx->xa, index, file)
- io_uring_del_task_file(file);
+ xa_for_each(&tctx->xa, index, node)
+ io_uring_del_task_file(index);
+ if (tctx->io_wq) {
+ io_wq_put_and_exit(tctx->io_wq);
+ tctx->io_wq = NULL;
+ }
+}
+
+static s64 tctx_inflight(struct io_uring_task *tctx)
+{
+ return percpu_counter_sum(&tctx->inflight);
+}
+
+static void io_sqpoll_cancel_cb(struct callback_head *cb)
+{
+ struct io_tctx_exit *work = container_of(cb, struct io_tctx_exit, task_work);
+ struct io_ring_ctx *ctx = work->ctx;
+ struct io_sq_data *sqd = ctx->sq_data;
+
+ if (sqd->thread)
+ io_uring_cancel_sqpoll(ctx);
+ complete(&work->completion);
+}
+
+static void io_sqpoll_cancel_sync(struct io_ring_ctx *ctx)
+{
+ struct io_sq_data *sqd = ctx->sq_data;
+ struct io_tctx_exit work = { .ctx = ctx, };
+ struct task_struct *task;
+
+ io_sq_thread_park(sqd);
+ list_del_init(&ctx->sqd_list);
+ io_sqd_update_thread_idle(sqd);
+ task = sqd->thread;
+ if (task) {
+ init_completion(&work.completion);
+ init_task_work(&work.task_work, io_sqpoll_cancel_cb);
+ WARN_ON_ONCE(task_work_add(task, &work.task_work, TWA_SIGNAL));
+ wake_up_process(task);
+ }
+ io_sq_thread_unpark(sqd);
+
+ if (task)
+ wait_for_completion(&work.completion);
}
void __io_uring_files_cancel(struct files_struct *files)
{
struct io_uring_task *tctx = current->io_uring;
- struct file *file;
+ struct io_tctx_node *node;
unsigned long index;
/* make sure overflow events are dropped */
atomic_inc(&tctx->in_idle);
- xa_for_each(&tctx->xa, index, file)
- io_uring_cancel_task_requests(file->private_data, files);
- atomic_dec(&tctx->in_idle);
+ xa_for_each(&tctx->xa, index, node) {
+ struct io_ring_ctx *ctx = node->ctx;
- if (files) {
- io_uring_remove_task_files(tctx);
- if (tctx->io_wq) {
- io_wq_put(tctx->io_wq);
- tctx->io_wq = NULL;
+ if (ctx->sq_data) {
+ io_sqpoll_cancel_sync(ctx);
+ continue;
}
+ io_uring_cancel_files(ctx, current, files);
+ if (!files)
+ io_uring_try_cancel_requests(ctx, current, NULL);
}
-}
+ atomic_dec(&tctx->in_idle);
-static s64 tctx_inflight(struct io_uring_task *tctx)
-{
- return percpu_counter_sum(&tctx->inflight);
+ if (files)
+ io_uring_clean_tctx(tctx);
}
+/* should only be called by SQPOLL task */
static void io_uring_cancel_sqpoll(struct io_ring_ctx *ctx)
{
struct io_sq_data *sqd = ctx->sq_data;
- struct io_uring_task *tctx;
+ struct io_uring_task *tctx = current->io_uring;
s64 inflight;
DEFINE_WAIT(wait);
- if (!sqd)
- return;
- io_disable_sqo_submit(ctx);
- if (!io_sq_thread_park(sqd))
- return;
- tctx = ctx->sq_data->thread->io_uring;
+ WARN_ON_ONCE(!sqd || ctx->sq_data->thread != current);
atomic_inc(&tctx->in_idle);
do {
@@ -8874,7 +8929,7 @@ static void io_uring_cancel_sqpoll(struct io_ring_ctx *ctx)
inflight = tctx_inflight(tctx);
if (!inflight)
break;
- io_uring_cancel_task_requests(ctx, NULL);
+ io_uring_try_cancel_requests(ctx, current, NULL);
prepare_to_wait(&tctx->wait, &wait, TASK_UNINTERRUPTIBLE);
/*
@@ -8887,7 +8942,6 @@ static void io_uring_cancel_sqpoll(struct io_ring_ctx *ctx)
finish_wait(&tctx->wait, &wait);
} while (1);
atomic_dec(&tctx->in_idle);
- io_sq_thread_unpark(sqd);
}
/*
@@ -8902,16 +8956,6 @@ void __io_uring_task_cancel(void)
/* make sure overflow events are dropped */
atomic_inc(&tctx->in_idle);
-
- /* trigger io_disable_sqo_submit() */
- if (tctx->sqpoll) {
- struct file *file;
- unsigned long index;
-
- xa_for_each(&tctx->xa, index, file)
- io_uring_cancel_sqpoll(file->private_data);
- }
-
do {
/* read completions before cancelations */
inflight = tctx_inflight(tctx);
@@ -8933,53 +8977,9 @@ void __io_uring_task_cancel(void)
atomic_dec(&tctx->in_idle);
- io_uring_remove_task_files(tctx);
-}
-
-static int io_uring_flush(struct file *file, void *data)
-{
- struct io_uring_task *tctx = current->io_uring;
- struct io_ring_ctx *ctx = file->private_data;
-
- /* Ignore helper thread files exit */
- if (current->flags & PF_IO_WORKER)
- return 0;
-
- if (fatal_signal_pending(current) || (current->flags & PF_EXITING)) {
- io_uring_cancel_task_requests(ctx, NULL);
- io_req_caches_free(ctx, current);
- }
-
- io_run_ctx_fallback(ctx);
-
- if (!tctx)
- return 0;
-
- /* we should have cancelled and erased it before PF_EXITING */
- WARN_ON_ONCE((current->flags & PF_EXITING) &&
- xa_load(&tctx->xa, (unsigned long)file));
-
- /*
- * fput() is pending, will be 2 if the only other ref is our potential
- * task file note. If the task is exiting, drop regardless of count.
