Swapfiles use bmap to build a list of extents belonging to the file,
and they assume these extents won't change over the life of the file.
They also use resulting list to do IO directly to the block device.
This causes problems for btrfs in a few ways:
btrfs returns logical block numbers through bmap, and these are not suitable
for IO. They might translate to different devices, raid etc.
COW means that file block mappings are going to change frequently.
Using swapfiles on btrfs will lead to corruption, so we're avoiding the
problem for now by dropping bmap support entirely. A later commit
will add fiemap support for people that really want to know how
a file is laid out.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
To improve performance, btrfs_sync_log merges tree log sync
requests. But it wrongly merges sync requests for different
tree logs. If multiple tree logs are synced at the same time,
only one of them actually gets synced.
This patch has following changes to fix the bug:
Move most tree log related fields in btrfs_fs_info to
btrfs_root. This allows merging sync requests separately
for each tree log.
Don't insert root item into the log root tree immediately
after log tree is allocated. Root item for log tree is
inserted when log tree get synced for the first time. This
allows syncing the log root tree without first syncing all
log trees.
At tree-log sync, btrfs_sync_log first sync the log tree;
then updates corresponding root item in the log root tree;
sync the log root tree; then update the super block.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
replace_one_extent searches tree leaves for references to a given extent. It
stops searching if it goes beyond the last possible position.
The last possible position is computed by adding the starting offset of a found
file extent to the full size of the extent. The code uses physical size of the
extent as the full size. This is incorrect when compression is used.
The fix is get the full size from ram_bytes field of file extent item.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Change one typedef to a regular enum, and remove an unused one.
Signed-off-by: Jan Engelhardt <jengelh@medozas.de>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_extent_post_op calls finish_current_insert and del_pending_extents. They
both may enter infinite loops.
finish_current_insert enters infinite loop if it only finds some backrefs to
update. The fix is to check for pending backref updates before restarting the
loop.
The infinite loop in del_pending_extents is due to a the skipped variable
not being properly reset before looping around.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Merge list_for_each* and list_entry to list_for_each_entry*
Signed-off-by: Qinghuang Feng <qhfeng.kernel@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
kthread_run() returns the kthread or ERR_PTR(-ENOMEM), not NULL.
Signed-off-by: Qinghuang Feng <qhfeng.kernel@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The "devid <xxx> transid <xxx>" printk in btrfs_scan_one_device()
actually follows another printk that doesn't end in a newline (since the
intention is for the two printks to make one line of output), so the
KERN_INFO just ends up messing up the output:
device label exp <6>devid 1 transid 9 /dev/sda5
Fix this by changing the extra KERN_INFO to KERN_CONT.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Andrew's review of the xattr code revealed some minor issues that this patch
addresses. Just an error return fix, got rid of a useless statement and
commented one of the trickier parts of __btrfs_getxattr.
Signed-off-by: Josef Bacik <jbacik@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
- Remove the unused local variable 'len';
- Check return value of kmalloc().
Signed-off-by: Wang Cong <wangcong@zeuux.org>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The structure used to send device in btrfs ioctl calls was not
properly aligned, and so 32 bit ioctls would not work properly on
64 bit kernels.
We could fix this with compat ioctls, but we're just one byte away
and it doesn't make sense at this stage to carry about the compat ioctls
forever at this stage in the project.
This patch brings the ioctl arg up to an evenly aligned 4k.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs maintains a queue of async bio submissions so the checksumming
threads don't have to wait on get_request_wait. In order to avoid
extra wakeups, this code has a running_pending flag that is used
to tell new submissions they don't need to wake the thread.
When the threads notice congestion on a single device, they
may decide to requeue the job and move on to other devices. This
makes sure the running_pending flag is cleared before the
job is requeued.
It should help avoid IO stalls by making sure the task is woken up
when new submissions come in.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Each subvolume has an extent_state_tree used to mark metadata
that needs to be sent to disk while syncing the tree. This is
used in addition to the dirty bits on the pages themselves so that
a single subvolume can be sent to disk efficiently in disk order.
Normally this marking happens in btrfs_alloc_free_block, which also does
special recording of dirty tree blocks for the tree log roots.
