| Message ID | f0a4ae9168940bf1756f89a140cabedb8972e0d1.1604065695.git.naohiro.aota@wdc.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | btrfs: zoned block device support | expand |
On 10/30/20 9:51 AM, Naohiro Aota wrote: > As updates to the space cache v1 are in-place, the space cache cannot be > located over sequential zones and there is no guarantees that the device > will have enough conventional zones to store this cache. Resolve this > problem by disabling completely the space cache v1. This does not > introduces any problems with sequential block groups: all the free space is > located after the allocation pointer and no free space before the pointer. > There is no need to have such cache. > > Note: we can technically use free-space-tree (space cache v2) on ZONED > mode. But, since ZONED mode now always allocate extents in a block group > sequentially regardless of underlying device zone type, it's no use to > enable and maintain the tree. > > For the same reason, NODATACOW is also disabled. > > Also INODE_MAP_CACHE is also disabled to avoid preallocation in the > INODE_MAP_CACHE inode. > > In summary, ZONED will disable: > > | Disabled features | Reason | > |-------------------+-----------------------------------------------------| > | RAID/Dup | Cannot handle two zone append writes to different | > | | zones | > |-------------------+-----------------------------------------------------| > | space_cache (v1) | In-place updating | > | NODATACOW | In-place updating | > |-------------------+-----------------------------------------------------| > | fallocate | Reserved extent will be a write hole | > | INODE_MAP_CACHE | Need pre-allocation. (and will be deprecated?) | > |-------------------+-----------------------------------------------------| > | MIXED_BG | Allocated metadata region will be write holes for | > | | data writes | > > Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> > --- > fs/btrfs/super.c | 12 ++++++++++-- > fs/btrfs/zoned.c | 18 ++++++++++++++++++ > fs/btrfs/zoned.h | 5 +++++ > 3 files changed, 33 insertions(+), 2 deletions(-) > > diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c > index 3312fe08168f..9064ca62b0a0 100644 > --- a/fs/btrfs/super.c > +++ b/fs/btrfs/super.c > @@ -525,8 +525,14 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options, > cache_gen = btrfs_super_cache_generation(info->super_copy); > if (btrfs_fs_compat_ro(info, FREE_SPACE_TREE)) > btrfs_set_opt(info->mount_opt, FREE_SPACE_TREE); > - else if (cache_gen) > - btrfs_set_opt(info->mount_opt, SPACE_CACHE); > + else if (cache_gen) { > + if (btrfs_is_zoned(info)) { > + btrfs_info(info, > + "clearring existing space cache in ZONED mode"); 'clearing', and then you can add Reviewed-by: Josef Bacik <josef@toxicpanda.com> Thanks, Josef
On 02/11/2020 18:02, Josef Bacik wrote:
> 'clearing', and then you can add
Fixed up, thanks
On Fri, Oct 30, 2020 at 10:51:14PM +0900, Naohiro Aota wrote: > INODE_MAP_CACHE inode. > > In summary, ZONED will disable: > > | Disabled features | Reason | > |-------------------+-----------------------------------------------------| > | RAID/Dup | Cannot handle two zone append writes to different | > | | zones | > |-------------------+-----------------------------------------------------| > | space_cache (v1) | In-place updating | > | NODATACOW | In-place updating | > |-------------------+-----------------------------------------------------| > | fallocate | Reserved extent will be a write hole | > | INODE_MAP_CACHE | Need pre-allocation. (and will be deprecated?) | space_cache is deprecated and actually in current dev cycle (5.11) > |-------------------+-----------------------------------------------------| > | MIXED_BG | Allocated metadata region will be write holes for | > | | data writes | > > Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> > --- > fs/btrfs/super.c | 12 ++++++++++-- > fs/btrfs/zoned.c | 18 ++++++++++++++++++ > fs/btrfs/zoned.h | 5 +++++ > 3 files changed, 33 insertions(+), 2 deletions(-) > > diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c > index 3312fe08168f..9064ca62b0a0 100644 > --- a/fs/btrfs/super.c > +++ b/fs/btrfs/super.c > @@ -525,8 +525,14 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options, > cache_gen = btrfs_super_cache_generation(info->super_copy); > if (btrfs_fs_compat_ro(info, FREE_SPACE_TREE)) > btrfs_set_opt(info->mount_opt, FREE_SPACE_TREE); > - else if (cache_gen) > - btrfs_set_opt(info->mount_opt, SPACE_CACHE); > + else if (cache_gen) { > + if (btrfs_is_zoned(info)) { > + btrfs_info(info, > + "clearring existing space cache in ZONED mode"); "zoned: clearing existing space cache" Is it clearing or just invalidating it? We have the same problem with enabling v2 so this could share some code once Boris' patches are merged. > + btrfs_set_super_cache_generation(info->super_copy, 0); > + } else } else { > + btrfs_set_opt(info->mount_opt, SPACE_CACHE); } > + } > > /* > * Even the options are empty, we still need to do extra check > --- a/fs/btrfs/zoned.c > +++ b/fs/btrfs/zoned.c > @@ -265,3 +265,21 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info) > out: > return ret; > } > + > +int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info) > +{ > + if (!btrfs_is_zoned(info)) > + return 0; > + > + /* > + * SPACE CACHE writing is not CoWed. Disable that to avoid write * Space cache writing is nod COWed. ... > + * errors in sequential zones. > + */ > + if (btrfs_test_opt(info, SPACE_CACHE)) { > + btrfs_err(info, > + "space cache v1 not supportted in ZONED mode"); "zoned: space cache v1 is not supported" > + return -EOPNOTSUPP; This should be EINVAL, like invalid cobination. EOPNOTSUPP is for cases where it's not supported but could be if implemented. > + } > + > + return 0; > +} > diff --git a/fs/btrfs/zoned.h b/fs/btrfs/zoned.h > index a63f6177f9ee..0b7756a7104d 100644 > --- a/fs/btrfs/zoned.h > +++ b/fs/btrfs/zoned.h > @@ -24,6 +24,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos, > int btrfs_get_dev_zone_info(struct btrfs_device *device); > void btrfs_destroy_dev_zone_info(struct btrfs_device *device); > int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info); > +int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info); > #else /* CONFIG_BLK_DEV_ZONED */ > static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos, > struct blk_zone *zone) > @@ -43,6 +44,10 @@ static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info) > btrfs_err(fs_info, "Zoned block devices support is not enabled"); > return -EOPNOTSUPP; > } newline > +static inline int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info) > +{ > + return 0; > +} newline > #endif > > static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
On Tue, Nov 03, 2020 at 01:48:34PM +0100, David Sterba wrote: >On Fri, Oct 30, 2020 at 10:51:14PM +0900, Naohiro Aota wrote: >> INODE_MAP_CACHE inode. >> >> In summary, ZONED will disable: >> >> | Disabled features | Reason | >> |-------------------+-----------------------------------------------------| >> | RAID/Dup | Cannot handle two zone append writes to different | >> | | zones | >> |-------------------+-----------------------------------------------------| >> | space_cache (v1) | In-place updating | >> | NODATACOW | In-place updating | >> |-------------------+-----------------------------------------------------| >> | fallocate | Reserved extent will be a write hole | >> | INODE_MAP_CACHE | Need pre-allocation. (and will be deprecated?) | > >space_cache is deprecated and actually in current dev cycle (5.11) > >> |-------------------+-----------------------------------------------------| >> | MIXED_BG | Allocated metadata region will be write holes for | >> | | data writes | >> >> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> >> --- >> fs/btrfs/super.c | 12 ++++++++++-- >> fs/btrfs/zoned.c | 18 ++++++++++++++++++ >> fs/btrfs/zoned.h | 5 +++++ >> 3 files changed, 33 insertions(+), 2 deletions(-) >> >> diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c >> index 3312fe08168f..9064ca62b0a0 100644 >> --- a/fs/btrfs/super.c >> +++ b/fs/btrfs/super.c >> @@ -525,8 +525,14 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options, >> cache_gen = btrfs_super_cache_generation(info->super_copy); >> if (btrfs_fs_compat_ro(info, FREE_SPACE_TREE)) >> btrfs_set_opt(info->mount_opt, FREE_SPACE_TREE); >> - else if (cache_gen) >> - btrfs_set_opt(info->mount_opt, SPACE_CACHE); >> + else if (cache_gen) { >> + if (btrfs_is_zoned(info)) { >> + btrfs_info(info, >> + "clearring existing space cache in ZONED mode"); > > "zoned: clearing existing space cache" > >Is it clearing or just invalidating it? We have the same problem with >enabling v2 so this could share some code once Boris' patches are >merged. Yep, this is just invalidating it. I'll use btrfs_set_free_space_cache_v1_active here once the patches are merged.