- */
- if (atomic_long_read(&file->f_count) != 2)
- return 0;
-
- if (ctx->flags & IORING_SETUP_SQPOLL) {
- /* there is only one file note, which is owned by sqo_task */
- WARN_ON_ONCE(ctx->sqo_task != current &&
- xa_load(&tctx->xa, (unsigned long)file));
- /* sqo_dead check is for when this happens after cancellation */
- WARN_ON_ONCE(ctx->sqo_task == current && !ctx->sqo_dead &&
- !xa_load(&tctx->xa, (unsigned long)file));
-
- io_disable_sqo_submit(ctx);
- }
-
- if (!(ctx->flags & IORING_SETUP_SQPOLL) || ctx->sqo_task == current)
- io_uring_del_task_file(file);
- return 0;
+ io_uring_clean_tctx(tctx);
+ /* all current's requests should be gone, we can kill tctx */
+ __io_uring_free(current);
}
static void *io_uring_validate_mmap_request(struct file *file,
@@ -9054,29 +9054,20 @@ static unsigned long io_uring_nommu_get_unmapped_area(struct file *file,
static int io_sqpoll_wait_sq(struct io_ring_ctx *ctx)
{
- int ret = 0;
DEFINE_WAIT(wait);
do {
if (!io_sqring_full(ctx))
break;
-
prepare_to_wait(&ctx->sqo_sq_wait, &wait, TASK_INTERRUPTIBLE);
- if (unlikely(ctx->sqo_dead)) {
- ret = -EOWNERDEAD;
- goto out;
- }
-
if (!io_sqring_full(ctx))
break;
-
schedule();
} while (!signal_pending(current));
finish_wait(&ctx->sqo_sq_wait, &wait);
-out:
- return ret;
+ return 0;
}
static int io_get_ext_arg(unsigned flags, const void __user *argp, size_t *argsz,
@@ -9150,15 +9141,10 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned int, fd, u32, to_submit,
if (ctx->flags & IORING_SETUP_SQPOLL) {
io_cqring_overflow_flush(ctx, false, NULL, NULL);
- if (unlikely(ctx->sqo_exec)) {
- ret = io_sq_thread_fork(ctx->sq_data, ctx);
- if (ret)
- goto out;
- ctx->sqo_exec = 0;
- }
ret = -EOWNERDEAD;
- if (unlikely(ctx->sqo_dead))
+ if (unlikely(ctx->sq_data->thread == NULL)) {
goto out;
+ }
if (flags & IORING_ENTER_SQ_WAKEUP)
wake_up(&ctx->sq_data->wait);
if (flags & IORING_ENTER_SQ_WAIT) {
@@ -9168,7 +9154,7 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned int, fd, u32, to_submit,
}
submitted = to_submit;
} else if (to_submit) {
- ret = io_uring_add_task_file(ctx, f.file);
+ ret = io_uring_add_task_file(ctx);
if (unlikely(ret))
goto out;
mutex_lock(&ctx->uring_lock);
@@ -9210,10 +9196,9 @@ out_fput:
}
#ifdef CONFIG_PROC_FS
-static int io_uring_show_cred(int id, void *p, void *data)
+static int io_uring_show_cred(struct seq_file *m, unsigned int id,
+ const struct cred *cred)
{
- const struct cred *cred = p;
- struct seq_file *m = data;
struct user_namespace *uns = seq_user_ns(m);
struct group_info *gi;
kernel_cap_t cap;
@@ -9281,9 +9266,13 @@ static void __io_uring_show_fdinfo(struct io_ring_ctx *ctx, struct seq_file *m)
seq_printf(m, "%5u: 0x%llx/%u\n", i, buf->ubuf,
(unsigned int) buf->len);
}
- if (has_lock && !idr_is_empty(&ctx->personality_idr)) {
+ if (has_lock && !xa_empty(&ctx->personalities)) {
+ unsigned long index;
+ const struct cred *cred;
+
seq_printf(m, "Personalities:\n");
- idr_for_each(&ctx->personality_idr, io_uring_show_cred, m);
+ xa_for_each(&ctx->personalities, index, cred)
+ io_uring_show_cred(m, index, cred);
}
seq_printf(m, "PollList:\n");
spin_lock_irq(&ctx->completion_lock);
@@ -9313,7 +9302,6 @@ static void io_uring_show_fdinfo(struct seq_file *m, struct file *f)
static const struct file_operations io_uring_fops = {
.release = io_uring_release,
- .flush = io_uring_flush,
.mmap = io_uring_mmap,
#ifndef CONFIG_MMU
