struct bhv_vnode is on it's way out, so move the trace buffer to the XFS
inode. Note that this makes the tracing macros rather misnamed, but this
kind of fallout will be fixed up incrementally later on.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29498a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
struct bhv_vnode is on it's way out, so move the I/O count to the XFS
inode.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29497a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
All flags previously handled at the vnode level are not in the xfs_inode
where we already have a flags mechanisms and free bits for flags
previously in the vnode.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29495a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
One of the perpetual scaling problems XFS has is indexing it's incore
inodes. We currently uses hashes and the default hash sizes chosen can
only ever be a tradeoff between memory consumption and the maximum
realistic size of the cache.
As a result, anyone who has millions of inodes cached on a filesystem
needs to tunes the size of the cache via the ihashsize mount option to
allow decent scalability with inode cache operations.
A further problem is the separate inode cluster hash, whose size is based
on the ihashsize but is smaller, and so under certain conditions (sparse
cluster cache population) this can become a limitation long before the
inode hash is causing issues.
The following patchset removes the inode hash and cluster hash and
replaces them with radix trees to avoid the scalability limitations of the
hashes. It also reduces the size of the inodes by 3 pointers....
SGI-PV: 969561
SGI-Modid: xfs-linux-melb:xfs-kern:29481a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Biggest bit is duplicating the dinode structure so we have one annotated for
native endianess and one for disk endianess. The other significant change
is that xfs_xlate_dinode_core is split into one helper per direction to
allow for proper annotations, everything else is trivial.
As a sidenode splitting out the incore dinode means we can move it into
xfs_inode.h in a later patch and severely improving on the include hell in
xfs.
SGI-PV: 968563
SGI-Modid: xfs-linux-melb:xfs-kern:29476a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
currently xfs_bmbt_rec_t is used both for ondisk extents as well as
host-endian ones. This patch adds a new xfs_bmbt_rec_host_t for the native
endian ones and cleans up the fallout. There have been various endianess
issues in the tracing / debug printf code that are fixed by this patch.
SGI-PV: 968563
SGI-Modid: xfs-linux-melb:xfs-kern:29318a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
In media spaces, video is often stored in a frame-per-file format. When
dealing with uncompressed realtime HD video streams in this format, it is
crucial that files do not get fragmented and that multiple files a placed
contiguously on disk.
When multiple streams are being ingested and played out at the same time,
it is critical that the filesystem does not cross the streams and
interleave them together as this creates seek and readahead cache miss
latency and prevents both ingest and playout from meeting frame rate
targets.
This patch set creates a "stream of files" concept into the allocator to
place all the data from a single stream contiguously on disk so that RAID
array readahead can be used effectively. Each additional stream gets
placed in different allocation groups within the filesystem, thereby
ensuring that we don't cross any streams. When an AG fills up, we select a
new AG for the stream that is not in use.
The core of the functionality is the stream tracking - each inode that we
create in a directory needs to be associated with the directories' stream.
Hence every time we create a file, we look up the directories' stream
object and associate the new file with that object.
Once we have a stream object for a file, we use the AG that the stream
object point to for allocations. If we can't allocate in that AG (e.g. it
is full) we move the entire stream to another AG. Other inodes in the same
stream are moved to the new AG on their next allocation (i.e. lazy
update).
Stream objects are kept in a cache and hold a reference on the inode.
Hence the inode cannot be reclaimed while there is an outstanding stream
reference. This means that on unlink we need to remove the stream
association and we also need to flush all the associations on certain
events that want to reclaim all unreferenced inodes (e.g. filesystem
freeze).
SGI-PV: 964469
SGI-Modid: xfs-linux-melb:xfs-kern:29096a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Barry Naujok <bnaujok@sgi.com>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Signed-off-by: Vlad Apostolov <vapo@sgi.com>
The problem that has been addressed is that of synchronising updates of
the file size with writes that extend a file. Without the fix the update
of a file's size, as a result of a write beyond eof, is independent of
when the cached data is flushed to disk. Often the file size update would
be written to the filesystem log before the data is flushed to disk. When
a system crashes between these two events and the filesystem log is
replayed on mount the file's size will be set but since the contents never
made it to disk the file is full of holes. If some of the cached data was
flushed to disk then it may just be a section of the file at the end that
has holes.
