platform/upstream/kernel-adaptation-pc.git
18 years ago[PATCH] WAN: ioremap() failure checks in drivers
Krzysztof Halasa [Thu, 22 Jun 2006 20:29:28 +0000 (22:29 +0200)]
[PATCH] WAN: ioremap() failure checks in drivers

Eric Sesterhenn found that pci200syn initialization lacks return
statement in ioremap() error path (coverity bug id #195). It looks
like more WAN drivers have problems with ioremap().

Signed-off-by: Krzysztof Halasa <khc@pm.waw.pl>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] WAN: register_hdlc_device() doesn't need dev_alloc_name()
Krzysztof Halasa [Thu, 22 Jun 2006 20:20:19 +0000 (22:20 +0200)]
[PATCH] WAN: register_hdlc_device() doesn't need dev_alloc_name()

David Boggs noticed that register_hdlc_device() no longer needs
to call dev_alloc_name() as it's called by register_netdev().
register_hdlc_device() is currently equivalent to register_netdev().

hdlc_setup() is now EXPORTed as per David's request.

Signed-off-by: Krzysztof Halasa <khc@pm.waw.pl>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] skb_padto()-area fixes in 8390, wavelan
Alan Cox [Thu, 22 Jun 2006 13:25:34 +0000 (14:25 +0100)]
[PATCH] skb_padto()-area fixes in 8390, wavelan

Ar Iau, 2006-06-22 am 21:29 +1000, ysgrifennodd Herbert Xu:
> Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
> >
> > The 8390 change (corrected version) also makes 8390.c faster so should
> > be applied anyway, and the orinoco one fixes some code that isn't even
> > needed and someone forgot to remove long ago. Otherwise the skb_padto
>
> Yeah I agree totally.  However, I haven't actually seen the fixed 8390
> version being posted yet or at least not to netdev :)

Ah the resounding clang of a subtle hint ;)

Signed-off-by: Alan Cox <alan@redhat.com>
- Return 8390.c to the old way of handling short packets (which is also
faster)

- Remove the skb_padto from orinoco. This got left in when the padding bad
write patch was added and is actually not needed. This is fixing a merge
error way back when.

- Wavelan can also use the stack based buffer trick if you want
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] make drivers/net/forcedeth.c:nv_update_pause() static
Adrian Bunk [Thu, 22 Jun 2006 10:03:29 +0000 (12:03 +0200)]
[PATCH] make drivers/net/forcedeth.c:nv_update_pause() static

This patch makes the needlessly global nv_update_pause() static.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] network driver for Hilscher netx
Sascha Hauer [Thu, 22 Jun 2006 05:11:13 +0000 (07:11 +0200)]
[PATCH] network driver for Hilscher netx

This is a patch for the Hilscher netx builtin ethernet ports. The
netx board support was merged into 2.6.17-git2.
The netx is a arm926 based SoC.

Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
--
 drivers/net/Kconfig             |   11
 drivers/net/Makefile            |    1
 drivers/net/netx-eth.c          |  516 ++++++++++++++++++++++++++++++++++++++++
 include/asm-arm/arch-netx/eth.h |   27 ++
 4 files changed, 555 insertions(+)
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] Dereference in tokenring/olympic.c
Eric Sesterhenn [Wed, 21 Jun 2006 14:17:17 +0000 (16:17 +0200)]
[PATCH] Dereference in tokenring/olympic.c

hi,

coverity found (bug id #225) that we might call free_netdev()
with NULL argument, when alloc_trdev() fails. This patch
changes the goto, so we dont call free_netdev() for
dev == NULL.

Signed-off-by: Eric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] Array overrun in drivers/net/wireless/wavelan.c
Eric Sesterhenn [Wed, 21 Jun 2006 14:40:24 +0000 (16:40 +0200)]
[PATCH] Array overrun in drivers/net/wireless/wavelan.c

hi,

this is another array overrun spotted by coverity (#id 507)
we should check the index against array size before using it.
Not sure why the driver doesnt use ARRAY_SIZE instead of its
own macro.

Signed-off-by: Eric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] Remove useless check in drivers/net/pcmcia/xirc2ps_cs.c
Eric Sesterhenn [Wed, 21 Jun 2006 14:10:48 +0000 (16:10 +0200)]
[PATCH] Remove useless check in drivers/net/pcmcia/xirc2ps_cs.c

hi,

coverity choked at this check (id #223), assuming that
skb might be NULL and used anyways later. Since
start_hard_xmit() always gets called with a valid
skb, the check is useless and this patch removes it.

Signed-off-by: Eric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] 8139cp: add ethtool eeprom support
Philip Craig [Wed, 21 Jun 2006 01:33:27 +0000 (11:33 +1000)]
[PATCH] 8139cp: add ethtool eeprom support

Implement the ethtool eeprom operations for the 8139cp driver.
Tested on x86 and big-endian ARM.

Signed-off-by: Philip Craig <philipc@snapgear.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] 8139cp: fix eeprom read command length
Philip Craig [Wed, 21 Jun 2006 01:33:26 +0000 (11:33 +1000)]
[PATCH] 8139cp: fix eeprom read command length

The read command for the 93C46/93C56 EEPROMS should be 3 bits plus
the address.  This doesn't appear to affect the operation of the
read command, but similar errors for write commands do cause failures.

Signed-off-by: Philip Craig <philipc@snapgear.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] b44: update b44 Kconfig entry
Gary Zambrano [Tue, 20 Jun 2006 22:34:46 +0000 (15:34 -0700)]
[PATCH] b44: update b44 Kconfig entry

Deleted "EXPERIMENTAL" from b44 entry in Kconfig.

Signed-off-by: Gary Zambrano <zambrano@broadcom.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] b44: update version to 1.01
Gary Zambrano [Tue, 20 Jun 2006 22:34:40 +0000 (15:34 -0700)]
[PATCH] b44: update version to 1.01

Update the driver version to 1.01

Signed-off-by: Gary Zambrano <zambrano@broadcom.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] b44: add wol for old nic
Gary Zambrano [Tue, 20 Jun 2006 22:34:36 +0000 (15:34 -0700)]
[PATCH] b44: add wol for old nic

This patch adds wol support for the older 440x nics that use pattern matching.
This patch is a redo thanks to feedback from Michael Chan and Francois Romieu.

Signed-off-by: Gary Zambrano <zambrano@broadcom.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] b44: add parameter
Gary Zambrano [Tue, 20 Jun 2006 22:34:26 +0000 (15:34 -0700)]
[PATCH] b44: add parameter

This patch adds a parameter to init_hw() to not completely initialize
the nic for wol.

