[S390] sclp_tty: Fix scheduling while atomic bug.

Finally fixes a possible scheduling while in atomic context bug. The driver
used to wait on a waitqueue if no empty buffer was available. This could
lead to a deadlock if the driver was called from non-schedulable context.
So fix this. The write operation may fail now. It returns the number of
characters accepted. put_char will never fail, since it writes characters
to an intermediate buffer which gets flushed as soon as it is full.
That means the driver now can busy wait if something is in the intermediate
buffer and a write_string operation follows. Seems to be an acceptable
compromise, since that shouldn't happen too often.

Cc: Peter Oberparleiter <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Cc: Martin Schwidefsky <[email protected]>

[S390] sclp_tty: remove ioctl interface.

After all we came to the conclusion that this interface doesn't make any
sense. Besides that the ioctl number used was never registered, the header
file isn't exported, and we doubt there is even a single user.
So remove this interface, since it eases maintenance.

Cc: Peter Oberparleiter <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Cc: Martin Schwidefsky <[email protected]>

[S390] Remove P390 support.

Most likely it is broken anyway because of the changes in memory
detection. Since we can't test it and there are probably better ways
that using a P390 card, remove support for it.

Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] Cleanup vmcp printk messages.

Signed-off-by: Christian Borntraeger <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup lcs printk messages.

Cc: Jeff Garzik <[email protected]>
Signed-off-by: Klaus-D. Wacker <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup kprobes printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup vmwatch printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup dcssblk printk messages.

Signed-off-by: Hongjie Yang <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup zfcp dumper printk messages.

Signed-off-by: Michael Holzheu <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup vmlogrdr printk messages.

The message descriptions are still missing though ..

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup s390 debug feature print messages.

Signed-off-by: Michael Holzheu <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup monreader printk messages.

Signed-off-by: Gerald Schaefer <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup appldata printk messages.

Signed-off-by: Gerald Schaefer <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup smsgiucv printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup cpacf printk messages.

Signed-off-by: Jan Glauber <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup qeth print messages.

Cc: Jeff Garzik <[email protected]>
Signed-off-by: Frank Blaschka <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup netiucv printk messages.

Cc: Jeff Garzik <[email protected]>
Signed-off-by: Ursula Braun <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] Cleanup iucv printk messages.

Cc: David S. Miller <[email protected]>
Signed-off-by: Ursula Braun <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup sclp printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup zcrypt printk messages.

Signed-off-by: Felix Beck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup 3215 printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup 3270 printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup xpram printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup vtime printk messages.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Cleanup cio printk messages.

Unnecessary dev_info, dev_warn and printk messages are removed.

Signed-off-by: Michael Ernst <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Move memory detection code to own file.

Move memory detection code to own file and also simplify it.
Also add an interface which can be called at any time to get the
current memory layout. This interface is needed by our kernel
internal system dumper.

Cc: Peter Oberparleiter <[email protected]>
Cc: Michael Holzheu <[email protected]>
Cc: Frank Munzert <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] cio: fix double unregistering of subchannels

In some cases where the channel subsystem decides to drop a subchannel
device device_unregister may be called twice, which results in an oops.
The patch prevents this by only unregistering registered devices.

Signed-off-by: Sebastian Ott <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] Add support for memory hot-add via sclp.

Cc: Peter Oberparleiter <[email protected]>
Cc: Gerald Schaefer <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] Add support for memory hot-add.

