linux-2.6
15 years agoBtrfs: don't use spin_is_contended
Chris Mason [Mon, 9 Feb 2009 21:22:03 +0000 (16:22 -0500)] 
Btrfs: don't use spin_is_contended

Btrfs was using spin_is_contended to see if it should drop locks before
doing extent allocations during btrfs_search_slot.  The idea was to avoid
expensive searches in the tree unless the lock was actually contended.

But, spin_is_contended is specific to the ticket spinlocks on x86, so this
is causing compile errors everywhere else.

In practice, the contention could easily appear some time after we started
doing the extent allocation, and it makes more sense to always drop the lock
instead.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Make sure dir is non-null before doing S_ISGID checks
Chris Mason [Fri, 6 Feb 2009 16:35:57 +0000 (11:35 -0500)] 
Btrfs: Make sure dir is non-null before doing S_ISGID checks

The S_ISGID check in btrfs_new_inode caused an oops during subvol creation
because sometimes the dir is null.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Fix memory leak in cache_drop_leaf_ref
Chris Mason [Thu, 5 Feb 2009 14:08:14 +0000 (09:08 -0500)] 
Btrfs: Fix memory leak in cache_drop_leaf_ref

The code wasn't doing a kfree on the sorted array

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: don't return congestion in write_cache_pages as often
Chris Mason [Wed, 4 Feb 2009 14:33:00 +0000 (09:33 -0500)] 
Btrfs: don't return congestion in write_cache_pages as often

On fast devices that go from congested to uncongested very quickly, pdflush
is waiting too often in congestion_wait, and the FS is backing off to
easily in write_cache_pages.

For now, fix this on the btrfs side by only checking congestion after
some bios have already gone down.  Longer term a real fix is needed
for pdflush, but that is a larger project.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Only prep for btree deletion balances when nodes are mostly empty
Chris Mason [Wed, 4 Feb 2009 14:12:46 +0000 (09:12 -0500)] 
Btrfs: Only prep for btree deletion balances when nodes are mostly empty

Whenever an item deletion is done, we need to balance all the nodes
in the tree to make sure we don't end up with an empty node if a pointer
is deleted.  This balance prep happens from the root of the tree down
so we can drop our locks as we go.

reada_for_balance was triggering read-ahead on neighboring nodes even
when no balancing was required.  This adds an extra check to avoid
calling balance_level() and avoid reada_for_balance() when a balance
won't be required.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: fix btrfs_unlock_up_safe to walk the entire path
Chris Mason [Wed, 4 Feb 2009 14:31:42 +0000 (09:31 -0500)] 
Btrfs: fix btrfs_unlock_up_safe to walk the entire path

btrfs_unlock_up_safe would break out at the first NULL node entry or
unlocked node it found in the path.

Some of the callers have missing nodes at the lower levels of the path, so this
commit fixes things to check all the nodes in the path before returning.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: change btrfs_del_leaf to drop locks earlier
Chris Mason [Wed, 4 Feb 2009 14:31:28 +0000 (09:31 -0500)] 
Btrfs: change btrfs_del_leaf to drop locks earlier

btrfs_del_leaf does two things.  First it removes the pointer in the
parent, and then it frees the block that has the leaf.  It has the
parent node locked for both operations.

But, it only needs the parent locked while it is deleting the pointer.
After that it can safely free the block without the parent locked.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Change btrfs_truncate_inode_items to stop when it hits the inode
Chris Mason [Wed, 4 Feb 2009 14:30:58 +0000 (09:30 -0500)] 
Btrfs: Change btrfs_truncate_inode_items to stop when it hits the inode

btrfs_truncate_inode_items is setup to stop doing btree searches when
it has finished removing the items for the inode.  It used to detect the
end of the inode by looking for an objectid that didn't match the
one we were searching for.

But, this would result in an extra search through the btree, which
adds extra balancing and cow costs to the operation.

This commit adds a check to see if we found the inode item, which means
we can stop searching early.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Don't try to compress pages past i_size
Chris Mason [Wed, 4 Feb 2009 14:31:06 +0000 (09:31 -0500)] 
Btrfs: Don't try to compress pages past i_size

The compression code had some checks to make sure we were only
compressing bytes inside of i_size, but it wasn't catching every
case.  To make things worse, some incorrect math about the number
of bytes remaining would make it try to compress more pages than the
file really had.

The fix used here is to fall back to the non-compression code in this
case, which does all the proper cleanup of delalloc and other accounting.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: join the transaction in __btrfs_setxattr
Josef Bacik [Wed, 4 Feb 2009 14:18:33 +0000 (09:18 -0500)] 
Btrfs: join the transaction in __btrfs_setxattr

With selinux on we end up calling __btrfs_setxattr when we create an inode,
which calls btrfs_start_transaction().  The problem is we've already called
that in btrfs_new_inode, and in btrfs_start_transaction we end up doing a
wait_current_trans().  If btrfs-transaction has started committing it will wait
for all handles to finish, while the other process is waiting for the
transaction to commit.  This is fixed by using btrfs_join_transaction, which
won't wait for the transaction to commit.  Thanks,

Signed-off-by: Josef Bacik <jbacik@redhat.com>
15 years agoBtrfs: Handle SGID bit when creating inodes
Chris Ball [Wed, 4 Feb 2009 14:29:54 +0000 (09:29 -0500)] 
Btrfs: Handle SGID bit when creating inodes
Before this patch, new files/dirs would ignore the SGID bit on their
parent directory and always be owned by the creating user's uid/gid.

Signed-off-by: Chris Ball <cjb@laptop.org>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Make btrfs_drop_snapshot work in larger and more efficient chunks
Chris Mason [Wed, 4 Feb 2009 14:27:02 +0000 (09:27 -0500)] 
Btrfs: Make btrfs_drop_snapshot work in larger and more efficient chunks

Every transaction in btrfs creates a new snapshot, and then schedules the
snapshot from the last transaction for deletion.  Snapshot deletion
works by walking down the btree and dropping the reference counts
on each btree block during the walk.

If if a given leaf or node has a reference count greater than one,
the reference count is decremented and the subtree pointed to by that
node is ignored.

If the reference count is one, walking continues down into that node
or leaf, and the references of everything it points to are decremented.

The old code would try to work in small pieces, walking down the tree
until it found the lowest leaf or node to free and then returning.  This
was very friendly to the rest of the FS because it didn't have a huge
impact on other operations.

But it wouldn't always keep up with the rate that new commits added new
snapshots for deletion, and it wasn't very optimal for the extent
allocation tree because it wasn't finding leaves that were close together
on disk and processing them at the same time.

This changes things to walk down to a level 1 node and then process it
in bulk.  All the leaf pointers are sorted and the leaves are dropped
in order based on their extent number.

The extent allocation tree and commit code are now fast enough for
this kind of bulk processing to work without slowing the rest of the FS
down.  Overall it does less IO and is better able to keep up with
snapshot deletions under high load.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: Change btree locking to use explicit blocking points
Chris Mason [Wed, 4 Feb 2009 14:25:08 +0000 (09:25 -0500)] 
Btrfs: Change btree locking to use explicit blocking points

Most of the btrfs metadata operations can be protected by a spinlock,
but some operations still need to schedule.

So far, btrfs has been using a mutex along with a trylock loop,
most of the time it is able to avoid going for the full mutex, so
the trylock loop is a big performance gain.

This commit is step one for getting rid of the blocking locks entirely.
btrfs_tree_lock takes a spinlock, and the code explicitly switches
to a blocking lock when it starts an operation that can schedule.