There are existing fixes to help alleviate this problem, particularly in
the case where a file has been truncated, that force cached data to be
flushed to disk when the file is closed. If the system crashes while the
file(s) are still open then this flushing will never occur.
The fix that we have implemented is to introduce a second file size,
called the in-memory file size, that represents the current file size as
viewed by the user. The existing file size, called the on-disk file size,
is the one that get's written to the filesystem log and we only update it
when it is safe to do so. When we write to a file beyond eof we only
update the in- memory file size in the write operation. Later when the I/O
operation, that flushes the cached data to disk completes, an I/O
completion routine will update the on-disk file size. The on-disk file
size will be updated to the maximum offset of the I/O or to the value of
the in-memory file size if the I/O includes eof.
SGI-PV: 958522
SGI-Modid: xfs-linux-melb:xfs-kern:28322a
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
This patch handles error return values in fs_flush_pages and
fs_flushinval_pages. It changes the prototype of fs_flushinval_pages so we
can propogate the errors and handle them at higher layers. I also modified
xfs_itruncate_start so that it could propogate the error further.
SGI-PV: 961990
SGI-Modid: xfs-linux-melb:xfs-kern:28231a
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Stewart Smith <stewart@flamingspork.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
The previous attempts to fix the linux inode use-after-free in xfs_iunpin
simply made the problem harder to hit. We actually need complete exclusion
between xfs_reclaim and xfs_iunpin, as well as ensuring that the i_flags
are consistent during both of these functions. Introduce a new spinlock
for exclusion and the i_flags, and fix up xfs_iunpin to use igrab before
marking the inode dirty.
SGI-PV: 952967
SGI-Modid: xfs-linux-melb:xfs-kern:26964a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
millions of inodes cached and has sparse cluster population, removing
inodes from the cluster hash consumes excessive amounts of CPU time.
Reduce the CPU cost by making removal O(1) via use of a double linked list
for the hash chains.
SGI-PV: 951551
SGI-Modid: xfs-linux-melb:xfs-kern:25683a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Nathan Scott <nathans@sgi.com>
xfs_bmap_search_multi_extents() wrapper function that I introduced in mod
xfs-linux:xfs-kern:207393a. The function was added as a wrapper around
xfs_bmap_do_search_extents() to avoid breaking the top-of-tree CXFS
interface. The idea of the function was basically to extract the target
extent buffer (if muli- level extent allocation mode), then call
xfs_bmap_do_search_extents() with either a pointer to the first extent in
the target buffer or a pointer to the first extent in the file, depending
on which extent mode was being used. However, in addition to locating the
target extent record for block bno, xfs_bmap_do_search_extents() also sets
four parameters needed by the caller: *lastx, *eofp, *gotp, *prevp.
Passing only the target extent buffer to xfs_bmap_do_search_extents()
causes *eofp to be set incorrectly if the extent is at the end of the
target list but there are actually more extents in the next er_extbuf.
Likewise, if the extent is the first one in the buffer but NOT the first
in the file, *prevp is incorrectly set to NULL. Adding the needed
functionality to xfs_bmap_search_multi_extents() to re-set any incorrectly
set fields is redundant and makes the call to xfs_bmap_do_search_extents()
not make much sense when multi-level extent allocation mode is being used.
This mod basically extracts the two functional components from
xfs_bmap_do_search_extents(), with the intent of obsoleting/removing
xfs_bmap_do_search_extents() after the CXFS mult-level in-core extent
changes are checked in. The two components are: 1) The binary search to
locate the target extent record, and 2) Setting the four parameters needed
by the caller (*lastx, *eofp, *gotp, *prevp). Component 1: I created a
new function in xfs_inode.c called xfs_iext_bno_to_ext(), which executes
the binary search to find the target extent record.
xfs_bmap_search_multi_extents() has been modified to call
xfs_iext_bno_to_ext() rather than xfs_bmap_do_search_extents(). Component
2: The parameter setting functionality has been added to
xfs_bmap_search_multi_extents(), eliminating the need for
xfs_bmap_do_search_extents(). These changes make the removal of
xfs_bmap_do_search_extents() trival once the CXFS changes are in place.