Cc: Gerald Schaefer <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] s390: use memory_read_from_buffer()

Signed-off-by: Akinobu Mita <[email protected]>
Cc: Michael Holzheu <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] s390: use simple_read_from_buffer()

Signed-off-by: Akinobu Mita <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] s390/cio: use memory_read_from_buffer()

Signed-off-by: Akinobu Mita <[email protected]>
Cc: Peter Oberparleiter <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] s390/vmcp: use simple_read_from_buffer()

Signed-off-by: Akinobu Mita <[email protected]>
Cc: Christian Borntraeger <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] drivers/s390: Eliminate NULL test and memset after alloc_bootmem

As noted by Akinobu Mita in patch b1fceac2b9e04d278316b2faddf276015fc06e3b,
alloc_bootmem and related functions never return NULL and always return a
zeroed region of memory.  Thus a NULL test or memset after calls to these
functions is unnecessary.

 drivers/s390/char/raw3270.c  |   11 +----------
 drivers/s390/char/sclp_con.c |    2 --
 2 files changed, 1 insertion(+), 12 deletions(-)

This was fixed using the following semantic patch.
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@@
expression E;
statement S;
@@

E = \(alloc_bootmem\|alloc_bootmem_low\|alloc_bootmem_pages\|alloc_bootmem_low_pages\)(...)
... when != E
(
- BUG_ON (E == NULL);
|
- if (E == NULL) S
)

@@
expression E,E1;
@@

E = \(alloc_bootmem\|alloc_bootmem_low\|alloc_bootmem_pages\|alloc_bootmem_low_pages\)(...)
... when != E
- memset(E,0,E1);
// </smpl>

Signed-off-by: Julia Lawall <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] arch/s390: Eliminate NULL test and memset after alloc_bootmem

As noted by Akinobu Mita in patch b1fceac2b9e04d278316b2faddf276015fc06e3b,
alloc_bootmem and related functions never return NULL and always return a
zeroed region of memory.  Thus a NULL test or memset after calls to these
functions is unnecessary.

 arch/s390/kernel/topology.c |    2 --
 1 file changed, 2 deletions(-)

This was fixed using the following semantic patch.
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@@
expression E;
statement S;
@@

E = \(alloc_bootmem\|alloc_bootmem_low\|alloc_bootmem_pages\|alloc_bootmem_low_pages\)(...)
... when != E
(
- BUG_ON (E == NULL);
|
- if (E == NULL) S
)

@@
expression E,E1;
@@

E = \(alloc_bootmem\|alloc_bootmem_low\|alloc_bootmem_pages\|alloc_bootmem_low_pages\)(...)
... when != E
- memset(E,0,E1);
// </smpl>

Signed-off-by: Julia Lawall <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] Add sched.h include to asm-s390/pgtable.h.

Some macros in pgtable.h access members from struct task_struct.
Currently always works since sched.h seems always to be included
before asm/pgtable.h. Unfortunately that is not anymore true with
Jeremy Fitzhardinge's ptep_modify_prot transaction abstraction patch.
So fix this.

Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] zcrypt: Add additional card IDs to CEX2C and CEX2A

Add support for new micro code load of CEX2C and CEX2A adapters,
which uses different IDs.  This patch just adds the IDs to the
existing drivers.

Signed-off-by: Ralph Wuerthner <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Extra Kernel Parameters via VMPARM

Now it is possible to specify additional kernel parameters on the IPL
command line using the IPL PARM option.
If the Linux system is already running, the new reipl sysfs attribute
'parm' can be used to change kernel parameters for the next reboot.
Examples:
  IPL    C PARM dasd=1234 root=/dev/dasda1
  IPL 1234 PARM savesys=mylnxnss
  echo "init=/bin/bash" > /sys/firmware/reipl/ccw/parm

Signed-off-by: Hendrik Brueckner <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] ap: Use high-resolution timer for polling

The ap poll mechanism is converted to use a high-resolution timer for
polling. This allows more specific polling. With this a new sysfs
attribute is introduced to specify the polling rate in nanoseconds.

Signed-off-by: Felix Beck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] sclp: simplify vt220 cleanup logic

Fix a number of sclp_vt220 cleanup problems:
* fix list_empty check after list_del()
* mark init-only flag as __initdata
* remove implicit dependency between slab_available() and num_pages
* straighten multiple init handling (use init count)

Signed-off-by: Peter Oberparleiter <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] idle: remove idle notifier chain.