We'll be able get rid of the blocking locks in smaller pieces over time.
Tracing allows us to find the most common cause of blocking, so we
can start with the hot spots first.

The basic idea is:

btrfs_tree_lock() returns with the spin lock held

btrfs_set_lock_blocking() sets the EXTENT_BUFFER_BLOCKING bit in
the extent buffer flags, and then drops the spin lock.  The buffer is
still considered locked by all of the btrfs code.

If btrfs_tree_lock gets the spinlock but finds the blocking bit set, it drops
the spin lock and waits on a wait queue for the blocking bit to go away.

Much of the code that needs to set the blocking bit finishes without actually
blocking a good percentage of the time.  So, an adaptive spin is still
used against the blocking bit to avoid very high context switch rates.

btrfs_clear_lock_blocking() clears the blocking bit and returns
with the spinlock held again.

btrfs_tree_unlock() can be called on either blocking or spinning locks,
it does the right thing based on the blocking bit.

ctree.c has a helper function to set/clear all the locked buffers in a
path as blocking.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: hash_lock is no longer needed
Chris Mason [Wed, 4 Feb 2009 14:24:25 +0000 (09:24 -0500)] 
Btrfs: hash_lock is no longer needed

Before metadata is written to disk, it is updated to reflect that writeout
has begun.  Once this update is done, the block must be cow'd before it
can be modified again.

This update was originally synchronized by using a per-fs spinlock.  Today
the buffers for the metadata blocks are locked before writeout begins,
and everyone that tests the flag has the buffer locked as well.

So, the per-fs spinlock (called hash_lock for no good reason) is no
longer required.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: disable leak debugging checks in extent_io.c
Chris Mason [Wed, 4 Feb 2009 14:24:05 +0000 (09:24 -0500)] 
Btrfs: disable leak debugging checks in extent_io.c

extent_io.c has debugging code to report and free leaked extent_state
and extent_buffer objects at rmmod time.  This helps track down
leaks and it saves you from rebooting just to properly remove the
kmem_cache object.

But, the code runs under a fairly expensive spinlock and the checks to
see if it is currently enabled are not entirely consistent.  Some use
#ifdef and some #if.

This changes everything to #if and disables the leak checking.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: sort references by byte number during btrfs_inc_ref
Chris Mason [Wed, 4 Feb 2009 14:23:45 +0000 (09:23 -0500)] 
Btrfs: sort references by byte number during btrfs_inc_ref

When a block goes through cow, we update the reference counts of
everything that block points to.  The internal pointers of the block
can be in just about any order, and it is likely to have clusters of
things that are close together and clusters of things that are not.

To help reduce the seeks that come with updating all of these reference
counts, sort them by byte number before actual updates are done.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: async threads should try harder to find work
Chris Mason [Wed, 4 Feb 2009 14:23:24 +0000 (09:23 -0500)] 
Btrfs: async threads should try harder to find work

Tracing shows the delay between when an async thread goes to sleep
and when more work is added is often very short.  This commit adds
a little bit of delay and extra checking to the code right before
we schedule out.

It allows more work to be added to the worker
without requiring notifications from other procs.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: selinux support
Jim Owens [Wed, 4 Feb 2009 14:29:13 +0000 (09:29 -0500)] 
Btrfs: selinux support
Add call to LSM security initialization and save
resulting security xattr for new inodes.

Add xattr support to symlink inode ops.

Set inode->i_op for existing special files.

Signed-off-by: jim owens <jowens@hp.com>
15 years agoBtrfs: make btrfs acls selectable
Christian Hesse [Wed, 4 Feb 2009 14:28:28 +0000 (09:28 -0500)] 
Btrfs: make btrfs acls selectable
This patch adds a menu entry to kconfig to enable acls for btrfs.
This allows you to enable FS_POSIX_ACL at kernel compile time.

(updated by Jeff Mahoney to make the changes in fs/btrfs/Kconfig instead)

Signed-off-by: Christian Hesse <mail@earthworm.de>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
15 years agoBtrfs: Catch missed bios in the async bio submission thread
Chris Mason [Wed, 4 Feb 2009 14:19:41 +0000 (09:19 -0500)] 
Btrfs: Catch missed bios in the async bio submission thread

The async bio submission thread was missing some bios that were
added after it had decided there was no work left to do.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoBtrfs: fix readdir on 32 bit machines
Chris Mason [Wed, 28 Jan 2009 20:34:27 +0000 (15:34 -0500)] 
Btrfs: fix readdir on 32 bit machines

After btrfs_readdir has gone through all the directory items, it
sets the directory f_pos to the largest possible int.  This way
applications that mix readdir with creating new files don't
end up in an endless loop finding the new directory items as they go.

It was a workaround for a bug in git, but the assumption was that if git
could make this looping mistake than it would be a common problem.

The largest possible int chosen was INT_LIMIT(typeof(file->f_pos),
and it is possible for that to be a larger number than 32 bit glibc
expects to come out of readdir.

This patches switches that to INT_LIMIT(off_t), which should keep
applications happy on 32 and 64 bit machines.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
15 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs...
Chris Mason [Thu, 29 Jan 2009 01:29:43 +0000 (20:29 -0500)] 
Merge branch 'master' of git://git./linux/kernel/git/mason/btrfs-unstable

Fix fs/btrfs/super.c conflict around #includes

15 years agoLinux 2.6.29-rc3 v2.6.29-rc3
Linus Torvalds [Wed, 28 Jan 2009 18:49:30 +0000 (10:49 -0800)] 
Linux 2.6.29-rc3

15 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Wed, 28 Jan 2009 17:01:42 +0000 (09:01 -0800)] 
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  powerpc/mm: Fix handling of _PAGE_COHERENT in BAT setup code
  powerpc/pseries: Correct VIO bus accounting problem in CMO env.
  powerpc: More printing warning fixes for the l64 to ll64 conversion
  powerpc: Remove arch/ppc cruft from Kconfig
  powerpc: Printing fix for l64 to ll64 conversion: phyp_dump.c
  powerpc/embedded6xx: Update defconfigs
  powerpc/8xx: Update defconfigs
  powerpc/86xx: Update defconfigs
  powerpc/83xx: Update defconfigs
  powerpc/85xx: Update defconfigs
  powerpc/mpc8313erdb: fix kernel panic because mdio device is not probed
  powerpc/4xx: Update multi-board PowerPC 4xx defconfigs
  powerpc/44x: Update PowerPC 44x defconfigs
  powerpc/40x: Update PowerPC 40x defconfigs
  powerpc/85xx: Fix typo in mpc8572ds dts
  powerpc/44x: Warp patches for the new NDFC driver
  powerpc/4xx: DTS: Add Add'l SDRAM0 Compatible and Interrupt Info

15 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gerg/m68knommu
Linus Torvalds [Wed, 28 Jan 2009 16:41:57 +0000 (08:41 -0800)] 
Merge branch 'for-linus' of git://git./linux/kernel/git/gerg/m68knommu

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gerg/m68knommu:
  m68knommu: fix 5329 ColdFire periphal addressing
  uclinux: add process name to allocation error message
  m68knommu: correct the mii calculations for 532x ColdFire FEC
  m68knommu: add ColdFire M532x to the FEC configuration options
  m68knommu: fix syscall restarting
  m68knommu: remove the obsolete and long unused comempci chip support
  m68knommu: remove the no longer used PCI support option
  m68knommu: remove obsolete and unused eLIA board
  m68knommu: set NO_DMA
  m68knommu: fix cache flushing for the 527x ColdFire processors
  m68knommu: fix ColdFire 5272 serial baud rates in mcf.c
  m68knommu: use one exist from execption

