Randy Dunlap [Wed, 11 Aug 2010 21:17:30 +0000 (14:17 -0700)]
pc8736x_gpio: depends on X86_32
Fix kconfig dependency warning for PC8736x_GPIO by restricting it to
X86_32.
warning: (SCx200_GPIO && SCx200 || PC8736x_GPIO && X86) selects NSC_GPIO which has unmet direct dependencies (X86_32)
NSC_GPIO is X86_32 only. The other driver (SCx200_GPIO) that selects
NSC_GPIO is X86_32 only (indirectly, since SCx200 depends on X86_32), so
limit this driver also.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: Jordan Crouse <jordan.crouse@amd.com> Cc: Jim Cromie <jim.cromie@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Randy Dunlap [Wed, 11 Aug 2010 21:17:30 +0000 (14:17 -0700)]
mm: fix fatal kernel-doc error
Fix a fatal kernel-doc error due to a #define coming between a function's
kernel-doc notation and the function signature. (kernel-doc cannot handle
this)
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Thomas Gleixner [Wed, 11 Aug 2010 21:17:29 +0000 (14:17 -0700)]
acpi: fix bogus preemption logic
The ACPI_PREEMPTION_POINT() logic was introduced in commit 8bd108d
(ACPICA: add preemption point after each opcode parse). The follow up
commits abe1dfab6, 138d15692, c084ca70 tried to fix the preemption logic
back and forth, but nobody noticed that the usage of
in_atomic_preempt_off() in that context is wrong.
The check which guards the call of cond_resched() is:
if (!in_atomic_preempt_off() && !irqs_disabled())
in_atomic_preempt_off() is not intended for general use as the comment
above the macro definition clearly says:
* Check whether we were atomic before we did preempt_disable():
* (used by the scheduler, *after* releasing the kernel lock)
On a CONFIG_PREEMPT=n kernel the usage of in_atomic_preempt_off() works by
accident, but with CONFIG_PREEMPT=y it's just broken.
The whole purpose of the ACPI_PREEMPTION_POINT() is to reduce the latency
on a CONFIG_PREEMPT=n kernel, so make ACPI_PREEMPTION_POINT() depend on
CONFIG_PREEMPT=n and remove the in_atomic_preempt_off() check.
[akpm@linux-foundation.org: fix build] Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Len Brown <lenb@kernel.org> Cc: Francois Valenduc <francois.valenduc@tvcablenet.be> Cc: Lin Ming <ming.m.lin@intel.com> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Stefani Seibold [Wed, 11 Aug 2010 21:17:27 +0000 (14:17 -0700)]
kernel/kfifo.c: add handling of chained scatterlists
The current kfifo scatterlist implementation will not work with chained
scatterlists. It assumes that struct scatterlist arrays are allocated
contiguously, which is not the case when chained scatterlists (struct
sg_table) are in use.
Axel Lin [Mon, 9 Aug 2010 06:52:16 +0000 (14:52 +0800)]
mfd: Fix incorrect kfree(i2c) in wm8994-core i2c_driver probe
The i2c_client received in probe() should not be kfree()'d.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Axel Lin [Mon, 9 Aug 2010 06:51:15 +0000 (14:51 +0800)]
mfd: Fix incorrect kfree(i2c) in wm831x-core i2c_driver probe
The i2c_client received in probe() should not be kfree()'d.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Mike Rapoport [Tue, 10 Aug 2010 23:11:04 +0000 (01:11 +0200)]
mfd: Add TPS6586x driver
Add mfd core driver for TPS6586x PMICs family.
The driver provides I/O access for the sub-device drivers and performs
regstration of the sub-devices based on the platform requirements.
In addition it implements GPIOlib interface for the chip GPIOs.
TODO:
- add interrupt support
- add platform data for PWM, backlight leds and charger
Signed-off-by: Mike Rapoport <mike@compulab.co.il> Signed-off-by: Mike Rapoport <mike.rapoport@gmail.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Jarkko Nikula [Sun, 8 Aug 2010 17:05:24 +0000 (20:05 +0300)]
mfd: Use macros instead of some constant magic numbers for menelaus
This patch is originally done by Carlos Eduardo Aguiar. Original fix is
commit 3305829b2816072b9c8ed01374b205ae4de74027 in
git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6.git
Author modified the fix for mainline version of menelaus.
Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Carlos Eduardo Aguiar <carlos.aguiar@indt.org.br> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Jarkko Nikula [Sun, 8 Aug 2010 17:05:23 +0000 (20:05 +0300)]
mfd: Fix menelaus mmc slot 2 misconfiguration
We are modifying register value instead of return value.
This fix is originally done by Carlos Eduardo Aguiar. Original fix is
commit bb4e91722e29efe31587d2cc664b6def645aecd9 in
git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6.git
Author modified the fix for mainline version of menelaus.
Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Carlos Eduardo Aguiar <carlos.aguiar@indt.org.br> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Axel Lin [Wed, 4 Aug 2010 01:54:21 +0000 (09:54 +0800)]
mfd: Fix wrong wm8350-core kfree in error path
This patch includes below fixes:
1. fix wm8350_create_cache error path
make sure wm8350->reg_cache is freed in error path.
2. fix wm8350_device_init error path
no need to kfree(wm8350->reg_cache) in the case of goto out.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Axel Lin [Wed, 4 Aug 2010 01:44:47 +0000 (09:44 +0800)]
mfd: Fix wm8994_device_init() return value
wm8994_device_init() will return 0 in the case of kzalloc fail
in current implementation.
This patch fixes the return value.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Axel Lin [Tue, 3 Aug 2010 08:34:13 +0000 (16:34 +0800)]
mfd: Avoid calling platform_device_put() twice in ucb1400 probe error path
In the case of goto err2, what we want is to call
platform_device_del() instead of platform_device_unregister().
Otherwise, we call platform_device_put() twice.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Mark Brown [Wed, 21 Jul 2010 13:23:37 +0000 (14:23 +0100)]
mfd: Add WM8321 support
The WM8321 is a PMIC for low power, high performance applications. From a
software point of view the device is identical to the WM8320, all the
differences between the two devices are visible only in hardware.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Hemanth V [Wed, 14 Jul 2010 10:04:56 +0000 (12:04 +0200)]
mfd: Add support for TWL6030 PWM
TWL6030 supports PWM (Pulse Width Modulator) which is used
to control charging LED. PWM allows for controlling brightness.
This patch implements the APIs required by leds-pwm driver.
Signed-off-by: Hemanth V <hemanthv@ti.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
This patch adds a MFD driver for the JZ4740 ADC unit. The driver is used to
demultiplex IRQs and synchronize access to shared registers between the
battery, hwmon and (future) touchscreen driver.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Rabin Vincent [Fri, 2 Jul 2010 11:22:09 +0000 (16:52 +0530)]
gpio: Add STMPE GPIO driver
Add support for the GPIOs on STMPE I/O Expanders.
[l.fu@pengutronix.de: fix set direction input]
[l.fu@pengutronix.de: set GPIO alternate function while requesting] Acked-by: Luotao Fu <l.fu@pengutronix.de> Acked-by: Linus Walleij <linus.walleij@stericsson.com> Signed-off-by: Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Rabin Vincent [Fri, 2 Jul 2010 11:22:08 +0000 (16:52 +0530)]
mfd: Add STMPE I/O Expander support
Add support for the STMPE family of I/O Expanders from
STMicroelectronics. These devices include upto 24 gpios and a varying
selection of blocks, including PWM, keypad, and touchscreen controllers.
This patch adds the MFD core.
[l.fu@pengutronix.de: fix stmpe811 enable hook]
[l.fu@pengutronix.de: add touchscreen platform data] Acked-by: Luotao Fu <l.fu@pengutronix.de> Acked-by: Linus Walleij <linus.walleij@stericsson.com> Signed-off-by: Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Axel Lin [Thu, 1 Jul 2010 09:43:50 +0000 (17:43 +0800)]
mfd: Properly free t7l66xb clk32k clock source
This patch includes below fixes to properly free clk32k clock source:
1. remove a redundant clk_put in t7l66xb_probe error path
2. add missing clk_disable(t7l66xb->clk32k) and clk_put(t7l66xb->clk32k)
to properly free the clock source.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Samuel Ortiz [Sat, 19 Jun 2010 00:29:24 +0000 (02:29 +0200)]
mfd: More verbose MFD Kconfig entry
For people to be able to intellingibly decide if they want to enable MFD
drivers or not, we have to give them a much better description of what they
are.