They also allow us to maintain the current XFS interface, using the new
search function introduced in mod xfs-linux:xfs-kern:207393a.
SGI-PV: 928864
SGI-Modid: xfs-linux-melb:xfs-kern:207866a
Signed-off-by: Mandy Kirkconnell <alkirkco@sgi.com>
Signed-off-by: Nathan Scott <nathans@sgi.com>
functionality, building upon the new layout introduced in mod
xfs-linux:xfs-kern:207390a. The new multi-level extent allocations are
only required for heavily fragmented files, so the old-style linear extent
list is used on files until the extents reach a pre-determined size of 4k.
4k buffers are used because this is the system page size on Linux i386 and
systems with larger page sizes don't seem to gain much, if anything, by
using their native page size as the extent buffer size. Also, using 4k
extent buffers everywhere provides a consistent interface for CXFS across
different platforms. The 4k extent buffers are managed by an indirection
array (xfs_ext_irec_t) which is basically just a pointer array with a bit
of extra information to keep track of the number of extents in each buffer
as well as the extent offset of each buffer. Major changes include: -
Add multi-level in-core file extent functionality to the xfs_iext_
subroutines introduced in mod: xfs-linux:xfs-kern:207390a - Introduce 13
new subroutines which add functionality for multi-level in-core file
extents: xfs_iext_add_indirect_multi()
xfs_iext_remove_indirect() xfs_iext_realloc_indirect()
xfs_iext_indirect_to_direct() xfs_iext_bno_to_irec()
xfs_iext_idx_to_irec() xfs_iext_irec_init()
xfs_iext_irec_new() xfs_iext_irec_remove()
xfs_iext_irec_compact() xfs_iext_irec_compact_pages()
xfs_iext_irec_compact_full() xfs_iext_irec_update_extoffs()
SGI-PV: 928864
SGI-Modid: xfs-linux-melb:xfs-kern:207393a
Signed-off-by: Mandy Kirkconnell <alkirkco@sgi.com>
Signed-off-by: Nathan Scott <nathans@sgi.com>
code to prepare for an upcoming mod which will introduce multi-level
in-core extent allocations. Although the in-core extent management is
using a new code path in this mod, the functionality remains the same.
Major changes include: - Introduce 10 new subroutines which re-orgainze
the existing code but do NOT change functionality:
xfs_iext_get_ext() xfs_iext_insert() xfs_iext_add()
xfs_iext_remove() xfs_iext_remove_inline()
xfs_iext_remove_direct() xfs_iext_realloc_direct()
xfs_iext_direct_to_inline() xfs_iext_inline_to_direct()
xfs_iext_destroy() - Remove 2 subroutines (functionality moved to new
subroutines above): xfs_iext_realloc() -replaced by xfs_iext_add()
and xfs_iext_remove() xfs_bmap_insert_exlist() - replaced by
xfs_iext_insert() xfs_bmap_delete_exlist() - replaced by
xfs_iext_remove() - Replace all hard-coded (indexed) extent assignments
with a call to xfs_iext_get_ext() - Replace all extent record pointer
arithmetic (ep++, ep--, base + lastx,..) with calls to
xfs_iext_get_ext() - Update comments to remove the idea of a single
"extent list" and introduce "extent record" terminology instead
SGI-PV: 928864
SGI-Modid: xfs-linux-melb:xfs-kern:207390a
Signed-off-by: Mandy Kirkconnell <alkirkco@sgi.com>
Signed-off-by: Nathan Scott <nathans@sgi.com>
updates and only sync back to the xfs inode when nessecary
SGI-PV: 946679
SGI-Modid: xfs-linux-melb:xfs-kern:203362a
Signed-off-by: Christoph Hellwig <hch@sgi.com>
Signed-off-by: Nathan Scott <nathans@sgi.com>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
Christoph Hellwig
David Chinner
Lachlan McIlroy
Nathan Scott
Mandy Kirkconnell
Christoph Hellwig
Linus Torvalds