15 years agodmi: Fix build breakage
Kumar Gala [Wed, 28 Jan 2009 06:07:20 +0000 (00:07 -0600)] 
dmi: Fix build breakage

Commit d7b1956fed33d30c4815e848fd7a143722916868 ("DMI: Introduce
dmi_first_match to make the interface more flexible") introduced compile
errors like the following when !CONFIG_DMI

    drivers/ata/sata_sil.c: In function 'sil_broken_system_poweroff':
    drivers/ata/sata_sil.c:713: error: implicit declaration of function 'dmi_first_match'
    drivers/ata/sata_sil.c:713: warning: initialization makes pointer from integer without a cast

We just need a dummy version of dmi_first_match() to fix this all up.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agodrm: Rip out the racy, unused vblank signal code.
Eric Anholt [Wed, 28 Jan 2009 05:19:41 +0000 (21:19 -0800)] 
drm: Rip out the racy, unused vblank signal code.

Schedule a vblank signal, kill the process, and we'll go walking over freed
memory.  Given that no open-source userland exists using this, nor have I
ever heard of a consumer, just let this code die.

Signed-off-by: Eric Anholt <eric@anholt.net>
Requested-by: Linus Torvalds <torvalds@linux-foundation.org>
Acked-by: Dave Airlie <airlied@linux.ie>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopowerpc/mm: Fix handling of _PAGE_COHERENT in BAT setup code
Gerhard Pircher [Fri, 23 Jan 2009 06:51:28 +0000 (06:51 +0000)] 
powerpc/mm: Fix handling of _PAGE_COHERENT in BAT setup code

_PAGE_COHERENT is now always set in _PAGE_RAM resp. PAGE_KERNEL.
Thus it has to be masked out, if the BAT mapping should be non
cacheable or CPU_FTR_NEED_COHERENT is not set.

This will work on normal SMP setups because we force-set
CPU_FTR_NEED_COHERENT as part of CPU_FTR_COMMON on SMP.

Signed-off-by: Gerhard Pircher <gerhard_pircher@gmx.net>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
15 years agopowerpc/pseries: Correct VIO bus accounting problem in CMO env.
Robert Jennings [Thu, 22 Jan 2009 09:40:00 +0000 (09:40 +0000)] 
powerpc/pseries: Correct VIO bus accounting problem in CMO env.

In the VIO bus code the wrappers for dma alloc_coherent and free_coherent
calls are rounding to IOMMU_PAGE_SIZE.  Taking a look at the underlying
calls, the actual mapping is promoted to PAGE_SIZE.  Changing the
rounding in these two functions fixes under-reporting the entitlement
used by the system.  Without this change, the system could run out of
entitlement before it believes it has and incur mapping failures at the
firmware level.

Also in the VIO bus code, the wrapper for dma map_sg is not exiting in
an error path where it should.  Rather than fall through to code for the
success case, this patch adds the return that is needed in the error path.

Signed-off-by: Robert Jennings <rcj@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
15 years agopowerpc: More printing warning fixes for the l64 to ll64 conversion
Stephen Rothwell [Wed, 21 Jan 2009 13:16:28 +0000 (13:16 +0000)] 
powerpc: More printing warning fixes for the l64 to ll64 conversion

These are all powerpc specific drivers.

res.start in fsl_elbc_nand.c needs to be cast since it may be either 32
or 64 bit.  Thanks to Scott Wood for noticing.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Arnd Bergmann <arnd@arndb.de> call_edac bits in particular
Acked-by: Olof Johansson <olof@lixom.net> pasemi_nand peices
Acked-by: Scott Wood <scottwood@freescale.com> fsl_elbc fixes
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
15 years agopowerpc: Remove arch/ppc cruft from Kconfig
Josh Boyer [Tue, 20 Jan 2009 05:16:36 +0000 (05:16 +0000)] 
powerpc: Remove arch/ppc cruft from Kconfig

Remove some leftover cruft from the arch/ppc days

Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
15 years agopowerpc: Printing fix for l64 to ll64 conversion: phyp_dump.c
Stephen Rothwell [Sun, 18 Jan 2009 23:33:38 +0000 (23:33 +0000)] 
powerpc: Printing fix for l64 to ll64 conversion: phyp_dump.c

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
15 years agoMerge commit 'jwb/jwb-merge' into merge
Benjamin Herrenschmidt [Wed, 28 Jan 2009 06:15:34 +0000 (17:15 +1100)] 
Merge commit 'jwb/jwb-merge' into merge

Manual merge of:
arch/powerpc/configs/44x/warp_defconfig

15 years agoMerge commit 'kumar/kumar-merge' into merge
Benjamin Herrenschmidt [Wed, 28 Jan 2009 06:14:54 +0000 (17:14 +1100)] 
Merge commit 'kumar/kumar-merge' into merge

15 years agosmackfs load append mode fix
Casey Schaufler [Wed, 28 Jan 2009 03:56:30 +0000 (19:56 -0800)] 
smackfs load append mode fix

Given just how hard it is to find the code that uses MAY_APPEND
it's probably not a big surprise that this went unnoticed for so
long. The Smack rules loading code is incorrectly setting the
MAY_READ bit when MAY_APPEND is requested.

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Reviewed-by: James Morris <jmorris@namei.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6
Linus Torvalds [Wed, 28 Jan 2009 00:08:04 +0000 (16:08 -0800)] 
Merge git://git./linux/kernel/git/gregkh/usb-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6: (36 commits)
  USB: Driver for Freescale QUICC Engine USB Host Controller
  USB: option: add QUANTA HSDPA Data Card device ids
  USB: storage: Add another unusual_dev for off-by-one bug
  USB: unusual_dev: usb-storage needs to ignore a device
  USB: GADGET: fix !x & y
  USB: new id for ti_usb_3410_5052 driver
  USB: cdc-acm: Add another conexant modem to the quirks
  USB: 'option' driver - onda device MT503HS has wrong id
  USB: Remove ZTE modem from unusual_devices
  USB: storage: support of Dane-Elec MediaTouch USB device
  USB: usbmon: Implement compat_ioctl
  USB: add kernel-doc for wusb_dev in struct usb_device
  USB: ftdi_sio driver support of bar code scanner from Diebold
  USB: ftdi_sio: added Alti-2 VID and Neptune 3 PID
  USB: cp2101 device
  USB: usblp.c: add USBLP_QUIRK_BIDIR to Brother HL-1440
  USB: remove vernier labpro from ldusb
  USB: CDC-ACM quirk for MTK GPS
  USB: cdc-acm: support some gps data loggers
  USB: composite: Fix bug: low byte of w_index is the usb interface number not the whole 2 bytes of w_index
  ...

15 years agoUSB: Driver for Freescale QUICC Engine USB Host Controller
Anton Vorontsov [Sat, 10 Jan 2009 02:03:21 +0000 (05:03 +0300)] 
USB: Driver for Freescale QUICC Engine USB Host Controller

This patch adds support for the FHCI USB controller, as found
in the Freescale MPC836x and MPC832x processors. It can support
Full or Low speed modes.

Quite a lot the hardware is doing by itself (SOF generation, CRC
generation and checking), though scheduling and retransmission is on
software's shoulders.