.get_unmapped_area = io_uring_nommu_get_unmapped_area,
@@ -9378,7 +9366,7 @@ static int io_uring_install_fd(struct io_ring_ctx *ctx, struct file *file)
if (fd < 0)
return fd;
- ret = io_uring_add_task_file(ctx, file);
+ ret = io_uring_add_task_file(ctx);
if (ret) {
put_unused_fd(fd);
return ret;
@@ -9468,7 +9456,6 @@ static int io_uring_create(unsigned entries, struct io_uring_params *p,
ctx->compat = in_compat_syscall();
if (!capable(CAP_IPC_LOCK))
ctx->user = get_uid(current_user());
- ctx->sqo_task = current;
/*
* This is just grabbed for accounting purposes. When a process exits,
@@ -9487,9 +9474,6 @@ static int io_uring_create(unsigned entries, struct io_uring_params *p,
if (ret)
goto err;
- if (!(p->flags & IORING_SETUP_R_DISABLED))
- io_sq_offload_start(ctx);
-
memset(&p->sq_off, 0, sizeof(p->sq_off));
p->sq_off.head = offsetof(struct io_rings, sq.head);
p->sq_off.tail = offsetof(struct io_rings, sq.tail);
@@ -9531,7 +9515,6 @@ static int io_uring_create(unsigned entries, struct io_uring_params *p,
*/
ret = io_uring_install_fd(ctx, file);
if (ret < 0) {
- io_disable_sqo_submit(ctx);
/* fput will clean it up */
fput(file);
return ret;
@@ -9540,7 +9523,6 @@ static int io_uring_create(unsigned entries, struct io_uring_params *p,
trace_io_uring_create(ret, ctx, p->sq_entries, p->cq_entries, p->flags);
return ret;
err:
- io_disable_sqo_submit(ctx);
io_ring_ctx_wait_and_kill(ctx);
return ret;
}
@@ -9619,14 +9601,16 @@ out:
static int io_register_personality(struct io_ring_ctx *ctx)
{
const struct cred *creds;
+ u32 id;
int ret;
creds = get_current_cred();
- ret = idr_alloc_cyclic(&ctx->personality_idr, (void *) creds, 1,
- USHRT_MAX, GFP_KERNEL);
- if (ret < 0)
- put_cred(creds);
+ ret = xa_alloc_cyclic(&ctx->personalities, &id, (void *)creds,
+ XA_LIMIT(0, USHRT_MAX), &ctx->pers_next, GFP_KERNEL);
+ if (!ret)
+ return id;
+ put_cred(creds);
return ret;
}
@@ -9709,9 +9693,8 @@ static int io_register_enable_rings(struct io_ring_ctx *ctx)
ctx->restricted = 1;
ctx->flags &= ~IORING_SETUP_R_DISABLED;
-
- io_sq_offload_start(ctx);
-
+ if (ctx->sq_data && wq_has_sleeper(&ctx->sq_data->wait))
+ wake_up(&ctx->sq_data->wait);
return 0;
}
diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 7ffcd7ef33d4..414769a6ad11 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -1221,7 +1221,7 @@ iomap_alloc_ioend(struct inode *inode, struct iomap_writepage_ctx *wpc,
struct iomap_ioend *ioend;
struct bio *bio;
- bio = bio_alloc_bioset(GFP_NOFS, BIO_MAX_PAGES, &iomap_ioend_bioset);
+ bio = bio_alloc_bioset(GFP_NOFS, BIO_MAX_VECS, &iomap_ioend_bioset);
bio_set_dev(bio, wpc->iomap.bdev);
bio->bi_iter.bi_sector = sector;
bio->bi_opf = REQ_OP_WRITE | wbc_to_write_flags(wbc);
@@ -1252,7 +1252,7 @@ iomap_chain_bio(struct bio *prev)
{
struct bio *new;
- new = bio_alloc(GFP_NOFS, BIO_MAX_PAGES);
+ new = bio_alloc(GFP_NOFS, BIO_MAX_VECS);
bio_copy_dev(new, prev);/* also copies over blkcg information */
new->bi_iter.bi_sector = bio_end_sector(prev);
new->bi_opf = prev->bi_opf;
diff --git a/fs/iomap/direct-io.c b/fs/iomap/direct-io.c
index e2c4991833b8..bdd0d89bbf0a 100644
--- a/fs/iomap/direct-io.c
+++ b/fs/iomap/direct-io.c
@@ -296,7 +296,7 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
*/
bio_opf = iomap_dio_bio_opflags(dio, iomap, use_fua);
- nr_pages = bio_iov_vecs_to_alloc(dio->submit.iter, BIO_MAX_PAGES);
+ nr_pages = bio_iov_vecs_to_alloc(dio->submit.iter, BIO_MAX_VECS);
