Linus Torvalds [Thu, 27 May 2010 18:34:20 +0000 (11:34 -0700)]
Merge branch 'for-linus' of git://git.o-hand.com/linux-rpurdie-leds
* 'for-linus' of git://git.o-hand.com/linux-rpurdie-leds:
leds: Add mx31moboard MC13783 led support
leds: Add mc13783 LED support
leds: leds-ss4200: fix led_classdev_unregister twice in error handling
leds: leds-lp3944: properly handle lp3944_configure fail in lp3944_probe
leds: led-class: set permissions on max_brightness file to 0444
leds: leds-gpio: Change blink_set callback to be able to turn off blinking
leds: Add LED driver for the Soekris net5501 board
leds: 88pm860x - fix checking in probe function
Linus Torvalds [Thu, 27 May 2010 18:33:46 +0000 (11:33 -0700)]
Merge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging
* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging: (23 commits)
hwmon: (lm75) Add support for the Texas Instruments TMP105
hwmon: (ltc4245) Read only one GPIO pin
hwmon: (dme1737) Add SCH5127 support
hwmon: (tmp102) Don't always stop chip at exit
hwmon: (tmp102) Fix suspend and resume functions
hwmon: (tmp102) Various fixes
hwmon: Driver for TI TMP102 temperature sensor
hwmon: EMC1403 thermal sensor support
hwmon: (applesmc) Add temperature sensor labels to sysfs interface
hwmon: (applesmc) Add generic support for MacBook Pro 7
hwmon: (applesmc) Add generic support for MacBook Pro 6
hwmon: (applesmc) Add support for MacBook Pro 5,3 and 5,4
hwmon: (tmp401) Reorganize code to get rid of static forward declarations
hwmon: (tmp401) Use constants for sysfs file permissions
hwmon: (adm1031) Allow setting update rate
hwmon: Add description of the update_rate sysfs attribute
hwmon: (lm90) Use programmed update rate
hwmon: (f71882fg) Acquire I/O regions while we're working with them
hwmon: (f71882fg) Code cleanup
hwmon: (f71882fg) Use strict_stro(l|ul) instead of simple_strto$1
...
Ira W. Snyder [Thu, 27 May 2010 17:59:02 +0000 (19:59 +0200)]
hwmon: (ltc4245) Read only one GPIO pin
Read only one of the GPIO pins as an analog voltage. The ADC can be
switched to a different GPIO pin at runtime, but this is not supported.
Previously, this driver would report the analog voltage of the currently
selected GPIO pin as all three GPIO voltages: in9_input, in10_input and
in11_input.
Jean Delvare [Thu, 27 May 2010 17:58:57 +0000 (19:58 +0200)]
hwmon: (tmp102) Various fixes
Fixes from my driver review:
http://lists.lm-sensors.org/pipermail/lm-sensors/2010-March/028051.html
Only the small changes are in there, more important changes will come
later separately as time permits.
* Drop the remnants of the now gone detect function
* The TMP102 has no known compatible chip
* Include the right header files
* Clarify why byte swapping of register values is needed
* Strip resolution info bit from temperature register value
* Set cache lifetime to 1/3 second
* Don't arbitrarily reject limit values; clamp as needed
* Make limit writing unconditional
* Don't check for transaction types the driver doesn't use
* Properly check for error when setting configuration
* Report error on failed probe
* Make the driver load automatically where needed
* Various other minor fixes
Steven King [Thu, 27 May 2010 17:58:56 +0000 (19:58 +0200)]
hwmon: Driver for TI TMP102 temperature sensor
Driver for the TI TMP102.
The TI TMP102 is similar to the LM75. It differs from the LM75 by
having a 16-bit conf register and the temp registers have a minimum
resolution of 12 bits; the extended conf register can select 13-bit
resolution (which this driver does) and also change the update rate
(which this driver currently doesn't use).
[JD: Fix tmp102_exit tag, must be __exit, not __init.]