Signed-off-by: Gary Zambrano <zambrano@broadcom.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] b44: add wol
Gary Zambrano [Tue, 20 Jun 2006 22:34:23 +0000 (15:34 -0700)]
[PATCH] b44: add wol

Adds wol to the driver.
This is a redo of a previous patch thanks to feedback from Francois Romieu.

Signed-off-by Gary Zambrano <zambrano@broadcom.com>

Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] b44: fix manual speed/duplex/autoneg settings
Gary Zambrano [Tue, 20 Jun 2006 22:34:15 +0000 (15:34 -0700)]
[PATCH] b44: fix manual speed/duplex/autoneg settings

Fixes for speed/duplex/autoneg settings and driver settings info.
This is a redo of a previous patch thanks to feedback from Jeff Garzik.

Signed-off-by: Gary Zambrano <zambrano@broadcom.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] AT91RM9200 Ethernet #4: Suspend/Resume
Andrew Victor [Tue, 20 Jun 2006 10:19:13 +0000 (12:19 +0200)]
[PATCH] AT91RM9200 Ethernet #4: Suspend/Resume

Adds power-management (suspend/resume) support to the AT91RM9200
Ethernet driver.
Patch from David Brownell.

Signed-off-by: Andrew Victor <andrew@sanpeople.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] AT91RM9200 Ethernet #3: Cleanup
Andrew Victor [Tue, 20 Jun 2006 10:10:57 +0000 (12:10 +0200)]
[PATCH] AT91RM9200 Ethernet #3: Cleanup

Moved global ether_clk variable into controller data structure.
Patch from David Brownell.

Davicom 9161 PHY was being incorrectly displayed as "9196".
Patch from Brian Stafford.

clk_get() doesn't return NULL on error, so the return value needs to be
tested with IS_ERR().

Whitespace cleanup.

Signed-off-by: Andrew Victor <andrew@sanpeople.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] AT91RM9200 Ethernet #2: MII interface
Andrew Victor [Tue, 20 Jun 2006 09:59:05 +0000 (11:59 +0200)]
[PATCH] AT91RM9200 Ethernet #2: MII interface

Adds support for the MII ioctls via generic_mii_ioctl().
Patch from Brian Stafford.

Set the mii.phy_id to the detected PHY address, otherwise ethtool cannot
access PHYs other than 0.
Patch from Roman Kolesnikov.

Signed-off-by: Andrew Victor <andrew@sanpeople.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] AT91RM9200 Ethernet #1: Link poll
Andrew Victor [Tue, 20 Jun 2006 09:50:23 +0000 (11:50 +0200)]
[PATCH] AT91RM9200 Ethernet #1: Link poll

For Ethernet PHYs that don't have an IRQ pin or boards that don't
connect the IRQ pin to the processor, we enable a timer to poll the
PHY's link state.

Patch originally supplied by Eric Benard and Roman Kolesnikov.

Signed-off-by: Andrew Victor <andrew@sanpeople.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] IP27: Really set PCI64_ATTR_VIRTUAL, not PCI64_ATTR_PREC.
Ralf Baechle [Sat, 17 Jun 2006 17:57:39 +0000 (18:57 +0100)]
[PATCH] IP27: Really set PCI64_ATTR_VIRTUAL, not PCI64_ATTR_PREC.

IOC3's homegrown DMA mapping functions that are used to optimize things
a little on IP27 set the wrong bit.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years agoMerge branch 'master' into upstream
Jeff Garzik [Fri, 23 Jun 2006 02:51:46 +0000 (22:51 -0400)]
Merge branch 'master' into upstream

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/i2c-2.6
Linus Torvalds [Thu, 22 Jun 2006 22:08:56 +0000 (15:08 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/i2c-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/i2c-2.6: (44 commits)
  [PATCH] I2C: I2C controllers go into right place on sysfs
  [PATCH] hwmon-vid: Add support for Intel Core and Conroe
  [PATCH] lm70: New hardware monitoring driver
  [PATCH] hwmon: Fix the Kconfig header
  [PATCH] i2c-i801: Merge setup function
  [PATCH] i2c-i801: Better pci subsystem integration
  [PATCH] i2c-i801: Cleanups
  [PATCH] i2c-i801: Remove PCI function check
  [PATCH] i2c-i801: Remove force_addr parameter
  [PATCH] i2c-i801: Fix block transaction poll loops
  [PATCH] scx200_acb: Documentation update
  [PATCH] scx200_acb: Mark scx200_acb_probe __init
  [PATCH] scx200_acb: Use PCI I/O resource when appropriate
  [PATCH] i2c: Mark block write buffers as const
  [PATCH] i2c-ocores: Minor cleanups
  [PATCH] abituguru: Fix fan detection
  [PATCH] abituguru: Review fixes
  [PATCH] abituguru: New hardware monitoring driver
  [PATCH] w83792d: Add missing data access locks
  [PATCH] w83792d: Fix setting the PWM value
  ...

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/w1-2.6
Linus Torvalds [Thu, 22 Jun 2006 22:08:34 +0000 (15:08 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/w1-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/w1-2.6:
  [PATCH] w1: warning fix
  [PATCH] w1: clean up W1_CON dependency.
  [PATCH] drivers/w1/w1.c: fix a compile error
  [PATCH] W1: fix dependencies of W1_SLAVE_DS2433_CRC
  [PATCH] W1: possible cleanups
  [PATCH] W1: cleanups
  [PATCH] w1 exports
  [PATCH] w1: Use mutexes instead of semaphores.
  [PATCH] w1: Make w1 connector notifications depend on connector.
  [PATCH] w1: netlink: Mark netlink group 1 as unused.
  [PATCH] w1: Move w1-connector definitions into linux/include/connector.h
  [PATCH] w1: Userspace communication protocol over connector.
  [PATCH] w1: Replace dscore and ds_w1_bridge with ds2490 driver.
  [PATCH] w1: Added default generic read/write operations.

