tjh.dev
commits
Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Fix section warning:
WARNING: arch/arm/mach-at91/built-in.o(.text+0xd74): Section mismatch in reference
from the function init_programmable_clock()
to the function .init.text:at91_css_to_clk()
The function init_programmable_clock() references
the function __init at91_css_to_clk().
This is often because init_programmable_clock lacks a __init
annotation or the annotation of at91_css_to_clk is wrong.
In this case the only calls to and from init_programmable_clock()
are from code marked as "__init", so this fix is trivially correct.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Acked-by: Uwe Kleine-Knig <Uwe.Kleine-Koenig@digi.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
According to PXA300/310 and PXA320 Developer manuals,
the ASCR[RDH] "bit needs to be cleared as part of the software
initialization coming out of any reset and coming out of D3".
The latter requirement is addressed by commit
"c4d1fb627ff3072", as for the former (coming out of any reset),
the kernel relies on boot loaders and assumes that RDH bit
is cleared there. Though, not all bootloaders follow the rule
so we have to clear the bit in kernel.
We clear the RDH bit in pxa3xx_init() function since
it is always invoked after any reset. We also preserve D1S, D2S
and D3S bits from being cleared in case we invoke pxa3xx_init()
function not from normal hardware reset (e.g. kexec scenario),
so these bits can be properly referenced later.
Signed-off-by: Dmitry Krivoschekov <dmitry.krivoschekov@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Convert special mapping install from nopage to fault.
Because the "vm_file" is NULL for the special mapping, the generic VM
code has messed up "vm_pgoff" thinking that it's an anonymous mapping
and the offset does't matter. For that reason, we need to undo the
vm_pgoff offset that got added into vmf->pgoff.
[ We _really_ should clean that up - either by making this whole special
mapping code just use a real file entry rather than that ugly array of
"struct page" pointers, or by just making the VM code realize that
even if vm_file is NULL it may not be a regular anonymous mmap.
- Linus ]
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: linux-mm@kvack.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
[IA64] Fix large MCA bootmem allocation
[IA64] Simplify cpu_idle_wait
[IA64] Synchronize RBS on PTRACE_ATTACH
[IA64] Synchronize kernel RSE to user-space and back
[IA64] Rename TIF_PERFMON_WORK back to TIF_NOTIFY_RESUME
[IA64] Wire up timerfd_{create,settime,gettime} syscalls
Commit 2f569afd9ced9ebec9a6eb3dbf6f83429be0a7b4 ("CONFIG_HIGHPTE vs.
sub-page page tables") caused some build breakage due to pgtable_t only
getting declared in the CONFIG_X86_PAE case.
Move the declaration outside the PAE section.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The MCA code allocates bootmem memory for NR_CPUS, regardless
of how many cpus the system actually has. This change allocates
memory only for cpus that actually exist.
On my test system with NR_CPUS = 1024, reserved memory was reduced by 130944k.
Before: Memory: 27886976k/28111168k available (8282k code, 242304k reserved, 5928k data, 1792k init)
After: Memory: 28017920k/28111168k available (8282k code, 111360k reserved, 5928k data, 1792k init)
Signed-off-by: Russ Anderson <rja@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/core: Remove unused struct ib_device.flags member
IB/core: Add IP checksum offload support
IPoIB: Add send gather support
IPoIB: Add high DMA feature flag
IB/mlx4: Use multiple WQ blocks to post smaller send WQEs
mlx4_core: Clean up struct mlx4_buf
mlx4_core: For 64-bit systems, vmap() kernel queue buffers
IB/mlx4: Consolidate code to get an entry from a struct mlx4_buf
This is just Venki's patch[*] for x86 ported to ia64.
* http://marc.info/?l=linux-kernel&m=120249201318159&w=2
Acked-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/blackfin-2.6: (24 commits)
[Blackfin] arch: import defines for BF547 -- it is just like the BF548, but no CAN
[Blackfin] arch: fix build fails only include header files when enabled
[Blackfin] arch: declare default INSTALL_PATH for Blackfin ports
[Blackfin] arch: Encourage users to use the spidev character driver: Provide platform support
[Blackfin] arch: Enable UART2 and UART3 for bf548
[Blackfin] arch: Enable NET2272 on BF561-EZkit - remove request_mem_region
[Blackfin] arch:Fix BUG [#3876] pfbutton test for BTN3 on bf533 don't show complete info
[Blackfin] arch: remove duplicated definitions of the line discipline numbers N_* in asm-blackfin/termios.h
[Blackfin] arch: fix building with mtd uclinux by putting the mtd_phys option into the function it actually gets used in
[Blackfin] arch: simpler header and update dates
[Blackfin] arch: move the init sections to the end of memory
[Blackfin] arch: change the trace buffer control start/stop logic in the exception handlers
[Blackfin] arch: fix typo in printk message
[Blackfin] arch: this is an ezkit, not a stamp, so fixup the init function name
[Blackfin] arch: add slightly better help text for CPLB_INFO
[Blackfin] arch: Fix BUG - Enable ISP1362 driver to work ok with BF561
[Blackfin] arch: Fix header file information
[Blackfin] arch: Add Support for ISP1362
[Blackfin] arch: add support for cmdline partitioning to the BF533-STAMP flash map driver and enable it as a module by default
[Blackfin] arch: hook up set_irq_wake in Blackfin's irq code
...
Avoid confusion about what it might mean, since it's never initialized.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When attaching to a stopped process, the RSE must be explicitly
synced to user-space, so the debugger can read the correct values.
Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
CC: Roland McGrath <roland@redhat.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Thanks to Kay for keeping us honest.
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Cc: "Williams, Dan J" <dan.j.williams@intel.com>
Acked-by: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Add a device capability to show when it can handle checksum offload.
