- 27 Jul, 2020 40 commits
-
-
Nikolay Borisov authored
It needs btrfs_inode so take it as a parameter directly. Signed-off-by:
Nikolay Borisov <nborisov@suse.com> Reviewed-by:
David Sterba <dsterba@suse.com> Signed-off-by:
-
Nikolay Borisov authored
It only uses btrfs_inode internally so take it as a parameter. Signed-off-by:
-
Nikolay Borisov authored
No point in taking an inode only to get btrfs_fs_info from it, instead take btrfs_fs_info directly. Signed-off-by:
-
Nikolay Borisov authored
There's only a single use of vfs_inode in a tracepoint so let's take btrfs_inode directly. Signed-off-by:
-
Nikolay Borisov authored
There is a single use of the generic vfs_inode so let's take btrfs_inode as a parameter and remove couple of redundant BTRFS_I() calls. Signed-off-by:
-
Nikolay Borisov authored
Preparation to make btrfs_dirty_pages take btrfs_inode as parameter. Signed-off-by:
-
Nikolay Borisov authored
Only find_lock_delalloc_range uses vfs_inode so let's take the btrfs_inode as a parameter. Signed-off-by:
-
Nikolay Borisov authored
It has only a single use for a generic vfs inode vs 3 for btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
This function really needs a btrfs_inode and not a generic vfs one. Take it as a parameter and get rid of superfluous BTRFS_I() calls. Signed-off-by:
-
Nikolay Borisov authored
Take btrfs_inode directly and stop using superfulous BTRFS_I calls. Signed-off-by:
-
Nikolay Borisov authored
Simply forwards its argument so let's get rid of one extra BTRFS_I by taking btrfs_inode directly. Signed-off-by:
-
Nikolay Borisov authored
All children now take btrfs_inode so convert it to taking it as a parameter as well. Signed-off-by:
-
Nikolay Borisov authored
Gets rid of superfulous BTRFS_I() calls and prepare for converting btrfs_run_delalloc_range to using btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
Simply gets rid of superfluous BTRFS_I() calls. Signed-off-by:
-
Nikolay Borisov authored
Gets rid of superfluous BTRFS_I() calls. Signed-off-by:
-
Nikolay Borisov authored
Preparation to converting btrfs_run_delalloc_range to using btrfs_inode without BTRFS_I() calls. Signed-off-by:
-
Nikolay Borisov authored
It really wants btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
It doesn't really need vfs_inode but btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
It only uses vfs inode for assigning it to the async_chunk function. Signed-off-by:
-
Nikolay Borisov authored
It only really uses btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
It really wants btrfs_inode and is prepration to converting run_delalloc_nocow to taking btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
Signed-off-by:
-
Nikolay Borisov authored
It just forwards its argument to __btrfs_qgroup_release_data. Signed-off-by:
-
Nikolay Borisov authored
All but 3 uses require vfs_inode so convert the logic to have btrfs_inode be the main inode struct. Signed-off-by:
-
Nikolay Borisov authored
Majority of its uses are for btrfs_inode so take it as an argument directly. Signed-off-by:
-
Nikolay Borisov authored
It simpy forwards its inode argument to __btrfs_add_ordered_extent which already takes btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
All its children functions take btrfs_inode so convert it to taking btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
Preparation to converting its callers to taking btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
It has only 2 uses for the vfs_inode - insert_inline_extent and i_size_read. On the flipside it will allow converting its callers to btrfs_inode, so convert it to taking btrfs_inode. Signed-off-by:
-
Nikolay Borisov authored
It passes btrfs_inode to its callee so change the interface. Signed-off-by:
-
Nikolay Borisov authored
It uses vfs_inode only for a tracepoint so convert its interface to take btrfs_inode directly. Signed-off-by:
-
Nikolay Borisov authored
It only uses btrfs_inode so can just as easily take it as an argument. Signed-off-by:
-
Nikolay Borisov authored
Only 6 out of all flush states were being printed correctly since only they were exported via the TRACE_DEFINE_ENUM macro. This patch converts all flush states to use the newly introduced EM macro so that they can all be printed correctly. Signed-off-by:
-
Nikolay Borisov authored
This fixes correct pint out of the extent io tree owner in btrfs_set_extent_bit/btrfs_clear_extent_bit/btrfs_convert_extent_bit tracepoints. Signed-off-by:
-
Nikolay Borisov authored
Since qgroup's reservation types are define in a macro they must be exported to user space in order for user space tools to convert raw binary data to symbolic names. Currently trace-cmd report produces the following output: kworker/u8:2-459 [003] 1208.543587: qgroup_update_reserve: 2b742cae-e0e5-4def-9ef7-28a9b34a951e: qgid=5 type=0x2 cur_reserved=54870016 diff=-32768 With this fix the output is: kworker/u8:2-459 [003] 1208.543587: qgroup_update_reserve: 2b742cae-e0e5-4def-9ef7-28a9b34a951e: qgid=5 type=BTRFS_QGROUP_RSV_META_PREALLOC cur_reserved=54870016 diff=-32768 Signed-off-by:
-
Nikolay Borisov authored
Since all enums used in btrfs' tracepoints are going to be redefined to allow proper parsing of their values by userspace tools let's rearrange when they are defined. This will allow to use only a single set of #define EM/#undef EM sequence. No functional changes. Signed-off-by:
-
Nikolay Borisov authored
extent's type is an enum and this requires that the enum values be exported to user space so that user space tools can correctly map raw binary data to the symbolic name. Currently tracepoints using btrfs__file_extent_item_regular or btrfs__file_extent_item_inline result in the following output: fio-443 [002] 586.609450: btrfs_get_extent_show_fi_regular: f0c3bf8e-0174-4bcc-92aa-6c2d62430420:i root=5(FS_TREE) inode=258 size=2136457216 disk_isize=0 file extent range=[2126946304 2136457216] (num_bytes=9510912 ram_bytes=9510912 disk_bytenr=0 disk_num_bytes=0 extent_offset=0 type=0x1 compression=0 E.g type is 0x1 . With this patch applie the output is: <ommitted for brevity> disk_bytenr=141348864 disk_num_bytes=4096 extent_offset=0 type=REG compression=0 Signed-off-by:
-
Nikolay Borisov authored
When tracepoints use __print_symbolic to print textual representation of a value that comes from an ENUM each enum value needs to be exported to user space so that user space tools can convert the binary value data to the trings as user space does not know what those enums are about. Doing a trace-cmd record && trace-cmd report currently results in: kworker/u8:1-61 [000] 66.299527: btrfs_flush_space: 5302ee13-c65e-45bb-98ef-8fe3835bd943: state=3(0x3) flags=4(METADATA) num_bytes=2621440 ret=0 I.e state is not translated to its symbolic counterpart. With this patch applied the output is: fio-370 [002] 56.762402: btrfs_trigger_flush: d04cd7ac-38e2-452f-a7f5-8157529fd5f0: preempt: flush=3(BTRFS_RESERVE_FLUSH_ALL) flags=4(METADATA) bytes=655360 See also 190f0b76 ("mm: tracing: Export enums in tracepoints to user space"). Signed-off-by:
-
David Sterba authored
On a filesystem with exhausted metadata, but still enough to start balance, it's possible to hit this error: [324402.053842] BTRFS info (device loop0): 1 enospc errors during balance [324402.060769] BTRFS info (device loop0): balance: ended with status: -28 [324402.172295] BTRFS: error (device loop0) in reset_balance_state:3321: errno=-28 No space left It fails inside reset_balance_state and turns the filesystem to read-only, which is unnecessary and should be fixed too, but the problem is caused by lack for space when the balance item is deleted. This is a one-time operation and from the same rank as unlink that is allowed to use the global block reserve. So do the same for the balance item. Status of the filesystem (100GiB) just after the balance fails: $ btrfs fi df mnt Data, single: total=80.01GiB, used=38.58GiB System, single: total=4.00MiB, used=16.00KiB Metadata, single: total=19.99GiB, used=19.48GiB GlobalReserve, single: total=512.00MiB, used=50.11MiB CC: stable@vger.kernel.org # 4.4+ Reviewed-by:
Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by:
-
Qu Wenruo authored
The function btrfs_check_can_nocow() now has two completely different call patterns. For nowait variant, callers don't need to do any cleanup. While for wait variant, callers need to release the lock if they can do nocow write. This is somehow confusing, and is already a problem for the exported btrfs_check_can_nocow(). So this patch will separate the different patterns into different functions. For nowait variant, the function will be called check_nocow_nolock(). For wait variant, the function pair will be btrfs_check_nocow_lock() btrfs_check_nocow_unlock(). Reviewed-by:
Anand Jain <anand.jain@oracle.com> Signed-off-by:
Qu Wenruo <wqu@suse.com> Reviewed-by:
-