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6
Linus Torvalds [Thu, 22 Jun 2006 22:07:59 +0000 (15:07 -0700)]
Merge /pub/scm/linux/kernel/git/gregkh/pci-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6: (27 commits)
  [PATCH] PCI: nVidia quirk to make AER PCI-E extended capability visible
  [PATCH] PCI: fix issues with extended conf space when MMCONFIG disabled because of e820
  [PATCH] PCI: Bus Parity Status sysfs interface
  [PATCH] PCI: fix memory leak in MMCONFIG error path
  [PATCH] PCI: fix error with pci_get_device() call in the mpc85xx driver
  [PATCH] PCI: MSI-K8T-Neo2-Fir: run only where needed
  [PATCH] PCI: fix race with pci_walk_bus and pci_destroy_dev
  [PATCH] PCI: clean up pci documentation to be more specific
  [PATCH] PCI: remove unneeded msi code
  [PATCH] PCI: don't move ioapics below PCI bridge
  [PATCH] PCI: cleanup unused variable about msi driver
  [PATCH] PCI: disable msi mode in pci_disable_device
  [PATCH] PCI: Allow MSI to work on kexec kernel
  [PATCH] PCI: AMD 8131 MSI quirk called too late, bus_flags not inherited ?
  [PATCH] PCI: Move various PCI IDs to header file
  [PATCH] PCI Bus Parity Status-broken hardware attribute, EDAC foundation
  [PATCH] PCI: i386/x86_84: disable PCI resource decode on device disable
  [PATCH] PCI ACPI: Rename the functions to avoid multiple instances.
  [PATCH] PCI: don't enable device if already enabled
  [PATCH] PCI: Add a "enable" sysfs attribute to the pci devices to allow userspace (Xorg) to enable devices without doing foul direct access
  ...

18 years ago[PATCH] x86_64: use select for GART_IOMMU to enable AGP
Roman Zippel [Thu, 22 Jun 2006 21:47:35 +0000 (14:47 -0700)]
[PATCH] x86_64: use select for GART_IOMMU to enable AGP

The AGP default doesn't work well with other selects, so use a select for
GART_IOMMU as well.  Remove a redundant default for SWIOTLB as well.

Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Andi Kleen <ak@muc.de>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Dave Airlie <airlied@linux.ie>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] zlib_inflate: Upgrade library code to a recent version
Richard Purdie [Thu, 22 Jun 2006 21:47:34 +0000 (14:47 -0700)]
[PATCH] zlib_inflate: Upgrade library code to a recent version

Upgrade the zlib_inflate implementation in the kernel from a patched
version 1.1.3/4 to a patched 1.2.3.

The code in the kernel is about seven years old and I noticed that the
external zlib library's inflate performance was significantly faster (~50%)
than the code in the kernel on ARM (and faster again on x86_32).

For comparison the newer deflate code is 20% slower on ARM and 50% slower
on x86_32 but gives an approx 1% compression ratio improvement.  I don't
consider this to be an improvement for kernel use so have no plans to
change the zlib_deflate code.

Various changes have been made to the zlib code in the kernel, the most
significant being the extra functions/flush option used by ppp_deflate.
This update reimplements the features PPP needs to ensure it continues to
work.

This code has been tested on ARM under both JFFS2 (with zlib compression
enabled) and ppp_deflate and on x86_32.  JFFS2 sees an approx.  10% real
world file read speed improvement.

This patch also removes ZLIB_VERSION as it no longer has a correct value.
We don't need version checks anyway as the kernel's module handling will
take care of that for us.  This removal is also more in keeping with the
zlib author's wishes (http://www.zlib.net/zlib_faq.html#faq24) and I've
added something to the zlib.h header to note its a modified version.

Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
Acked-by: Joern Engel <joern@wh.fh-wedel.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] vgacon: make VGA_MAP_MEM take size, remove extra use
Bjorn Helgaas [Thu, 22 Jun 2006 21:47:32 +0000 (14:47 -0700)]
[PATCH] vgacon: make VGA_MAP_MEM take size, remove extra use

VGA_MAP_MEM translates to ioremap() on some architectures.  It makes sense
to do this to vga_vram_base, because we're going to access memory between
vga_vram_base and vga_vram_end.

But it doesn't really make sense to map starting at vga_vram_end, because
we aren't going to access memory starting there.  On ia64, which always has
to be different, ioremapping vga_vram_end gives you something completely
incompatible with ioremapped vga_vram_start, so vga_vram_size ends up being
nonsense.

As a bonus, we often know the size up front, so we can use ioremap()
correctly, rather than giving it a zero size.

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] prune_one_dentry() tweaks
Andrew Morton [Thu, 22 Jun 2006 21:47:31 +0000 (14:47 -0700)]
[PATCH] prune_one_dentry() tweaks

- Add description of d_lock handling to comments over prune_one_dentry().

- It has three callsites - uninline it, saving 200 bytes of text.

Cc: Jan Blunck <jblunck@suse.de>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Olaf Hering <olh@suse.de>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Fix dcache race during umount
NeilBrown [Thu, 22 Jun 2006 21:47:28 +0000 (14:47 -0700)]
[PATCH] Fix dcache race during umount

The race is that the shrink_dcache_memory shrinker could get called while a
filesystem is being unmounted, and could try to prune a dentry belonging to
that filesystem.

If it does, then it will call in to iput on the inode while the dentry is
no longer able to be found by the umounting process.  If iput takes a
while, generic_shutdown_super could get all the way though
shrink_dcache_parent and shrink_dcache_anon and invalidate_inodes without
ever waiting on this particular inode.

Eventually the superblock gets freed anyway and if the iput tried to touch
it (which some filesystems certainly do), it will lose.  The promised
"Self-destruct in 5 seconds" doesn't lead to a nice day.

The race is closed by holding s_umount while calling prune_one_dentry on
someone else's dentry.  As a down_read_trylock is used,
shrink_dcache_memory will no longer try to prune the dentry of a filesystem
that is being unmounted, and unmount will not be able to start until any
such active prune_one_dentry completes.

This requires that prune_dcache *knows* which filesystem (if any) it is
doing the prune on behalf of so that it can be careful of other
filesystems.  shrink_dcache_memory isn't called it on behalf of any
filesystem, and so is careful of everything.

shrink_dcache_anon is now passed a super_block rather than the s_anon list
out of the superblock, so it can get the s_anon list itself, and can pass
the superblock down to prune_dcache.

If prune_dcache finds a dentry that it cannot free, it leaves it where it
is (at the tail of the list) and exits, on the assumption that some other
thread will be removing that dentry soon.  To try to make sure that some
work gets done, a limited number of dnetries which are untouchable are
skipped over while choosing the dentry to work on.

I believe this race was first found by Kirill Korotaev.

Cc: Jan Blunck <jblunck@suse.de>
Acked-by: Kirill Korotaev <dev@openvz.org>
Cc: Olaf Hering <olh@suse.de>
Acked-by: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Balbir Singh <balbir@in.ibm.com>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] avoid tasklist_lock at getrusage for multithreaded case too
Ravikiran G Thirumalai [Thu, 22 Jun 2006 21:47:26 +0000 (14:47 -0700)]
[PATCH] avoid tasklist_lock at getrusage for multithreaded case too

Avoid taking tasklist_lock for at getrusage for the multithreaded case too.
We don't need to take the tasklist lock for thread traversal of a process
since Oleg's do-__unhash_process-under-siglock.patch and related work.