Also add a send flag for inserting checksums and a csum_ok field to
the completion record.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This is base kernel patch for ptrace RSE bug. It's basically a backport
from the utrace RSE patch I sent out several weeks ago. please review.
when a thread is stopped (ptraced), debugger might change thread's user
stack (change memory directly), and we must avoid the RSE stored in
kernel to override user stack (user space's RSE is newer than kernel's
in the case). To workaround the issue, we copy kernel RSE to user RSE
before the task is stopped, so user RSE has updated data. we then copy
user RSE to kernel after the task is resummed from traced stop and
kernel will use the newer RSE to return to user.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
CC: Roland McGrath <roland@redhat.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
x25_asy does not take an ldisc reference before calling the flush method.
Fix it to use the helper function we provide.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Krzysztof Halasa <khc@pm.waw.pl>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
This patch acts as a preparation for using checksum offload for IB
devices capable of inserting/verifying checksum in IP packets. The
patch does not actaully turn on NETIF_F_SG - we defer that to the
patches adding checksum offload capabilities.
We only add support for send gathers for datagram mode, since existing
HW does not support checksum offload on connected QPs.
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since the RSE synchronization will need a TIF_ flag, but all
work-to-be-done bits are already used, so we have to multiplex
TIF_NOTIFY_RESUME again.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
Signed-off-by: Tony Luck <tony.luck@intel.com>
lib/scatterlist.c is needed by drivers/media/video/videobuf-dma-sg.c, and
we would like to be able to use the latter without PCI too, for example, on
PXA270 ARM CPU. It is then possible to create a configuration with
CONFIG_BLOCK=n, where only module code will need scatterlist.c. Therefore
it must be in obj-y.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@pengutronix.de>
Acked-by: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
All current InfiniBand devices can handle all DMA addresses, and it's
hard to imagine anyone would be silly enough to build a new device
that couldn't. Therefore, enable the NETIF_F_HIGHDMA feature for IPoIB.
This has no effect for no, but is needed when we enable gather/scatter
support and checksum stateless offloads.
Signed-off-by: Eli Cohen <eli@mellnaox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add ia64 hooks for the new syscalls that were added in
commit 4d672e7ac79b5ec5cdc90e450823441e20464691
Signed-off-by: Tony Luck <tony.luck@intel.com>
In file included from /home/bunk/linux/kernel-2.6/git/linux-2.6/arch/xtensa/kernel/syscall.c:39:
include2/asm/unistd.h:681: error: 'sys_timerfd' undeclared here (not in a function)
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Cc: Christian Zankel <chris@zankel.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
- Enable kernel generic spidev driver for blackfin SPI ADC
- spi_adc driver, document and test sample not synced
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
ConnectX HCA supports shrinking WQEs, so that a single work request
can be made of multiple units of wqe_shift. This way, WRs can differ
in size, and do not have to be a power of 2 in size, saving memory and
speeding up send WR posting. Unfortunately, if we do this then the
wqe_index field in CQEs can't be used to look up the WR ID anymore, so
our implementation does this only if selective signaling is off.
Further, on 32-bit platforms, we can't use vmap() to make the QP
buffer virtually contigious. Thus we have to use constant-sized WRs to
make sure a WR is always fully within a single page-sized chunk.
Finally, we use WRs with the NOP opcode to avoid wrapping around the
queue buffer in the middle of posting a WR, and we set the
NoErrorCompletion bit to avoid getting completions with error for NOP
WRs. However, NEC is only supported starting with firmware 2.2.232,
so we use constant-sized WRs for older firmware. And, since MLX QPs
only support SEND, we use constant-sized WRs in this case.
When stamping during NOP posting, do stamping following setting of the
NOP WQE valid bit.
Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-2.6-dm: (44 commits)
dm raid1: report fault status
dm raid1: handle read failures
dm raid1: fix EIO after log failure
dm raid1: handle recovery failures
dm raid1: handle write failures
dm snapshot: combine consecutive exceptions in memory
dm: stripe enhanced status return
dm: stripe trigger event on failure
dm log: auto load modules
dm: move deferred bio flushing to workqueue
dm crypt: use async crypto
dm crypt: prepare async callback fn
dm crypt: add completion for async
dm crypt: add async request mempool
dm crypt: extract scatterlist processing
dm crypt: tidy io ref counting
dm crypt: introduce crypt_write_io_loop
dm crypt: abstract crypt_write_done
dm crypt: store sector mapping in dm_crypt_io
dm crypt: move queue functions
...
Fix the FRV cmpxchg_local by breaking the following header dependency loop :
linux/kernel.h -> linux/bitops.h -> asm-frv/bitops.h -> asm-frv/atomic.h
-> asm-frv/system.h ->
asm-generic/cmpxchg_local.h -> typecheck() defined in linux/kernel.h
and
linux/kernel.h -> linux/bitops.h -> asm-frv/bitops.h -> asm-frv/atomic.h ->
asm-generic/cmpxchg_local.h -> typecheck() defined in linux/kernel.h
In order to fix this :
- Move the atomic_test_and_ *_mask inlines from asm-frv/atomic.h (why are they
there at all anyway ? They are not touching atomic_t variables!) to
asm-frv/bitops.h.
Also fix a build issue with cmpxchg : it does not cast to (unsigned long *)
like other architectures, to deal with it in the cmpxchg_local macro.
FRV builds fine with this patch.
Thanks to Adrian Bunk <bunk@kernel.org> for spotting this bug.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Cc: Adrian Bunk <bunk@kernel.org>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Now that struct mlx4_buf.u is a struct instead of a union because of
the vmap() changes, there's no point in having a struct at all. So
move .direct and .page_list directly into struct mlx4_buf and get rid
of a bunch of unnecessary ".u"s.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'release' of git://lm-sensors.org/kernel/mhoffman/hwmon-2.6: (59 commits)
hwmon: (lm80) Add individual alarm files
hwmon: (lm80) De-macro the sysfs callbacks
hwmon: (lm80) Various cleanups
hwmon: (w83627hf) Refactor beep enable handling
hwmon: (w83627hf) Add individual alarm and beep files
hwmon: (w83627hf) Enable VBAT monitoring
hwmon: (w83627ehf) The W83627DHG has 8 VID pins
hwmon: (asb100) Add individual alarm files
hwmon: (asb100) De-macro the sysfs callbacks
hwmon: (asb100) Various cleanups
hwmon: VRM is not written to registers
hwmon: (dme1737) fix Super-IO device ID override
hwmon: (dme1737) fix divide-by-0
hwmon: (abituguru3) Add AUX4 fan input for Abit IP35 Pro
hwmon: Add support for Texas Instruments/Burr-Brown ADS7828
hwmon: (adm9240) Add individual alarm files
hwmon: (lm77) Add individual alarm files
hwmon: Discard useless I2C driver IDs
hwmon: (lm85) Make the pwmN_enable files writable
hwmon: (lm85) Return standard values in pwmN_enable
...