Kalhan Trisal [Thu, 27 May 2010 17:58:56 +0000 (19:58 +0200)]
hwmon: EMC1403 thermal sensor support
Provides support for the EMC1403 thermal sensor. Only reporting of values
is supported. The various Moorestown specific extras to do with thermal
alerts and the like are not in this version of the driver.
Considerably edited and tidied up by Alan Cox, plus fixes and detection
bits from Jean Delvare.
Alex Murray [Thu, 27 May 2010 17:58:54 +0000 (19:58 +0200)]
hwmon: (applesmc) Add temperature sensor labels to sysfs interface
The Apple SMC uses a systematic labeling scheme for the hardware
temperature sensors. This scheme is currently hidden from
userland. Since the sensor set, and consequently the numbering,
differs between models, an extensive database of configurations is
required for an application such as fan control. This patch adds the
SMC labels to the hwmon sysfs interface, allowing applications to use
the sensors more intelligibly.
Jean Delvare [Thu, 27 May 2010 17:58:46 +0000 (19:58 +0200)]
hwmon: (adm1031) Allow setting update rate
Based on earlier work by Ira W. Snyder.
The adm1031 chip is capable of using a runtime configurable sampling rate,
using the fan filter register. Add support for reading and setting the
update rate via sysfs.
Ira W. Snyder [Thu, 27 May 2010 17:58:44 +0000 (19:58 +0200)]
hwmon: (lm90) Use programmed update rate
The lm90 driver programs the sensor chip to update its readings at 2 Hz
(500 ms between readings). However, the driver only does reads from the
chip at intervals of 2 * HZ (2000 ms between readings). Change the driver
update rate to the programmed update rate.
Matthew Garrett [Thu, 27 May 2010 17:58:38 +0000 (19:58 +0200)]
hwmon: (lm63) Add basic support for LM64
The LM64 appears to be an LM63 with added GPIO lines. Add support for the
hwmon functionality - GPIO can be added at some later stage if someone
has a need for them.
Jean Delvare [Thu, 27 May 2010 17:58:37 +0000 (19:58 +0200)]
hwmon: (asus_atk0110) Don't load if ACPI resources aren't enforced
When the user passes the kernel parameter acpi_enforce_resources=lax,
the ACPI resources are no longer protected, so a native driver can
make use of them. In that case, we do not want the asus_atk0110 to be
loaded. Unfortunately, this driver loads automatically due to its
MODULE_DEVICE_TABLE, so the user ends up with two drivers loaded for
the same device - this is bad.
So I suggest that we prevent the asus_atk0110 driver from loading if
acpi_enforce_resources=lax.
Linus Torvalds [Thu, 27 May 2010 17:47:41 +0000 (10:47 -0700)]
Merge branch 'sfi-release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-sfi-2.6
* 'sfi-release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-sfi-2.6:
SFI: add sysfs interface for SFI tables.
SFI: add support for v0.81 spec
* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable: (27 commits)
Btrfs: add more error checking to btrfs_dirty_inode
Btrfs: allow unaligned DIO
Btrfs: drop verbose enospc printk
Btrfs: Fix block generation verification race
Btrfs: fix preallocation and nodatacow checks in O_DIRECT
Btrfs: avoid ENOSPC errors in btrfs_dirty_inode
Btrfs: move O_DIRECT space reservation to btrfs_direct_IO
Btrfs: rework O_DIRECT enospc handling
Btrfs: use async helpers for DIO write checksumming
Btrfs: don't walk around with task->state != TASK_RUNNING
Btrfs: do aio_write instead of write
Btrfs: add basic DIO read/write support
direct-io: do not merge logically non-contiguous requests
direct-io: add a hook for the fs to provide its own submit_bio function
fs: allow short direct-io reads to be completed via buffered IO
Btrfs: Metadata ENOSPC handling for balance
Btrfs: Pre-allocate space for data relocation
Btrfs: Metadata ENOSPC handling for tree log
Btrfs: Metadata reservation for orphan inodes
Btrfs: Introduce global metadata reservation
...
