At writepage time, we have the page locked and we have the
extent_map entry for this extent pinned in the extent_map tree.
So, the page can't go away and its mapping can't change.
There is no need for the extra extent_state lock bits during writepage.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Many of the btrfs extent state tree users follow the same pattern.
They lock an extent range in the tree, do some operation and then
unlock.
This translates to at least 2 rbtree searches, and maybe more if they
are doing operations on the extent state tree. A locked extent
in the tree isn't going to be merged or changed, and so we can
safely return the extent state structure as a cached handle.
This changes set_extent_bit to give back a cached handle, and also
changes both set_extent_bit and clear_extent_bit to use the cached
handle if it is available.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs is currently mirroring some of the page state bits into
its extent state tree. The goal behind this was to use it in supporting
blocksizes other than the page size.
But, we don't currently support that, and we're using quite a lot of CPU
on the rb tree and its spin lock. This commit starts a series of
cleanups to reduce the amount of work done in the extent state tree as
part of each IO.
This commit:
* Adds the ability to lock an extent in the state tree and also set
other bits. The idea is to do locking and delalloc in one call
* Removes the EXTENT_WRITEBACK and EXTENT_DIRTY bits. Btrfs is using
a combination of the page bits and the ordered write code for this
instead.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
As the extent state tree is manipulated, there are call backs
that are used to take extra actions when different state bits are set
or cleared. One example of this is a counter for the total number
of delayed allocation bytes in a single inode and in the whole FS.
When new states are inserted, this callback is being done before we
properly setup the new state. This hasn't caused problems before
because the lock bit was always done first, and the existing call backs
don't care about the lock bit.
This patch makes sure the state is properly setup before using the
callback, which is important for later optimizations that do more work
without using the lock bit.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
There are two main users of the extent_map tree. The
first is regular file inodes, where it is evenly spread
between readers and writers.
The second is the chunk allocation tree, which maps blocks from
logical addresses to phyiscal ones, and it is 99.99% reads.
The mapping tree is a point of lock contention during heavy IO
workloads, so this commit switches things to a rw lock.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The btrfs io submission thread tries to back off congested devices in
favor of rotating off to another disk.
But, it tries to make sure it submits at least some IO before rotating
on (the others may be congested too), and so it has a magic number of
requests it tries to write before it hops.
This makes the magic number smaller. Testing shows that we're spending
too much time on congested devices and leaving the other devices idle.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When btrfs fills a large delayed allocation extent, it is a good idea
to try and convince the write_cache_pages caller to go ahead and
write a good chunk of that extent. The extra IO is basically free
because we know it is contiguous.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This changes the btrfs worker threads to batch work items
into a local list. It allows us to pull work items in
large chunks and significantly reduces the number of times we
need to take the worker thread spinlock.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The btrfs worker thread spinlock was being used both for the
queueing of IO and for the processing of ordered events.
The ordered events never happen from end_io handlers, and so they
don't need to use the _irq version of spinlocks. This adds a
dedicated lock to the ordered lists so they don't have to run
with irqs off.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The Btrfs set_extent_bit call currently searches the rbtree
every time it needs to find more extent_state objects to fill
the requested operation.
This adds a simple test with rb_next to see if the next object
in the tree was adjacent to the one we just found. If so,
we skip the search and just use the next object.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The Btrfs worker threads don't currently die off after they have
been idle for a while, leading to a lot of threads sitting around
doing nothing for each mount.
Also, they are unable to start atomically (from end_io hanlders).
This commit reworks the worker threads so they can be started
from end_io handlers (just setting a flag that asks for a thread
to be added at a later date) and so they can exit if they
have been idle for a long time.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Splice should update the modification and access times on regular
files just like read and write. Not updating mtime will confuse
backup tools, etc...
This patch only adds the time updates for regular files. For pipes
and other special files that splice touches the need for updating the
times is less clear. Let's discuss and fix that separately.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Get rid of any functions that test for these bits and make callers
use bio_rw_flagged() directly. Then it is at least directly apparent
what variable and flag they check.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Also a debugging aid. We want to catch dirty inodes being added to
backing devices that don't do writeback.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This enables us to track who does what and print info. Its main use
is catching dirty inodes on the default_backing_dev_info, so we can
fix that up.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This gets rid of pdflush for bdi writeout and kupdated style cleaning.