do {
size_t n;
if (dio->error) {
@@ -338,7 +338,7 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
copied += n;
nr_pages = bio_iov_vecs_to_alloc(dio->submit.iter,
- BIO_MAX_PAGES);
+ BIO_MAX_VECS);
iomap_dio_submit_bio(dio, iomap, bio, pos);
pos += n;
} while (nr_pages);
diff --git a/fs/locks.c b/fs/locks.c
index 99ca97e81b7a..6125d2de39b8 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -1808,9 +1808,6 @@ check_conflicting_open(struct file *filp, const long arg, int flags)
if (flags & FL_LAYOUT)
return 0;
- if (flags & FL_DELEG)
- /* We leave these checks to the caller. */
- return 0;
if (arg == F_RDLCK)
return inode_is_open_for_write(inode) ? -EAGAIN : 0;
diff --git a/fs/mpage.c b/fs/mpage.c
index 961234d68779..334e7d09aa65 100644
--- a/fs/mpage.c
+++ b/fs/mpage.c
@@ -616,7 +616,7 @@ alloc_new:
goto out;
}
bio = mpage_alloc(bdev, blocks[0] << (blkbits - 9),
- BIO_MAX_PAGES, GFP_NOFS|__GFP_HIGH);
+ BIO_MAX_VECS, GFP_NOFS|__GFP_HIGH);
if (bio == NULL)
goto confused;
diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig
index e2a488d403a6..14a72224b657 100644
--- a/fs/nfs/Kconfig
+++ b/fs/nfs/Kconfig
@@ -127,7 +127,7 @@ config PNFS_BLOCK
config PNFS_FLEXFILE_LAYOUT
tristate
depends on NFS_V4_1 && NFS_V3
- default m
+ default NFS_V4
config NFS_V4_1_IMPLEMENTATION_ID_DOMAIN
string "NFSv4.1 Implementation ID Domain"
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 19a9f434442f..fc4f490f2d78 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -81,8 +81,9 @@ static struct nfs_open_dir_context *alloc_nfs_open_dir_context(struct inode *dir
spin_lock(&dir->i_lock);
if (list_empty(&nfsi->open_files) &&
(nfsi->cache_validity & NFS_INO_DATA_INVAL_DEFER))
- nfsi->cache_validity |= NFS_INO_INVALID_DATA |
- NFS_INO_REVAL_FORCED;
+ nfs_set_cache_invalid(dir,
+ NFS_INO_INVALID_DATA |
+ NFS_INO_REVAL_FORCED);
list_add(&ctx->list, &nfsi->open_files);
spin_unlock(&dir->i_lock);
return ctx;
@@ -1401,6 +1402,13 @@ out_force:
goto out;
}
+static void nfs_mark_dir_for_revalidate(struct inode *inode)
+{
+ spin_lock(&inode->i_lock);
+ nfs_set_cache_invalid(inode, NFS_INO_REVAL_PAGECACHE);
+ spin_unlock(&inode->i_lock);
+}
+
/*
* We judge how long we want to trust negative
* dentries by looking at the parent inode mtime.
@@ -1435,19 +1443,14 @@ nfs_lookup_revalidate_done(struct inode *dir, struct dentry *dentry,
__func__, dentry);
return 1;
case 0:
- nfs_mark_for_revalidate(dir);
- if (inode && S_ISDIR(inode->i_mode)) {
- /* Purge readdir caches. */
- nfs_zap_caches(inode);
- /*
- * We can't d_drop the root of a disconnected tree:
- * its d_hash is on the s_anon list and d_drop() would hide
- * it from shrink_dcache_for_unmount(), leading to busy
- * inodes on unmount and further oopses.
- */
- if (IS_ROOT(dentry))
- return 1;
- }
+ /*
+ * We can't d_drop the root of a disconnected tree:
+ * its d_hash is on the s_anon list and d_drop() would hide
+ * it from shrink_dcache_for_unmount(), leading to busy
+ * inodes on unmount and further oopses.
+ */
+ if (inode && IS_ROOT(dentry))
+ return 1;
dfprintk(LOOKUPCACHE, "NFS: %s(%pd2) is invalid\n",
__func__, dentry);
return 0;
@@ -1525,6 +1528,13 @@ out:
nfs_free_fattr(fattr);
nfs_free_fhandle(fhandle);
nfs4_label_free(label);
+
+ /*
+ * If the lookup failed despite the dentry change attribute being
+ * a match, then we should revalidate the directory cache.