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6:
[SCSI] fix race in scsi_target_reap
[SCSI] aacraid: Eliminate use after free
[SCSI] arcmsr: Support HW reset for EH and polling scheme for scsi device
[SCSI] bfa: fix system crash when reading sysfs fc_host statistics
[SCSI] iscsi_tcp: remove sk_sleep check
[SCSI] ipr: improve interrupt service routine performance
[SCSI] ipr: set the data list length in the request control block
[SCSI] ipr: fix a register read to use the correct address for 64 bit adapters
[SCSI] ipr: include the resource path in the IOA status area structure
[SCSI] ipr: implement fixes for 64 bit adapter support
[SCSI] be2iscsi: correct return value in mgmt_invalidate_icds()
Linus Torvalds [Thu, 27 May 2010 17:26:37 +0000 (10:26 -0700)]
Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (40 commits)
ext4: Make fsync sync new parent directories in no-journal mode
ext4: Drop whitespace at end of lines
ext4: Fix compat EXT4_IOC_ADD_GROUP
ext4: Conditionally define compat ioctl numbers
tracing: Convert more ext4 events to DEFINE_EVENT
ext4: Add new tracepoints to track mballoc's buddy bitmap loads
ext4: Add a missing trace hook
ext4: restart ext4_ext_remove_space() after transaction restart
ext4: Clear the EXT4_EOFBLOCKS_FL flag only when warranted
ext4: Avoid crashing on NULL ptr dereference on a filesystem error
ext4: Use bitops to read/modify i_flags in struct ext4_inode_info
ext4: Convert calls of ext4_error() to EXT4_ERROR_INODE()
ext4: Convert callers of ext4_get_blocks() to use ext4_map_blocks()
ext4: Add new abstraction ext4_map_blocks() underneath ext4_get_blocks()
ext4: Use our own write_cache_pages()
ext4: Show journal_checksum option
ext4: Fix for ext4_mb_collect_stats()
ext4: check for a good block group before loading buddy pages
ext4: Prevent creation of files larger than RLIMIT_FSIZE using fallocate
ext4: Remove extraneous newlines in ext4_msg() calls
...
Linus Torvalds [Thu, 27 May 2010 17:22:06 +0000 (10:22 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
ieee1394: schedule for removal
firewire: core: use separate timeout for each transaction
firewire: core: Fix tlabel exhaustion problem
firewire: core: make transaction label allocation more robust
firewire: core: clean up config ROM related defined constants
ieee1394: mark char device files as not seekable
firewire: cdev: mark char device files as not seekable
firewire: ohci: cleanups and fix for nonstandard build without debug facility
firewire: ohci: wait for PHY register accesses to complete
firewire: ohci: fix up configuration of TI chips
firewire: ohci: enable 1394a enhancements
firewire: ohci: do not clear PHY interrupt status inadvertently
firewire: ohci: add a function for reading PHY registers
Trivial conflicts in Documentation/feature-removal-schedule.txt
Linus Torvalds [Thu, 27 May 2010 17:18:44 +0000 (10:18 -0700)]
Merge branch 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6
* 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
NFS: Fix another nfs_wb_page() deadlock
NFS: Ensure that we mark the inode as dirty if we exit early from commit
NFS: Fix a lock imbalance typo in nfs_access_cache_shrinker
sunrpc: fix leak on error on socket xprt setup
Linus Torvalds [Thu, 27 May 2010 16:19:55 +0000 (09:19 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
Input: usbtouchscreen - support bigger iNexio touchscreens
Input: ads7846 - return error on regulator_get() failure
Input: twl4030-vibra - correct the power down sequence
Input: enable onkey driver of max8925
Input: use ABS_CNT rather than (ABS_MAX + 1)
Lee Schermerhorn [Wed, 26 May 2010 21:45:06 +0000 (14:45 -0700)]
numa: update Documentation/vm/numa, add memoryless node info
Kamezawa Hiroyuki requested documentation for the numa_mem_id() and slab
related changes. He suggested Documentation/vm/numa for this
documentation. Looking at this file, it seems to me to be hopelessly out
of date relative to current Linux NUMA support. At the risk of going down
a rathole, I have made an attempt to rewrite the doc at a slightly higher
level [I think] and provide pointers to other in-tree documents and
out-of-tree man pages that cover the details.