Signed-off-by: Ravikiran Thirumalai <kiran@scalex86.org>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] remove steal_locks()
Miklos Szeredi [Thu, 22 Jun 2006 21:47:22 +0000 (14:47 -0700)]
[PATCH] remove steal_locks()

This patch removes the steal_locks() function.

steal_locks() doesn't work correctly with any filesystem that does it's own
lock management, including NFS, CIFS, etc.

In addition it has weird semantics on local filesystems in case tasks
sharing file-descriptor tables are doing POSIX locking operations in
parallel to execve().

The steal_locks() function has an effect on applications doing:

clone(CLONE_FILES)
  /* in child */
  lock
  execve
  lock

POSIX locks acquired before execve (by "child", "parent" or any further
task sharing files_struct) will after the execve be owned exclusively by
"child".

According to Chris Wright some LSB/LTP kind of suite triggers without the
stealing behavior, but there's no known real-world application that would
also fail.

Apps using NPTL are not affected, since all other threads are killed before
execve.

Apps using LinuxThreads are only affected if they

  - have multiple threads during exec (LinuxThreads doesn't kill other
    threads, the app may do it with pthread_kill_other_threads_np())
  - rely on POSIX locks being inherited across exec

Both conditions are documented, but not their interaction.

Apps using clone() natively are affected if they

  - use clone(CLONE_FILES)
  - rely on POSIX locks being inherited across exec

The above scenarios are unlikely, but possible.

If the patch is vetoed, there's a plan B, that involves mostly keeping the
weird stealing semantics, but changing the way lock ownership is handled so
that network and local filesystems work consistently.

That would add more complexity though, so this solution seems to be
preferred by most people.

Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: Matthew Wilcox <willy@debian.org>
Cc: Chris Wright <chrisw@sous-sol.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Steven French <sfrench@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Fix a race condition between ->i_mapping and iput()
OGAWA Hirofumi [Thu, 22 Jun 2006 21:47:21 +0000 (14:47 -0700)]
[PATCH] Fix a race condition between ->i_mapping and iput()

This race became a cause of oops, and can reproduce by the following.

    while true; do
dd if=/dev/zero of=/dev/.static/dev/hdg1 bs=512 count=1000 & sync
    done

This race condition was between __sync_single_inode() and iput().

          cpu0 (fs's inode)                 cpu1 (bdev's inode)
          -----------------                 -------------------
                                       close("/dev/hda2")
                                       [...]
__sync_single_inode()
   /* copy the bdev's ->i_mapping */
   mapping = inode->i_mapping;

                                       generic_forget_inode()
                                          bdev_clear_inode()
     /* restre the fs's ->i_mapping */
             inode->i_mapping = &inode->i_data;
          /* bdev's inode was freed */
                                          destroy_inode(inode);

   if (wait) {
      /* dereference a freed bdev's mapping->host */
      filemap_fdatawait(mapping);  /* Oops */

Since __sync_single_inode() is only taking a ref-count of fs's inode, the
another process can be close() and freeing the bdev's inode while writing
fs's inode.  So, __sync_signle_inode() accesses the freed ->i_mapping,
oops.

This patch takes a ref-count on the bdev's inode for the fs's inode before
setting a ->i_mapping, and the clear_inode() of the fs's inode does iput() on
the bdev's inode.  So if the fs's inode is still living, bdev's inode
shouldn't be freed.

Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] PCI: Add PCI_CAP_ID_VNDR
Brice Goglin [Thu, 22 Jun 2006 21:47:20 +0000 (14:47 -0700)]
[PATCH] PCI: Add PCI_CAP_ID_VNDR

Add the vendor-specific extended capability PCI_CAP_ID_VNDR.  It is required
by the Myri-10G Ethernet driver.

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] myri10ge build fix
Andrew Morton [Thu, 22 Jun 2006 21:47:19 +0000 (14:47 -0700)]
[PATCH] myri10ge build fix

Someone changed skb_linearize().

Cc: Brice Goglin <bgoglin@myri.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] suspend_console() warning fix
Andrew Morton [Thu, 22 Jun 2006 21:47:18 +0000 (14:47 -0700)]
[PATCH] suspend_console() warning fix

kernel/power/main.c: In function 'suspend_prepare':
kernel/power/main.c:89: warning: implicit declaration of function 'suspend_console'
kernel/power/main.c: In function 'suspend_finish':
kernel/power/main.c:137: warning: implicit declaration of function 'resume_console'

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Keys: Fix race between two instantiators of a key
David Howells [Thu, 22 Jun 2006 21:47:18 +0000 (14:47 -0700)]
[PATCH] Keys: Fix race between two instantiators of a key

Add a revocation notification method to the key type and calls it whilst
the key's semaphore is still write-locked after setting the revocation
flag.

The patch then uses this to maintain a reference on the task_struct of the
process that calls request_key() for as long as the authorisation key
remains unrevoked.

This fixes a potential race between two processes both of which have
assumed the authority to instantiate a key (one may have forked the other
for example).  The problem is that there's no locking around the check for
revocation of the auth key and the use of the task_struct it points to, nor
does the auth key keep a reference on the task_struct.

Access to the "context" pointer in the auth key must thenceforth be done
with the auth key semaphore held.  The revocation method is called with the
target key semaphore held write-locked and the search of the context
process's keyrings is done with the auth key semaphore read-locked.

The check for the revocation state of the auth key just prior to searching
it is done after the auth key is read-locked for the search.  This ensures
that the auth key can't be revoked between the check and the search.

The revocation notification method is added so that the context task_struct
can be released as soon as instantiation happens rather than waiting for
the auth key to be destroyed, thus avoiding the unnecessary pinning of the
requesting process.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] selinux: add hooks for key subsystem
Michael LeMay [Thu, 22 Jun 2006 21:47:17 +0000 (14:47 -0700)]
[PATCH] selinux: add hooks for key subsystem

Introduce SELinux hooks to support the access key retention subsystem
within the kernel.  Incorporate new flask headers from a modified version
of the SELinux reference policy, with support for the new security class
representing retained keys.  Extend the "key_alloc" security hook with a
task parameter representing the intended ownership context for the key
being allocated.  Attach security information to root's default keyrings
within the SELinux initialization routine.

Has passed David's testsuite.