This controller does not integrate the root hub, so this driver also
fakes one-port hub. External hub is required to support more than
one device.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: option: add QUANTA HSDPA Data Card device ids
Alex.Cheng@quantatw.com [Thu, 22 Jan 2009 08:01:57 +0000 (16:01 +0800)] 
USB: option: add QUANTA HSDPA Data Card device ids

This patch adds the support  for the QUANTA Q101 series HSDPA Data Card.
With the vendor and product IDs are set properly,
the data card can be detected and works fine.

Signed-off-by: Alex Cheng <alex.cheng@quantatw.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: storage: Add another unusual_dev for off-by-one bug
Phil Dibowitz [Tue, 20 Jan 2009 22:42:52 +0000 (23:42 +0100)] 
USB: storage: Add another unusual_dev for off-by-one bug

Argosy has released another device with the off-by-one sector. This is a
harddrive with an internal cardreader which is affected.

Based on a patch written by Martijn Hijdra <martijn.hijdra@gmail.com>

Signed-off-by: Phil Dibowitz <phil@ipom.com>
Cc: Martijn Hijdra <martijn.hijdra@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: unusual_dev: usb-storage needs to ignore a device
Phil Dibowitz [Tue, 20 Jan 2009 22:48:36 +0000 (23:48 +0100)] 
USB: unusual_dev: usb-storage needs to ignore a device

This patch adds an unusual_devs entry for a Sony Ericsson modem. Like many
other modems, we have to ignore the storage device in order to access the
modem.

At this time usb_modeswitch does not work with this device.

Reported-by: The Solutor <thesolutor@gmail.com>.
Signed-off-by: Phil Dibowitz <phil@ipom.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: GADGET: fix !x & y
Roel Kluin [Sat, 17 Jan 2009 15:52:17 +0000 (16:52 +0100)] 
USB: GADGET: fix !x & y

! has a higher precedence than &

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Acked-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: new id for ti_usb_3410_5052 driver
Oliver Neukum [Mon, 12 Jan 2009 12:31:16 +0000 (13:31 +0100)] 
USB: new id for ti_usb_3410_5052 driver

This adds a new device id

Signed-off-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: cdc-acm: Add another conexant modem to the quirks
Alan Cox [Sun, 11 Jan 2009 19:53:10 +0000 (19:53 +0000)] 
USB: cdc-acm: Add another conexant modem to the quirks

Another Conexant, another device with the same quirk

Signed-off-by: Alan Cox <alan@redhat.com>
Acked-by: Oliver Neukum <oliver@neukum.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: 'option' driver - onda device MT503HS has wrong id
Marcel Sebek [Sun, 28 Dec 2008 13:06:50 +0000 (14:06 +0100)] 
USB: 'option' driver - onda device MT503HS has wrong id

While trying to make GSM modem Onda MT503HS working, I found a mismatch
between device id in the driver code (0x0200) and id in the lsusb
output (0x2000).

This patch fixed it for me, but I don't know if the original device id was
also correct and the new ID should be added instead of replacing the
old one.

Signed-off-by: Marcel Sebek <sebek64@post.cz>
Acked-by: Domenico Riccio <domenico.riccio@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: Remove ZTE modem from unusual_devices
Phil Dibowitz [Sun, 11 Jan 2009 17:46:20 +0000 (18:46 +0100)] 
USB: Remove ZTE modem from unusual_devices

The ZTE modem entry causes usb-storage to ignore the device, but for some
versions of the device, usb-storage mode is required to get to modem ode. For
both kinds the tool: http://www.draisberghof.de/usb_modeswitch/ should work.

Note that the various versions of the device have the same ProductId,
VendorId, and bcdDevice number, so we cannot have the entry for some and not
others.

Signed-off-by: Phil Dibowitz <phil@ipom.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: storage: support of Dane-Elec MediaTouch USB device
Jean-Baptiste Onofre [Sat, 20 Dec 2008 19:11:55 +0000 (20:11 +0100)] 
USB: storage: support of Dane-Elec MediaTouch USB device

This adds another unusual_devs.h entry for a device that can't handle more
than 64k reads/writes in a single command.

Signed-off-by: Jean-Baptiste Onofre <jb@nanthrax.net>
Signed-off-by: Phil Dibowitz <phil@ipom.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: usbmon: Implement compat_ioctl
Pete Zaitcev [Sat, 20 Dec 2008 19:56:08 +0000 (12:56 -0700)] 
USB: usbmon: Implement compat_ioctl

Running a 32-bit usbmon(8) on 2.6.28-rc9 produces the following:
ioctl32(usbmon:28563): Unknown cmd fd(3) cmd(400c9206){t:ffffff92;sz:12} arg(ffd3f458) on /dev/usbmon0

It happens because the compatibility mode was implemented for 2.6.18
and not updated for the fsops.compat_ioctl API.

This patch relocates the pieces from under #ifdef CONFIG_COMPAT into
compat_ioctl with no other changes except one new whitespace.

Signed-off-by: Pete Zaitcev <zaitcev@redhat.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: add kernel-doc for wusb_dev in struct usb_device
Inaky Perez-Gonzalez [Thu, 8 Jan 2009 20:52:19 +0000 (12:52 -0800)] 
USB: add kernel-doc for wusb_dev in struct usb_device

Reported by Randy Dunlap from a warning on the v2.6.29 merge window.

Signed-off-by: Inaky Perez-Gonzalez <inaky@linux.intel.com>
Cc: David Vrabel <david.vrabel@csr.com>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: ftdi_sio driver support of bar code scanner from Diebold
Mhayk Whandson [Fri, 9 Jan 2009 10:48:16 +0000 (06:48 -0400)] 
USB: ftdi_sio driver support of bar code scanner from Diebold

Added the product id of bcs(bar code scanner) from Diebold Procomp Brazil.

Signed-off-by: Mhayk Whandson <eu@mhayk.com.br>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: ftdi_sio: added Alti-2 VID and Neptune 3 PID
Robie Basak [Mon, 12 Jan 2009 23:05:59 +0000 (23:05 +0000)] 
USB: ftdi_sio: added Alti-2 VID and Neptune 3 PID

This patch adds the vendor and product ID for the Alti-2 Neptune 3
(http://www.alti-2.com) which uses the FTDI chip.

Signed-off-by: Robie Basak <rb-oss-1@justgohome.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: cp2101 device
Tomasz K [Sun, 4 Jan 2009 11:47:11 +0000 (12:47 +0100)] 
USB: cp2101 device

My girl use modem GSM (EDGE) Commanader 2 on iPlus Polsih provider,

PLEASE add this vendor=0x10C4 and product=0x822B to USB serial driver cp2101.c

From: Tomasz K <eros81@o2.pl>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: usblp.c: add USBLP_QUIRK_BIDIR to Brother HL-1440
Brandon Philips [Mon, 5 Jan 2009 16:30:39 +0000 (08:30 -0800)] 
USB: usblp.c: add USBLP_QUIRK_BIDIR to Brother HL-1440

My Brother HL-1440 would print one document before CUPS would stop
printing with the error "Printer not connected; will retry in 30
seconds...". I traced this down to the CUPS usb backend getting an EIO
out of usblp on the IOCNR_GET_DEVICE_ID IOCTL. Adding the
USBLP_QUIRK_BIDIR fixes the problem but is it the right solution?

output from strace /usr/lib/cups/backend/usb after printing a document
(Note: SNDCTL_DSP_SYNC == IOCNR_GET_DEVICE_ID):

before patch

open("/dev/usb/lp0", O_RDWR|O_EXCL)     = 3
ioctl(3, SNDCTL_DSP_SYNC, 0x7fff2478cef0) = -1 EIO (Input/output error)

after patch

open("/dev/usb/lp0", O_RDWR|O_EXCL)     = 3
ioctl(3, SNDCTL_DSP_SYNC, 0x7fffb8d474c0) = 0

Possibly related bug: https://bugs.launchpad.net/ubuntu/+source/cupsys/+bug/35638

Signed-off-by: Brandon Philips <bphilips@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: remove vernier labpro from ldusb
sware [Wed, 7 Jan 2009 23:35:55 +0000 (15:35 -0800)] 
USB: remove vernier labpro from ldusb

Labpro device is in both ldusb and vstusb device tables.
Should only be a vstusb device.

