David Chinner [Wed, 6 Feb 2008 02:37:40 +0000 (13:37 +1100)]
[XFS] Fix oops in xfs_file_readdir()
When xfs_file_readdir() exactly fills a buffer, it can move it's index
past the end of the buffer and dereference it even though the result of
the dereference is never used. On some platforms this causes an oops.
[XFS] keep i_nlink updated and use proper accessors
To get the read-only bind mounts in -mm to work correctly with XFS we need
to call the drop_nlink and inc_nlink helpers to monitor the link count.
Add calls to these to xfs_bumplink and xfs_droplink and stop copying over
di_nlink to i_nlink in xfs_validate_fields and vn_revalidate.
The VFS doesn't use i_blocks, it's only used by generic_fillattr and the
generic quota code which XFS doesn't use. In XFS there is one use to check
whether we have an inline or out of line sumlink, but we can replace that
with a check of the XFS_IFINLINE inode flag.
David Chinner [Tue, 5 Feb 2008 01:13:38 +0000 (12:13 +1100)]
[XFS] Make xfs_ail_check check less by default
Checking the entire AIL on every insert and remove is prohibitively
expensive - the sustained sequntial create rate on a single disk drops
from about 1800/s to 60/s because of this checking resulting in the
xfslogd becoming cpu bound.
By default on debug builds, only check the next and previous entries in
the list to ensure they are ordered correctly. If you really want, define
XFS_TRANS_DEBUG to use the old behaviour.
David Chinner [Tue, 5 Feb 2008 01:13:32 +0000 (12:13 +1100)]
[XFS] Move AIL pushing into it's own thread
When many hundreds to thousands of threads all try to do simultaneous
transactions and the log is in a tail-pushing situation (i.e. full), we
can get multiple threads walking the AIL list and contending on the AIL
lock.
The AIL push is, in effect, a simple I/O dispatch algorithm complicated by
the ordering constraints placed on it by the transaction subsystem. It
really does not need multiple threads to push on it - even when only a
single CPU is pushing the AIL, it can push the I/O out far faster that
pretty much any disk subsystem can handle.
So, to avoid contention problems stemming from multiple list walkers, move
the list walk off into another thread and simply provide a "target" to
push to. When a thread requires a push, it sets the target and wakes the
push thread, then goes to sleep waiting for the required amount of space
to become available in the log.
This mechanism should also be a lot fairer under heavy load as the waiters
will queue in arrival order, rather than queuing in "who completed a push
first" order.
Also, by moving the pushing to a separate thread we can do more
effectively overload detection and prevention as we can keep context from
loop iteration to loop iteration. That is, we can push only part of the
list each loop and not have to loop back to the start of the list every
time we run. This should also help by reducing the number of items we try
to lock and/or push items that we cannot move.
Note that this patch is not intended to solve the inefficiencies in the
AIL structure and the associated issues with extremely large list
contents. That needs to be addresses separately; parallel access would
cause problems to any new structure as well, so I'm only aiming to isolate
the structure from unbounded parallelism here.
Now that all direct caller of xfs_iaccess are gone we can kill xfs_iaccess
and xfs_access and just use generic_permission with a check_acl callback.
This is required for the per-mount read-only patchset in -mm to work
properly with XFS.
xfs_swapext should simplify check if we have a writeable file descriptor
instead of re-checking the permissions using xfs_iaccess. Add an
additional check to refuse O_APPEND file descriptors because swapext is
not an append-only write operation.
[XFS] remove permission check from xfs_change_file_space
Both callers of xfs_change_file_space alreaedy do the file->f_mode &
FMODE_WRITE check to ensure we have a file descriptor that has been opened
for write mode, so there is no need to re-check that with xfs_iaccess.
Especially as the later might wrongly deny it for corner cases like file
descriptor passing through unix domain sockets.
Lachlan McIlroy [Thu, 10 Jan 2008 05:43:36 +0000 (16:43 +1100)]
[XFS] prevent panic during log recovery due to bogus op_hdr length
A problem was reported where a system panicked in log recovery due to a
corrupt log record. The cause of the corruption is not known but this
change will at least prevent a crash for this specific scenario. Log
recovery definitely needs some more work in this area.
- merge xfs_fid2 into it's only caller xfs_dm_inode_to_fh.
- remove xfs_vget and opencode it in the two callers, simplifying
both of them by avoiding the awkward calling convetion.
- assign directly to the dm_fid_t members in various places in the
dmapi code instead of casting them to xfs_fid_t first (which
is identical to dm_fid_t)
[XFS] Remove CFORK macros and use code directly in IFORK and DFORK macros.