The idle notifier chain consists of at most one element. So there's
no point in having a notifier chain. Remove it and directly call the
function.

Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] cio: Add chsc subchannel driver.

This patch adds a driver for subchannels of type chsc.

A device /dev/chsc is created which may be used to issue ioctls to:
- obtain information about the machine's I/O configuration
- dynamically change the machine's I/O configuration via
  asynchronous chsc commands

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: suppress chpid event in case of configure error

Do not send CHP_ONLINE/CHP_OFFLINE events to subchannel drivers when a
channel-path configure request failed.

Signed-off-by: Peter Oberparleiter <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] css: Use css_device_id for bus matching.

css_device_id exists, so use it for determining the right driver
(and add a match_flags which is always 1 for valid types).

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Repair chpid event handling.

Passing the affected chpid in chp_event() worked only by
chance since chpid is the first element in res_acc_data.
Make it work properly by generalizing res_acc_data as
chp_link and always passing around a properly filled out
chp_link structure in chp_event().

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Use isc_{register,unregister}.

Use the new isc registration functions for all places that need
an I/O interruption subclass.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: introduce isc_(un)register functions.

This interface makes it easy for drivers to register usage of different
I/O interruption subclasses without needing to worry about possible
other users of the same isc.

Signed-off-by: Sebastian Ott <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Allow adapter interrupt handlers per isc.

Enhance the adapter interruption API so that device drivers can
register a handler for a specific interruption subclass. This
will allow different device drivers to move to differently
prioritized subclasses in order to avoid congestion.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Introduce abstract isc definitions.

Replace the numeric values for I/O interruption subclass usage
with abstract definitions and collect them all in asm/isc.h.
This gives us a better overview of which iscs are actually used
and makes it possible to better spread out isc usage in the
future.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Get rid of css_characteristics_avail.

Checking for the validity of the css_*_characteristics is
superfluous since they stay 0 for non-successful scsc.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] stp support.

Add support for clock synchronization with the server time protocol.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] initrd vs bss section clearing.

In case the initrd is located within the bss section it will be
overwritten when the section is cleared. To prevent this just move
the initrd right behind the bss section if it starts within the
section.
The current code already moves the initrd if the bootmem allocator
bitmap would overwrite it. With this patch we should be safe against
initrd corruptions.

Cc: Peter Oberparleiter <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] Introduce user_regset accessors for s390

Add the user_regset definitions for normal and compat processes, replace
the dump_regs core dump cruft with the generic CORE_DUMP_USER_REGSET and
replace binfmt_elf32.c with the generic compat_binfmt_elf.c implementation.

Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: remove lock from ccw_device_oper_notify.

Remove unnecessary ccw device locking inside ccw_device_oper_notify.

Signed-off-by: Sebastian Ott <[email protected]>
Acked-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: provide helper functions for fcx enabled I/O

Provide functions which can be used to incrementally construct fcx
enabled I/O control blocks.

Signed-off-by: Peter Oberparleiter <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: provide functions for fcx enabled I/O

Provide functions for assembling and starting fcx enabled I/O request
blocks.

Signed-off-by: Peter Oberparleiter <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: introduce fcx enabled scsw format

Extend the scsw data structure to the format required by fcx. Also
provide helper functions for easier access to fields which are present
in both the traditional as well as the modified format.

Signed-off-by: Peter Oberparleiter <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: introduce fcx bit to chsc characteristics

Introduce fcx bit to chsc characteristics.

Signed-off-by: Peter Oberparleiter <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Base message subchannel handling.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Export some symbols for modular css drivers.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Cleanup crw interface.

Eliminate the need for the machine check handler to call into
the common I/O layer directly by introducing an interface to
register handlers for crws per rsc.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Rework css driver.

Rework the css driver methods to provide sane callbacks for
subchannels of all types.