Signed-off-by: stephen ware <stephen.ware@eqware.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: CDC-ACM quirk for MTK GPS
Andrew Lunn [Tue, 23 Dec 2008 16:31:23 +0000 (17:31 +0100)] 
USB: CDC-ACM quirk for MTK GPS

This patch adds a device quirk for a MediaTek Inc GPS chipset. The
device implements USB CDC ACM, but is missing the union descriptor, so
the ACM class driver fails to probe the device.

I've tested this patch with an iBlue A+ GPS which uses this chipset
and using kernel 2.6.28-rc9.

Signed-off-by: Andrew Lunn, <andrew@lunn.ch>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: cdc-acm: support some gps data loggers
James A. Treacy [Sun, 25 Jan 2009 04:37:43 +0000 (23:37 -0500)] 
USB: cdc-acm: support some gps data loggers

Below is a patch which allows a number of GPS loggers to work
under linux. It is known to support the i-Blue 747 (all models),
i-Blue 757, Qstarz BT-Q1000, i.Trek Z1, Konet BGL-32, and the Holux
M-241.

From: James A. Treacy <treacy@debian.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: composite: Fix bug: low byte of w_index is the usb interface number not the...
Bryan Wu [Wed, 7 Jan 2009 16:21:19 +0000 (00:21 +0800)] 
USB: composite: Fix bug: low byte of w_index is the usb interface number not the whole 2 bytes of w_index

In some usb gadget driver, for example usb audio class device, the high
byte of w_index is the entity id and low byte is the interface number.
If we use the 2 bytes of w_index as the array number, we will get a
wrong pointer or NULL pointer.

This patch fixes this issue.

Signed-off-by: Bryan Wu <cooloney@kernel.org>
Acked-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: composite: Fix bug: should test set_alt function pointer before use it
Bryan Wu [Wed, 7 Jan 2009 16:21:18 +0000 (00:21 +0800)] 
USB: composite: Fix bug: should test set_alt function pointer before use it

Signed-off-by: Bryan Wu <cooloney@kernel.org>
Acked-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb: Kconfig fix
David Brownell [Mon, 26 Jan 2009 10:05:43 +0000 (02:05 -0800)] 
USB: musb: Kconfig fix

The Blackfin MUSB Kconfig text didn't properly parenthesise its
dependencies.  This was visible in non-Blackfin configs by the
way the user interfaces lost track of dependencies, when doing
a bunch of test builds.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb cppi dma fix
Swaminathan S [Sun, 25 Jan 2009 01:57:37 +0000 (17:57 -0800)] 
USB: musb cppi dma fix

Initializes the actual_len field to 0 before every DMA transaction.

Signed-off-by: Swaminathan S <swami.iyer@ti.com>
Acked-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb cppi bugfixes
Hugo Villeneuve [Sun, 25 Jan 2009 01:57:30 +0000 (17:57 -0800)] 
USB: musb cppi bugfixes

These compilation errors are related to incorrect
debugging macro and variable names and generated the
following errors:

  drivers/usb/musb/cppi_dma.c:437:5: warning: "MUSB_DEBUG" is not defined
  drivers/usb/musb/cppi_dma.c: In function 'cppi_next_rx_segment':
  drivers/usb/musb/cppi_dma.c:884: error: 'debug' undeclared (first use in this function)

Signed-off-by: Hugo Villeneuve <hugo@hugovil.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb uses endpoint functions
Julia Lawall [Sun, 25 Jan 2009 01:57:24 +0000 (17:57 -0800)] 
USB: musb uses endpoint functions

This set of patches introduces calls to the following set of functions:

usb_endpoint_dir_in(epd)
usb_endpoint_dir_out(epd)
usb_endpoint_is_bulk_in(epd)
usb_endpoint_is_bulk_out(epd)
usb_endpoint_is_int_in(epd)
usb_endpoint_is_int_out(epd)
usb_endpoint_num(epd)
usb_endpoint_type(epd)
usb_endpoint_xfer_bulk(epd)
usb_endpoint_xfer_control(epd)
usb_endpoint_xfer_int(epd)
usb_endpoint_xfer_isoc(epd)

In some cases, introducing one of these functions is not possible, and it
just replaces an explicit integer value by one of the following constants:

USB_ENDPOINT_XFER_BULK
USB_ENDPOINT_XFER_CONTROL
USB_ENDPOINT_XFER_INT
USB_ENDPOINT_XFER_ISOC

An extract of the semantic patch that makes these changes is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@r1@ struct usb_endpoint_descriptor *epd; @@

- ((epd->bmAttributes & \(USB_ENDPOINT_XFERTYPE_MASK\|3\)) ==
- \(USB_ENDPOINT_XFER_CONTROL\|0\))
+ usb_endpoint_xfer_control(epd)

@r5@ struct usb_endpoint_descriptor *epd; @@

- ((epd->bEndpointAddress & \(USB_ENDPOINT_DIR_MASK\|0x80\)) ==
-  \(USB_DIR_IN\|0x80\))
+ usb_endpoint_dir_in(epd)

@inc@
@@

#include <linux/usb.h>

@depends on !inc && (r1||r5)@
@@

+ #include <linux/usb.h>
  #include <linux/usb/...>
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Acked-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb: tusb6010 buildfix
Kalle Valo [Sun, 25 Jan 2009 01:57:15 +0000 (17:57 -0800)] 
USB: musb: tusb6010 buildfix

drivers/usb/musb/tusb6010_omap.c:18:26: error: asm/arch/dma.h:
  No such file or directory
drivers/usb/musb/tusb6010_omap.c:19:26: error: asm/arch/mux.h:
  No such file or directory

Signed-off-by: Kalle Valo <kalle.valo@nokia.com>
Acked-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb free_irq bugfix
Ajay Kumar Gupta [Sun, 25 Jan 2009 01:56:39 +0000 (17:56 -0800)] 
USB: musb free_irq bugfix

Fixes insert module failure as free_irq() was not
done in previous rmmod.

Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
Acked-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb_hdrc: another davinci buildfix (otg related)
David Brownell [Sun, 25 Jan 2009 01:56:25 +0000 (17:56 -0800)] 
USB: musb_hdrc: another davinci buildfix (otg related)

The DaVinci code had an implementation of the OTG transceiver glue
too; make it use the new-standard one.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Acked-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: musb davinci buildfix
David Brownell [Sun, 25 Jan 2009 01:56:17 +0000 (17:56 -0800)] 
USB: musb davinci buildfix

Trying once more to get this merged.  The original was submitted
for 2.6.27-rc2 or so, and never got correctly merged.  Neither
were any of the numerous subsequent resends.  Sigh.