pdflush writeout suffers from lack of locality and also requires more
threads to handle the same workload, since it has to work in a
non-blocking fashion against each queue. This also introduces lumpy
behaviour and potential request starvation, since pdflush can be starved
for queue access if others are accessing it. A sample ffsb workload that
does random writes to files is about 8% faster here on a simple SATA drive
during the benchmark phase. File layout also seems a LOT more smooth in
vmstat:
r b swpd free buff cache si so bi bo in cs us sy id wa
0 1 0 608848 2652 375372 0 0 0 71024 604 24 1 10 48 42
0 1 0 549644 2712 433736 0 0 0 60692 505 27 1 8 48 44
1 0 0 476928 2784 505192 0 0 4 29540 553 24 0 9 53 37
0 1 0 457972 2808 524008 0 0 0 54876 331 16 0 4 38 58
0 1 0 366128 2928 614284 0 0 4 92168 710 58 0 13 53 34
0 1 0 295092 3000 684140 0 0 0 62924 572 23 0 9 53 37
0 1 0 236592 3064 741704 0 0 4 58256 523 17 0 8 48 44
0 1 0 165608 3132 811464 0 0 0 57460 560 21 0 8 54 38
0 1 0 102952 3200 873164 0 0 4 74748 540 29 1 10 48 41
0 1 0 48604 3252 926472 0 0 0 53248 469 29 0 7 47 45
where vanilla tends to fluctuate a lot in the creation phase:
r b swpd free buff cache si so bi bo in cs us sy id wa
1 1 0 678716 5792 303380 0 0 0 74064 565 50 1 11 52 36
1 0 0 662488 5864 319396 0 0 4 352 302 329 0 2 47 51
0 1 0 599312 5924 381468 0 0 0 78164 516 55 0 9 51 40
0 1 0 519952 6008 459516 0 0 4 78156 622 56 1 11 52 37
1 1 0 436640 6092 541632 0 0 0 82244 622 54 0 11 48 41
0 1 0 436640 6092 541660 0 0 0 8 152 39 0 0 51 49
0 1 0 332224 6200 644252 0 0 4 102800 728 46 1 13 49 36
1 0 0 274492 6260 701056 0 0 4 12328 459 49 0 7 50 43
0 1 0 211220 6324 763356 0 0 0 106940 515 37 1 10 51 39
1 0 0 160412 6376 813468 0 0 0 8224 415 43 0 6 49 45
1 1 0 85980 6452 886556 0 0 4 113516 575 39 1 11 54 34
0 2 0 85968 6452 886620 0 0 0 1640 158 211 0 0 46 54
A 10 disk test with btrfs performs 26% faster with per-bdi flushing. A
SSD based writeback test on XFS performs over 20% better as well, with
the throughput being very stable around 1GB/sec, where pdflush only
manages 750MB/sec and fluctuates wildly while doing so. Random buffered
writes to many files behave a lot better as well, as does random mmap'ed
writes.
A separate thread is added to sync the super blocks. In the long term,
adding sync_supers_bdi() functionality could get rid of this thread again.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This is a first step at introducing per-bdi flusher threads. We should
have no change in behaviour, although sb_has_dirty_inodes() is now
ridiculously expensive, as there's no easy way to answer that question.
Not a huge problem, since it'll be deleted in subsequent patches.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This adds two new exported functions:
- writeback_inodes_sb(), which only attempts to writeback dirty inodes on
this super_block, for WB_SYNC_NONE writeout.
- sync_inodes_sb(), which writes out all dirty inodes on this super_block
and also waits for the IO to complete.
Acked-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
When ext4_dx_add_entry() has to split an index node, it has to ensure that
name_len of dx_node's fake_dirent is also zero, because otherwise e2fsck
won't recognise it as an intermediate htree node and consider the htree to
be corrupted.
Signed-off-by: Andreas Schlick <schlick@lavabit.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Previously the journal_async_commit mount option was equivalent to
using barrier=0 (and just as unsafe). This patch fixes it so that we
eliminate the barrier before the commit block (by not using ordered
mode), and explicitly issuing an empty barrier bio after writing the
commit block. Because of the journal checksum, it is safe to do this;
if the journal blocks are not all written before a power failure, the
checksum in the commit block will prevent the last transaction from
being replayed.
Using the fs_mark benchmark, using journal_async_commit shows a 50%
improvement:
FSUse% Count Size Files/sec App Overhead
8 1000 10240 30.5 28242
vs.
FSUse% Count Size Files/sec App Overhead
8 1000 10240 45.8 28620
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This avoids updating the superblock write time when we are mounting
the root file system read/only but we need to replay the journal; at
that point, for people who are east of GMT and who make their clock
tick in localtime for Windows bug-for-bug compatibility, and this will
cause e2fsck to complain and force a full file system check.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
In fs/binfmt_elf.c, load_elf_interp() calls padzero() for .bss even if
the PT_LOAD has no PROT_WRITE and no .bss. This generates EFAULT.
Here is a small test case. (Yes, there are other, useful PT_INTERP
which have only .text and no .data/.bss.)
----- ptinterp.S
_start: .globl _start
nop
int3
-----
$ gcc -m32 -nostartfiles -nostdlib -o ptinterp ptinterp.S
$ gcc -m32 -Wl,--dynamic-linker=ptinterp -o hello hello.c
$ ./hello
Segmentation fault # during execve() itself
After applying the patch:
$ ./hello
Trace trap # user-mode execution after execve() finishes
If the ELF headers are actually self-inconsistent, then dying is fine.