+ */
+ if (!ret && nfs_verify_change_attribute(dir, dentry->d_time))
+ nfs_mark_dir_for_revalidate(dir);
return nfs_lookup_revalidate_done(dir, dentry, inode, ret);
}
@@ -1567,7 +1577,7 @@ nfs_do_lookup_revalidate(struct inode *dir, struct dentry *dentry,
error = nfs_lookup_verify_inode(inode, flags);
if (error) {
if (error == -ESTALE)
- nfs_zap_caches(dir);
+ nfs_mark_dir_for_revalidate(dir);
goto out_bad;
}
nfs_advise_use_readdirplus(dir);
@@ -1691,10 +1701,9 @@ static void nfs_drop_nlink(struct inode *inode)
if (inode->i_nlink > 0)
drop_nlink(inode);
NFS_I(inode)->attr_gencount = nfs_inc_attr_generation_counter();
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
- | NFS_INO_INVALID_CTIME
- | NFS_INO_INVALID_OTHER
- | NFS_INO_REVAL_FORCED;
+ nfs_set_cache_invalid(
+ inode, NFS_INO_INVALID_CHANGE | NFS_INO_INVALID_CTIME |
+ NFS_INO_INVALID_OTHER | NFS_INO_REVAL_FORCED);
spin_unlock(&inode->i_lock);
}
@@ -1706,7 +1715,7 @@ static void nfs_dentry_iput(struct dentry *dentry, struct inode *inode)
{
if (S_ISDIR(inode->i_mode))
/* drop any readdir cache as it could easily be old */
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_DATA;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA);
if (dentry->d_flags & DCACHE_NFSFS_RENAMED) {
nfs_complete_unlink(dentry, inode);
@@ -2064,7 +2073,6 @@ out:
dput(parent);
return d;
out_error:
- nfs_mark_for_revalidate(dir);
d = ERR_PTR(error);
goto out;
}
@@ -2473,9 +2481,9 @@ int nfs_rename(struct user_namespace *mnt_userns, struct inode *old_dir,
if (error == 0) {
spin_lock(&old_inode->i_lock);
NFS_I(old_inode)->attr_gencount = nfs_inc_attr_generation_counter();
- NFS_I(old_inode)->cache_validity |= NFS_INO_INVALID_CHANGE
- | NFS_INO_INVALID_CTIME
- | NFS_INO_REVAL_FORCED;
+ nfs_set_cache_invalid(old_inode, NFS_INO_INVALID_CHANGE |
+ NFS_INO_INVALID_CTIME |
+ NFS_INO_REVAL_FORCED);
spin_unlock(&old_inode->i_lock);
}
out:
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 749bbea14d99..a7fb076a5f44 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -207,7 +207,7 @@ static bool nfs_has_xattr_cache(const struct nfs_inode *nfsi)
}
#endif
-static void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
+void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
{
struct nfs_inode *nfsi = NFS_I(inode);
bool have_delegation = NFS_PROTO(inode)->have_delegation(inode, FMODE_READ);
@@ -229,6 +229,7 @@ static void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
if (flags & NFS_INO_INVALID_DATA)
nfs_fscache_invalidate(inode);
}
+EXPORT_SYMBOL_GPL(nfs_set_cache_invalid);
/*
* Invalidate the local caches
@@ -1067,8 +1068,8 @@ void nfs_inode_attach_open_context(struct nfs_open_context *ctx)
spin_lock(&inode->i_lock);
if (list_empty(&nfsi->open_files) &&
(nfsi->cache_validity & NFS_INO_DATA_INVAL_DEFER))
- nfsi->cache_validity |= NFS_INO_INVALID_DATA |
- NFS_INO_REVAL_FORCED;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA |
+ NFS_INO_REVAL_FORCED);
list_add_tail_rcu(&ctx->list, &nfsi->open_files);
spin_unlock(&inode->i_lock);
}
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 25fb43b69e5a..7b644d6c09e4 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -411,7 +411,8 @@ extern int nfs_write_inode(struct inode *, struct writeback_control *);
extern int nfs_drop_inode(struct inode *);
extern void nfs_clear_inode(struct inode *);
extern void nfs_evict_inode(struct inode *);
-void nfs_zap_acl_cache(struct inode *inode);
+extern void nfs_zap_acl_cache(struct inode *inode);
+extern void nfs_set_cache_invalid(struct inode *inode, unsigned long flags);
extern bool nfs_check_cache_invalid(struct inode *, unsigned long);
extern int nfs_wait_bit_killable(struct wait_bit_key *key, int mode);
extern int nfs_wait_atomic_killable(atomic_t *p, unsigned int mode);
diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c
index ca10072644ff..ed1c83738c30 100644
--- a/fs/nfs/nfs3xdr.c
+++ b/fs/nfs/nfs3xdr.c
@@ -36,6 +36,7 @@
#define NFS3_pagepad_sz (1) /* Page padding */
#define NFS3_fhandle_sz (1+16)
#define NFS3_fh_sz (NFS3_fhandle_sz) /* shorthand */