Signed-off-by: Michael LeMay <mdlemay@epoch.ncsc.mil>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: James Morris <jmorris@namei.org>
Acked-by: Chris Wright <chrisw@sous-sol.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] NTFS: Critical bug fix (affects MIPS and possibly others)
Anton Altaparmakov [Thu, 22 Jun 2006 21:47:15 +0000 (14:47 -0700)]
[PATCH] NTFS: Critical bug fix (affects MIPS and possibly others)

Many thanks to Pauline Ng for the detailed bug report and analysis!

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] sparc build breakage
Al Viro [Thu, 22 Jun 2006 21:47:14 +0000 (14:47 -0700)]
[PATCH] sparc build breakage

rd_prompt et.al.  depend on CONFIG_BLK_DEV_RAM, not CONFIG_BLK_INITRD; now
that those are independent, setup.c blows with INITRD on and BLK_DEV_RAM
off.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: William Lee Irwin III <wli@holomorphy.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] UML: fix wall_to_monotonic initialization
Jeff Dike [Thu, 22 Jun 2006 21:47:09 +0000 (14:47 -0700)]
[PATCH] UML: fix wall_to_monotonic initialization

Change a variable from unsigned to signed in order to get sign-extension
when the thing is negated.  Without this, uptime is horribly confused.

Signed-off-by: Jeff Dike <jdike@addtoit.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[ALSA] version 1.0.12rc1
Jaroslav Kysela [Thu, 22 Jun 2006 19:35:11 +0000 (21:35 +0200)]
[ALSA] version 1.0.12rc1

18 years ago[ALSA] aoa driver - Kconfig - remove spaces for SND!=n
Jaroslav Kysela [Thu, 22 Jun 2006 15:49:58 +0000 (17:49 +0200)]
[ALSA] aoa driver - Kconfig - remove spaces for SND!=n

Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] hda-codec - Show EAPD and pin-detection capabilities in proc
Takashi Iwai [Wed, 21 Jun 2006 17:19:25 +0000 (19:19 +0200)]
[ALSA] hda-codec - Show EAPD and pin-detection capabilities in proc

Show EAPD and pin-detection capabilities in proc files.
They are often required to support the proper audio functionality.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Remove nested mutexes in seq_ports.c
Takashi Iwai [Wed, 21 Jun 2006 14:37:54 +0000 (16:37 +0200)]
[ALSA] Remove nested mutexes in seq_ports.c

Removed nested mutexes in the removal routine of port connections.
The port is guaranteed to be offline before calling it, so no mutex
is needed.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Remove ppc/toonie.c
Takashi Iwai [Wed, 21 Jun 2006 14:23:31 +0000 (16:23 +0200)]
[ALSA] Remove ppc/toonie.c

Remove obsoleted ppc/toonie.c.
The function is replaced with new snd-aoa driver.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] snd-powermac: no longer handle anything with a layout-id property
Johannes Berg [Wed, 21 Jun 2006 13:43:44 +0000 (15:43 +0200)]
[ALSA] snd-powermac: no longer handle anything with a layout-id property

This patch removes from snd-powermac the code that check for the layout-id
and instead adds code that makes it refuse loading when a layout-id property
is present, nothing that snd-aoa should be used.
It also removes the 'toonie' codec from snd-powermac which was only ever
used on the mac mini which has a layout-id property.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] snd-aoa: add snd-aoa
Johannes Berg [Wed, 21 Jun 2006 13:42:43 +0000 (15:42 +0200)]
[ALSA] snd-aoa: add snd-aoa

This large patch adds all of snd-aoa.
Consisting of many modules, it currently replaces snd-powermac
for all layout-id based machines and handles many more (for
example new powerbooks and powermacs with digital output that
previously couldn't be used at all).
It also has support for all layout-IDs that Apple has (judging
from their Info.plist file) but not all are tested.
The driver currently has 2 known regressions over snd-powermac:
 * it doesn't handle powermac 7,2 and 7,3
 * it doesn't have a DRC control on snapper-based machines
I will fix those during the 2.6.18 development cycle.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda-codec - Use 3stack model for ASUS P5RD2-VM / P5GPL-X SE
Takashi Iwai [Wed, 21 Jun 2006 10:14:40 +0000 (12:14 +0200)]
[ALSA] hda-codec - Use 3stack model for ASUS P5RD2-VM / P5GPL-X SE

Use 3stack model as default for ASUS P5RD2-VM and P5GPL-X SE
boards with AD1986A codec (ALSA bug#2103).

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] HDA - Lenovo 3000 N100-07684JU - enable laptop-eapd by default
Daniel T Chen [Wed, 21 Jun 2006 06:51:07 +0000 (08:51 +0200)]
[ALSA] HDA - Lenovo 3000 N100-07684JU - enable laptop-eapd by default

Justin Sunseri reports that sound is audible on his Lenovo 3000
N100-07684JU by passing 'model=laptop-eapd' to modprobe, so this
patch adds the pci ids for his sound device to patch_analog.c .
This commit closes LP#39517. Alexey Parshin also confirmed the fix at
http://bugs.gentoo.org/137245
TODO: Mute onboard speakers when device is plugged into the headphone
jack. Muting the 'External Amplifier' mixer element while a device is
plugged into the headphone jack allows sound to be played only from the
headphone jack.
From: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Daniel T Chen <crimsun at ubuntu.com>
Signed-off-by: Daniel Drake <dsd@gentoo.org>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] USB midi: Remove duplicate CS_AUDIO_* #defines
Ben Williamson [Mon, 19 Jun 2006 15:20:09 +0000 (17:20 +0200)]
[ALSA] USB midi: Remove duplicate CS_AUDIO_* #defines

Removed the CS_AUDIO_* #defines, which were duplicates of the
class-specific USB_DT_CS_* #defines in <linux/usb_ch9.h>.

Signed-off-by: Ben Williamson <ben.williamson@greyinnovation.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] AD1888 suspend/resume fix
Jaya Kumar [Mon, 19 Jun 2006 13:06:14 +0000 (15:06 +0200)]
[ALSA] AD1888 suspend/resume fix

This patch adds a write to an undocumented register, 0x60 Extended
Codec Register Page in the AD1888 codec. It is neccessary in order
to make suspend/resume work with the AD1888.