  CC      drivers/usb/musb/davinci.o
drivers/usb/musb/davinci.c:35:32: error: mach/arch/hardware.h: No such file or directory
drivers/usb/musb/davinci.c:36:30: error: mach/arch/memory.h: No such file or directory
drivers/usb/musb/davinci.c:37:28: error: mach/arch/gpio.h: No such file or directory
drivers/usb/musb/davinci.c:373: error: redefinition of 'musb_platform_set_mode'
drivers/usb/musb/davinci.c:368: error: previous definition of 'musb_platform_set_mode' was here

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Acked-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: omap1 ohci buildfix (otg related)
David Brownell [Sun, 25 Jan 2009 01:55:57 +0000 (17:55 -0800)] 
USB: omap1 ohci buildfix (otg related)

> > drivers/built-in.o: In function `ohci_omap_init':
 > > hid-quirks.c:(.text+0x6c608): undefined reference to `otg_get_transceiver'
 > > drivers/built-in.o: In function `omap_udc_probe':
 > > hid-quirks.c:(.init.text+0x34c0): undefined reference to `otg_get_transceiver'
 > > hid-quirks.c:(.init.text+0x3d40): undefined reference to `otg_put_transceiver'

Reported-by: Russell King <linux@arm.linux.org.uk>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: fix char-device disconnect handling
Alan Stern [Tue, 13 Jan 2009 16:33:42 +0000 (11:33 -0500)] 
USB: fix char-device disconnect handling

This patch (as1198) fixes a conceptual bug: Somewhere along the line
we managed to confuse USB class devices with USB char devices.  As a
result, the code to send a disconnect signal to userspace would not be
built if both CONFIG_USB_DEVICE_CLASS and CONFIG_USB_DEVICEFS were
disabled.

The usb_fs_classdev_common_remove() routine has been renamed to
usbdev_remove() and it is now called whenever any USB device is
removed, not just when a class device is unregistered.  The notifier
registration and unregistration calls are no longer conditionally
compiled.  And since the common removal code will always be called as
part of the char device interface, there's no need to call it again as
part of the usbfs interface; thus the invocation of
usb_fs_classdev_common_remove() has been taken out of
usbfs_remove_device().

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-by: Alon Bar-Lev <alon.barlev@gmail.com>
Tested-by: Alon Bar-Lev <alon.barlev@gmail.com>
Cc: stable <stable@kernel.org>
15 years agoUSB: Fix suspend-resume of PCI USB controllers
Rafael J. Wysocki [Tue, 20 Jan 2009 00:26:56 +0000 (01:26 +0100)] 
USB: Fix suspend-resume of PCI USB controllers

Commit a0d4922da2e4ccb0973095d8d29f36f6b1b5f703
(USB: fix up suspend and resume for PCI host controllers) attempted
to fix the suspend-resume of PCI USB controllers, but unfortunately
it did that incorrectly and interrupts are left enabled by the USB
controllers' ->suspend_late() callback as a result.  This leads to
serious problems during suspend which are very difficult to debug.

Fix the issue by removing the ->suspend_late() callback of PCI
USB controllers and moving the code from there to the ->suspend()
callback executed with interrupts enabled.  Additionally, make
the ->resume() callback of PCI USB controllers execute
pci_enable_wake(dev, PCI_D0, false) to disable wake-up from the
full power state (PCI_D0).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Tested-by: Andrey Borzenkov <arvidjaar@mail.ru>
Tested-by: "Jeff Chua" <jeff.chua.linux@gmail.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: "Zdenek Kabelac" <zdenek.kabelac@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: don't enable wakeup by default for PCI host controllers
Alan Stern [Tue, 13 Jan 2009 16:35:54 +0000 (11:35 -0500)] 
USB: don't enable wakeup by default for PCI host controllers

This patch (as1199) changes the initial wakeup settings for PCI USB
host controllers.  The controllers are marked as capable of waking the
system, but wakeup is not enabled by default.

It turns out that enabling wakeup for USB host controllers has a lot
of bad consequences.  As the simplest example, if a USB mouse or
keyboard is unplugged immediately after the computer is put to sleep,
the unplug will cause the system to wake back up again!  We are better
off marking them as wakeup-capable and leaving wakeup disabled.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-by: Rafael J. Wysocki <rjw@sisk.pl>
CC: David Brownell <david-b@pacbell.net>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: storage: add unusual devs entry
Oliver Neukum [Wed, 14 Jan 2009 15:17:19 +0000 (16:17 +0100)] 
USB: storage: add unusual devs entry

This adds an unusual devs entry for 2116:0320

Signed-off-by: Oliver Neukum <oneukum@suse.de>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: fix toggle mismatch in disable_endpoint paths
Alan Stern [Thu, 15 Jan 2009 22:03:33 +0000 (17:03 -0500)] 
USB: fix toggle mismatch in disable_endpoint paths

This patch (as1200) finishes some fixes that were left incomplete by
an earlier patch.

Although nobody has addressed this issue in the past, it turns out
that we need to distinguish between two different modes of disabling
and enabling endpoints.  In one mode only the data structures in
usbcore are affected, and in the other mode the host controller and
device hardware states are affected as well.

The earlier patch added an extra argument to the routines in the
enable_endpoint pathways to reflect this difference.  This patch adds
corresponding arguments to the disable_endpoint pathways.  Without
this change, the endpoint toggle state can get out of sync between
the host and the device.  The exact mechanism depends on the details
of the host controller (whether or not it stores its own copy of the
toggle values).

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-by: Dan Streetman <ddstreet@ieee.org>
Tested-by: Dan Streetman <ddstreet@ieee.org>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoUSB: cp2101: add fasttrax GPS evaluation kit vendor/product ID
Wolfgang Glas [Thu, 15 Jan 2009 22:11:53 +0000 (23:11 +0100)] 
USB: cp2101: add fasttrax GPS evaluation kit vendor/product ID

This adds the vendor/product ID of the fasttrax GPS evaluation kit from:
http://www.fastraxgps.com/products/evaluationtools/evaluationkit/

to the cp2101 module since this device is actually equipped with a
CP210x USB to serial bridge.

The vendor/product ID is: 0x10c4/0x826b.

Signed-off-by: Wolfgang Glas <wolfgang.glas@ev-i.at>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
15 years agoMerge branch 'for-2.6.29' of git://linux-nfs.org/~bfields/linux
Linus Torvalds [Wed, 28 Jan 2009 00:07:44 +0000 (16:07 -0800)] 
Merge branch 'for-2.6.29' of git://linux-nfs.org/~bfields/linux

* 'for-2.6.29' of git://linux-nfs.org/~bfields/linux:
  nfsd: only set file_lock.fl_lmops in nfsd4_lockt if a stateowner is found
  nfsd: fix cred leak on every rpc
  nfsd: fix null dereference on error path
  nfs: note that CONFIG_SUNRPC_XPRT_RDMA turns on server side support too
  update port number in NFS/RDMA documentation