Currently XFS_IFORK_* and XFS_DFORK* are implemented by means of
XFS_CFORK* macros. But given that XFS_IFORK_* operates on an xfs_inode
that embedds and xfs_icdinode_core and XFS_DFORK_* operates on an
xfs_dinode that embedds a xfs_dinode_core one will have to do endian
swapping while the other doesn't. Instead of having the current mess with
the CFORK macros that have byteswapping and non-byteswapping version
(which are inconsistantly named while we're at it) just define each family
of the macros to stand by itself and simplify the whole matter.
A few direct references to the CFORK variants were cleaned up to use IFORK
or DFORK to make this possible.
There is no need to lock any page in xfs_buf.c because we operate on our
own address_space and all locking is covered by the buffer semaphore. If
we ever switch back to main blockdeive address_space as suggested e.g. for
fsblock with a similar scheme the locking will have to be totally revised
anyway because the current scheme is neither correct nor coherent with
itself.
Tim Shimmin [Fri, 23 Nov 2007 05:30:42 +0000 (16:30 +1100)]
[XFS] Remove the BPCSHIFT and NB* based macros from XFS.
The BPCSHIFT based macros, btoc*, ctob*, offtoc* and ctooff are either not
used or don't need to be used. The NDPP, NDPP, NBBY macros don't need to
be used but instead are replaced directly by PAGE_SIZE and PAGE_CACHE_SIZE
where appropriate. Initial patch and motivation from Nicolas Kaiser.
Niv Sardi [Tue, 27 Nov 2007 06:01:13 +0000 (17:01 +1100)]
[XFS] Remove bogus assert
This assert is bogus. We can have a forced shutdown occur
between the check for the XLOG_FORCED_SHUTDOWN and the ASSERT. Also, the
logging system shouldn't care about the state of XFS_FORCED_SHUTDOWN, it
should only check XLOG_FORCED_SHUTDOWN. The logging system has it's own
forced shutdown flag so, for the case of a forced shutdown that's not due
to a logging error, we can flush the log.
David Chinner [Fri, 23 Nov 2007 05:29:18 +0000 (16:29 +1100)]
[XFS] Fix inode allocation latency
The log force added in xfs_iget_core() has been a performance issue since
it was introduced for tight loops that allocate then unlink a single file.
under heavy writeback, this can introduce unnecessary latency due tothe
log I/o getting stuck behind bulk data writes.
Fix this latency problem by avoinding the need for the log force by moving
the place we mark linux inode dirty to the transaction commit rather than
on transaction completion.
This also closes a potential hole in the sync code where a linux inode is
not dirty between the time it is modified and the time the log buffer has
been written to disk.
David Chinner [Fri, 23 Nov 2007 05:29:11 +0000 (16:29 +1100)]
[XFS] Fix transaction overrun during writeback.
Prevent transaction overrun in xfs_iomap_write_allocate() if we race with
a truncate that overlaps the delalloc range we were planning to allocate.
If we race, we may allocate into a hole and that requires block
allocation. At this point in time we don't have a reservation for block
allocation (apart from metadata blocks) and so allocating into a hole
rather than a delalloc region results in overflowing the transaction block
reservation.
Fix it by only allowing a single extent to be allocated at a time.
David Chinner [Fri, 23 Nov 2007 05:28:17 +0000 (16:28 +1100)]
[XFS] Fix sparse warning in xlog_recover_do_efd_trans.
Sparse trips over the locking order in xlog_recover_do_efd_trans() when
xfs_trans_delete_ail() drops the ail lock. Because the unlock is
conditional, we need to either annotate with a "fake unlock" or change the
structure of the code so sparse thinks the function always unlocks.
Show C1 idle time in /sysfs cpuidle interface. C1 idle time may not
be entirely accurate in all cases. It includes the time spent
in the interrupt handler after wakeup with "hlt" based C1. But, it will
be accurate with "mwait" based C1.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
changed this behavior and made xfs_bulkstat() to filter all unlinked
inodes including those that are not destroyed yet but held by reference.
The attached patch returns back to the original behavior by marking the
on-disk inode buffer "dirty" when di_mode is cleared (at that time both
inode link and reference counter are zero).
There is no need to lock any page in xfs_buf.c because we operate on our
own address_space and all locking is covered by the buffer semaphore. If
we ever switch back to main blockdeive address_space as suggested e.g. for
fsblock with a similar scheme the locking will have to be totally revised
anyway because the current scheme is neither correct nor coherent with
itself.