As a bonus, this cleans up and simplyfies the machine check
handling for I/O subchannels a lot.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Introduce modalias for css bus.

Add modalias and subchannel type attributes for all subchannels.
I/O subchannel specific attributes are now created in
io_subchannel_probe(). modalias and subchannel type are also
added to the uevent for the css bus. Also make the css modalias
known.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Register all subchannels.

Register all valid subchannels, not only I/O subchannels.
Move I/O subchannel specific initialization to io_subchannel_probe().

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] Remove ipldelay kernel parameter.

Using the ipldelay kernel parameter leads to a crash at IPL time.
Since this is broken since a long time it looks like nobody is using
it anymore. So remove it instead of fixing it.

Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] sclp: fix possible deadlock on cpu rescan.

smp_rescan_cpus() calls get_online_cpus() from a multithreaded
workqueue context. This may deadlock. This is the same bug as in
arch/s390/kernel/topology.c. This patch can be reverted as soon as
Oleg's patch gets merged.

Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] make appldata compile w/o CONFIG_SWAP

Avoid compile error by using EXPORT_SYMBOL_GPL(si_swapinfo) only if
CONFIG_SWAP is set.

Signed-off-by: Gerald Schaefer <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] sclp: keep facility mask up to date.

In case the supported sclp facilities change
the new mask should be saved.

Cc: Peter Oberparleiter <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>

[S390] qdio: Repair timeout handling for qdio_shutdown

If qdio shutdown runs in parallel with a channel error,
the qdio_timeout_handler might not be triggered.
In this case neither state INACTIVE nor state ERR
is reached and the following wait_event hangs forever.
Solution: do not make use of ccw_device_set_timeout(),
but add a timeout to the following wait_event.
And make sure, wake_up is called in case of an
i/o error on the qdio-device.

Signed-off-by: Ursula Braun <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Use locks when accessing /sys/firmware/cpi data.

Mutex locks are used to avoid problems when /sys/firmware/cpi data
are accessed to in parallel.

Signed-off-by: Michael Ernst <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

[S390] cio: Clear correct bit in cio_release_console().

Fallout from the console isc 7 -> 1 change.

Signed-off-by: Cornelia Huck <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>

Linux 2.6.26

devcgroup: fix permission check when adding entry to child cgroup

 # cat devices.list
 c 1:3 r
 # echo 'c 1:3 w' > sub/devices.allow
 # cat sub/devices.list
 c 1:3 w

As illustrated, the parent group has no write permission to /dev/null, so
it's child should not be allowed to add this write permission.

Signed-off-by: Li Zefan <[email protected]>
Acked-by: Serge Hallyn <[email protected]>
Cc: Serge Hallyn <[email protected]>
Cc: Paul Menage <[email protected]>
Cc: Pavel Emelyanov <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

devcgroup: always show positive major/minor num

 # echo "b $((0x7fffffff)):$((0x80000000)) rwm" > devices.allow
 # cat devices.list
 b 214748364:-21474836 rwm

though a major/minor number of 0x800000000 is meaningless, we
should not cast it to a negative value.

Signed-off-by: Li Zefan <[email protected]>
Acked-by: Serge Hallyn <[email protected]>
Cc: Serge Hallyn <[email protected]>
Cc: Paul Menage <[email protected]>
Cc: Pavel Emelyanov <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

Documentation/HOWTO: correct wrong kernel bugzilla FAQ URL

Signed-off-by: Jiri Pirko <[email protected]>
Cc: Randy Dunlap <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

Merge branch 'sched-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  cpusets, hotplug, scheduler: fix scheduler domain breakage

cpusets, hotplug, scheduler: fix scheduler domain breakage

Commit f18f982ab ("sched: CPU hotplug events must not destroy scheduler
domains created by the cpusets") introduced a hotplug-related problem as
described below:

Upon CPU_DOWN_PREPARE,

  update_sched_domains() -> detach_destroy_domains(&cpu_online_map)

does the following:

/*
 * Force a reinitialization of the sched domains hierarchy. The domains
 * and groups cannot be updated in place without racing with the balancing
 * code, so we temporarily attach all running cpus to the NULL domain
 * which will prevent rebalancing while the sched domains are recalculated.
 */

The sched-domains should be rebuilt when a CPU_DOWN ops. has been
completed, effectively either upon CPU_DEAD{_FROZEN} (upon success) or
CPU_DOWN_FAILED{_FROZEN} (upon failure -- restore the things to their
initial state). That's what update_sched_domains() also does but only
for !CPUSETS case.

With f18f982ab, sched-domains' reinitialization is delegated to
CPUSETS code:

cpuset_handle_cpuhp() -> common_cpu_mem_hotplug_unplug() ->
rebuild_sched_domains()

Being called for CPU_UP_PREPARE and if its callback is called after
update_sched_domains()), it just negates all the work done by
update_sched_domains() -- i.e. a soon-to-be-offline cpu is included in
the sched-domains and that makes it visible for the load-balancer
while the CPU_DOWN ops. is in progress.

__migrate_live_tasks() moves the tasks off a 'dead' cpu (it's already
"offline" when this function is called).

try_to_wake_up() is called for one of these tasks from another CPU ->
the load-balancer (wake_idle()) picks up a "dead" CPU and places the
task on it. Then e.g. BUG_ON(rq->nr_running) detects this a bit later
-> oops.

Signed-off-by: Dmitry Adamushko <[email protected]>
Tested-by: Vegard Nossum <[email protected]>
Cc: Paul Menage <[email protected]>
Cc: Max Krasnyansky <[email protected]>
Cc: Paul Jackson <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: Linus Torvalds <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>

Merge branch 'x86-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86: fix ldt limit for 64 bit

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] bsg: fix oops on remove
  [SCSI] fusion: default MSI to disabled for SPI and FC controllers
  [SCSI] ipr: Fix HDIO_GET_IDENTITY oops for SATA devices
  [SCSI] mptspi: fix oops in mptspi_dv_renegotiate_work()
  [SCSI] erase invalid data returned by device

cifs: fix wksidarr declaration to be big-endian friendly

The current definition of wksidarr works fine on little endian arches
(since cpu_to_le32 is a no-op there), but on big-endian arches, it fails
to compile with this error:

error: braced-group within expression allowed only inside a function

The problem is that this static declaration has cpu_to_le32 embedded
within it, and that expands into a function macro.  We need to use
__constant_cpu_to_le32() instead.

Signed-off-by: Jeff Layton <[email protected]>
Cc: Steven French <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

cifs: fix inode leak in cifs_get_inode_info_unix

Try this:

    mount a share with unix extensions
    create a file on it
    umount the share

You'll get the following message in the ring buffer:

VFS: Busy inodes after unmount of cifs. Self-destruct in 5 seconds.  Have a
nice day...

...the problem is that cifs_get_inode_info_unix is creating and hashing
a new inode even when it's going to return error anyway. The first
lookup when creating a file returns an error so we end up leaking this
inode before we do the actual create. This appears to be a regression
caused by commit 0e4bbde94fdc33f5b3d793166b21bf768ca3e098.

The following patch seems to fix it for me, and fixes a minor
formatting nit as well.