Signed-off-by: Jaya Kumar <jayakumar.alsa@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Disable AC97 AUX and VIDEO controls for WM9705 touchscreen
Rodolfo Giometti [Mon, 19 Jun 2006 13:04:54 +0000 (15:04 +0200)]
[ALSA] Disable AC97 AUX and VIDEO controls for WM9705 touchscreen

This patch by Rodolfo Giometti disables the AC97 AUX and VIDEO controls
on the WM9705 when the touchscreen is selected as the AUX and VIDEO
lines are shared with the touch controller.
Changes:-
 o Added AC97_HAS_NO_AUX flag
 o Test for AC97_HAS_NO_AUX flag in snd_ac97_mixer_build()
 o Sets AC97_HAS_NO_VIDEO and AC97_HAS_NO_AUX in patch_wolfson05() when
WM9705 touch driver is selected.

Signed-off-by: Rodolfo Giometti <giometti@linux.it>
Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] via82xx - Default to variable samplerate enabled for MSI K8T Neo2-FI
Karsten Wiese [Mon, 19 Jun 2006 11:22:52 +0000 (13:22 +0200)]
[ALSA] via82xx - Default to variable samplerate enabled for MSI K8T Neo2-FI

Default to variable samplerate enabled for MSI K8T Neo2-FI
No crackles here with 44100.

Signed-off-by: Karsten Wiese <annabellesgarden@yahoo.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Add hp_only quirk for pci id [161f:2032] to via82xx
Daniel T Chen [Fri, 16 Jun 2006 14:21:54 +0000 (16:21 +0200)]
[ALSA] Add hp_only quirk for pci id [161f:2032] to via82xx

http://www.kernel.org/git/?p=linux/kernel/git/bcollins/ubuntu-dapper.git;a=commitdiff;h=eae2cc78de39502595f67b7fc1f821f5963bb8ae
UpstreamStatus: Not merged
Christian Bjalevik reports in LP#38546 that his sound chipset requires
the 'hp_only' quirk to allow him to control sound volume correctly when
headphones are inserted. This patch adds the appropriate pci id to the
via82xx ALSA driver so that the quirk is applied automatically, thereby
removing the need for users to modify /etc/modprobe.d/alsa-base (or to
unload and reload snd-via82xx with ac97_quirk=hp_only).
This patch closes LP#38546.

Signed-off-by: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Ben Collins <bcollins@ubuntu.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] sound/pci/: Add hp_only quirk for Dell D800 laptops
Daniel T Chen [Fri, 16 Jun 2006 10:13:00 +0000 (12:13 +0200)]
[ALSA] sound/pci/: Add hp_only quirk for Dell D800 laptops

http://www.kernel.org/git/?p=linux/kernel/git/bcollins/ubuntu-dapper.git;a=commitdiff;h=9ad787cd9670c3f3b8f3db235e84baf00a2ea526
Anders Ostling comments in Malone #41015 that his Dell D800 laptop's
volume control works correctly when the hp_only quirk is passed to
modprobe. This commit adds his hardware's sub{vendor,device} ids to
the quirk list for the intel8x0 driver.

Signed-off-by: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Ben Collins <bcollins@ubuntu.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] cs5535audio - trivial debug printk
Jaya Kumar [Tue, 13 Jun 2006 10:01:14 +0000 (12:01 +0200)]
[ALSA] cs5535audio - trivial debug printk

Following is a trivial patch to get more info for boards where
the AC97_VENDOR_ID2 register (or others) time out.

Signed-off-by: Jaya Kumar <jayakumar.alsa@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] fix potential NULL pointer deref in snd_sb8dsp_midi_interrupt()
Jesper Juhl [Tue, 13 Jun 2006 09:58:12 +0000 (11:58 +0200)]
[ALSA] fix potential NULL pointer deref in snd_sb8dsp_midi_interrupt()

First testing if a pointer is NULL and if it is (or might be), proceeding
with code that dereferences that same pointer is clearly a mistake.
This happens in sound/isa/sb/sb8_midi.c::snd_sb8dsp_midi_interrupt()
The patch below reworks the code so this unfortunate case doesn't happen.
Also remove some blank comments.
Found by the Coverity checker as bug #367
Patch is compile testted only due to lack of hardware.

Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda-codec - Add SPDIF support to Thinkpad T/X/Z60
Takashi Iwai [Tue, 13 Jun 2006 09:57:22 +0000 (11:57 +0200)]
[ALSA] hda-codec - Add SPDIF support to Thinkpad T/X/Z60

Added IEC958 (SPDIF) output support to Thinkpad T/X/Z60 with
AD1981HD codec.  The spdif jack is on docking station.
Also, renamed 'IEC958 Playback Route' to 'IEC958 Playback Source'
to avoid the mixer name confliction with IEC958 switch.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] ac97_codec - fix duplicate control creation in AC97
Jaya Kumar [Mon, 12 Jun 2006 08:08:02 +0000 (10:08 +0200)]
[ALSA] ac97_codec - fix duplicate control creation in AC97

This patch conditions AC97 control creation by whether or not the
codec is an AD18xx codec. This fixes the case where the default control
would get created and then snd_ac97_mixer_build fails out when creation
of ad18xx specific control would get attempted. This problem was found
and debuged by Marcelo Tosatti.

Signed-off-by: Jaya Kumar <jayakumar.alsa@gmail.com>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] RME HDSP - fixed proc interface (missing {})
Remy Bruno [Mon, 12 Jun 2006 07:25:22 +0000 (09:25 +0200)]
[ALSA] RME HDSP - fixed proc interface (missing {})

From: Remy Bruno <remy.bruno@trinnov.com>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] snd-ca0106: Update playback to 24bit. Fix typo is comment.
James Courtier-Dutton [Sat, 10 Jun 2006 08:16:49 +0000 (09:16 +0100)]
[ALSA] snd-ca0106: Update playback to 24bit. Fix typo is comment.