Mark Brown [Wed, 9 Jun 2010 13:54:54 +0000 (14:54 +0100)]
mfd: Staticise ab3550 register access functions
These are now exported via an ops table rather than referenced
directly and so should be staticised.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Linus Walleij <linus.walleij@stericsson.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Linus Walleij [Fri, 28 May 2010 00:59:44 +0000 (02:59 +0200)]
mfd: AB8500 mask off irrelevant bits from the SPI message
The registers on the AB8500 are only 8 bits wide, so the content
of the remaining bits is undefined. Let's mask off the undefined
stuff when returning a register in an SPI read.
Acked-by: Rabin Vincent <rabin.vincent@stericsson.com> Signed-off-by: Linus Walleij <linus.walleij@stericsson.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Dan Carpenter [Wed, 26 May 2010 22:54:09 +0000 (00:54 +0200)]
mfd: Fix 88pm860x uninitialized variable and clean up
The original code had a compile warning:
drivers/mfd/88pm860x-core.c:431: warning: ‘ret’ may be used
uninitialized in this function
It seems like the warning is valid if either pdata or pdata->touch is
NULL.
This patch checks pdata and pdata->touch at the beginning of the
function. That means everything can be pulled in one indent level.
Now all the statements fit within the 80 character limit.
Also at that point the "use_gpadc" variable isn't needed and removing
it simplifies the logic.
Signed-off-by: Dan Carpenter <error27@gmail.com> Acked-by: Haojian Zhuang <hzhuang1@marvell.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Axel Lin [Tue, 25 May 2010 06:49:51 +0000 (14:49 +0800)]
mfd: Fix memory leak in ab3100_otp_probe
In current implementation, there is a memory leak if ab3100_otp_read fail.
And in the case of ab3100_otp_init_debugfs fail, it does not properly remove
sysfs entries.
This patch properly handle above failure cases.
Signed-off-by: Axel Lin <axel.lin@gmail.com> Acked-by: Linus Walleij <linus.walleij@stericsson.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Sam Ravnborg [Thu, 12 Aug 2010 07:11:52 +0000 (09:11 +0200)]
kconfig: fix tristate choice with minimal config
If a minimal config did not specify the value
of all choice values, the resulting configuration
could have wrong values.
Consider following example:
config M
def_bool y
option modules
choice
prompt "choice list"
config A
tristate "a"
config B
tristate "b"
endchoice
With a defconfig like this:
CONFIG_M=y
CONFIG_A=y
The resulting configuration would have
CONFIG_A=m
which was unexpected.
The problem was not not all choice values were set and thus
kconfig calculated a wrong value.
The fix is to set all choice values when we
read a defconfig files.
conf_set_all_new_symbols() is refactored such that
random choice values are now handled by a dedicated function.
And new choice values are set by set_all_choice_values().
This was not the minimal fix, but the fix that resulted
in the most readable code.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Reported-by: Arve Hjønnevåg <arve@android.com> Tested-by: Arve Hjønnevåg <arve@android.com> Signed-off-by: Michal Marek <mmarek@suse.cz>
Randy Dunlap [Wed, 11 Aug 2010 13:26:57 +0000 (13:26 +0000)]
phylib: available for any speed ethernet
Several gigabit network drivers (SB1250_MAC, TIGON3, FSL, GIANFAR,
UCC_GETH, MV643XX_ETH, XILINX_LL_TEMAC, S6GMAC, STMMAC_ETH, PASEMI_MAC,
and OCTEON_ETHERNET) select PHYLIB. These drivers are not under
NET_ETHERNET (10/100 mbit), so this warning is generated (long, irrelevant
parts are omitted):
Mike Snitzer [Thu, 12 Aug 2010 03:14:32 +0000 (04:14 +0100)]
dm mpath: support discard
Enable discard support in the DM multipath target.
This discard support depends on a few discard-specific fixes to the
block layer's request stacking driver methods.
Discard requests are optional so don't allow a failed discard to trigger
path failures. If there is a real problem with a given path the
barriers associated with the discard (either before or after the
discard) will cause path failure. That said, unconditionally passing
discard failures up the stack is not ideal. This must be fixed once DM
has more information about the nature of the underlying storage failure.
Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com> Cc: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Mikulas Patocka [Thu, 12 Aug 2010 03:14:26 +0000 (04:14 +0100)]
dm stripe: support discards
The DM core will submit a discard bio to the stripe target for each
stripe in a striped DM device. The stripe target will determine
stripe-specific portions of the supplied bio to be remapped into
individual (at most 'num_discard_requests' extents). If a given
stripe-specific discard bio doesn't touch a particular stripe the bio
will be dropped.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mike Snitzer [Thu, 12 Aug 2010 03:14:24 +0000 (04:14 +0100)]
dm: split discard requests on target boundaries
Update __clone_and_map_discard to loop across all targets in a DM
device's table when it processes a discard bio. If a discard crosses a
target boundary it must be split accordingly.
Update __issue_target_requests and __issue_target_request to allow a
cloned discard bio to have a custom start sector and size.
Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mike Snitzer [Thu, 12 Aug 2010 03:14:10 +0000 (04:14 +0100)]
dm: factor out max_io_len_target_boundary
Split max_io_len_target_boundary out of max_io_len so that the discard
support can make use of it without duplicating max_io_len code.
Avoiding max_io_len's split_io logic enables DM's discard support to
submit the entire discard request to a target. But discards must still
be split on target boundaries.
Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mike Snitzer [Thu, 12 Aug 2010 03:14:08 +0000 (04:14 +0100)]
dm: linear support discard
Allow discards to be passed through to linear mappings if at least one
underlying device supports it. Discards will be forwarded only to
devices that support them.
A target that supports discards should set num_discard_requests to
indicate how many times each discard request must be submitted to it.
Verify table's underlying devices support discards prior to setting the
associated DM device as capable of discards (via QUEUE_FLAG_DISCARD).
Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Reviewed-by: Joe Thornber <thornber@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Milan Broz [Thu, 12 Aug 2010 03:14:07 +0000 (04:14 +0100)]
dm crypt: simplify crypt_ctr
Allocate cipher strings indpendently of struct crypt_config and move
cipher parsing and allocation into a separate function to prepare for
supporting the cryptoapi format e.g. "xts(aes)".
No functional change in this patch.
Signed-off-by: Milan Broz <mbroz@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Will Drewry [Thu, 12 Aug 2010 03:14:03 +0000 (04:14 +0100)]
dm ioctl: refactor dm_table_complete
This change unifies the various checks and finalization that occurs on a
table prior to use. By doing so, it allows table construction without
traversing the dm-ioctl interface.
Signed-off-by: Will Drewry <wad@chromium.org> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mikulas Patocka [Thu, 12 Aug 2010 03:14:02 +0000 (04:14 +0100)]
dm snapshot: implement merge
Implement merge method for the snapshot origin to improve read
performance.
Without merge method, dm asks the upper layers to submit smallest possible
bios --- one page. Submitting such small bios impacts performance negatively
when reading or writing the origin device.
Without this patch, CPU consumption when reading the origin on lvm on md-raid0
was 6 to 12%, with this patch, it drops to 1 to 4%.
Note: in my testing, it actually degraded performance in some settings, I
traced it to Maxtor disks having problems with > 512-sector requests.
Reducing the number of sectors to /sys/block/sd*/queue/max_sectors_kb to
256 fixed the read performance. I think we don't have to care about weird
disks that actually degrade performance because of large requests being
sent to them.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mike Snitzer [Thu, 12 Aug 2010 03:14:02 +0000 (04:14 +0100)]
dm: do not initialise full request queue when bio based
Change bio-based mapped devices no longer to have a fully initialized
request_queue (request_fn, elevator, etc). This means bio-based DM
devices no longer register elevator sysfs attributes ('iosched/' tree
or 'scheduler' other than "none").
In contrast, a request-based DM device will continue to have a full
request_queue and will register elevator sysfs attributes. Therefore
a user can determine a DM device's type by checking if elevator sysfs
attributes exist.
First allocate a minimalist request_queue structure for a DM device
(needed for both bio and request-based DM).
Initialization of a full request_queue is deferred until it is known
that the DM device is request-based, at the end of the table load
sequence.
Factor DM device's request_queue initialization:
- common to both request-based and bio-based into dm_init_md_queue().
- specific to request-based into dm_init_request_based_queue().
The md->type_lock mutex is used to protect md->queue, in addition to
md->type, during table_load().