Lee Schermerhorn [Wed, 26 May 2010 21:45:04 +0000 (14:45 -0700)]
numa: in-kernel profiling: use cpu_to_mem() for per cpu allocations
In kernel profiling requires that we be able to allocate "local" memory
for each cpu. Use "cpu_to_mem()" instead of "cpu_to_node()" to support
memoryless nodes.
Lee Schermerhorn [Wed, 26 May 2010 21:45:03 +0000 (14:45 -0700)]
numa: slab: use numa_mem_id() for slab local memory node
Example usage of generic "numa_mem_id()":
The mainline slab code, since ~ 2.6.19, does not handle memoryless nodes
well. Specifically, the "fast path"--____cache_alloc()--will never
succeed as slab doesn't cache offnode object on the per cpu queues, and
for memoryless nodes, all memory will be "off node" relative to
numa_node_id(). This adds significant overhead to all kmem cache
allocations, incurring a significant regression relative to earlier
kernels [from before slab.c was reorganized].
This patch uses the generic topology function "numa_mem_id()" to return
the "effective local memory node" for the calling context. This is the
first node in the local node's generic fallback zonelist-- the same node
that "local" mempolicy-based allocations would use. This lets slab cache
these "local" allocations and avoid fallback/refill on every allocation.
N.B.: Slab will need to handle node and memory hotplug events that could
change the value returned by numa_mem_id() for any given node if recent
changes to address memory hotplug don't already address this. E.g., flush
all per cpu slab queues before rebuilding the zonelists while the
"machine" is held in the stopped state.
Performance impact on "hackbench 400 process 200"
2.6.34-rc3-mmotm-100405-1609 no-patch this-patch
ia64 no memoryless nodes [avg of 10]: 11.713 11.637 ~0.65 diff
ia64 cpus all on memless nodes [10]: 228.259 26.484 ~8.6x speedup
The slowdown of the patched kernel from ~12 sec to ~28 seconds when
configured with memoryless nodes is the result of all cpus allocating from
a single node's mm pagepool. The cache lines of the single node are
distributed/interleaved over the memory of the real physical nodes, but
the zone lock, list heads, ... of the single node with memory still each
live in a single cache line that is accessed from all processors.
Lee Schermerhorn [Wed, 26 May 2010 21:45:01 +0000 (14:45 -0700)]
numa: ia64: support numa_mem_id() for memoryless nodes
Enable 'HAVE_MEMORYLESS_NODES' by default when NUMA configured on ia64.
Initialize percpu 'numa_mem' variable when starting secondary cpus.
Generic initialization will handle the boot cpu.
Nothing uses 'numa_mem_id()' yet. Subsequent patch with modify slab to
use this.
Lee Schermerhorn [Wed, 26 May 2010 21:45:00 +0000 (14:45 -0700)]
numa: introduce numa_mem_id()- effective local memory node id
Introduce numa_mem_id(), based on generic percpu variable infrastructure
to track "nearest node with memory" for archs that support memoryless
nodes.
Define API in <linux/topology.h> when CONFIG_HAVE_MEMORYLESS_NODES
defined, else stubs. Architectures will define HAVE_MEMORYLESS_NODES
if/when they support them.
Archs can override definitions of:
numa_mem_id() - returns node number of "local memory" node
set_numa_mem() - initialize [this cpus'] per cpu variable 'numa_mem'
cpu_to_mem() - return numa_mem for specified cpu; may be used as lvalue
Generic initialization of 'numa_mem' occurs in __build_all_zonelists().