Yan Zheng noticed that when the root of the log tree is allocated, it is added
to the wrong writeback list. The fix used here is to explicitly set
it dirty as part of tree log creation.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Checksum verification happens in a helper thread, and there is no
need to mess with interrupts. This switches to kmap() instead.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch contains following things.
1) Limit the max size of btrfs_ordered_sum structure to PAGE_SIZE. This
struct is kmalloced so we want to keep it reasonable.
2) Replace copy_extent_csums by btrfs_lookup_csums_range. This was
duplicated code in tree-log.c
3) Remove replay_one_csum. csum items are replayed at the same time as
replaying file extents. This guarantees we only replay useful csums.
4) nbytes accounting fix.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
btrfs_drop_extents doesn't change file extent's ram_bytes
in the case of booked extent. To be consistent, we should
also not change ram_bytes when truncating existing extent.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Snapshot creation happens at a specific time during transaction commit. We
need to make sure the code called by snapshot creation doesn't wait
for the running transaction to commit.
This changes btrfs_delete_inode and finish_pending_snaps to use
btrfs_join_transaction instead of btrfs_start_transaction to avoid deadlocks.
It would be better if btrfs_delete_inode didn't use the join, but the
call path that triggers it is:
btrfs_commit_transaction->create_pending_snapshots->
create_pending_snapshot->btrfs_lookup_dentry->
fixup_tree_root_location->btrfs_read_fs_root->
btrfs_read_fs_root_no_name->btrfs_orphan_cleanup->iput
This will be fixed in a later patch by moving the orphan cleanup to the
cleaner thread.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This is a patch to fix discard semantic to make Btrfs work with FTL and SSD.
We can improve FTL's performance by telling it which sectors are freed by file
system. But if we don't tell FTL the information of free sectors in proper
time, the transaction mechanism of Btrfs will be destroyed and Btrfs could not
roll back the previous transaction under the power loss condition.
There are some problems in the old implementation:
1, In __free_extent(), the pinned down extents should not be discarded.
2, In free_extents(), the free extents are all pinned, so they need to
be discarded in transaction committing time instead of free_extents().
3, The reserved extent used by log tree should be discard too.
This patch change discard behavior as follows:
1, For the extents which need to be free at once,
we discard them in update_block_group().
2, Delay discarding the pinned extent in btrfs_finish_extent_commit()
when committing transaction.
3, Remove discarding from free_extents() and __free_extent()
4, Add discard interface into btrfs_free_reserved_extent()
5, Discard sectors before updating the free space cache, otherwise,
FTL will destroy file system data.
drop_one_dir_item does not properly update inode's link count. It can be
reproduced by executing following commands:
#touch test
#sync
#rm -f test
#dd if=/dev/zero bs=4k count=1 of=test conv=fsync
#echo b > /proc/sysrq-trigger
This fixes it by adding an BTRFS_ORPHAN_ITEM_KEY for the inode
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
The data in fs_info->super_for_commit are zeros before the
first transaction commit. If tree log sync and system crash
both occur before the first transaction commit, super block
will get corrupted.
This fixes it by properly filling in the super_for_commit field at
open time.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
In clear_state_cb, we should check 'tree->ops->clear_bit_hook' instead
of 'tree->ops->set_bit_hook'.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Only root can add/remove devices
Only root can defrag subtrees
Only files open for writing can be defragged
Only files open for writing can be the destination for a clone
Signed-off-by: Chris Mason <chris.mason@oracle.com>
there's a new ptrace arch level feature in .28:
config X86_PTRACE_BTS
bool "Branch Trace Store"
it has broken fork() handling: the old DS area gets copied over into
a new task without clearing it.
Fixes exist but they came too late:
c5dee61: x86, bts: memory accounting
bf53de9: x86, bts: add fork and exit handling
and are queued up for v2.6.29. This shows that the facility is still not
tested well enough to release into a stable kernel - disable it for now and
reactivate in .29. In .29 the hardware-branch-tracer will use the DS/BTS
facilities too - hopefully resulting in better code.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
flush_tlb_mm's "optimized" uniprocessor case of allocating a new
context for userspace is exposing a race where we can suddely return
to a syscall with the protection id and space id out of sync, trapping
on the next userspace access.