A DM device's first table_load will establish the immutable md->type.
But md->queue initialization, based on md->type, may fail at that time
(because blk_init_allocated_queue cannot allocate memory). Therefore
any subsequent table_load must (re)try dm_setup_md_queue independently of
establishing md->type.
Signed-off-by: Mike Snitzer <snitzer@redhat.com> Acked-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mike Snitzer [Thu, 12 Aug 2010 03:14:01 +0000 (04:14 +0100)]
dm ioctl: make bio or request based device type immutable
Determine whether a mapped device is bio-based or request-based when
loading its first (inactive) table and don't allow that to be changed
later.
This patch performs different device initialisation in each of the two
cases. (We don't think it's necessary to add code to support changing
between the two types.)
Allowed md->type transitions:
DM_TYPE_NONE to DM_TYPE_BIO_BASED
DM_TYPE_NONE to DM_TYPE_REQUEST_BASED
We now prevent table_load from replacing the inactive table with a
conflicting type of table even after an explicit table_clear.
Introduce 'type_lock' into the struct mapped_device to protect md->type
and to prepare for the next patch that will change the queue
initialization and allocate memory while md->type_lock is held.
Mikulas Patocka [Thu, 12 Aug 2010 03:14:00 +0000 (04:14 +0100)]
dm: skip second flush on bio unsupported error
When processing barriers, skip the second flush if processing the bio
failed with -EOPNOTSUPP. This can happen with discard+barrier requests.
If the device doesn't support discard, there would be two useless
SYNCHRONIZE CACHE commands. The first dm_flush cannot be so easily
optimized out, so we leave it there.
Previously, -EOPNOTSUPP could be received in dec_pending only with empty
barriers and we ignored that error, assuming the device not supporting
cache flushes has cache always consistent. With the addition of discard
barriers, this -EOPNOTSUPP can also be generated by discards and we
must record it in md->barrier_error for process_barrier.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Tomohiro Kusumi [Thu, 12 Aug 2010 03:13:59 +0000 (04:13 +0100)]
dm snapshot: persistent use define for disk header chunk size
This patch fixes hard-coded value for the size of a chunk that includes
disk header for persistent snapshot. It should be changed to existing
macro NUM_SNAPSHOT_HDR_CHUNKS instead of using hard-coded value 1.
Arnd Bergmann [Thu, 12 Aug 2010 03:13:57 +0000 (04:13 +0100)]
dm ioctl: use nonseekable_open
The dm control device does not implement read/write, so it has no use for
seeking. Using no_llseek prevents falling back to default_llseek, which
requires the BKL.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Kiyoshi Ueda [Thu, 12 Aug 2010 03:13:56 +0000 (04:13 +0100)]
dm: separate device deletion from dm_put
This patch separates the device deletion code from dm_put()
to make sure the deletion happens in the process context.
By this patch, device deletion always occurs in an ioctl (process)
context and dm_put() can be called in interrupt context.
As a result, the request-based dm's bad dm_put() usage pointed out
by Mikulas below disappears.
http://marc.info/?l=dm-devel&m=126699981019735&w=2
Without this patch, I confirmed there is a case to crash the system:
dm_put() => dm_table_destroy() => vfree() => BUG_ON(in_interrupt())
Some more backgrounds and details:
In request-based dm, a device opener can remove a mapped_device
while the last request is still completing, because bios in the last
request complete first and then the device opener can close and remove
the mapped_device before the last request completes:
CPU0 CPU1
=================================================================
<<INTERRUPT>>
blk_end_request_all(clone_rq)
blk_update_request(clone_rq)
bio_endio(clone_bio) == end_clone_bio
blk_update_request(orig_rq)
bio_endio(orig_bio)
<<I/O completed>>
dm_blk_close()
dev_remove()
dm_put(md)
<<Free md>>
blk_finish_request(clone_rq)
....
dm_end_request(clone_rq)
free_rq_clone(clone_rq)
blk_end_request_all(orig_rq)
rq_completed(md)
So request-based dm used dm_get()/dm_put() to hold md for each I/O
until its request completion handling is fully done.
However, the final dm_put() can call the device deletion code which
must not be run in interrupt context and may cause kernel panic.