15 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6
Linus Torvalds [Wed, 28 Jan 2009 00:07:24 +0000 (16:07 -0800)] 
Merge git://git./linux/kernel/git/jejb/scsi-rc-fixes-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6:
  [SCSI] ibmvfc: Fix DMA mapping leak on memory allocation failure
  [SCSI] qla2xxx: Update version number to 8.03.00-k2.
  [SCSI] qla2xxx: Add checks for a valid fcport in dev-loss-tmo/terminate_rport_io callbacks.
  [SCSI] qla2xxx: Correct regression in DMA-mask setting prior to allocations.
  [SCSI] qla2xxx: Correct descriptions in flash manipulation routines.
  [SCSI] qla2xxx: Correct regression in EH abort handling.
  [SCSI] qla2xxx: Correct endianness issue during flash manipulation.
  [SCSI] qla2xxx: Correct MSI-X vector allocation for single queue mode.
  [SCSI] qla2xxx: Modify firmware-load order precedence for ISP81XX parts.
  [SCSI] qla2xxx: Always serialize mailbox command execution.
  [SCSI] qla2xxx: Ensure RISC-interrupt-enabled consistency for IS_NOPOLLING_TYPE() ISPs.
  [SCSI] qla2xxx: Simplify sector-mask calculation in preparation for larger flash parts.
  [SCSI] qla2xxx: Fix memory leak in error path
  [SCSI] qla4xxx: do not reuse session when connecting to different target port
  [SCSI] libiscsi: fix iscsi pool leak

15 years agocpumask fallout: Initialize irq_default_affinity earlier
David Daney [Tue, 27 Jan 2009 17:53:22 +0000 (09:53 -0800)] 
cpumask fallout: Initialize irq_default_affinity earlier

Move the initialization of irq_default_affinity to early_irq_init as
core_initcall is too late.

irq_default_affinity can be used in init_IRQ and potentially timer and
SMP init as well.  All of these happen before core_initcall.  Moving
the initialization to early_irq_init ensures that it is initialized
before it is used.

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Acked-by: Mike Travis <travis@sgi.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoMake irq_*_affinity depend on CONFIG_GENERIC_HARDIRQS too.
David Daney [Tue, 27 Jan 2009 17:53:21 +0000 (09:53 -0800)] 
Make irq_*_affinity depend on CONFIG_GENERIC_HARDIRQS too.

In interrupt.h these functions are declared only if
CONFIG_GENERIC_HARDIRQS is set.  We should define them under identical
conditions.

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoFix longstanding "error: storage size of '__mod_dmi_device_table' isn't known"
Alexey Dobriyan [Wed, 21 Jan 2009 22:58:36 +0000 (01:58 +0300)] 
Fix longstanding "error: storage size of '__mod_dmi_device_table' isn't known"

gcc 3.4.6 doesn't like MODULE_DEVICE_TABLE(dmi, x) expansion enough to
error out.  Shut it up in a most simple way.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agonfsd: only set file_lock.fl_lmops in nfsd4_lockt if a stateowner is found
Jeff Layton [Thu, 22 Jan 2009 19:16:04 +0000 (14:16 -0500)] 
nfsd: only set file_lock.fl_lmops in nfsd4_lockt if a stateowner is found

nfsd4_lockt does a search for a lockstateowner when building the lock
struct to test. If one is found, it'll set fl_owner to it. Regardless of
whether that happens, it'll also set fl_lmops. Given that this lock is
basically a "lightweight" lock that's just used for checking conflicts,
setting fl_lmops is probably not appropriate for it.

This behavior exposed a bug in DLM's GETLK implementation where it
wasn't clearing out the fields in the file_lock before filling in
conflicting lock info. While we were able to fix this in DLM, it
still seems pointless and dangerous to set the fl_lmops this way
when we may have a NULL lockstateowner.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@pig.fieldses.org>
15 years agonfsd: fix cred leak on every rpc
J. Bruce Fields [Wed, 21 Jan 2009 00:34:22 +0000 (19:34 -0500)] 
nfsd: fix cred leak on every rpc

Since override_creds() took its own reference on new, we need to release
our own reference.

(Note the put_cred on the return value puts the *old* value of
current->creds, not the new passed-in value).

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
15 years agonfsd: fix null dereference on error path
J. Bruce Fields [Wed, 21 Jan 2009 00:32:59 +0000 (19:32 -0500)] 
nfsd: fix null dereference on error path

We're forgetting to check the return value from groups_alloc().

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
15 years agonfs: note that CONFIG_SUNRPC_XPRT_RDMA turns on server side support too
J. Bruce Fields [Sun, 2 Nov 2008 21:18:08 +0000 (16:18 -0500)] 
nfs: note that CONFIG_SUNRPC_XPRT_RDMA turns on server side support too

We forgot to update this when adding server-side support.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
15 years agoupdate port number in NFS/RDMA documentation
James Lentini [Thu, 8 Jan 2009 18:13:26 +0000 (13:13 -0500)] 
update port number in NFS/RDMA documentation

Update the NFS/RDMA documentation to use the new port number assigned
by IANA.

Signed-off-by: James Lentini <jlentini@netapp.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
15 years agoMerge branch 'hibern_fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik...
Linus Torvalds [Tue, 27 Jan 2009 15:50:41 +0000 (07:50 -0800)] 
Merge branch 'hibern_fixes' of git://git./linux/kernel/git/jgarzik/libata-dev

* 'hibern_fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  SATA PIIX: Blacklist system that spins off disks during ACPI power off
  SATA Sil: Blacklist system that spins off disks during ACPI power off
  SATA AHCI: Blacklist system that spins off disks during ACPI power off
  SATA: Blacklisting of systems that spin off disks during ACPI power off
  DMI: Introduce dmi_first_match to make the interface more flexible
  Hibernation: Introduce system_entering_hibernation

15 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzi...
Linus Torvalds [Tue, 27 Jan 2009 15:47:22 +0000 (07:47 -0800)] 
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  pata-rb532-cf: remove set_irq_type from finish_io
  [libata] pata_via: support VX855, future chips whose IDE controller use 0x0571
  sata_mv: no longer experimental (v2)
  sata_mv: msi masking fix (v2)
  sata_mv: Properly initialize main irq mask
  sata_mv: remove bogus nsect restriction
  sata_mv: don't read hc_irq_cause
  sata_mv: fix 8-port timeouts on 508x/6081 chips
  sata_nv: fix MCP5x reset
  sata_nv: rename nv_nf2_hardreset()
  libata: set NODEV_HINT for 0x7f status
  libata-sff: fix incorrect EH message

15 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Tue, 27 Jan 2009 15:46:29 +0000 (07:46 -0800)] 
Merge git://git./linux/kernel/git/herbert/crypto-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: ccm - Fix handling of null assoc data
  crypto: blkcipher - Fix WARN_ON handling in walk_done
  crypto: authenc - Fix zero-length IV crash

15 years agoAdd support for '8-port RS-232 MIC-3620 from advantech'
Michael Bramer [Tue, 27 Jan 2009 11:51:16 +0000 (11:51 +0000)] 
Add support for '8-port RS-232 MIC-3620 from advantech'

This Patch add the device information for the
MIC-3620 8-port RS-232 cPCI card from Advantech Co. Ltd.

Signed-off-by: Michael Bramer <grisu@deb-support.de>
Signed-off-by: Alan Cox <number6@the-village.bc.nu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoMove jsm_remove_one to .devexit.text
Uwe Kleine-König [Tue, 27 Jan 2009 11:51:06 +0000 (11:51 +0000)] 
Move jsm_remove_one to .devexit.text

The function jsm_remove_one is used only wrapped by __devexit_p so define
it using __devexit.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Alan Cox <number6@the-village.bc.nu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoMIPS: enable serial UART support on PNX833X devices.
Ihar Hrachyshka [Tue, 27 Jan 2009 11:50:46 +0000 (11:50 +0000)] 
MIPS: enable serial UART support on PNX833X devices.