This will initialize the boot cpu at boot time, and all cpus on change of
numa_zonelist_order, or when node or memory hot-plug requires zonelist
rebuild. Archs that support memoryless nodes will need to initialize
'numa_mem' for secondary cpus as they're brought on-line.
Lee Schermerhorn [Wed, 26 May 2010 21:44:59 +0000 (14:44 -0700)]
numa: ia64: use generic percpu var numa_node_id() implementation
ia64: Use generic percpu implementation of numa_node_id()
+ intialize per cpu 'numa_node'
+ remove ia64 cpu_to_node() macro; use generic
+ define CONFIG_USE_PERCPU_NUMA_NODE_ID when NUMA configured
Lee Schermerhorn [Wed, 26 May 2010 21:44:58 +0000 (14:44 -0700)]
numa: x86_64: use generic percpu var numa_node_id() implementation
x86 arch specific changes to use generic numa_node_id() based on generic
percpu variable infrastructure. Back out x86's custom version of
numa_node_id()
Lee Schermerhorn [Wed, 26 May 2010 21:44:56 +0000 (14:44 -0700)]
numa: add generic percpu var numa_node_id() implementation
Rework the generic version of the numa_node_id() function to use the new
generic percpu variable infrastructure.
Guard the new implementation with a new config option:
CONFIG_USE_PERCPU_NUMA_NODE_ID.
Archs which support this new implemention will default this option to 'y'
when NUMA is configured. This config option could be removed if/when all
archs switch over to the generic percpu implementation of numa_node_id().
Arch support involves:
1) converting any existing per cpu variable implementations to use
this implementation. x86_64 is an instance of such an arch.
2) archs that don't use a per cpu variable for numa_node_id() will
need to initialize the new per cpu variable "numa_node" as cpus
are brought on-line. ia64 is an example.
3) Defining USE_PERCPU_NUMA_NODE_ID in arch dependent Kconfig--e.g.,
when NUMA is configured. This is required because I have
retained the old implementation by default to allow archs to
be modified incrementally, as desired.
Subsequent patches will convert x86_64 and ia64 to use this implemenation.
Jan Blunck [Wed, 26 May 2010 21:44:54 +0000 (14:44 -0700)]
Documentation/filesystems/Locking: update documentation on llseek() wrt BKL
The inode's i_size is not protected by the big kernel lock. Therefore it
does not make sense to recommend taking the BKL in filesystems llseek
operations. Instead it should use the inode's mutex or use just use
i_size_read() instead. Add a note that this is not protecting
file->f_pos.
jan Blunck [Wed, 26 May 2010 21:44:53 +0000 (14:44 -0700)]
fs/: do not fallback to default_llseek() when readdir() uses BKL
Do not use the fallback default_llseek() if the readdir operation of the
filesystem still uses the big kernel lock.
Since llseek() modifies
file->f_pos of the directory directly it may need locking to not confuse
readdir which usually uses file->f_pos directly as well
Since the special characteristics of the BKL (unlocked on schedule) are
not necessary in this case, the inode mutex can be used for locking as
provided by generic_file_llseek(). This is only possible since all
filesystems, except reiserfs, either use a directory as a flat file or
with disk address offsets. Reiserfs on the other hand uses a 32bit hash
off the filename as the offset so generic_file_llseek() can get used as
well since the hash is always smaller than sb->s_maxbytes (= (512 << 32) -
blocksize).
Jan Blunck [Wed, 26 May 2010 21:44:51 +0000 (14:44 -0700)]
st: use noop_llseek() instead of default_llseek()
st_open() suggests that llseek() doesn't work: "We really want to do
nonseekable_open(inode, filp); here, but some versions of tar incorrectly
call lseek on tapes and bail out if that fails. So we disallow pread()
and pwrite(), but permit lseeks."