To solve the problem, this patch moves the device deletion code,
dm_destroy(), to predetermined places that is actually deleting
the mapped_device in ioctl (process) context, and changes dm_put()
just to decrement the reference count of the mapped_device.
By this change, dm_put() can be used in any context and the symmetric
model below is introduced:
dm_create(): create a mapped_device
dm_destroy(): destroy a mapped_device
dm_get(): increment the reference count of a mapped_device
dm_put(): decrement the reference count of a mapped_device
dm_destroy() waits for all references of the mapped_device to disappear,
then deletes the mapped_device.
dm_destroy() uses active waiting with msleep(1), since deleting
the mapped_device isn't performance-critical task.
And since at this point, nobody opens the mapped_device and no new
reference will be taken, the pending counts are just for racing
completing activity and will eventually decrease to zero.
For the unlikely case of the forced module unload, dm_destroy_immediate(),
which doesn't wait and forcibly deletes the mapped_device, is also
introduced and used in dm_hash_remove_all(). Otherwise, "rmmod -f"
may be stuck and never return.
And now, because the mapped_device is deleted at this point, subsequent
accesses to the mapped_device may cause NULL pointer references.
Kiyoshi Ueda [Thu, 12 Aug 2010 03:13:55 +0000 (04:13 +0100)]
dm ioctl: release _hash_lock between devices in remove_all
This patch changes dm_hash_remove_all() to release _hash_lock when
removing a device. After removing the device, dm_hash_remove_all()
takes _hash_lock and searches the hash from scratch again.
This patch is a preparation for the next patch, which changes device
deletion code to wait for md reference to be 0. Without this patch,
the wait in the next patch may cause AB-BA deadlock:
CPU0 CPU1
-----------------------------------------------------------------------
dm_hash_remove_all()
down_write(_hash_lock)
table_status()
md = find_device()
dm_get(md)
<increment md->holders>
dm_get_live_or_inactive_table()
dm_get_inactive_table()
down_write(_hash_lock)
<in the md deletion code>
<wait for md->holders to be 0>
Kiyoshi Ueda [Thu, 12 Aug 2010 03:13:54 +0000 (04:13 +0100)]
dm: prevent access to md being deleted
This patch prevents access to mapped_device which is being deleted.
Currently, even after a mapped_device has been removed from the hash,
it could be accessed through idr_find() using minor number.
That could cause a race and NULL pointer reference below:
CPU0 CPU1
------------------------------------------------------------------
dev_remove(param)
down_write(_hash_lock)
dm_lock_for_deletion(md)
spin_lock(_minor_lock)
set_bit(DMF_DELETING)
spin_unlock(_minor_lock)
__hash_remove(hc)
up_write(_hash_lock)
dev_status(param)
md = find_device(param)
down_read(_hash_lock)
__find_device_hash_cell(param)
dm_get_md(param->dev)
md = dm_find_md(dev)
spin_lock(_minor_lock)
md = idr_find(MINOR(dev))
spin_unlock(_minor_lock)
dm_put(md)
free_dev(md)
dm_get(md)
up_read(_hash_lock)
__dev_status(md, param)
dm_put(md)
Peter Rajnoha [Thu, 12 Aug 2010 03:13:53 +0000 (04:13 +0100)]
dm ioctl: return uevent flag after rename
All the dm ioctls that generate uevents set the DM_UEVENT_GENERATED flag so
that userspace knows whether or not to wait for a uevent to be processed
before continuing,
The dm rename ioctl sets this flag but was not structured to return it
to userspace. This patch restructures the rename ioctl processing to
behave like the other ioctls that return data and so fix this.
Signed-off-by: Peter Rajnoha <prajnoha@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Peter Rajnoha [Thu, 12 Aug 2010 03:13:52 +0000 (04:13 +0100)]
dm ioctl: remove __dev_status from geometry and target message
Remove useless __dev_status call while processing an ioctl that sets up
device geometry and target message. The data is not returned to
userspace so there is no point collecting it and in the case of
target_message it is collected before processing the message so if it
did return it might be stale.
Signed-off-by: Peter Rajnoha <prajnoha@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mikulas Patocka [Thu, 12 Aug 2010 03:13:51 +0000 (04:13 +0100)]
dm snapshot: test chunk size against both origin and snapshot
Validate chunk size against both origin and snapshot sector size
Don't allow chunk size smaller than either origin or snapshot logical
sector size. Reading or writing data not aligned to sector size is not
allowed and causes immediate errors.