Signed-off-by: James Courtier-Dutton <James@superbug.co.uk>
18 years ago[ALSA] Fix invalid __init in ALSA ISA drivers
Takashi Iwai [Fri, 9 Jun 2006 13:28:07 +0000 (15:28 +0200)]
[ALSA] Fix invalid __init in ALSA ISA drivers

Replaced invalid __init with __devinit in snd-sbawe and snd-opl3sa2
drivers.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] sound/vxpocket: fix printk warning
Randy Dunlap [Thu, 8 Jun 2006 10:01:44 +0000 (12:01 +0200)]
[ALSA] sound/vxpocket: fix printk warning

Fix printk format warning:
sound/pcmcia/vx/vxp_ops.c:205: warning: format '%x' expects type 'unsigned int', but argument 5 has type 'size_t'

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Remove bogus check of mmap_count in snd_pcm_release()
Takashi Iwai [Wed, 7 Jun 2006 16:12:31 +0000 (18:12 +0200)]
[ALSA] Remove bogus check of mmap_count in snd_pcm_release()

Removed a bogus check of mmap_count in snd_pcm_release().
This is no longer true for the shared streams.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Fix races in irq handler and ioremap
Takashi Iwai [Tue, 6 Jun 2006 13:44:34 +0000 (15:44 +0200)]
[ALSA] Fix races in irq handler and ioremap

Call ioremap before request_irq for avoiding possible races
in the irq handler.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] virmidi: revert erroneous removal of zero initialization
Clemens Ladisch [Fri, 2 Jun 2006 07:15:44 +0000 (09:15 +0200)]
[ALSA] virmidi: revert erroneous removal of zero initialization

The last patch that tried to remove zero initializations of static
variables accidentally removed a not-quite-zero initialization too.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
18 years ago[ALSA] hda-codec - Add model entry for HP nx6320
Takashi Iwai [Thu, 1 Jun 2006 19:08:53 +0000 (21:08 +0200)]
[ALSA] hda-codec - Add model entry for HP nx6320

Added a model entry for HP nx6320 with AD1981HD codec.
It wasn't covered by the generic HP entry because of a hardware
bug (the SSID is reversed).

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] ac97 - Add Thinkpad T41p to AD1981 jack-sense blacklist
Takashi Iwai [Thu, 1 Jun 2006 15:16:41 +0000 (17:16 +0200)]
[ALSA] ac97 - Add Thinkpad T41p to AD1981 jack-sense blacklist

Added Thinkpad T41p to the blacklist to disable HP/line jack-sensing
with AD1981B.  The jack-sensing is just harmful on this laptop.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Fix possible races in PCI driver removal
Takashi Iwai [Thu, 1 Jun 2006 12:47:29 +0000 (14:47 +0200)]
[ALSA] Fix possible races in PCI driver removal

Call free_irq() before releasing others to avoid races when
shared irq is issued.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda-intel - Fix race in remove
Takashi Iwai [Thu, 1 Jun 2006 09:42:14 +0000 (11:42 +0200)]
[ALSA] hda-intel - Fix race in remove

Call iounmap after free_irq to avoid invalid accesses in the
shared irq.  The patch is taken from
https://bugzilla.novell.com/show_bug.cgi?id=167869

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] bt87x: add Voodoo TV 200 whitelist entry
Clemens Ladisch [Thu, 1 Jun 2006 06:33:48 +0000 (08:33 +0200)]
[ALSA] bt87x: add Voodoo TV 200 whitelist entry

This adds a whitelist entry for the digital audio input of the Voodoo TV 200.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
18 years ago[ALSA] NULL pointer dereference in sound/synth/emux/soundfont.c
Eric Sesterhenn [Wed, 31 May 2006 09:55:17 +0000 (11:55 +0200)]
[ALSA] NULL pointer dereference in sound/synth/emux/soundfont.c

this is about coverity id #100.
It seems the if statement is negated, since the else branch calls
remove_info() with sflist->currsf as a parameter where it gets
dereferenced.

Signed-off-by: Eric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] cmipci - Fix a typo in 'PC Speaker Playback Switch' control
Takashi Iwai [Mon, 29 May 2006 17:05:28 +0000 (19:05 +0200)]
[ALSA] cmipci - Fix a typo in 'PC Speaker Playback Switch' control

Fixed a typo in  'PC Speaker Playback Switch' control name.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] ice1712 - Set mpu401 info flags from _card_info
Alan Horstmann [Fri, 26 May 2006 15:19:34 +0000 (17:19 +0200)]
[ALSA] ice1712 - Set mpu401 info flags from _card_info

To permit use, in ice1712, of the mpu401 info flags recently added to
mpu401_uart, adds info_flags in snd_ice1712_card_info so that additional
flags can be set, if desired.  'MPU401_INFO_INTEGRATED' is always set with
the ice1712.  The flags are passed on to snd_mpu401_uart_new().
_INFO_OUTPUT is set for DMX6fire mpu2.

Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda-codec - Fix model for HP dc7600
Takashi Iwai [Fri, 26 May 2006 12:58:29 +0000 (14:58 +0200)]
[ALSA] hda-codec - Fix model for HP dc7600

Changed the assigned model for HP dc7600 with ALC260 codec
to match better with the actual I/O assignment.
Patch taken from ALSA bug#2157.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] ice1724 - Add functionality for Audiotrak Prodigy 7.1 LT
Takashi Iwai [Tue, 23 May 2006 13:46:10 +0000 (15:46 +0200)]
[ALSA] ice1724 - Add functionality for Audiotrak Prodigy 7.1 LT

This patch adds support for useable front audio channels,
user controllable headphone channel and optical output.
From: Anho Ki
Signed-off-by: Matt Taylor <mwtaylor@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] au88x0 - Fix 64bit address of MPU401 MMIO port
Takashi Iwai [Tue, 23 May 2006 11:30:59 +0000 (13:30 +0200)]
[ALSA] au88x0 - Fix 64bit address of MPU401 MMIO port

Fix 64bit address of MPU401 MMIO port on au88x0 chip.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] ice1712 - Disable AC97 for DMX6fire
Alan Horstmann [Tue, 23 May 2006 11:29:51 +0000 (13:29 +0200)]
[ALSA] ice1712 - Disable AC97 for DMX6fire

Consumer AC97 is not used by the Terratec DMX6fire, but eeprom bit indicates
it is; change the stored value to disable failing consumer mode.

Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Change an arugment of snd_mpu401_uart_new() to bit flags
Takashi Iwai [Tue, 23 May 2006 11:24:30 +0000 (13:24 +0200)]
[ALSA] Change an arugment of snd_mpu401_uart_new() to bit flags

Change the 5th argument of snd_mpu401_uart_new() to bit flags
instead of a boolean.  The argument takes bits that consist of
MPU401_INFO_XXX flags.
The callers that used the value 1 there are replaced with
MPU401_INFO_INTEGRATED.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hdsp - Fix compilation with hdsp driver built in kernel
Takashi Iwai [Mon, 22 May 2006 12:31:57 +0000 (14:31 +0200)]
[ALSA] hdsp - Fix compilation with hdsp driver built in kernel

Fixed the compilation with hdsp driver built in kernel.
The traditional hwdep loader is used in this case.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Fix description of cs5535audio driver in ALSA-Configuration.txt
Takashi Iwai [Fri, 19 May 2006 17:22:34 +0000 (19:22 +0200)]
[ALSA] Fix description of cs5535audio driver in ALSA-Configuration.txt

Fix the description of cs5535audio driver in ALSA-Configuraiton.txt.
Now it supports only single device.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Single variables for cs5535audio
Jaya Kumar [Fri, 19 May 2006 10:04:22 +0000 (12:04 +0200)]
[ALSA] Single variables for cs5535audio

As per Takashi's feedback, this is a cleanup to make cs5535audio be
single device per system. The diff is against 2.6.17-rc4 with
Takashi's patch adding the module_params for index, id and enable.