But having no PROT_WRITE segment is perfectly normal and correct if
there is no segment with p_memsz > p_filesz (i.e. bss). John Reiser
suggested checking for PROT_WRITE in the bss logic. I think it makes
most sense to simply apply the bss logic only when there is bss.
This patch looks less trivial than it is due to some reindentation.
It just moves the "if (last_bss > elf_bss) {" test up to include the
partial-page bss logic as well as the more-pages bss logic.
Reported-by: John Reiser <jreiser@bitwagon.com>
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: James Morris <jmorris@namei.org>
This patch amends and nicifies commentaries in file.c, as well as
fixes some spelling problems.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
The 'ubifs_scan()' function returns -EUCLEAN if something is corrupted
and recovery is needed, otherwise it returns other error codes. However,
in few places UBIFS does not check the error codes and runs recovery.
This patch changes this behavior and makes UBIFS start recovery only
on -EUCLEAN errors.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Reviewed-by: Adrian Hunter <Adrian.Hunter@nokia.com>
At the moment UBIFS print large and scary error messages and
flash dumps in case of nearly any corruption, even if it is
a recoverable corruption. For example, if the master node is
corrupted, ubifs_scan() prints error dumps, then UBIFS recovers
just fine and goes on.
This patch makes UBIFS print scary error messages only in
real cases, which are not recoverable. It adds 'quiet' argument
to the 'ubifs_scan()' function, so the caller may ask 'ubi_scan()'
not to print error messages if the caller is able to do recovery.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Reviewed-by: Adrian Hunter <Adrian.Hunter@nokia.com>
Add one more check to UBIFS - a check that makes sure that there
are no data nodes beyond inode size. And few commantaries fixes
along the line.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Reviewed-by: Adrian Hunter <Adrian.Hunter@nokia.com>
We don't need to take the alloc_sem lock when we are adding new
groups, since mballoc won't see the new group added until we bump
sbi->s_groups_count.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
We should check for need init flag with the group's alloc_sem held, to
make sure while we are loading the buddy cache and holding a reference
to it, a file system resize can't add new blocks to same group.
The patch also drops the need init flag check in
ext4_mb_regular_allocator() because doing the check without holding
alloc_sem is racy.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
This moves the function around so that it can be called from
ext4_mb_load_buddy().
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
In fs/binfmt_elf.c, load_elf_interp() calls padzero() for .bss even if
the PT_LOAD has no PROT_WRITE and no .bss. This generates EFAULT.
Here is a small test case. (Yes, there are other, useful PT_INTERP
which have only .text and no .data/.bss.)
----- ptinterp.S
_start: .globl _start
nop
int3
-----
$ gcc -m32 -nostartfiles -nostdlib -o ptinterp ptinterp.S
$ gcc -m32 -Wl,--dynamic-linker=ptinterp -o hello hello.c
$ ./hello
Segmentation fault # during execve() itself
After applying the patch:
$ ./hello
Trace trap # user-mode execution after execve() finishes
If the ELF headers are actually self-inconsistent, then dying is fine.
But having no PROT_WRITE segment is perfectly normal and correct if
there is no segment with p_memsz > p_filesz (i.e. bss). John Reiser
suggested checking for PROT_WRITE in the bss logic. I think it makes
most sense to simply apply the bss logic only when there is bss.
This patch looks less trivial than it is due to some reindentation.
It just moves the "if (last_bss > elf_bss) {" test up to include the
partial-page bss logic as well as the more-pages bss logic.
Reported-by: John Reiser <jreiser@bitwagon.com>
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Teach ext4_write_inode() and ext4_do_update_inode() about non-journal
mode: If we're not using a journal, ext4_write_inode() now calls
ext4_do_update_inode() (after getting the iloc via ext4_get_inode_loc())
with a new "do_sync" parameter. If that parameter is nonzero _and_ we're
not using a journal, ext4_do_update_inode() calls sync_dirty_buffer()
instead of ext4_handle_dirty_metadata().
This problem was found in power-fail testing, checking the amount of
loss of files and blocks after a power failure when using fsync() and
when not using fsync(). It turned out that using fsync() was actually
worse than not doing so, possibly because it increased the likelihood
that the inodes would remain unflushed and would therefore be lost at
the power failure.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When there is no journal present, we must attach buffer heads
associated with extent tree and indirect blocks to the inode's
mapping->private_list via mark_buffer_dirty_inode() so that
ext4_sync_file() --- which is called to service fsync() and
fdatasync() system calls --- can write out the inode's metadata blocks
by calling sync_mapping_buffers().