Signed-off-by: Jeff Layton <[email protected]>
Acked-by: Steven French <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

frv: fix irqs_disabled() to return an int, not an unsigned long

Fix FRV irqs_disabled() to return an int, not an unsigned long to avoid
this warning:

kernel/sched.c: In function '__might_sleep':
kernel/sched.c:8198: warning: format '%d' expects type 'int', but argument 3 has type 'long unsigned int'

Signed-off-by: David Howells <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

OProfile kernel maintainership changes

Cc: Philippe Elie <[email protected]>
Cc: John Levon <[email protected]>
Cc: Maynard Johnson <[email protected]>
Cc: Richard Purdie <[email protected]>
Cc: Daniel Hansel <[email protected]>
Cc: Jason Yeh <[email protected]>
Cc: Andrew Morton <[email protected]>
Signed-off-by: Robert Richter <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

rtc-pcf8563: add chip id

Add the rtc8564 chip entry

Signed-off-by: Jon Smirl <[email protected]>
Signed-off-by: Alessandro Zummo <[email protected]>
Cc: David Brownell <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

rtc-fm3130: fix chip naming

Fix chip naming from fm3031-rtc to fm3031

Signed-off-by: Alessandro Zummo <[email protected]>
Cc: Sergey Lapin <[email protected]>
Cc: David Brownell <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

ov7670: clean up ov7670_read semantics

Cortland Setlow pointed out a bug in ov7670.c where the result from
ov7670_read() was just being checked for !0, rather than <0.  This made me
realize that ov7670_read's semantics were rather confusing; it both fills
in 'value' with the result, and returns it.  This is goes against general
kernel convention; so rather than fixing callers, let's fix the function.

This makes ov7670_read return <0 in the case of an error, and 0 upon
success. Thus, code like:

res = ov7670_read(...);
if (!res)
goto error;

..will work properly.

Signed-off-by: Cortland Setlow <[email protected]>
Signed-off-by: Andres Salomon <[email protected]>
Acked-by: Jonathan Corbet <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

serial8250: sanity check nr_uarts on all paths.

I had 8250.nr_uarts=16 in the boot line of a test kernel and I had a weird
mysterious crash in sysfs.  After taking an in-depth look I realized that
CONFIG_SERIAL_8250_NR_UARTS was set to 4 and I was walking off the end of
the serial8250_ports array.

Ouch!!!

Don't let this happen to someone else.

Signed-off-by: Eric W. Biederman <[email protected]>
Acked-by: Alan Cox <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

fbdev: bugfix for multiprocess defio

This patch is a bugfix for how defio handles multiple processes manipulating
the same framebuffer.

Thanks to Bernard Blackham for identifying this bug.

It occurs when two applications mmap the same framebuffer and concurrently
write to the same page.  Normally, this doesn't occur since only a single
process mmaps the framebuffer.  The symptom of the bug is that the mapping
applications will hang.  The cause is that defio incorrectly tries to add the
same page twice to the pagelist.  The solution I have is to walk the pagelist
and check for a duplicate before adding.  Since I needed to walk the pagelist,
I now also keep the pagelist in sorted order.

Signed-off-by: Jaya Kumar <[email protected]>
Cc: Bernard Blackham <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

drivers/isdn/i4l/isdn_common.c fix small resource leak

Coverity CID: 1356 RESOURCE_LEAK

I found a very old patch for this that was Acked but did not get applied
https://lists.linux-foundation.org/pipermail/kernel-janitors/2006-September/016362.html

There looks to be a small leak in isdn_writebuf_stub() in isdn_common.c, when
copy_from_user() returns an un-copied data length (length != 0).  The below
patch should be a minimally invasive fix.

Signed-off-by: Darren Jenkins <darrenrjenkins@gmailcom>
Acked-by: Karsten Keil <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

drivers/char/pcmcia/ipwireless/hardware.c fix resource leak

Coverity CID: 2172 RESOURCE_LEAK

When pool_allocate() tries to enlarge a packet, if it can not allocate enough
memory, it returns NULL without first freeing the old packet.

This patch just frees the packet first.