This patch adds extra information to the mirror status output, so that
it can be determined which device(s) have failed. For each mirror device,
a character is printed indicating the most severe error encountered. The
characters are:
* A => Alive - No failures
* D => Dead - A write failure occurred leaving mirror out-of-sync
* S => Sync - A sychronization failure occurred, mirror out-of-sync
* R => Read - A read failure occurred, mirror data unaffected
This allows userspace to properly reconfigure the mirror set.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Revert commit c6caeb7c4544608e8ae62731334661fc396c7f85 ("proc: fix the
threaded /proc/self"), since Eric says "The patch really is wrong.
There is at least one corner case in procps that cares."
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "Guillaume Chazarain" <guichaz@yahoo.fr>
Cc: "Pavel Emelyanov" <xemul@openvz.org>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
Since kernel virtual memory is not a problem on 64-bit systems, there
is no reason to use our own 2-layer page mapping scheme for large
kernel queue buffers on such systems. Instead, map the page list to a
single virtually contiguous buffer with vmap(), so that can we access
buffer memory via direct indexing.
Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://oss.sgi.com:8090/xfs/xfs-2.6: (62 commits)
[XFS] add __init/__exit mark to specific init/cleanup functions
[XFS] Fix oops in xfs_file_readdir()
[XFS] kill xfs_root
[XFS] keep i_nlink updated and use proper accessors
[XFS] stop updating inode->i_blocks
[XFS] Make xfs_ail_check check less by default
[XFS] Move AIL pushing into it's own thread
[XFS] use generic_permission
[XFS] stop re-checking permissions in xfs_swapext
[XFS] clean up xfs_swapext
[XFS] remove permission check from xfs_change_file_space
[XFS] prevent panic during log recovery due to bogus op_hdr length
[XFS] Cleanup various fid related bits:
[XFS] Fix xfs_lowbit64
[XFS] Remove CFORK macros and use code directly in IFORK and DFORK macros.
[XFS] kill superflous buffer locking (2nd attempt)
[XFS] Use kernel-supplied "roundup_pow_of_two" for simplicity
[XFS] Remove the BPCSHIFT and NB* based macros from XFS.
[XFS] Remove bogus assert
[XFS] optimize XFS_IS_REALTIME_INODE w/o realtime config
...
The new libsensors needs these individual alarm files.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Mark M. Hoffman <mhoffman@lightlink.com>
This patch gives the ability to respond-to/record device failures
that happen during read operations. It also adds the ability to
read from mirror devices that are not the primary if they are
in-sync.
There are essentially two read paths in mirroring; the direct path
and the queued path. When a read request is mapped, if the region
is 'in-sync' the direct path is taken; otherwise the queued path
is taken.
If the direct path is taken, we must record bio information so that
if the read fails we can retry it. We then discover the status of
a direct read through mirror_end_io. If the read has failed, we will
mark the device from which the read was attempted as failed (so we
don't try to read from it again), restore the bio and try again.
If the queued path is taken, we discover the results of the read
from 'read_callback'. If the device failed, we will mark the device
as failed and attempt the read again if there is another device
where this region is known to be 'in-sync'.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'cris' of git://www.jni.nu/cris: (158 commits)
CRIS v32: Remove hwregs/timer_defs.h, it is now architecture specific.
CRIS v32: Change drivers/i2c.c locking.
CRIS v32: Rewrite ARTPEC-3 gpio driver to avoid volatiles and general cleanup.
CRIS: Add new timerfd syscall entries.
MAINTAINERS: Add my information for the CRIS port.
CRIS v32: Correct spelling of bandwidth in function name.
CRIS v32: Clean up nandflash.c for ARTPEC-3 and ETRAX FS.
CRIS v10: Cleanup of drivers/gpio.c
CRIS v10: drivers/net/cris/eth_v10.c rename LED defines to CRIS_LED to avoid name clash.
CRIS: Make io_pwm_set_period members unsigned in etraxgpio.h
CRIS: Move ETRAX_AXISFLASHMAP to common Kconfig file.
CRIS: Drop regs parameter from call to profile_tick in kernel/time.c
CRIS v32: Fix minor formatting issue in mach-a3/io.c
CRIS v32: Initialize GIO even if we're rambooting in kernel/head.S
CRIS v32: Remove kernel/arbiter.c, it now exists in machine dependent directory.
CRIS v32: Minor changes to avoid errors in asm-cris/arch-v32/hwregs/reg_rdwr.h
CRIS v32: arch-v32/hwregs/intr_vect_defs.h moved to machine dependent directory.
CRIS v32: Correct offset for TASK_pid in asm-cris/arch-v32/offset.h
CRIS v32: Move register map header to machine dependent directory.
CRIS v32: Let compiler know that memory is clobbered after a break op.
...
- Buttons on the BF533-STAMP board are not inverted
- Fix spurious GPIO Interrupt caused during set irq_type for edge triggered interrupts
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
We use struct mlx4_buf for kernel QP, CQ and SRQ buffers, and the code
to look up an entry is duplicated in get_cqe_from_buf() and the QP and
SRQ versions of get_wqe(). Factor this out into mlx4_buf_offset().