Eric Sandeen [Fri, 12 Oct 2007 01:03:40 +0000 (11:03 +1000)]
[XFS] Refactor xfs_mountfs
Refactoring xfs_mountfs() to call sub-functions for logical chunks can
help save a bit of stack, and can make it easier to read this long
function.
The mount path is one of the longest common callchains, easily getting to
within a few bytes of the end of a 4k stack when over lvm, quotas are
enabled, and quotacheck must be done.
With this change on top of the other stack-related changes I've sent, I
can get xfs to survive a normal xfsqa run on 4k stacks over lvm.
Mostly trivial conversion with one exceptions: h_num_logops was kept in
native endian previously and only converted to big endian in xlog_sync,
but we always keep it big endian now. With todays cpus fast byteswap
instructions that's not an issue but the new variant keeps the code clean
and maintainable.
- the various assign lsn macros are replaced by a single inline,
xlog_assign_lsn, which is equivalent to ASSIGN_ANY_LSN_HOST except
for a more sane calling convention. ASSIGN_LSN_DISK is replaced
by xlog_assign_lsn and a manual bytespap, and ASSIGN_LSN by the same,
except we pass the cycle and block arguments explicitly instead of a
log paramter. The latter two variants only had 2, respectively one
user anyway.
- the GET_CYCLE is replaced by a xlog_get_cycle inline with exactly the
same calling conventions.
- GET_CLIENT_ID is replaced by xlog_get_client_id which leaves away
the unused arch argument. Instead of conditional defintions
depending on host endianess we now do an unconditional swap and shift
then, which generates equal code.
- the unused XLOG_SET macro is removed.
- the various assign lsn macros are replaced by a single inline,
xlog_assign_lsn, which is equivalent to ASSIGN_ANY_LSN_HOST except
for a more sane calling convention. ASSIGN_LSN_DISK is replaced
by xlog_assign_lsn and a manual bytespap, and ASSIGN_LSN by the same,
except we pass the cycle and block arguments explicitly instead of a
log paramter. The latter two variants only had 2, respectively one
user anyway.
- the GET_CYCLE is replaced by a xlog_get_cycle inline with exactly the
same calling conventions.
- GET_CLIENT_ID is replaced by xlog_get_client_id which leaves away
the unused arch argument. Instead of conditional defintions
depending on host endianess we now do an unconditional swap and shift
then, which generates equal code.
- the unused XLOG_SET macro is removed.
Export acpi_check_resource_conflict(), sometimes drivers already have
a struct resource at hand so no need to use the wrappers to build a new
one.
Signed-off-by: Jean Delvare <jdelvare@suse.de> Cc: "Mark M. Hoffman" <mhoffman@lightlink.com> Cc: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
ACPI: track opregion names to avoid driver resource conflicts.
Small ACPICA extension to be able to store the name of operation regions in osl.c later
In ACPI, AML can define accesses to IO ports and System Memory by Operation
Regions. Those are not registered as done by PNPACPI using resource templates
(and _CRS/_SRS methods).
The IO ports and System Memory regions may get accessed by arbitrary AML code.
When native drivers are accessing the same resources bad things can happen
(e.g. a critical shutdown temperature of 3000 C every 2 months or so).
It is not really possible to register the operation regions via
request_resource, as they often overlap with pnp or other resources (e.g.
statically setup IO resources below 0x100).
This approach stores all Operation Region declarations (IO and System Memory
only) at ACPI table parse time. It offers a similar functionality like
request_region and let drivers which are known to possibly use the same IO
ports and Memory which are also often used by ACPI (hwmon and i2c) check for
ACPI interference.
A boot parameter acpi_enforce_resources=strict/lax/no is provided, which
is default set to lax:
- strict: let conflicting drivers fail to load with an error message
- lax: let conflicting driver work normal with a warning message
- no: no functional change at all
Depending on the feedback and the kind of interferences we see, this
should be set to strict at later time.
Goal of this patch set is:
- Identify ACPI interferences in bug reports (very hard to reproduce
and to identify)
- Find BIOSes for that an ACPI driver should exist for specific HW
instead of a native one.
- stability in general
Provide acpi_check_{mem_}region.
Drivers can additionally check against possible ACPI interference by also
invoking this shortly before they call request_region.
If -EBUSY is returned, the driver must not load.
Use acpi_enforce_resources=strict/lax/no options to:
- strict: let conflicting drivers fail to load with an error message
- lax: let conflicting driver work normal with a warning message
- no: no functional change at all
Cc: "Mark M. Hoffman" <mhoffman@lightlink.com> Cc: Jean Delvare <khali@linux-fr.org> Cc: Len Brown <lenb@kernel.org> Cc: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Thomas Renninger <trenn@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
xfs_ioctl.c passes around vnode pointers quite a lot, but all places
already have the Linux inode which is identical to the vnode these days.