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
When ext4 is using a journal, a metadata block which is deallocated
must be passed into the journal layer so it can be dropped from the
current transaction and/or revoked. This is done by calling the
functions ext4_journal_forget() and ext4_journal_revoke(), which call
jbd2_journal_forget(), and jbd2_journal_revoke(), respectively.
Since the jbd2_journal_forget() and jbd2_journal_revoke() call
bforget(), if ext4 is not using a journal, ext4_journal_forget() and
ext4_journal_revoke() must call bforget() to avoid a dirty metadata
block overwriting a block after it has been reallocated and reused for
another inode's data block.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
This patch adds a setxattr handler to the file, directory, and symlink
inode_operations structures for sysfs. The patch uses hooks introduced in the
previous patch to handle the getting and setting of security information for
the sysfs inodes. As was suggested by Eric Biederman the struct iattr in the
sysfs_dirent structure has been replaced by a structure which contains the
iattr, secdata and secdata length to allow the changes to persist in the event
that the inode representing the sysfs_dirent is evicted. Because sysfs only
stores this information when a change is made all the optional data is moved
into one dynamically allocated field.
This patch addresses an issue where SELinux was denying virtd access to the PCI
configuration entries in sysfs. The lack of setxattr handlers for sysfs
required that a single label be assigned to all entries in sysfs. Granting virtd
access to every entry in sysfs is not an acceptable solution so fine grained
labeling of sysfs is required such that individual entries can be labeled
appropriately.
[sds: Fixed compile-time warnings, coding style, and setting of inode security init flags.]
Signed-off-by: David P. Quigley <dpquigl@tycho.nsa.gov>
Signed-off-by: Stephen D. Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@namei.org>
This factors out the part of the vfs_setxattr function that performs the
setting of the xattr and its notification. This is needed so the SELinux
implementation of inode_setsecctx can handle the setting of the xattr while
maintaining the proper separation of layers.
Signed-off-by: David P. Quigley <dpquigl@tycho.nsa.gov>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: James Morris <jmorris@namei.org>
We added the ENOSPC handling patch in xfs_create just after it got mered
with xfs_mkdir. Change the log reservation to the variable for either
the create or mkdir value so it does the right thing if get here for creating
a directory.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
The /sys/fs/gfs2/<fsname>/lock_module/id file has been unused for
some time now, so we can remove it. We still accept the mount option
though, as userspace still sends that.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When mounting an "nfs" type file system, recognize "v4," "vers=4," or
"nfsvers=4" mount options, and convert the file system to "nfs4" under
the covers.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
[trondmy: fixed up binary mount code so it sets the 'version' field too]
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up: Refactor nfs4_get_sb() to allow its guts to be invoked by
nfs_get_sb().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up: Refactor the part of nfs4_validate_mount_options() that
handles text-based options, so we can call it from the NFSv2/v3
option validation function.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The meaning of not specifying the "port=" mount option is different
for "-t nfs" and "-t nfs4" mounts. The default port value for
NFSv2/v3 mounts is 0, but the default for NFSv4 mounts is 2049.
To support "-t nfs -o vers=4", the mount option parser must detect
when "port=" is missing so that the correct default port value can be
set depending on which NFS version is requested.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Hi Trond,
Recently we were observing the behaviour difference between a 2.4.x and
2.6.x kernel with respect to O_EXCL. A comment from 2.4.x era, "For now,
we don't implement O_EXCL." seems inaccurate in TOT.
If so, here's a patch to remove the comment.
This patch is against:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
Signed-off-by: Harshula Jayasuriya <harshula@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
This avoids an indirect call in the VFS for each path component lookup.
Well, at least as long as you own the directory in question, and the ACL
check is unnecessary.
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Don't implement per-filesystem 'extX_permission()' functions that have
to be called for every path component operation, and instead just expose
the actual ACL checking so that the VFS layer can now do it for us.
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is stage one in flattening out the callchains for the common
permission testing. Rather than have most filesystem implement their
own inode->i_op->permission function that just calls back down to the
VFS layers 'generic_permission()' with the per-filesystem ACL checking
function, the filesystem can just expose its 'check_acl' function
directly, and let the VFS layer do everything for it.
This is all just preparatory - no filesystem actually enables this yet.
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Don't call down to the generic inode_permission() function just to
call the inode-specific permission function - just do it directly.
The generic inode_permission() code does things like checking MAY_WRITE
and devcgroup_inode_permission(), neither of which are relevant for the
light pathname walk permission checks (we always do just MAY_EXEC, and
the inode is never a special device).
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This function is only called for path components that are already known
to be directories (they have a '->lookup' method). So don't bother
doing that whole S_ISDIR() testing, the whole point of the 'lite()'
version is that we know that we are looking at a directory component,
and that we're only checking name lookup permission.
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Instead of returning EAGAIN and having the caller do something
special for that case, just do the special case directly.
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Chris Mason