Signed-off-by: Darren Jenkins <[email protected]>
Acked-by: Jiri Kosina <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

[SCSI] bsg: fix oops on remove

If you do a modremove of any sas driver, you run into an oops on
shutdown when the host is removed (coming from the host bsg device).
The root cause seems to be that there's a use after free of the
bsg_class_device:  In bsg_kref_release_function, this is used (to do a
put_device(bcg->parent) after bcg->release has been called.  In sas (and
possibly many other things) bcd->release frees the queue which contains
the bsg_class_device, so we get a put_device on unreferenced memory.
Fix this by taking a copy of the pointer to the parent before releasing
bsg.

Acked-by: FUJITA Tomonori <[email protected]>
Signed-off-by: James Bottomley <[email protected]>

[SCSI] fusion: default MSI to disabled for SPI and FC controllers

There's a fault on the FC controllers that makes them not respond
correctly to MSI.  The SPI controllers are fine, but are likely to be
onboard on older motherboards which don't handle MSI correctly, so
default both these cases to disabled.  Enable by setting the module
parameter mpt_msi_enable=1.

For the SAS case, enable MSI by default, but it can be disabled by
setting the module parameter mpt_msi_enable=0.

Cc: "Prakash, Sathya" <[email protected]>
Signed-off-by: James Bottomley <[email protected]>

x86: fix ldt limit for 64 bit

Fix size of LDT entries. On x86-64, ldt_desc is a double-sized descriptor.

Signed-off-by: Michael Karcher <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>

Merge git://git./linux/kernel/git/wim/linux-2.6-watchdog

* git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog:
  [PATCH] IPMI: return correct value from ipmi_write

[PATCH] IPMI: return correct value from ipmi_write

This patch corrects the handling of write operations to the IPMI watchdog
to work as intended by returning the number of characters actually
processed. Without this patch, an "echo V >/dev/watchdog" enables the
watchdog if IPMI is providing the watchdog function.

Signed-off-by: Mark Rustad <[email protected]>
Signed-off-by: Corey Minyard <[email protected]>
Signed-off-by: Wim Van Sebroeck <[email protected]>

[SCSI] ipr: Fix HDIO_GET_IDENTITY oops for SATA devices

Currently, ipr does not support HDIO_GET_IDENTITY to SATA devices.
An oops occurs if userspace attempts to send the command. Since hald
issues the command, ensure we fail the ioctl in ipr. This is a
temporary solution to the oops. Once the ipr libata EH conversion
is upstream, ipr will fully support HDIO_GET_IDENTITY.

Tested-by: Milton Miller <[email protected]>
Signed-off-by: Brian King <[email protected]>
Signed-off-by: James Bottomley <[email protected]>

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:
  libata-acpi: don't call sleeping function from invalid context
  Added Targa Visionary 1000 IDE adapter to pata_sis.c
  libata-acpi: filter out DIPM enable

Fix reference counting race on log buffers

When we release the iclog, we do an atomic_dec_and_lock to determine if
we are the last reference and need to trigger update of log headers and
writeout.  However, in xlog_state_get_iclog_space() we also need to
check if we have the last reference count there.  If we do, we release
the log buffer, otherwise we decrement the reference count.

But the compare and decrement in xlog_state_get_iclog_space() is not
atomic, so both places can see a reference count of 2 and neither will
release the iclog.  That leads to a filesystem hang.

Close the race by replacing the atomic_read() and atomic_dec() pair with
atomic_add_unless() to ensure that they are executed atomically.

Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Tim Shimmin <[email protected]>
Tested-by: Eric Sandeen <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

libata-acpi: don't call sleeping function from invalid context

The problem is introduced by commit
664d080c41463570b95717b5ad86e79dc1be0877.

acpi_evaluate_integer is a sleeping function,
and it should not be called with spin_lock_irqsave.
https://bugzilla.redhat.com/show_bug.cgi?id=451399

Signed-off-by: Zhang Rui <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>

Added Targa Visionary 1000 IDE adapter to pata_sis.c

This enables short 40-wire detection for my laptop thus
enabling UDMA/100.

Signed-off-by: Jeff Garzik <[email protected]>