Clean the code up to always use the Linux inode.
We were already filling the Linux struct statfs anyway, and doing this
trivial task directly in xfs_fs_statfs makes the code quite a bit cleaner.
While I was at it I also moved copying attributes that don't change over
the lifetime of the filesystem outside the superblock lock.
xfs_fs_fill_super used to get the magic number and blocksize through
xfs_statvfs, but assigning them directly is a lot cleaner and will save
some stack space during mount.
ACPI: acpi_pci_irq_find_prt_entry(): use list_for_each_entry() instead of list_for_each()
Signed-off-by: Matthias Kaehlcke <matthias.kaehlcke@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
xfs_iocore_t is a structure embedded in xfs_inode. Except for one field it
just duplicates fields already in xfs_inode, and there is nothing this
abstraction buys us on XFS/Linux. This patch removes it and shrinks source
and binary size of xfs aswell as shrinking the size of xfs_inode by 60/44
bytes in debug/non-debug builds.
Eric Sandeen [Thu, 11 Oct 2007 07:43:56 +0000 (17:43 +1000)]
[XFS] Remove spin.h
remove spinlock init abstraction macro in spin.h, remove the callers, and
remove the file. Move no-op spinlock_destroy to xfs_linux.h Cleanup
spinlock locals in xfs_mount.c
Eric Sandeen [Thu, 11 Oct 2007 07:42:32 +0000 (17:42 +1000)]
[XFS] Unwrap XFS_SB_LOCK.
Un-obfuscate XFS_SB_LOCK, remove XFS_SB_LOCK->mutex_lock->spin_lock
macros, call spin_lock directly, remove extraneous cookie holdover from
old xfs code, and change lock type to spinlock_t.
Eric Sandeen [Thu, 11 Oct 2007 07:41:21 +0000 (17:41 +1000)]
[XFS] Unwrap xfs_dabuf_global_lock
Un-obfuscate dabuf_global_lock, remove mutex_lock->spin_lock macros, call
spin_lock directly, remove extraneous cookie holdover from old xfs code,
and change lock type to spinlock_t.
Eric Sandeen [Thu, 11 Oct 2007 07:38:28 +0000 (17:38 +1000)]
[XFS] Unwrap pagb_lock.
Un-obfuscate pagb_lock, remove mutex_lock->spin_lock macros, call
spin_lock directly, remove extraneous cookie holdover from old xfs code,
and change lock type to spinlock_t.
Luca Tettamanti [Tue, 5 Feb 2008 07:31:18 +0000 (23:31 -0800)]
asus_acpi: add support for F3Sa
Add support for ASUS F3Sa notebook. Features:
- LCD on/off
- Brightness
- Wifi kill
- Bluetooth kill
Signed-off-by: Luca Tettamanti <kronos.it@gmail.com> Cc: Corentin Chary <corentincj@iksaif.net> Cc: Karol Kozimor <sziwan@users.sourceforge.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
Eric Sandeen [Thu, 11 Oct 2007 07:38:18 +0000 (17:38 +1000)]
[XFS] Unwrap XFS_DQ_PINUNLOCK.
Un-obfuscate DQ_PINLOCK, remove DQ_PINLOCK->mutex_lock->spin_lock macros,
call spin_lock directly, remove extraneous cookie holdover from old xfs
code, and change lock type to spinlock_t.
Eric Sandeen [Thu, 11 Oct 2007 07:37:31 +0000 (17:37 +1000)]
[XFS] Unwrap GRANT_LOCK.
Un-obfuscate GRANT_LOCK, remove GRANT_LOCK->mutex_lock->spin_lock macros,
call spin_lock directly, remove extraneous cookie holdover from old xfs
code, and change lock type to spinlock_t.
Eric Sandeen [Thu, 11 Oct 2007 07:37:10 +0000 (17:37 +1000)]
[XFS] Unwrap LOG_LOCK.
Un-obfuscate LOG_LOCK, remove LOG_LOCK->mutex_lock->spin_lock macros, call
spin_lock directly, remove extraneous cookie holdover from old xfs code,
and change lock type to spinlock_t.
Lachlan McIlroy [Thu, 11 Oct 2007 07:34:33 +0000 (17:34 +1000)]
[XFS] kill unnessecary ioops indirection
Currently there is an indirection called ioops in the XFS data I/O path.