+#define NFS3_post_op_fh_sz (1+NFS3_fh_sz)
#define NFS3_sattr_sz (15)
#define NFS3_filename_sz (1+(NFS3_MAXNAMLEN>>2))
#define NFS3_path_sz (1+(NFS3_MAXPATHLEN>>2))
@@ -73,7 +74,7 @@
#define NFS3_readlinkres_sz (1+NFS3_post_op_attr_sz+1+NFS3_pagepad_sz)
#define NFS3_readres_sz (1+NFS3_post_op_attr_sz+3+NFS3_pagepad_sz)
#define NFS3_writeres_sz (1+NFS3_wcc_data_sz+4)
-#define NFS3_createres_sz (1+NFS3_fh_sz+NFS3_post_op_attr_sz+NFS3_wcc_data_sz)
+#define NFS3_createres_sz (1+NFS3_post_op_fh_sz+NFS3_post_op_attr_sz+NFS3_wcc_data_sz)
#define NFS3_renameres_sz (1+(2 * NFS3_wcc_data_sz))
#define NFS3_linkres_sz (1+NFS3_post_op_attr_sz+NFS3_wcc_data_sz)
#define NFS3_readdirres_sz (1+NFS3_post_op_attr_sz+2+NFS3_pagepad_sz)
diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c
index f3fd935620fc..094024b0aca1 100644
--- a/fs/nfs/nfs42proc.c
+++ b/fs/nfs/nfs42proc.c
@@ -357,13 +357,15 @@ static ssize_t _nfs42_proc_copy(struct file *src,
truncate_pagecache_range(dst_inode, pos_dst,
pos_dst + res->write_res.count);
spin_lock(&dst_inode->i_lock);
- NFS_I(dst_inode)->cache_validity |= (NFS_INO_REVAL_PAGECACHE |
- NFS_INO_REVAL_FORCED | NFS_INO_INVALID_SIZE |
- NFS_INO_INVALID_ATTR | NFS_INO_INVALID_DATA);
+ nfs_set_cache_invalid(
+ dst_inode, NFS_INO_REVAL_PAGECACHE | NFS_INO_REVAL_FORCED |
+ NFS_INO_INVALID_SIZE | NFS_INO_INVALID_ATTR |
+ NFS_INO_INVALID_DATA);
spin_unlock(&dst_inode->i_lock);
spin_lock(&src_inode->i_lock);
- NFS_I(src_inode)->cache_validity |= (NFS_INO_REVAL_PAGECACHE |
- NFS_INO_REVAL_FORCED | NFS_INO_INVALID_ATIME);
+ nfs_set_cache_invalid(src_inode, NFS_INO_REVAL_PAGECACHE |
+ NFS_INO_REVAL_FORCED |
+ NFS_INO_INVALID_ATIME);
spin_unlock(&src_inode->i_lock);
status = res->write_res.count;
out:
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 74bc5120013d..c65c4b41e2c1 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1169,14 +1169,14 @@ int nfs4_call_sync(struct rpc_clnt *clnt,
static void
nfs4_inc_nlink_locked(struct inode *inode)
{
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_OTHER;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
inc_nlink(inode);
}
static void
nfs4_dec_nlink_locked(struct inode *inode)
{
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_OTHER;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
drop_nlink(inode);
}
@@ -1187,35 +1187,31 @@ nfs4_update_changeattr_locked(struct inode *inode,
{
struct nfs_inode *nfsi = NFS_I(inode);
- nfsi->cache_validity |= NFS_INO_INVALID_CTIME
- | NFS_INO_INVALID_MTIME
- | cache_validity;
+ cache_validity |= NFS_INO_INVALID_CTIME | NFS_INO_INVALID_MTIME;
if (cinfo->atomic && cinfo->before == inode_peek_iversion_raw(inode)) {
nfsi->cache_validity &= ~NFS_INO_REVAL_PAGECACHE;
nfsi->attrtimeo_timestamp = jiffies;
} else {
if (S_ISDIR(inode->i_mode)) {
- nfsi->cache_validity |= NFS_INO_INVALID_DATA;
+ cache_validity |= NFS_INO_INVALID_DATA;
nfs_force_lookup_revalidate(inode);
} else {
if (!NFS_PROTO(inode)->have_delegation(inode,
FMODE_READ))
- nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE;
+ cache_validity |= NFS_INO_REVAL_PAGECACHE;
}
if (cinfo->before != inode_peek_iversion_raw(inode))
- nfsi->cache_validity |= NFS_INO_INVALID_ACCESS |
- NFS_INO_INVALID_ACL |
- NFS_INO_INVALID_XATTR;
+ cache_validity |= NFS_INO_INVALID_ACCESS |
+ NFS_INO_INVALID_ACL |
+ NFS_INO_INVALID_XATTR;
}
inode_set_iversion_raw(inode, cinfo->after);
nfsi->read_cache_jiffies = timestamp;
nfsi->attr_gencount = nfs_inc_attr_generation_counter();
+ nfs_set_cache_invalid(inode, cache_validity);
nfsi->cache_validity &= ~NFS_INO_INVALID_CHANGE;
-
- if (nfsi->cache_validity & NFS_INO_INVALID_DATA)
- nfs_fscache_invalidate(inode);
}
void
@@ -5893,6 +5889,9 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
unsigned int npages = DIV_ROUND_UP(buflen, PAGE_SIZE);
int ret, i;
+ /* You can't remove system.nfs4_acl: */
+ if (buflen == 0)
+ return -EINVAL;
if (!nfs4_server_supports_acls(server))
return -EOPNOTSUPP;
if (npages > ARRAY_SIZE(pages))
@@ -5915,9 +5914,9 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
* so mark the attribute cache invalid.