This will also make it easier to switch over to using vmap() for buffers.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Douglas Gilbert <dougg@torque.net>
Cc: James Bottomley <James.Bottomley@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:30459a
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Denis Cheng <crquan@gmail.com>
Use standard dynamic sysfs callbacks instead of macro-generated
functions. This makes the code more readable, and the binary smaller
(by about 34%).
As a side note, another benefit of this type of cleanup is that they
shrink the build time. For example, this cleanup saves about 29% of
the lm80 driver build time.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Mark M. Hoffman <mhoffman@lightlink.com>
This patch adds the ability to requeue write I/O to
core device-mapper when there is a log device failure.
If a write to the log produces and error, the pending writes are
put on the "failures" list. Since the log is marked as failed,
they will stay on the failures list until a suspend happens.
Suspends come in two phases, presuspend and postsuspend. We must
make sure that all the writes on the failures list are requeued
in the presuspend phase (a requirement of dm core). This means
that recovery must be complete (because writes may be delayed
behind it) and the failures list must be requeued before we
return from presuspend.
The mechanisms to ensure recovery is complete (or stopped) was
already in place, but needed to be moved from postsuspend to
presuspend. We rely on 'flush_workqueue' to ensure that the
mirror thread is complete and therefore, has requeued all writes
in the failures list.
Because we are using flush_workqueue, we must ensure that no
additional 'queue_work' calls will produce additional I/O
that we need to requeue (because once we return from
presuspend, we are unable to do anything about it). 'queue_work'
is called in response to the following functions:
- complete_resync_work = NA, recovery is stopped
- rh_dec (mirror_end_io) = NA, only calls 'queue_work' if it
is ready to recover the region
(recovery is stopped) or it needs
to clear the region in the log*
**this doesn't get called while
suspending**
- rh_recovery_end = NA, recovery is stopped
- rh_recovery_start = NA, recovery is stopped
- write_callback = 1) Writes w/o failures simply call
bio_endio -> mirror_end_io -> rh_dec
(see rh_dec above)
2) Writes with failures are put on
the failures list and queue_work is
called**
** write_callbacks don't happen
during suspend **
- do_failures = NA, 'queue_work' not called if suspending
- add_mirror (initialization) = NA, only done on mirror creation
- queue_bio = NA, 1) delayed I/O scheduled before flush_workqueue
is called. 2) No more I/Os are being issued.
3) Re-attempted READs can still be handled.
(Write completions are handled through rh_dec/
write_callback - mention above - and do not
use queue_bio.)
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
Enhanced partition statistics: documentation update
Enhanced partition statistics: remove old partition statistics
Enhanced partition statistics: procfs
Enhanced partition statistics: sysfs
Enhanced partition statistics: aoe fix
Enhanced partition statistics: update partition statitics
Enhanced partition statistics: core statistics
block: fixup rq_init() a bit
Manually fixed conflict in drivers/block/aoe/aoecmd.c due to statistics
support.
- File is moved to arch-v32/mach-fs/hwregs/timer_defs.h
The definitions of the line discipline numbers N_* have been moved
from asm-*/termios.h to linux/tty.h, but the Blackfin architecture
has somehow evaded that move. Bring it in line with the others.
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
[IA64] make pfm_get_task work with virtual pids
[IA64] honor notify_die() returning NOTIFY_STOP
[IA64] remove dead code: __cpu_{down,die} from !HOTPLUG_CPU
[IA64] Appoint kvm/ia64 Maintainers
[IA64] ia64_set_psr should use srlz.i
[IA64] Export three symbols for module use
[IA64] mca style cleanup
[IA64] sn_hwperf semaphore to mutex
[IA64] generalize attribute of fsyscall_gtod_data
[IA64] efi.c Add /* never reached */ annotation
[IA64] efi.c Spelling/punctuation fixes
[IA64] Make efi.c mostly fit in 80 columns
[IA64] aliasing-test: fix gcc warnings on non-ia64
[IA64] Slim-down __clear_bit_unlock
[IA64] Fix the order of atomic operations in restore_previous_kprobes on ia64
[IA64] constify function pointer tables
[IA64] fix userspace compile error in gcc_intrin.h
* 'slub-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/christoph/vm:
SLUB: fix checkpatch warnings
Use non atomic unlock
SLUB: Support for performance statistics
SLUB: Alternate fast paths using cmpxchg_local
SLUB: Use unique end pointer for each slab page.
SLUB: Deal with annoying gcc warning on kfree()
When xfs_file_readdir() exactly fills a buffer, it can move it's index
past the end of the buffer and dereference it even though the result of
the dereference is never used. On some platforms this causes an oops.
SGI-PV: 976923
SGI-Modid: xfs-linux-melb:xfs-kern:30458a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
* Drop trailing whitespace
* Fold a long line
* Rename new_client to client
* Drop redundant initializations to 0
* Drop bogus comment
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Mark M. Hoffman <mhoffman@lightlink.com>
This patch adds the calls to 'fail_mirror' if an error occurs during
mirror recovery (aka resynchronization). 'fail_mirror' is responsible
for recording the type of error by mirror device and ensuring an event
gets raised for the purpose of notifying userspace.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/dlm:
dlm: add __init and __exit marks to init and exit functions
dlm: eliminate astparam type casting
dlm: proper types for asts and basts
dlm: dlm/user.c input validation fixes
dlm: fix dlm_dir_lookup() handling of too long names
dlm: fix overflows when copying from ->m_extra to lvb
dlm: make find_rsb() fail gracefully when namelen is too large
dlm: receive_rcom_lock_args() overflow check
dlm: verify that places expecting rcom_lock have packet long enough
dlm: validate data in dlm_recover_directory()
dlm: missing length check in check_config()
dlm: use proper type for ->ls_recover_buf
dlm: do not byteswap rcom_config
dlm: do not byteswap rcom_lock
dlm: dlm_process_incoming_buffer() fixes
dlm: use proper C for dlm/requestqueue stuff (and fix alignment bug)
Update the documentation to reflect the change in userspace interface.