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 3312fe08168f..9064ca62b0a0 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -525,8 +525,14 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options, cache_gen = btrfs_super_cache_generation(info->super_copy); if (btrfs_fs_compat_ro(info, FREE_SPACE_TREE)) btrfs_set_opt(info->mount_opt, FREE_SPACE_TREE); - else if (cache_gen) - btrfs_set_opt(info->mount_opt, SPACE_CACHE); + else if (cache_gen) { + if (btrfs_is_zoned(info)) { + btrfs_info(info, + "clearring existing space cache in ZONED mode"); + btrfs_set_super_cache_generation(info->super_copy, 0); + } else + btrfs_set_opt(info->mount_opt, SPACE_CACHE); + } /* * Even the options are empty, we still need to do extra check @@ -985,6 +991,8 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options, ret = -EINVAL; } + if (!ret) + ret = btrfs_check_mountopts_zoned(info); if (!ret && btrfs_test_opt(info, SPACE_CACHE)) btrfs_info(info, "disk space caching is enabled"); if (!ret && btrfs_test_opt(info, FREE_SPACE_TREE)) diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index 1b42e13b8227..3885fa327049 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -265,3 +265,21 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info) out: return ret; } + +int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info) +{ + if (!btrfs_is_zoned(info)) + return 0; + + /* + * SPACE CACHE writing is not CoWed. Disable that to avoid write + * errors in sequential zones. + */ + if (btrfs_test_opt(info, SPACE_CACHE)) { + btrfs_err(info, + "space cache v1 not supportted in ZONED mode"); + return -EOPNOTSUPP; + } + + return 0; +} diff --git a/fs/btrfs/zoned.h b/fs/btrfs/zoned.h index a63f6177f9ee..0b7756a7104d 100644 --- a/fs/btrfs/zoned.h +++ b/fs/btrfs/zoned.h @@ -24,6 +24,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos, int btrfs_get_dev_zone_info(struct btrfs_device *device); void btrfs_destroy_dev_zone_info(struct btrfs_device *device); int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info); +int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info); #else /* CONFIG_BLK_DEV_ZONED */ static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos, struct blk_zone *zone) @@ -43,6 +44,10 @@ static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info) btrfs_err(fs_info, "Zoned block devices support is not enabled"); return -EOPNOTSUPP; } +static inline int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info) +{ + return 0; +} #endif static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
As updates to the space cache v1 are in-place, the space cache cannot be located over sequential zones and there is no guarantees that the device will have enough conventional zones to store this cache. Resolve this problem by disabling completely the space cache v1. This does not introduces any problems with sequential block groups: all the free space is located after the allocation pointer and no free space before the pointer. There is no need to have such cache. Note: we can technically use free-space-tree (space cache v2) on ZONED mode. But, since ZONED mode now always allocate extents in a block group sequentially regardless of underlying device zone type, it's no use to enable and maintain the tree. For the same reason, NODATACOW is also disabled. Also INODE_MAP_CACHE is also disabled to avoid preallocation in the INODE_MAP_CACHE inode. In summary, ZONED will disable: | Disabled features | Reason | |-------------------+-----------------------------------------------------| | RAID/Dup | Cannot handle two zone append writes to different | | | zones | |-------------------+-----------------------------------------------------| | space_cache (v1) | In-place updating | | NODATACOW | In-place updating | |-------------------+-----------------------------------------------------| | fallocate | Reserved extent will be a write hole | | INODE_MAP_CACHE | Need pre-allocation. (and will be deprecated?) | |-------------------+-----------------------------------------------------| | MIXED_BG | Allocated metadata region will be write holes for | | | data writes | Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> --- fs/btrfs/super.c | 12 ++++++++++-- fs/btrfs/zoned.c | 18 ++++++++++++++++++ fs/btrfs/zoned.h | 5 +++++ 3 files changed, 33 insertions(+), 2 deletions(-)