| Message ID | 36d4b7502a8654882718421a18472d41dbc1c83c.1697529529.git.naohiro.aota@wdc.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | btrfs: zoned: wait for data BG to be finished on direct IO allocation | expand |
On 17.10.23 10:01, Naohiro Aota wrote: > This happens because -EAGAIN error returned from btrfs_reserve_extent() > called from btrfs_new_extent_direct() is splling over to the userland. spilling ~^ Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
On Oct 17, 2023 / 17:00, Naohiro Aota wrote: > Running the fio command below on a ZNS device results in "Resource > temporarily unavailable" error. Thanks for the fix. The fio command is missing here. It should be like, $ sudo fio --name=w --directory=/mnt --filesize=1GB --bs=16MB --numjobs=16 \ --rw=write --ioengine=libaio --iodepth=128 --direct=1 > > fio: io_u error on file /mnt/w.2.0: Resource temporarily unavailable: write offset=117440512, buflen=16777216 > fio: io_u error on file /mnt/w.2.0: Resource temporarily unavailable: write offset=134217728, buflen=16777216 > ... > > This happens because -EAGAIN error returned from btrfs_reserve_extent() > called from btrfs_new_extent_direct() is splling over to the userland. > > btrfs_reserve_extent() returns -EAGAIN when there is no active zone > available. Then, the caller should wait for some other on-going IO to > finish a zone and retry the allocation. > > This logic is already implemented for buffered write in cow_file_range(), > but it is missing for the direct IO counterpart. Implement the same logic > for it. > > Fixes: 2ce543f47843 ("btrfs: zoned: wait until zone is finished when allocation didn't progress") > CC: stable@vger.kernel.org # 6.1+ > Reported-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> > Tested-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> > Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> > --- > fs/btrfs/inode.c | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c > index b388505c91cc..a979e964375d 100644 > --- a/fs/btrfs/inode.c > +++ b/fs/btrfs/inode.c > @@ -6979,8 +6979,16 @@ static struct extent_map *btrfs_new_extent_direct(struct btrfs_inode *inode, > int ret; > > alloc_hint = get_extent_allocation_hint(inode, start, len); > +again: > ret = btrfs_reserve_extent(root, len, len, fs_info->sectorsize, > 0, alloc_hint, &ins, 1, 1); > + if (ret == -EAGAIN) { > + ASSERT(btrfs_is_zoned(fs_info)); > + wait_on_bit_io(&inode->root->fs_info->flags, > + BTRFS_FS_NEED_ZONE_FINISH, > + TASK_UNINTERRUPTIBLE); > + goto again; > + } > if (ret) > return ERR_PTR(ret); > > -- > 2.42.0 >
On Tue, Oct 17, 2023 at 05:00:31PM +0900, Naohiro Aota wrote: > Running the fio command below on a ZNS device results in "Resource > temporarily unavailable" error. > > fio: io_u error on file /mnt/w.2.0: Resource temporarily unavailable: write offset=117440512, buflen=16777216 > fio: io_u error on file /mnt/w.2.0: Resource temporarily unavailable: write offset=134217728, buflen=16777216 > ... > > This happens because -EAGAIN error returned from btrfs_reserve_extent() > called from btrfs_new_extent_direct() is splling over to the userland. > > btrfs_reserve_extent() returns -EAGAIN when there is no active zone > available. Then, the caller should wait for some other on-going IO to > finish a zone and retry the allocation. > > This logic is already implemented for buffered write in cow_file_range(), > but it is missing for the direct IO counterpart. Implement the same logic > for it. > > Fixes: 2ce543f47843 ("btrfs: zoned: wait until zone is finished when allocation didn't progress") > CC: stable@vger.kernel.org # 6.1+ > Reported-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> > Tested-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> > Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> With fixed spelling and added fio command, patch added to misc-next, thanks.
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index b388505c91cc..a979e964375d 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6979,8 +6979,16 @@ static struct extent_map *btrfs_new_extent_direct(struct btrfs_inode *inode, int ret; alloc_hint = get_extent_allocation_hint(inode, start, len); +again: ret = btrfs_reserve_extent(root, len, len, fs_info->sectorsize, 0, alloc_hint, &ins, 1, 1); + if (ret == -EAGAIN) { + ASSERT(btrfs_is_zoned(fs_info)); + wait_on_bit_io(&inode->root->fs_info->flags, + BTRFS_FS_NEED_ZONE_FINISH, + TASK_UNINTERRUPTIBLE); + goto again; + } if (ret) return ERR_PTR(ret);
Running the fio command below on a ZNS device results in "Resource temporarily unavailable" error. fio: io_u error on file /mnt/w.2.0: Resource temporarily unavailable: write offset=117440512, buflen=16777216 fio: io_u error on file /mnt/w.2.0: Resource temporarily unavailable: write offset=134217728, buflen=16777216 ... This happens because -EAGAIN error returned from btrfs_reserve_extent() called from btrfs_new_extent_direct() is splling over to the userland. btrfs_reserve_extent() returns -EAGAIN when there is no active zone available. Then, the caller should wait for some other on-going IO to finish a zone and retry the allocation. This logic is already implemented for buffered write in cow_file_range(), but it is missing for the direct IO counterpart. Implement the same logic for it. Fixes: 2ce543f47843 ("btrfs: zoned: wait until zone is finished when allocation didn't progress") CC: stable@vger.kernel.org # 6.1+ Reported-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Tested-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> --- fs/btrfs/inode.c | 8 ++++++++ 1 file changed, 8 insertions(+)