Signed-off-by: Jerome Marchand <jmarchan@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
- Change spin_lock + local_irq_save into spin_lock_irqsave
- Change spin_unlock + local_irq_restore into spin_unlock_irqrestore
- Return ENOTTY if ioctl is not recognized as a cris ioctl.
- Make init functions static.
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <bryan.wu@analog.com>
* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6:
[S390] dcss: Initialize workqueue before using it.
[S390] Remove BUILD_BUG_ON() in vmem code.
[S390] sclp_tty/sclp_vt220: Fix scheduling while atomic
[S390] dasd: fix panic caused by alias device offline
[S390] dasd: add ifcc handling
[S390] latencytop s390 support.
[S390] Implement ext2_find_next_bit.
[S390] Cleanup & optimize bitops.
[S390] Define GENERIC_LOCKBREAK.
[S390] console: allow vt220 console to be the only console
[S390] Fix couple of section mismatches.
[S390] Fix smp_call_function_mask semantics.
[S390] Fix linker script.
[S390] DEBUG_PAGEALLOC support for s390.
[S390] cio: Add shutdown callback for ccwgroup.
[S390] cio: Update documentation.
[S390] cio: Clean up chsc response code handling.
[S390] cio: make sense id procedure work with partial hardware response
Fix section warning:
WARNING: arch/arm/mach-at91/built-in.o(.text+0xd74): Section mismatch in reference
from the function init_programmable_clock()
to the function .init.text:at91_css_to_clk()
The function init_programmable_clock() references
the function __init at91_css_to_clk().
This is often because init_programmable_clock lacks a __init
annotation or the annotation of at91_css_to_clk is wrong.
In this case the only calls to and from init_programmable_clock()
are from code marked as "__init", so this fix is trivially correct.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Acked-by: Uwe Kleine-Knig <Uwe.Kleine-Koenig@digi.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
According to PXA300/310 and PXA320 Developer manuals,
the ASCR[RDH] "bit needs to be cleared as part of the software
initialization coming out of any reset and coming out of D3".
The latter requirement is addressed by commit
"c4d1fb627ff3072", as for the former (coming out of any reset),
the kernel relies on boot loaders and assumes that RDH bit
is cleared there. Though, not all bootloaders follow the rule
so we have to clear the bit in kernel.
We clear the RDH bit in pxa3xx_init() function since
it is always invoked after any reset. We also preserve D1S, D2S
and D3S bits from being cleared in case we invoke pxa3xx_init()
function not from normal hardware reset (e.g. kexec scenario),
so these bits can be properly referenced later.
Signed-off-by: Dmitry Krivoschekov <dmitry.krivoschekov@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Convert special mapping install from nopage to fault.
Because the "vm_file" is NULL for the special mapping, the generic VM
code has messed up "vm_pgoff" thinking that it's an anonymous mapping
and the offset does't matter. For that reason, we need to undo the
vm_pgoff offset that got added into vmf->pgoff.
[ We _really_ should clean that up - either by making this whole special
mapping code just use a real file entry rather than that ugly array of
"struct page" pointers, or by just making the VM code realize that
even if vm_file is NULL it may not be a regular anonymous mmap.
- Linus ]
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: linux-mm@kvack.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
[IA64] Fix large MCA bootmem allocation
[IA64] Simplify cpu_idle_wait
[IA64] Synchronize RBS on PTRACE_ATTACH
[IA64] Synchronize kernel RSE to user-space and back
[IA64] Rename TIF_PERFMON_WORK back to TIF_NOTIFY_RESUME
[IA64] Wire up timerfd_{create,settime,gettime} syscalls
Commit 2f569afd9ced9ebec9a6eb3dbf6f83429be0a7b4 ("CONFIG_HIGHPTE vs.
sub-page page tables") caused some build breakage due to pgtable_t only
getting declared in the CONFIG_X86_PAE case.
Move the declaration outside the PAE section.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The MCA code allocates bootmem memory for NR_CPUS, regardless
of how many cpus the system actually has. This change allocates
memory only for cpus that actually exist.
On my test system with NR_CPUS = 1024, reserved memory was reduced by 130944k.
Before: Memory: 27886976k/28111168k available (8282k code, 242304k reserved, 5928k data, 1792k init)
After: Memory: 28017920k/28111168k available (8282k code, 111360k reserved, 5928k data, 1792k init)
Signed-off-by: Russ Anderson <rja@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/core: Remove unused struct ib_device.flags member
IB/core: Add IP checksum offload support
IPoIB: Add send gather support
IPoIB: Add high DMA feature flag
IB/mlx4: Use multiple WQ blocks to post smaller send WQEs
mlx4_core: Clean up struct mlx4_buf
mlx4_core: For 64-bit systems, vmap() kernel queue buffers
IB/mlx4: Consolidate code to get an entry from a struct mlx4_buf
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/blackfin-2.6: (24 commits)
[Blackfin] arch: import defines for BF547 -- it is just like the BF548, but no CAN
[Blackfin] arch: fix build fails only include header files when enabled
[Blackfin] arch: declare default INSTALL_PATH for Blackfin ports
[Blackfin] arch: Encourage users to use the spidev character driver: Provide platform support
[Blackfin] arch: Enable UART2 and UART3 for bf548
[Blackfin] arch: Enable NET2272 on BF561-EZkit - remove request_mem_region
[Blackfin] arch:Fix BUG [#3876] pfbutton test for BTN3 on bf533 don't show complete info
[Blackfin] arch: remove duplicated definitions of the line discipline numbers N_* in asm-blackfin/termios.h
[Blackfin] arch: fix building with mtd uclinux by putting the mtd_phys option into the function it actually gets used in
[Blackfin] arch: simpler header and update dates
[Blackfin] arch: move the init sections to the end of memory
[Blackfin] arch: change the trace buffer control start/stop logic in the exception handlers
[Blackfin] arch: fix typo in printk message
[Blackfin] arch: this is an ezkit, not a stamp, so fixup the init function name
[Blackfin] arch: add slightly better help text for CPLB_INFO
[Blackfin] arch: Fix BUG - Enable ISP1362 driver to work ok with BF561
[Blackfin] arch: Fix header file information
[Blackfin] arch: Add Support for ISP1362
[Blackfin] arch: add support for cmdline partitioning to the BF533-STAMP flash map driver and enable it as a module by default
[Blackfin] arch: hook up set_irq_wake in Blackfin's irq code
...