Various functions are called by functions pointers, but there is no
coherence in what this is for, and of course for XFS itself it's entirely
unused. This patch removes it instead and significantly reduces source and
binary size of XFS while making maintaince easier.
There is no reason to go through xfs_iomap for the BMAPI_UNWRITTEN because
it has nothing in common with the other cases. Instead check for the
shutdown filesystem in xfs_end_bio_unwritten and perform a direct call to
xfs_iomap_write_unwritten (which should be renamed to something more
sensible one day)
There is no reason to go into the iomap machinery just to get the right
block device for an inode. Instead look at the realtime flag in the inode
and grab the right device from the mount structure.
I created a new helper, xfs_find_bdev_for_inode instead of opencoding it
because I plan to use it in other places in the future.
[XFS] remove dead SYNC_BDFLUSH case in xfs_sync_inodes
A large part of xfs_sync_inodes is conditional on the SYNC_BDFLUSH which
is never passed to it. This patch removes it and adds an assert that
triggers in case some new code tries to pass SYNC_BDFLUSH to it.
Roland Dreier [Thu, 7 Feb 2008 05:17:59 +0000 (21:17 -0800)]
mlx4_core: Clean up struct mlx4_buf
Now that struct mlx4_buf.u is a struct instead of a union because of
the vmap() changes, there's no point in having a struct at all. So
move .direct and .page_list directly into struct mlx4_buf and get rid
of a bunch of unnecessary ".u"s.
Jack Morgenstein [Mon, 28 Jan 2008 08:40:51 +0000 (10:40 +0200)]
mlx4_core: For 64-bit systems, vmap() kernel queue buffers
Since kernel virtual memory is not a problem on 64-bit systems, there
is no reason to use our own 2-layer page mapping scheme for large
kernel queue buffers on such systems. Instead, map the page list to a
single virtually contiguous buffer with vmap(), so that can we access
buffer memory via direct indexing.
Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il> Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
Márton Németh [Thu, 24 Jan 2008 05:58:57 +0000 (15:58 +1000)]
drm: cleanup DRM_DEBUG() parameters
As DRM_DEBUG macro already prints out the __FUNCTION__ string (see
drivers/char/drm/drmP.h), it is not worth doing this again. At some
other places the ending "\n" was added.
airlied:- I cleaned up a few that this patch missed also
Jesper Juhl [Sun, 16 Dec 2007 23:47:17 +0000 (09:47 +1000)]
mga_dma: return 'err' not just zero from mga_do_cleanup_dma()
While reading some code I stumbled across the use of 'err' in
drivers/char/drm/mga_dma.c::mga_do_cleanup_dma() and I think there's a small
problem.
The variable is only used inside #if __OS_HAS_AGP which is fine, but all that
ever happens is an assignment to the variable - it is never actually used for
anything. The variable is nicely initialized to zero which is also what the
return statement at the end of function returns (always at the moment).
It looks to me like that function should be returning 'err' instead of always
just returning 0. Here's a patch to do that.
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Dave Airlie <airlied@linux.ie>
Dave Airlie [Sun, 16 Dec 2007 23:41:56 +0000 (09:41 +1000)]
drm: add _DRM_DRIVER flag, and re-order unload.
Allow drivers to addmaps that won't be removed by lastclose or unload.
The unload needs to be re-ordered to avoid removing the hashs before
the driver has removed the final maps.
Ian Romanick [Thu, 22 Nov 2007 07:02:08 +0000 (17:02 +1000)]
drm: Fix ioc32 compat layer
Previously any ioctls that weren't explicitly listed in the compat ioctl
table would fail with ENOTTY. If the incoming ioctl number is outside the
range of the table, assume that it Just Works, and pass it off to drm_ioctl.
This make the fence related ioctls work on 64-bit PowerPC.
Jesse Barnes [Thu, 22 Nov 2007 04:14:14 +0000 (14:14 +1000)]
i915: add suspend/resume support
Add suspend/resume support to the i915 driver. Moves some of the
initialization into the driver load routine, and fixes up places where we
assumed no dev_private existed in some of the cleanup paths. This allows
us to suspend/resume properly even if X isn't running.
Jesse Barnes [Thu, 22 Nov 2007 04:02:38 +0000 (14:02 +1000)]
drm: update DRM sysfs support
Make DRM devices use real Linux devices instead of class devices, which are
going away. While we're at it, clean up some of the interfaces to take
struct drm_device * or struct device * and use the global drm_class where
needed instead of passing it around.
projects / linux-2.6-omap-h63xx.git / log