Debugged-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Tested-by: Helge Deller <deller@gmx.de>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When deleting an edac device, we have to wait for its edac_dev.work to be
completed before deleting the whole edac_dev structure. Since we have no
idea which work in current edac_poller's workqueue is the work we are
conerned about, we wait for all work in the edac_poller's workqueue to be
proceseed. This is done via flush_cpu_workqueue() which inserts a
wq_barrier into the tail of the workqueue and then sleeping on the
completion of this wq_barrier. The edac_poller will wake up sleepers when
it is found.
EDAC core creates only one kernel worker thread, edac_poller, to run the
works of all current edac devices. They share the same callback function
of edac_device_workq_function(), which would grab the mutex of
device_ctls_mutex first before it checks the device. This is exactly
where edac_poller and rmmod would have a great chance to deadlock.
In below call trace of rmmod > ... >
edac_device_del_device >
edac_device_workq_teardown > flush_workqueue > flush_cpu_workqueue,
device_ctls_mutex would have already been grabbed by
edac_device_del_device(). So, on one hand rmmod would sleep on the
completion of a wq_barrier, holding device_ctls_mutex; on the other hand
edac_poller would be blocked on the same mutex when it's running any one
of works of existing edac evices(Note, this edac_dev.work is likely to be
totally irrelevant to the one that is being removed right now)and never
would have a chance to run the work of above wq_barrier to wake rmmod up.
edac_device_workq_teardown() should not be called within the critical
region of device_ctls_mutex. Just like is done in edac_pci_del_device()
and edac_mc_del_mc(), where edac_pci_workq_teardown() and
edac_mc_workq_teardown() are called after related mutex are released.
Moreover, an edac_dev.work should check first if it is being removed. If
this is the case, then it should bail out immediately. Since not all of
existing edac devices are to be removed, this "shutting flag" should be
contained to edac device being removed. The current edac_dev.op_state can
be used to serve this purpose.
The original deadlock problem and the solution have been witnessed and
tested on actual hardware. Without the solution, rmmod an edac driver
would result in below deadlock:
root@localhost:/root> rmmod mv64x60_edac
EDAC DEBUG: mv64x60_dma_err_remove()
EDAC DEBUG: edac_device_del_device()
EDAC DEBUG: find_edac_device_by_dev()
(hang for a moment)
INFO: task edac-poller:2030 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
edac-poller D 00000000 0 2030 2
Call Trace:
[df159dc0] [c0071e3c] free_hot_cold_page+0x17c/0x304 (unreliable)
[df159e80] [c000a024] __switch_to+0x6c/0xa0
[df159ea0] [c03587d8] schedule+0x2f4/0x4d8
[df159f00] [c03598a8] __mutex_lock_slowpath+0xa0/0x174
[df159f40] [e1030434] edac_device_workq_function+0x28/0xd8 [edac_core]
[df159f60] [c003beb4] run_workqueue+0x114/0x218
[df159f90] [c003c674] worker_thread+0x5c/0xc8
[df159fd0] [c004106c] kthread+0x5c/0xa0
[df159ff0] [c0013538] original_kernel_thread+0x44/0x60
INFO: task rmmod:2062 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
rmmod D 0ff2c9fc 0 2062 1839
Call Trace:
[df119c00] [c0437a74] 0xc0437a74 (unreliable)
[df119cc0] [c000a024] __switch_to+0x6c/0xa0
[df119ce0] [c03587d8] schedule+0x2f4/0x4d8
[df119d40] [c03591dc] schedule_timeout+0xb0/0xf4
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If cgroup_get_rootdir() failed, free_cg_links() will be called in the
failure path, but tmp_cg_links hasn't been initialized at that time.
I introduced this bug in the 2.6.27 merge window.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
During test of the w1-gpio driver i found that in "w1.c:679
w1_slave_found()" the device id is converted to little-endian with
"cpu_to_le64()", but its not converted back to cpu format in "w1_io.c:293
w1_reset_select_slave()".
Based on a patch created by Andreas Hummel.
[akpm@linux-foundation.org: remove unneeded cast]
Reported-by: Andreas Hummel <andi_hummel@gmx.de>
Signed-off-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