Thanks to Kay for keeping us honest.
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Cc: "Williams, Dan J" <dan.j.williams@intel.com>
Acked-by: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is base kernel patch for ptrace RSE bug. It's basically a backport
from the utrace RSE patch I sent out several weeks ago. please review.
when a thread is stopped (ptraced), debugger might change thread's user
stack (change memory directly), and we must avoid the RSE stored in
kernel to override user stack (user space's RSE is newer than kernel's
in the case). To workaround the issue, we copy kernel RSE to user RSE
before the task is stopped, so user RSE has updated data. we then copy
user RSE to kernel after the task is resummed from traced stop and
kernel will use the newer RSE to return to user.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
CC: Roland McGrath <roland@redhat.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
x25_asy does not take an ldisc reference before calling the flush method.
Fix it to use the helper function we provide.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Krzysztof Halasa <khc@pm.waw.pl>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch acts as a preparation for using checksum offload for IB
devices capable of inserting/verifying checksum in IP packets. The
patch does not actaully turn on NETIF_F_SG - we defer that to the
patches adding checksum offload capabilities.
We only add support for send gathers for datagram mode, since existing
HW does not support checksum offload on connected QPs.
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
lib/scatterlist.c is needed by drivers/media/video/videobuf-dma-sg.c, and
we would like to be able to use the latter without PCI too, for example, on
PXA270 ARM CPU. It is then possible to create a configuration with
CONFIG_BLOCK=n, where only module code will need scatterlist.c. Therefore
it must be in obj-y.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@pengutronix.de>
Acked-by: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
All current InfiniBand devices can handle all DMA addresses, and it's
hard to imagine anyone would be silly enough to build a new device
that couldn't. Therefore, enable the NETIF_F_HIGHDMA feature for IPoIB.
This has no effect for no, but is needed when we enable gather/scatter
support and checksum stateless offloads.
Signed-off-by: Eli Cohen <eli@mellnaox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
In file included from /home/bunk/linux/kernel-2.6/git/linux-2.6/arch/xtensa/kernel/syscall.c:39:
include2/asm/unistd.h:681: error: 'sys_timerfd' undeclared here (not in a function)
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Cc: Christian Zankel <chris@zankel.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
ConnectX HCA supports shrinking WQEs, so that a single work request
can be made of multiple units of wqe_shift. This way, WRs can differ
in size, and do not have to be a power of 2 in size, saving memory and
speeding up send WR posting. Unfortunately, if we do this then the
wqe_index field in CQEs can't be used to look up the WR ID anymore, so
our implementation does this only if selective signaling is off.
Further, on 32-bit platforms, we can't use vmap() to make the QP
buffer virtually contigious. Thus we have to use constant-sized WRs to
make sure a WR is always fully within a single page-sized chunk.
Finally, we use WRs with the NOP opcode to avoid wrapping around the
queue buffer in the middle of posting a WR, and we set the
NoErrorCompletion bit to avoid getting completions with error for NOP
WRs. However, NEC is only supported starting with firmware 2.2.232,
so we use constant-sized WRs for older firmware. And, since MLX QPs
only support SEND, we use constant-sized WRs in this case.
When stamping during NOP posting, do stamping following setting of the
NOP WQE valid bit.
Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-2.6-dm: (44 commits)
dm raid1: report fault status
dm raid1: handle read failures
dm raid1: fix EIO after log failure
dm raid1: handle recovery failures
dm raid1: handle write failures
dm snapshot: combine consecutive exceptions in memory
dm: stripe enhanced status return
dm: stripe trigger event on failure
dm log: auto load modules
dm: move deferred bio flushing to workqueue
dm crypt: use async crypto
dm crypt: prepare async callback fn
dm crypt: add completion for async
dm crypt: add async request mempool
dm crypt: extract scatterlist processing
dm crypt: tidy io ref counting
dm crypt: introduce crypt_write_io_loop
dm crypt: abstract crypt_write_done
dm crypt: store sector mapping in dm_crypt_io
dm crypt: move queue functions
...
Fix the FRV cmpxchg_local by breaking the following header dependency loop :
linux/kernel.h -> linux/bitops.h -> asm-frv/bitops.h -> asm-frv/atomic.h
-> asm-frv/system.h ->
asm-generic/cmpxchg_local.h -> typecheck() defined in linux/kernel.h
and
linux/kernel.h -> linux/bitops.h -> asm-frv/bitops.h -> asm-frv/atomic.h ->
asm-generic/cmpxchg_local.h -> typecheck() defined in linux/kernel.h
In order to fix this :
- Move the atomic_test_and_ *_mask inlines from asm-frv/atomic.h (why are they
there at all anyway ? They are not touching atomic_t variables!) to
asm-frv/bitops.h.
Also fix a build issue with cmpxchg : it does not cast to (unsigned long *)
like other architectures, to deal with it in the cmpxchg_local macro.
FRV builds fine with this patch.