Instead of using the fallback default_llseek() the driver should use
noop_llseek() which leaves the file->f_pos untouched but succeeds.
Jan Blunck [Wed, 26 May 2010 21:44:50 +0000 (14:44 -0700)]
osst: use noop_llseek() instead of default_llseek()
__os_scsi_tape_open() suggests that llseek() doesn't work: "We really want
to do nonseekable_open(inode, filp); here, but some versions of tar
incorrectly call lseek on tapes and bail out if that fails. So we
disallow pread() and pwrite(), but permit lseeks."
Instead of using the fallback default_llseek() the driver should use
noop_llseek() which leaves the file->f_pos untouched but succeeds.
jan Blunck [Wed, 26 May 2010 21:44:48 +0000 (14:44 -0700)]
vfs: introduce noop_llseek()
This is an implementation of ->llseek useable for the rare special case
when userspace expects the seek to succeed but the (device) file is
actually not able to perform the seek. In this case you use noop_llseek()
instead of falling back to the default implementation of ->llseek.
Jan Blunck [Wed, 26 May 2010 21:44:46 +0000 (14:44 -0700)]
mISDN: remove unnecessary test on f_pos
This test is not doing anything since it is always false if the
mISDN_read() is called from vfs_read(). Besides that the driver uses
nonseekable_open() and is not using off or file->f_pos anywhere.
Jan Blunck [Wed, 26 May 2010 21:44:44 +0000 (14:44 -0700)]
osst: update ppos instead of using file->f_pos
osst_read()/osst_write() modify file->f_pos directly instead of the ppos
given to them. The VFS later updates the file->f_pos and overwrites it
with the value of ppos.
FUJITA Tomonori [Wed, 26 May 2010 21:44:43 +0000 (14:44 -0700)]
parisc: use asm-generic/scatterlist.h
parisc uses iova and iova_length in scatterlist structure instead of
dma_address and dma_length. However, the accessor are used so we can
convert parisc to use asm-generic/scatterlist.h easily.
FUJITA Tomonori [Wed, 26 May 2010 21:44:34 +0000 (14:44 -0700)]
asm-generic: remove ARCH_HAS_SG_CHAIN in scatterlist.h
There are more architectures that don't support ARCH_HAS_SG_CHAIN than
those that support it. This removes removes ARCH_HAS_SG_CHAIN in
asm-generic/scatterlist.h and lets arhictectures to define it.
It's clearer than defining ARCH_HAS_SG_CHAIN asm-generic/scatterlist.h and
undefing it in arhictectures that don't support it.
FUJITA Tomonori [Wed, 26 May 2010 21:44:32 +0000 (14:44 -0700)]
asm-generic: add NEED_SG_DMA_LENGTH to define sg_dma_len()
There are only two ways to define sg_dma_len(); use sg->dma_length or
sg->length. This patch introduces NEED_SG_DMA_LENGTH that enables
architectures to choose sg->dma_length or sg->length.
lktdm: add support for hardlockup, softlockup and hung task crashes
This adds three new types of kernel "crashes" in the lkdtm driver to
trigger hardlockups, softlockups and task hung states at will.
The first two are useful to test the new generic lockup detector and check
its further regressions. The latter one is a bonus to check the hung task
detector regressions even though it's not currently in rework.
radix-tree: fix radix_tree_prev_hole() underflow case
radix_tree_prev_hole() used LONG_MAX to detect underflow; however,
ULONG_MAX is clearly what was intended, both here and by its only user
(count_history_pages at mm/readahead.c).
Jeff Moyer [Wed, 26 May 2010 21:44:26 +0000 (14:44 -0700)]
aio: fix the compat vectored operations
The aio compat code was not converting the struct iovecs from 32bit to
64bit pointers, causing either EINVAL to be returned from io_getevents, or
EFAULT as the result of the I/O. This patch passes a compat flag to
io_submit to signal that pointer conversion is necessary for a given iocb
array.