*/
spin_lock(&inode->i_lock);
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
- | NFS_INO_INVALID_CTIME
- | NFS_INO_REVAL_FORCED;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE |
+ NFS_INO_INVALID_CTIME |
+ NFS_INO_REVAL_FORCED);
spin_unlock(&inode->i_lock);
nfs_access_zap_cache(inode);
nfs_zap_acl_cache(inode);
@@ -5969,7 +5968,7 @@ static int _nfs4_get_security_label(struct inode *inode, void *buf,
return ret;
if (!(fattr.valid & NFS_ATTR_FATTR_V4_SECURITY_LABEL))
return -ENOENT;
- return 0;
+ return label.len;
}
static int nfs4_get_security_label(struct inode *inode, void *buf,
diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c
index b27ebdccef70..5fa11e1aca4c 100644
--- a/fs/nfs/unlink.c
+++ b/fs/nfs/unlink.c
@@ -500,9 +500,9 @@ nfs_sillyrename(struct inode *dir, struct dentry *dentry)
nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
spin_lock(&inode->i_lock);
NFS_I(inode)->attr_gencount = nfs_inc_attr_generation_counter();
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
- | NFS_INO_INVALID_CTIME
- | NFS_INO_REVAL_FORCED;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE |
+ NFS_INO_INVALID_CTIME |
+ NFS_INO_REVAL_FORCED);
spin_unlock(&inode->i_lock);
d_move(dentry, sdentry);
break;
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 82bdcb982186..f05a90338a76 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -303,9 +303,9 @@ static void nfs_set_pageerror(struct address_space *mapping)
nfs_zap_mapping(mapping->host, mapping);
/* Force file size revalidation */
spin_lock(&inode->i_lock);
- NFS_I(inode)->cache_validity |= NFS_INO_REVAL_FORCED |
- NFS_INO_REVAL_PAGECACHE |
- NFS_INO_INVALID_SIZE;
+ nfs_set_cache_invalid(inode, NFS_INO_REVAL_FORCED |
+ NFS_INO_REVAL_PAGECACHE |
+ NFS_INO_INVALID_SIZE);
spin_unlock(&inode->i_lock);
}
@@ -1604,7 +1604,7 @@ static int nfs_writeback_done(struct rpc_task *task,
/* Deal with the suid/sgid bit corner case */
if (nfs_should_remove_suid(inode)) {
spin_lock(&inode->i_lock);
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_OTHER;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
spin_unlock(&inode->i_lock);
}
return 0;
diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index 821e5913faee..d6cff5fbe705 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -73,6 +73,7 @@ config NFSD_V4
select NFSD_V3
select FS_POSIX_ACL
select SUNRPC_GSS
+ select CRYPTO
select CRYPTO_MD5
select CRYPTO_SHA256
select GRACE_PERIOD
diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c
index 53fcbf79bdca..7629248fdd53 100644
--- a/fs/nfsd/filecache.c
+++ b/fs/nfsd/filecache.c
@@ -898,6 +898,8 @@ nfsd_file_find_locked(struct inode *inode, unsigned int may_flags,
continue;
if (!nfsd_match_cred(nf->nf_cred, current_cred()))
continue;
+ if (!test_bit(NFSD_FILE_HASHED, &nf->nf_flags))
+ continue;
if (nfsd_file_get(nf) != NULL)
return nf;
}
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 052be5bf9ef5..7325592b456e 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -1189,6 +1189,7 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
switch (task->tk_status) {
case -EIO:
case -ETIMEDOUT:
+ case -EACCES:
nfsd4_mark_cb_down(clp, task->tk_status);
}
break;
diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index acdb3cd806a1..dd9f38d072dd 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -1302,7 +1302,7 @@ nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct nfsd_file *src,
struct nfsd_file *dst)
{
nfs42_ssc_close(src->nf_file);
- /* 'src' is freed by nfsd4_do_async_copy */
+ fput(src->nf_file);
nfsd_file_put(dst);
mntput(ss_mnt);
}
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 423fd6683f3a..97447a64bad0 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -4940,31 +4940,6 @@ static struct file_lock *nfs4_alloc_init_lease(struct nfs4_delegation *dp,
return fl;
}
-static int nfsd4_check_conflicting_opens(struct nfs4_client *clp,
- struct nfs4_file *fp)
-{
- struct nfs4_clnt_odstate *co;
- struct file *f = fp->fi_deleg_file->nf_file;
- struct inode *ino = locks_inode(f);
- int writes = atomic_read(&ino->i_writecount);
-
- if (fp->fi_fds[O_WRONLY])
- writes--;
- if (fp->fi_fds[O_RDWR])
- writes--;
- if (writes > 0)
- return -EAGAIN;
- spin_lock(&fp->fi_lock);
- list_for_each_entry(co, &fp->fi_clnt_odstate, co_perfile) {
- if (co->co_client != clp) {
- spin_unlock(&fp->fi_lock);
- return -EAGAIN;
- }
- }
- spin_unlock(&fp->fi_lock);
- return 0;
-}
-
static struct nfs4_delegation *
nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
struct nfs4_file *fp, struct nfs4_clnt_odstate *odstate)
@@ -4984,12 +4959,9 @@ nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
nf = find_readable_file(fp);
if (!nf) {
- /*
- * We probably could attempt another open and get a read
- * delegation, but for now, don't bother until the
- * client actually sends us one.