Thanks to Adrian Bunk <bunk@kernel.org> for spotting this bug.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Cc: Adrian Bunk <bunk@kernel.org>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'release' of git://lm-sensors.org/kernel/mhoffman/hwmon-2.6: (59 commits)
hwmon: (lm80) Add individual alarm files
hwmon: (lm80) De-macro the sysfs callbacks
hwmon: (lm80) Various cleanups
hwmon: (w83627hf) Refactor beep enable handling
hwmon: (w83627hf) Add individual alarm and beep files
hwmon: (w83627hf) Enable VBAT monitoring
hwmon: (w83627ehf) The W83627DHG has 8 VID pins
hwmon: (asb100) Add individual alarm files
hwmon: (asb100) De-macro the sysfs callbacks
hwmon: (asb100) Various cleanups
hwmon: VRM is not written to registers
hwmon: (dme1737) fix Super-IO device ID override
hwmon: (dme1737) fix divide-by-0
hwmon: (abituguru3) Add AUX4 fan input for Abit IP35 Pro
hwmon: Add support for Texas Instruments/Burr-Brown ADS7828
hwmon: (adm9240) Add individual alarm files
hwmon: (lm77) Add individual alarm files
hwmon: Discard useless I2C driver IDs
hwmon: (lm85) Make the pwmN_enable files writable
hwmon: (lm85) Return standard values in pwmN_enable
...
This patch adds extra information to the mirror status output, so that
it can be determined which device(s) have failed. For each mirror device,
a character is printed indicating the most severe error encountered. The
characters are:
* A => Alive - No failures
* D => Dead - A write failure occurred leaving mirror out-of-sync
* S => Sync - A sychronization failure occurred, mirror out-of-sync
* R => Read - A read failure occurred, mirror data unaffected
This allows userspace to properly reconfigure the mirror set.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Revert commit c6caeb7c4544608e8ae62731334661fc396c7f85 ("proc: fix the
threaded /proc/self"), since Eric says "The patch really is wrong.
There is at least one corner case in procps that cares."
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "Guillaume Chazarain" <guichaz@yahoo.fr>
Cc: "Pavel Emelyanov" <xemul@openvz.org>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since kernel virtual memory is not a problem on 64-bit systems, there
is no reason to use our own 2-layer page mapping scheme for large
kernel queue buffers on such systems. Instead, map the page list to a
single virtually contiguous buffer with vmap(), so that can we access
buffer memory via direct indexing.
Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://oss.sgi.com:8090/xfs/xfs-2.6: (62 commits)
[XFS] add __init/__exit mark to specific init/cleanup functions
[XFS] Fix oops in xfs_file_readdir()
[XFS] kill xfs_root
[XFS] keep i_nlink updated and use proper accessors
[XFS] stop updating inode->i_blocks
[XFS] Make xfs_ail_check check less by default
[XFS] Move AIL pushing into it's own thread
[XFS] use generic_permission
[XFS] stop re-checking permissions in xfs_swapext
[XFS] clean up xfs_swapext
[XFS] remove permission check from xfs_change_file_space
[XFS] prevent panic during log recovery due to bogus op_hdr length
[XFS] Cleanup various fid related bits:
[XFS] Fix xfs_lowbit64
[XFS] Remove CFORK macros and use code directly in IFORK and DFORK macros.
[XFS] kill superflous buffer locking (2nd attempt)
[XFS] Use kernel-supplied "roundup_pow_of_two" for simplicity
[XFS] Remove the BPCSHIFT and NB* based macros from XFS.
[XFS] Remove bogus assert
[XFS] optimize XFS_IS_REALTIME_INODE w/o realtime config
...
This patch gives the ability to respond-to/record device failures
that happen during read operations. It also adds the ability to
read from mirror devices that are not the primary if they are
in-sync.
There are essentially two read paths in mirroring; the direct path
and the queued path. When a read request is mapped, if the region
is 'in-sync' the direct path is taken; otherwise the queued path
is taken.
If the direct path is taken, we must record bio information so that
if the read fails we can retry it. We then discover the status of
a direct read through mirror_end_io. If the read has failed, we will
mark the device from which the read was attempted as failed (so we
don't try to read from it again), restore the bio and try again.
If the queued path is taken, we discover the results of the read
from 'read_callback'. If the device failed, we will mark the device
as failed and attempt the read again if there is another device
where this region is known to be 'in-sync'.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'cris' of git://www.jni.nu/cris: (158 commits)
CRIS v32: Remove hwregs/timer_defs.h, it is now architecture specific.
CRIS v32: Change drivers/i2c.c locking.
CRIS v32: Rewrite ARTPEC-3 gpio driver to avoid volatiles and general cleanup.
CRIS: Add new timerfd syscall entries.
MAINTAINERS: Add my information for the CRIS port.
CRIS v32: Correct spelling of bandwidth in function name.
CRIS v32: Clean up nandflash.c for ARTPEC-3 and ETRAX FS.
CRIS v10: Cleanup of drivers/gpio.c
CRIS v10: drivers/net/cris/eth_v10.c rename LED defines to CRIS_LED to avoid name clash.
CRIS: Make io_pwm_set_period members unsigned in etraxgpio.h
CRIS: Move ETRAX_AXISFLASHMAP to common Kconfig file.
CRIS: Drop regs parameter from call to profile_tick in kernel/time.c
CRIS v32: Fix minor formatting issue in mach-a3/io.c
CRIS v32: Initialize GIO even if we're rambooting in kernel/head.S
CRIS v32: Remove kernel/arbiter.c, it now exists in machine dependent directory.
CRIS v32: Minor changes to avoid errors in asm-cris/arch-v32/hwregs/reg_rdwr.h
CRIS v32: arch-v32/hwregs/intr_vect_defs.h moved to machine dependent directory.
CRIS v32: Correct offset for TASK_pid in asm-cris/arch-v32/offset.h
CRIS v32: Move register map header to machine dependent directory.
CRIS v32: Let compiler know that memory is clobbered after a break op.
...
We use struct mlx4_buf for kernel QP, CQ and SRQ buffers, and the code
to look up an entry is duplicated in get_cqe_from_buf() and the QP and
SRQ versions of get_wqe(). Factor this out into mlx4_buf_offset().
This will also make it easier to switch over to using vmap() for buffers.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use standard dynamic sysfs callbacks instead of macro-generated
functions. This makes the code more readable, and the binary smaller
(by about 34%).