A variant of this was tested by Michael Tokarev. I have also updated the
libaio test harness to exercise this code path with good success.
Further, I grabbed a copy of ltp and ran the
testcases/kernel/syscall/readv and writev tests there (compiled with -m32
on my 64bit system). All seems happy, but extra eyes on this would be
welcome.
Jeff Moyer [Wed, 26 May 2010 21:44:25 +0000 (14:44 -0700)]
compat: factor out compat_rw_copy_check_uvector from compat_do_readv_writev
It was reported in http://lkml.org/lkml/2010/3/8/309 that 32 bit readv and
writev AIO operations were not functioning properly. It turns out that
the code to convert the 32bit io vectors to 64 bits was never written.
The results of that can be pretty bad, but in my testing, it mostly ended
up in generating EFAULT as we walked off the list of I/O vectors provided.
This patch set fixes the problem in my environment. are greatly
appreciated.
This patch:
Factor out code that will be used by both compat_do_readv_writev and the
compat aio submission code paths.
Anton Blanchard [Wed, 26 May 2010 21:44:24 +0000 (14:44 -0700)]
panic: call console_verbose() in panic
Most distros turn the console verbosity down and that means a backtrace
after a panic never makes it to the console. I assume we haven't seen
this because a panic is often preceeded by an oops which will have called
console_verbose. There are however a lot of places we call panic
directly, and they are broken.
Use console_verbose like we do in the oops path to ensure a directly
called panic will print a backtrace.
FUJITA Tomonori [Wed, 26 May 2010 21:44:18 +0000 (14:44 -0700)]
dma-mapping: remove unnecessary sync_single_range_* in dma_map_ops
sync_single_range_for_cpu and sync_single_range_for_device hooks are
unnecessary because sync_single_for_cpu and sync_single_for_device can
be used instead.
swiotlb_sync_single_range_for_cpu and swiotlb_sync_single_range_for_device
are unnecessary because swiotlb_sync_single_for_cpu and
swiotlb_sync_single_for_device can be used instead.
FUJITA Tomonori [Wed, 26 May 2010 21:44:17 +0000 (14:44 -0700)]
powerpc: remove unnecessary sync_single_range_* in swiotlb_dma_ops
sync_single_range_for_cpu and sync_single_range_for_device hooks in
swiotlb_dma_ops are unnecessary because sync_single_for_cpu and
sync_single_for_device are used there.
FUJITA Tomonori [Wed, 26 May 2010 21:44:16 +0000 (14:44 -0700)]
x86: remove unnecessary sync_single_range_* in swiotlb_dma_ops
sync_single_range_for_cpu and sync_single_range_for_device hooks in
swiotlb_dma_ops are unnecessary because sync_single_for_cpu and
sync_single_for_device are used there.
FUJITA Tomonori [Wed, 26 May 2010 21:44:15 +0000 (14:44 -0700)]
ia64: remove unnecessary sync_single_range_* in swiotlb_dma_ops
sync_single_range_for_cpu and sync_single_range_for_device hooks in
swiotlb_dma_ops are unnecessary because sync_single_for_cpu and
sync_single_for_device are used there.
Joe Eykholt [Wed, 26 May 2010 21:44:13 +0000 (14:44 -0700)]
lib/random32: export pseudo-random number generator for modules
This patch moves the definition of struct rnd_state and the inline
__seed() function to linux/random.h. It renames the static __random32()
function to prandom32() and exports it for use in modules.
prandom32() is useful as a privately-seeded pseudo random number generator
that can give the same result every time it is initialized.
For FCoE FC-BB-6 VN2VN mode self-selected unique FC address generation, we
need an pseudo-random number generator seeded with the 64-bit world-wide
port name. A truly random generator or one seeded with randomness won't
do because the same sequence of numbers should be generated each time we
boot or the link comes up.
A prandom32_seed() inline function is added to the header file. It is
inlined not for speed, but so the function won't be expanded in the base
kernel, but only in the module that uses it.