Signed-off-by: Jaya Kumar <jayakumar.alsa@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] mpu401_uart - Fix coding style and code clean up
Takashi Iwai [Thu, 18 May 2006 12:48:26 +0000 (14:48 +0200)]
[ALSA] mpu401_uart - Fix coding style and code clean up

- fold lines and fix spaces to follow the standard style
- added some comments
- moved EXPORT_SYMBOL() near the definition
- some code clean up

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] usbaudio - Fix a typo
Takashi Iwai [Thu, 18 May 2006 12:47:03 +0000 (14:47 +0200)]
[ALSA] usbaudio - Fix a typo

Fix a typo introduced by the last fix.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Remove ENTER_UART from au88x0 init
Alan Horstmann [Thu, 18 May 2006 12:24:30 +0000 (14:24 +0200)]
[ALSA] Remove ENTER_UART from au88x0 init

Remove an unnecessary ENTER_UART instruction during au88x0 init as it makes
the first/subsequent midi open to fail.

Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] usb-audio: add workaround for CSR Bluetooth Headphones (Saitek A-250)
Clemens Ladisch [Thu, 18 May 2006 07:35:15 +0000 (09:35 +0200)]
[ALSA] usb-audio: add workaround for CSR Bluetooth Headphones (Saitek A-250)

Some devices don't have the required class-specific endpoint descriptors.
Instead of making this case an error, this patch makes the driver guess the
endpoint attributes.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
18 years ago[ALSA] hda-codec - Fix init verbs for ALC260 hp model
Takashi Iwai [Wed, 17 May 2006 15:17:43 +0000 (17:17 +0200)]
[ALSA] hda-codec - Fix init verbs for ALC260 hp model

Use the basic init verbs for ALC260 instead of hp init verbs
since hp init verbs seem incomplete and not working on some machines.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Remove zero-initialization of static variables
Takashi Iwai [Wed, 17 May 2006 15:14:51 +0000 (17:14 +0200)]
[ALSA] Remove zero-initialization of static variables

Removed zero-initializations of static variables.
A tiny optimization.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Fix pcm-draining of capture stream in PCM middle layer
Takashi Iwai [Wed, 17 May 2006 15:10:35 +0000 (17:10 +0200)]
[ALSA] Fix pcm-draining of capture stream in PCM middle layer

Fix the draining of PCM capture stream in the PCM middle layer.
snd_pcm_drain() ignored capture streams, but it should change
the state to SNDRV_PCM_DRAINING.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Fix description of snd-hda-intel driver in document
Takashi Iwai [Wed, 17 May 2006 09:26:39 +0000 (11:26 +0200)]
[ALSA] Fix description of snd-hda-intel driver in document

Fixed the description of snd-hda-intel driver in ALSA-Configuration.txt
document.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda - Add support for the ATI RS600 HDMI audio device
Felix Kuehling [Wed, 17 May 2006 09:22:21 +0000 (11:22 +0200)]
[ALSA] hda - Add support for the ATI RS600 HDMI audio device

Add support for the ATI RS600 HDMI audio device. It has a one-stream
pure digital stereo codec that isn't handled by the generic codec
support.

Signed-off-by: Felix Kuehling <fkuehlin@ati.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] azt3328.c: use kernel coding style
Andreas Mohr [Wed, 17 May 2006 09:04:19 +0000 (11:04 +0200)]
[ALSA] azt3328.c: use kernel coding style

Scope braces were not done the One True Kernel Way.

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] azt3328.c: add 3D sound mixer switch/rename controls
Andreas Mohr [Wed, 17 May 2006 09:03:16 +0000 (11:03 +0200)]
[ALSA] azt3328.c: add 3D sound mixer switch/rename controls

- add 3D sound pre-3D/post-3D switch, as seen in standard AC-97
- rename controls to shorter and more accurate strings

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] azt3328.c: add suspend/resume support
Andreas Mohr [Wed, 17 May 2006 09:02:24 +0000 (11:02 +0200)]
[ALSA] azt3328.c: add suspend/resume support

- add suspend/resume handlers
- fix problem (private_data members not set)
Playing a file while suspending will resume correctly with this patch,
so I assume the hardware to get fully correctly reinitialized with
this patch.

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Fix rwlock around snd_iprintf() in sound core
Takashi Iwai [Mon, 15 May 2006 17:49:05 +0000 (19:49 +0200)]
[ALSA] Fix rwlock around snd_iprintf() in sound core

Fixed rwlock around snd_iprintf() in sound core part.
Replaced with mutex.
Also, make mutex and flags static variables with addition of
snd_card_locked() function (just for sound.c).

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda-codec - Fix handling of capture controls on ALC882 3/6-stack models
Takashi Iwai [Thu, 11 May 2006 16:12:23 +0000 (18:12 +0200)]
[ALSA] hda-codec - Fix handling of capture controls on ALC882 3/6-stack models

Fixed the handling of capture controls on ALC882 3/6-stack models.
Now the driver checks the availability of NID 07h.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] Remove obsolete description from ALSA-Configuration.txt
Takashi Iwai [Thu, 11 May 2006 09:30:10 +0000 (11:30 +0200)]
[ALSA] Remove obsolete description from ALSA-Configuration.txt

Remove obsolete descriptions about the dependency on CONFIG_ISA
for vxpocket and pdaudiocf drivers.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] hda-codec - Add support for Sony Vaio VGN-S3HP
Takashi Iwai [Wed, 10 May 2006 14:33:11 +0000 (16:33 +0200)]
[ALSA] hda-codec - Add support for Sony Vaio VGN-S3HP

Added the missing support for Sony Vaio VGN-S3HP with ALC260 codec.
The patch taken from ALSA bug#2101.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
18 years ago[ALSA] rme96 - Fix OSS full-duplex
Takashi Iwai [Wed, 10 May 2006 14:24:42 +0000 (16:24 +0200)]
[ALSA] rme96 - Fix OSS full-duplex

Fixed a bug in rme96 driver that the full-duplex on OSS emulation
doesn't work due to the invalid period size parameter.

Signed-off-by: Takashi Iwai <tiwai@suse.de>