As a side note, another benefit of this type of cleanup is that they
shrink the build time. For example, this cleanup saves about 29% of
the lm80 driver build time.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Mark M. Hoffman <mhoffman@lightlink.com>
This patch adds the ability to requeue write I/O to
core device-mapper when there is a log device failure.
If a write to the log produces and error, the pending writes are
put on the "failures" list. Since the log is marked as failed,
they will stay on the failures list until a suspend happens.
Suspends come in two phases, presuspend and postsuspend. We must
make sure that all the writes on the failures list are requeued
in the presuspend phase (a requirement of dm core). This means
that recovery must be complete (because writes may be delayed
behind it) and the failures list must be requeued before we
return from presuspend.
The mechanisms to ensure recovery is complete (or stopped) was
already in place, but needed to be moved from postsuspend to
presuspend. We rely on 'flush_workqueue' to ensure that the
mirror thread is complete and therefore, has requeued all writes
in the failures list.
Because we are using flush_workqueue, we must ensure that no
additional 'queue_work' calls will produce additional I/O
that we need to requeue (because once we return from
presuspend, we are unable to do anything about it). 'queue_work'
is called in response to the following functions:
- complete_resync_work = NA, recovery is stopped
- rh_dec (mirror_end_io) = NA, only calls 'queue_work' if it
is ready to recover the region
(recovery is stopped) or it needs
to clear the region in the log*
**this doesn't get called while
suspending**
- rh_recovery_end = NA, recovery is stopped
- rh_recovery_start = NA, recovery is stopped
- write_callback = 1) Writes w/o failures simply call
bio_endio -> mirror_end_io -> rh_dec
(see rh_dec above)
2) Writes with failures are put on
the failures list and queue_work is
called**
** write_callbacks don't happen
during suspend **
- do_failures = NA, 'queue_work' not called if suspending
- add_mirror (initialization) = NA, only done on mirror creation
- queue_bio = NA, 1) delayed I/O scheduled before flush_workqueue
is called. 2) No more I/Os are being issued.
3) Re-attempted READs can still be handled.
(Write completions are handled through rh_dec/
write_callback - mention above - and do not
use queue_bio.)
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
Enhanced partition statistics: documentation update
Enhanced partition statistics: remove old partition statistics
Enhanced partition statistics: procfs
Enhanced partition statistics: sysfs
Enhanced partition statistics: aoe fix
Enhanced partition statistics: update partition statitics
Enhanced partition statistics: core statistics
block: fixup rq_init() a bit
Manually fixed conflict in drivers/block/aoe/aoecmd.c due to statistics
support.
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
[IA64] make pfm_get_task work with virtual pids
[IA64] honor notify_die() returning NOTIFY_STOP
[IA64] remove dead code: __cpu_{down,die} from !HOTPLUG_CPU
[IA64] Appoint kvm/ia64 Maintainers
[IA64] ia64_set_psr should use srlz.i
[IA64] Export three symbols for module use
[IA64] mca style cleanup
[IA64] sn_hwperf semaphore to mutex
[IA64] generalize attribute of fsyscall_gtod_data
[IA64] efi.c Add /* never reached */ annotation
[IA64] efi.c Spelling/punctuation fixes
[IA64] Make efi.c mostly fit in 80 columns
[IA64] aliasing-test: fix gcc warnings on non-ia64
[IA64] Slim-down __clear_bit_unlock
[IA64] Fix the order of atomic operations in restore_previous_kprobes on ia64
[IA64] constify function pointer tables
[IA64] fix userspace compile error in gcc_intrin.h
* 'slub-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/christoph/vm:
SLUB: fix checkpatch warnings
Use non atomic unlock
SLUB: Support for performance statistics
SLUB: Alternate fast paths using cmpxchg_local
SLUB: Use unique end pointer for each slab page.
SLUB: Deal with annoying gcc warning on kfree()
When xfs_file_readdir() exactly fills a buffer, it can move it's index
past the end of the buffer and dereference it even though the result of
the dereference is never used. On some platforms this causes an oops.
SGI-PV: 976923
SGI-Modid: xfs-linux-melb:xfs-kern:30458a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
This patch adds the calls to 'fail_mirror' if an error occurs during
mirror recovery (aka resynchronization). 'fail_mirror' is responsible
for recording the type of error by mirror device and ensuring an event
gets raised for the purpose of notifying userspace.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/dlm:
dlm: add __init and __exit marks to init and exit functions
dlm: eliminate astparam type casting
dlm: proper types for asts and basts
dlm: dlm/user.c input validation fixes
dlm: fix dlm_dir_lookup() handling of too long names
dlm: fix overflows when copying from ->m_extra to lvb
dlm: make find_rsb() fail gracefully when namelen is too large
dlm: receive_rcom_lock_args() overflow check
dlm: verify that places expecting rcom_lock have packet long enough
dlm: validate data in dlm_recover_directory()
dlm: missing length check in check_config()
dlm: use proper type for ->ls_recover_buf
dlm: do not byteswap rcom_config
dlm: do not byteswap rcom_lock
dlm: dlm_process_incoming_buffer() fixes
dlm: use proper C for dlm/requestqueue stuff (and fix alignment bug)
* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6:
[S390] dcss: Initialize workqueue before using it.
[S390] Remove BUILD_BUG_ON() in vmem code.
[S390] sclp_tty/sclp_vt220: Fix scheduling while atomic
[S390] dasd: fix panic caused by alias device offline
[S390] dasd: add ifcc handling
[S390] latencytop s390 support.
[S390] Implement ext2_find_next_bit.
[S390] Cleanup & optimize bitops.
[S390] Define GENERIC_LOCKBREAK.
[S390] console: allow vt220 console to be the only console
[S390] Fix couple of section mismatches.
[S390] Fix smp_call_function_mask semantics.
[S390] Fix linker script.
[S390] DEBUG_PAGEALLOC support for s390.
[S390] cio: Add shutdown callback for ccwgroup.
[S390] cio: Update documentation.
[S390] cio: Clean up chsc response code handling.
[S390] cio: make sense id procedure work with partial hardware response