This requires us to open the origin before initialising the
exception store and to export dm_snap_origin.
Cc: stable@kernel.org Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Reviewed-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Mikulas Patocka [Thu, 12 Aug 2010 03:13:50 +0000 (04:13 +0100)]
dm snapshot: iterate origin and cow devices
Iterate both origin and snapshot devices
iterate_devices method should call the callback for all the devices where
the bio may be remapped. Thus, snapshot_iterate_devices should call the callback
for both snapshot and origin underlying devices because it remaps some bios
to the snapshot and some to the origin.
snapshot_iterate_devices called the callback only for the origin device.
This led to badly calculated device limits if snapshot and origin were placed
on different types of disks.
Cc: stable@kernel.org Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Reviewed-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Chris Wilson [Wed, 11 Aug 2010 13:41:16 +0000 (14:41 +0100)]
drm: Remove count_lock for calling lastclose() after 58474713 (v2)
When removing of the BKL the locking around lastclose() was rearranged
and resulted in the holding of the open_count spinlock over the call
into drm_lastclose(). The drivers were not ready for this path to be
atomic - it may indeed involve long waits to release old objects and
cleanup the GPU - and so we ended up scheduling whilst atomic.
v2: The spinlock is actually superfluous as access to open_count is
entirely serialised by drm_global_mutex and so can be dropped. The
count_lock spinlock instead appears to be used to protect access to
dev->buf_alloc and dev->buf_use.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Dave Airlie <airlied@redhat.com>
Oliver Hartkopp [Wed, 11 Aug 2010 23:12:35 +0000 (16:12 -0700)]
can: add limit for nframes and clean up signed/unsigned variables
This patch adds a limit for nframes as the number of frames in TX_SETUP and
RX_SETUP are derived from a single byte multiplex value by default.
Use-cases that would require to send/filter more than 256 CAN frames should
be implemented in userspace for complexity reasons anyway.
Additionally the assignments of unsigned values from userspace to signed
values in kernelspace and vice versa are fixed by using unsigned values in
kernelspace consistently.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Reported-by: Ben Hawkes <hawkes@google.com> Acked-by: Urs Thuermann <urs.thuermann@volkswagen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
wanglei [Wed, 11 Aug 2010 08:38:04 +0000 (09:38 +0100)]
AFS: Implement an autocell mount capability [ver #2]
Implement the ability for the root directory of a mounted AFS filesystem to
accept lookups of arbitrary directory names, to interpet the names as the names
of cells, to look the cell names up in the DNS for AFSDB records and to mount
the root.cell volume of the nominated cell on the pseudo-directory created by
lookup.
This facility is requested by passing:
-o autocell
to the mountpoint for which this is desired, usually the /afs mount.
To use this facility, a DNS upcall program is required for AFSDB records. This
can be obtained from:
Wang Lei [Wed, 11 Aug 2010 08:37:58 +0000 (09:37 +0100)]
DNS: If the DNS server returns an error, allow that to be cached [ver #2]
If the DNS server returns an error, allow that to be cached in the DNS resolver
key in lieu of a value. Userspace passes the desired error number as an option
in the payload:
"#dnserror=<number>"
Userspace must map h_errno from the name resolution routines to an appropriate
Linux error before passing it up. Something like the following mapping is
recommended:
in lieu of Linux errors specifically for representing name service errors. The
filesystem must map these errors appropropriately before passing them to
userspace. AFS is made to map ENODATA and EAGAIN to EDESTADDRREQ for the
return to userspace; ECONNREFUSED is allowed to stand as is.
The error can be seen in /proc/keys as a negative number after the description
of the key. Compare, for example, the following key entries:
If the error option is supplied in the payload, the main part of the payload is
discarded. The key should have an expiry time set by userspace.
Signed-off-by: Wang Lei <wang840925@gmail.com> Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
Bryan Schumaker [Wed, 11 Aug 2010 08:37:53 +0000 (09:37 +0100)]
NFS: Use kernel DNS resolver [ver #2]
Use the kernel DNS resolver to translate hostnames to IP addresses. Create a
new config option to choose between the legacy DNS resolver and the new
resolver.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
git://bbs.cooldavid.org / net-next-2.6.git / log