Enabled serial UART driver for PNX833X devices.

Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
Signed-off-by: Alan Cox <number6@the-village.bc.nu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agotty_open can return to userspace holding tty_mutex
Eric Paris [Tue, 27 Jan 2009 11:50:37 +0000 (11:50 +0000)] 
tty_open can return to userspace holding tty_mutex

__tty_open could return (to userspace) holding the tty_mutex thanks to a
regression introduced by 4a2b5fddd53b80efcb3266ee36e23b8de28e761a ("Move
tty lookup/reopen to caller").

This was found by bisecting an fsfuzzer problem.  Admittedly I have no
idea how it managed to tickle this 100% reliably, but it is clearly a
regression and when hit leaves the box in a completely unusable state.
This patch lets the fsfuzzer test complete every time.

Signed-off-by: Eric Paris <eparis@redhat.com>
Signed-off-by: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoSATA PIIX: Blacklist system that spins off disks during ACPI power off
Rafael J. Wysocki [Mon, 19 Jan 2009 19:59:22 +0000 (20:59 +0100)] 
SATA PIIX: Blacklist system that spins off disks during ACPI power off

Some notebooks from HP have the problem that their BIOSes attempt to
spin down hard drives before entering ACPI system states S4 and S5.
This leads to a yo-yo effect during system power-off shutdown and the
last phase of hibernation when the disk is first spun down by the
kernel and then almost immediately turned on and off by the BIOS.
This, in turn, may result in shortening the disk's life times.

To prevent this from happening we can blacklist the affected systems
using DMI information.

Blacklist HP 2510p that uses the ata_piix driver.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agoSATA Sil: Blacklist system that spins off disks during ACPI power off
Rafael J. Wysocki [Mon, 19 Jan 2009 19:58:29 +0000 (20:58 +0100)] 
SATA Sil: Blacklist system that spins off disks during ACPI power off

Some notebooks from HP have the problem that their BIOSes attempt to
spin down hard drives before entering ACPI system states S4 and S5.
This leads to a yo-yo effect during system power-off shutdown and the
last phase of hibernation when the disk is first spun down by the
kernel and then almost immediately turned on and off by the BIOS.
This, in turn, may result in shortening the disk's life times.

To prevent this from happening we can blacklist the affected systems
using DMI information.

Blacklist HP nx6325 that uses the sata_sil driver.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agoSATA AHCI: Blacklist system that spins off disks during ACPI power off
Rafael J. Wysocki [Mon, 19 Jan 2009 19:57:36 +0000 (20:57 +0100)] 
SATA AHCI: Blacklist system that spins off disks during ACPI power off

Some notebooks from HP have the problem that their BIOSes attempt to
spin down hard drives before entering ACPI system states S4 and S5.
This leads to a yo-yo effect during system power-off shutdown and the
last phase of hibernation when the disk is first spun down by the
kernel and then almost immediately turned on and off by the BIOS.
This, in turn, may result in shortening the disk's life times.

To prevent this from happening we can blacklist the affected systems
using DMI information.

Blacklist HP nx6310 that uses the AHCI driver.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agoSATA: Blacklisting of systems that spin off disks during ACPI power off
Rafael J. Wysocki [Mon, 19 Jan 2009 19:56:43 +0000 (20:56 +0100)] 
SATA: Blacklisting of systems that spin off disks during ACPI power off

Introduce new libata flags ATA_FLAG_NO_POWEROFF_SPINDOWN and
ATA_FLAG_NO_HIBERNATE_SPINDOWN that, if set, will prevent disks from
being spun off during system power off and hibernation, respectively
(to handle the hibernation case we need the new system state
SYSTEM_HIBERNATE_ENTER that can be checked against by libata, in
analogy with SYSTEM_POWER_OFF).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agoDMI: Introduce dmi_first_match to make the interface more flexible
Rafael J. Wysocki [Mon, 19 Jan 2009 19:55:50 +0000 (20:55 +0100)] 
DMI: Introduce dmi_first_match to make the interface more flexible

Some notebooks from HP have the problem that their BIOSes attempt to
spin down hard drives before entering ACPI system states S4 and S5.
This leads to a yo-yo effect during system power-off shutdown and the
last phase of hibernation when the disk is first spun down by the
kernel and then almost immediately turned on and off by the BIOS.
This, in turn, may result in shortening the disk's life times.

To prevent this from happening we can blacklist the affected systems
using DMI information.  However, only the on-board controlles should
be blacklisted and their PCI slot numbers can be used for this
purpose.  Unfortunately the existing interface for checking DMI
information of the system is not very convenient for this purpose,
because to use it, we would have to define special callback functions
or create a separate struct dmi_system_id table for each blacklisted
system.

To overcome this difficulty introduce a new function
dmi_first_match() returning a pointer to the first entry in an array
of struct dmi_system_id elements that matches the system DMI
information.  Then, we can use this pointer to access the entry's
.driver_data field containing the additional information, such as
the PCI slot number, allowing us to do the desired blacklisting.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agoHibernation: Introduce system_entering_hibernation
Rafael J. Wysocki [Mon, 19 Jan 2009 19:54:54 +0000 (20:54 +0100)] 
Hibernation: Introduce system_entering_hibernation

Introduce boolean function system_entering_hibernation() returning
'true' during the last phase of hibernation, in which devices are
being put into low power states and the sleep state (for example,
ACPI S4) is finally entered.

Some device drivers need such a function to check if the system is
in the final phase of hibernation.  In particular, some SATA drivers
are going to use it for blacklisting systems in which the disks
should not be spun down during the last phase of hibernation (the
BIOS will do that anyway).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agopata-rb532-cf: remove set_irq_type from finish_io
Phil Sutter [Tue, 20 Jan 2009 16:40:56 +0000 (17:40 +0100)] 
pata-rb532-cf: remove set_irq_type from finish_io

The driver has been tested without the call to set_irq_type at this
point and occurs to work fine, so it should be safe to remove it.

Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
15 years agom68knommu: fix 5329 ColdFire periphal addressing
Matt Waddel [Sat, 20 Dec 2008 14:16:38 +0000 (07:16 -0700)] 
m68knommu: fix 5329 ColdFire periphal addressing

The 5329 ColdFire peripheral IO register addresses are not relative to
the MBAR register. So fix the serial platform setup array and IRQ acking
to use just the direct addresses.

Signed-off-by: Matt Waddel <Matt.Waddel@freescale.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
15 years agouclinux: add process name to allocation error message
Greg Ungerer [Tue, 13 Jan 2009 07:30:22 +0000 (17:30 +1000)] 
uclinux: add process name to allocation error message

This patch adds the name of the process to the bad allocation error
message on non-MMU systems.

Changed suggested by jsujjavanich@syntech-fuelmaster.com

Signed-off-by: Greg Ungerer <gerg@uclinux.org>
15 years agom68knommu: correct the mii calculations for 532x ColdFire FEC
Matt Waddel [Tue, 13 Jan 2009 07:14:20 +0000 (17:14 +1000)] 
m68knommu: correct the mii calculations for 532x ColdFire FEC

Signed-off-by: Matt Waddel <Matt.Waddel@freescale.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
15 years agom68knommu: add ColdFire M532x to the FEC configuration options
Matt Waddel [Tue, 13 Jan 2009 07:13:06 +0000 (17:13 +1000)] 
m68knommu: add ColdFire M532x to the FEC configuration options

Signed-off-by: Matt Waddel <Matt.Waddel@freescale.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>