- */
- return ERR_PTR(-EAGAIN);
+ /* We should always have a readable file here */
+ WARN_ON_ONCE(1);
+ return ERR_PTR(-EBADF);
}
spin_lock(&state_lock);
spin_lock(&fp->fi_lock);
@@ -5019,19 +4991,11 @@ nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
if (!fl)
goto out_clnt_odstate;
- status = nfsd4_check_conflicting_opens(clp, fp);
- if (status) {
- locks_free_lock(fl);
- goto out_clnt_odstate;
- }
status = vfs_setlease(fp->fi_deleg_file->nf_file, fl->fl_type, &fl, NULL);
if (fl)
locks_free_lock(fl);
if (status)
goto out_clnt_odstate;
- status = nfsd4_check_conflicting_opens(clp, fp);
- if (status)
- goto out_clnt_odstate;
spin_lock(&state_lock);
spin_lock(&fp->fi_lock);
@@ -5113,6 +5077,17 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open,
goto out_no_deleg;
if (!cb_up || !(oo->oo_flags & NFS4_OO_CONFIRMED))
goto out_no_deleg;
+ /*
+ * Also, if the file was opened for write or
+ * create, there's a good chance the client's
+ * about to write to it, resulting in an
+ * immediate recall (since we don't support
+ * write delegations):
+ */
+ if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE)
+ goto out_no_deleg;
+ if (open->op_create == NFS4_OPEN_CREATE)
+ goto out_no_deleg;
break;
default:
goto out_no_deleg;
@@ -5389,7 +5364,7 @@ nfs4_laundromat(struct nfsd_net *nn)
idr_for_each_entry(&nn->s2s_cp_stateids, cps_t, i) {
cps = container_of(cps_t, struct nfs4_cpntf_state, cp_stateid);
if (cps->cp_stateid.sc_type == NFS4_COPYNOTIFY_STID &&
- cps->cpntf_time > cutoff)
+ cps->cpntf_time < cutoff)
_free_cpntf_state_locked(nn, cps);
}
spin_unlock(&nn->s2s_cp_lock);
diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c
index 1e75417bfe6e..56872e93823d 100644
--- a/fs/nilfs2/segbuf.c
+++ b/fs/nilfs2/segbuf.c
@@ -399,7 +399,7 @@ static void nilfs_segbuf_prepare_write(struct nilfs_segment_buffer *segbuf,
{
wi->bio = NULL;
wi->rest_blocks = segbuf->sb_sum.nblocks;
- wi->max_pages = BIO_MAX_PAGES;
+ wi->max_pages = BIO_MAX_VECS;
wi->nr_vecs = min(wi->max_pages, wi->rest_blocks);
wi->start = wi->end = 0;
wi->blocknr = segbuf->sb_pseg_start;
diff --git a/fs/pnode.h b/fs/pnode.h
index 26f74e092bd9..988f1aa9b02a 100644
--- a/fs/pnode.h
+++ b/fs/pnode.h
@@ -12,7 +12,7 @@
#define IS_MNT_SHARED(m) ((m)->mnt.mnt_flags & MNT_SHARED)
#define IS_MNT_SLAVE(m) ((m)->mnt_master)
-#define IS_MNT_NEW(m) (!(m)->mnt_ns)
+#define IS_MNT_NEW(m) (!(m)->mnt_ns || is_anon_ns((m)->mnt_ns))
#define CLEAR_MNT_SHARED(m) ((m)->mnt.mnt_flags &= ~MNT_SHARED)
#define IS_MNT_UNBINDABLE(m) ((m)->mnt.mnt_flags & MNT_UNBINDABLE)
#define IS_MNT_MARKED(m) ((m)->mnt.mnt_flags & MNT_MARKED)
diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index 3cec6fbef725..e862cab69583 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -1036,8 +1036,6 @@ struct clear_refs_private {
#ifdef CONFIG_MEM_SOFT_DIRTY
-#define is_cow_mapping(flags) (((flags) & (VM_SHARED | VM_MAYWRITE)) == VM_MAYWRITE)
-
static inline bool pte_is_pinned(struct vm_area_struct *vma, unsigned long addr, pte_t pte)
{
struct page *page;
diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c
index 93a217e4f563..14658b009f1b 100644
--- a/fs/pstore/inode.c
+++ b/fs/pstore/inode.c
@@ -467,7 +467,7 @@ static struct dentry *pstore_mount(struct file_system_type *fs_type,
static void pstore_kill_sb(struct super_block *sb)
{
mutex_lock(&pstore_sb_lock);
- WARN_ON(pstore_sb != sb);
+ WARN_ON(pstore_sb && pstore_sb != sb);
kill_litter_super(sb);
pstore_sb = NULL;
diff --git a/fs/pstore/ram_core.c b/fs/pstore/ram_core.c
index aa8e0b65ff1a..fff363bfd484 100644
--- a/fs/pstore/ram_core.c
+++ b/fs/pstore/ram_core.c
@@ -246,7 +246,7 @@ static int persistent_ram_init_ecc(struct persistent_ram_zone *prz,
pr_info("error in header, %d\n", numerr);
prz->corrected_bytes += numerr;
} else if (numerr < 0) {
- pr_info("uncorrectable error in header\n");
+ pr_info_ratelimited("uncorrectable error in header\n");
prz->bad_blocks++;
}
diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c
index 45f44425d856..b9e87ebb1060 100644
--- a/fs/squashfs/block.c
+++ b/fs/squashfs/block.c
@@ -87,7 +87,7 @@ static int squashfs_bio_read(struct super_block *sb, u64 index, int length,
int error, i;
struct bio *bio;
- if (page_count <= BIO_MAX_PAGES)
+ if (page_count <= BIO_MAX_VECS)
bio = bio_alloc(GFP_NOIO, page_count);
else
bio = bio_kmalloc(GFP_NOIO, page_count);
diff --git a/fs/zonefs/super.c b/fs/zonefs/super.c
index b6ff4a21abac..0fe76f376dee 100644
--- a/fs/zonefs/super.c
+++ b/fs/zonefs/super.c
@@ -684,7 +684,7 @@ static ssize_t zonefs_file_dio_append(struct kiocb *iocb, struct iov_iter *from)
max = ALIGN_DOWN(max << SECTOR_SHIFT, inode->i_sb->s_blocksize);
iov_iter_truncate(from, max);
- nr_pages = iov_iter_npages(from, BIO_MAX_PAGES);
+ nr_pages = iov_iter_npages(from, BIO_MAX_VECS);
if (!nr_pages)
return 0;