Oleg Nesterov [Wed, 26 May 2010 21:44:11 +0000 (14:44 -0700)]
pids: fix fork_idle() to setup ->pids correctly
copy_process(pid => &init_struct_pid) doesn't do attach_pid/etc.
It shouldn't, but this means that the idle threads run with the wrong
pids copied from the caller's task_struct. In x86 case the caller is
either kernel_init() thread or keventd.
In particular, this means that after the series of cpu_up/cpu_down an
idle thread (which never exits) can run with .pid pointing to nowhere.
Change fork_idle() to initialize idle->pids[] correctly. We only set
.pid = &init_struct_pid but do not add .node to list, INIT_TASK() does
the same for the boot-cpu idle thread (swapper).
Oleg Nesterov [Wed, 26 May 2010 21:44:10 +0000 (14:44 -0700)]
pids: init_struct_pid.tasks should never see the swapper process
"statically initialize struct pid for swapper" commit 820e45db says:
Statically initialize a struct pid for the swapper process (pid_t == 0)
and attach it to init_task. This is needed so task_pid(), task_pgrp()
and task_session() interfaces work on the swapper process also.
OK, but:
- it doesn't make sense to add init_task.pids[].node into
init_struct_pid.tasks[], and in fact this just wrong.
idle threads are special, they shouldn't be visible on any
global list. In particular do_each_pid_task(init_struct_pid)
shouldn't see swapper.
This is the actual reason why kill(0, SIGKILL) from /sbin/init
(which starts with 0,0 special pids) crashes the kernel. The
signal sent to pgid/sid == 0 must never see idle threads, even
if the previous patch fixed the crash itself.
- we have other idle threads running on the non-boot CPUs, see
the next patch.
Change INIT_STRUCT_PID/INIT_PID_LINK to create the empty/unhashed
hlist_head/hlist_node. Like any other idle thread swapper can never exit,
so detach_pid()->__hlist_del() is not possible, but we could change
INIT_PID_LINK() to set pprev = &next if needed.
All we need is the valid swapper->pids[].pid == &init_struct_pid.
Oleg Nesterov [Wed, 26 May 2010 21:44:08 +0000 (14:44 -0700)]
INIT_TASK() should initialize ->thread_group list
The trivial /sbin/init doing
int main(void)
{
kill(0, SIGKILL)
}
crashes the kernel.
This happens because __kill_pgrp_info(init_struct_pid) also sends SIGKILL
to the swapper process which runs with the uninitialized ->thread_group.
Change INIT_TASK() to initialize ->thread_group properly.
Note: the real problem is that the swapper process must not be visible to
signals, see the next patch. But this change is right anyway and fixes
the crash.
Hedi Berriche [Wed, 26 May 2010 21:44:06 +0000 (14:44 -0700)]
pids: increase pid_max based on num_possible_cpus
On a system with a substantial number of processors, the early default
pid_max of 32k will not be enough. A system with 1664 CPU's, there are
25163 processes started before the login prompt. It's estimated that with
2048 CPU's we will pass the 32k limit. With 4096, we'll reach that limit
very early during the boot cycle, and processes would stall waiting for an
available pid.
This patch increases the early maximum number of pids available, and
increases the minimum number of pids that can be set during runtime.
Thomas Moll [Wed, 26 May 2010 21:44:05 +0000 (14:44 -0700)]
rapidio: fix maintenance access to higher memory areas
Fix the maintenance access functions to farend RapidIO devices.
1. Fixed shift of the given offset, to open the maintenance window
2. Mask offset to limit access to the opened maintenance window
3. Added extended destid part to rowtear register, required for 16bit mode
This method is matching maintenance transactions generation described
by Freescale in the appnote AN2932. With this modification full access
to a 16MB maintenance window is possible, this patch is required for
IDT cps switches. For easier handling of the access routines, the
access was limited to aligned memory regions. This should be no problem
because all registers are 32bit wide.
projects / linux.git / log