Linus Torvalds [Fri, 1 Mar 2013 01:43:09 +0000 (17:43 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client
Pull Ceph updates from Sage Weil:
"A few groups of patches here. Alex has been hard at work improving
the RBD code, layout groundwork for understanding the new formats and
doing layering. Most of the infrastructure is now in place for the
final bits that will come with the next window.
There are a few changes to the data layout. Jim Schutt's patch fixes
some non-ideal CRUSH behavior, and a set of patches from me updates
the client to speak a newer version of the protocol and implement an
improved hashing strategy across storage nodes (when the server side
supports it too).
A pair of patches from Sam Lang fix the atomicity of open+create
operations. Several patches from Yan, Zheng fix various mds/client
issues that turned up during multi-mds torture tests.
A final set of patches expose file layouts via virtual xattrs, and
allow the policies to be set on directories via xattrs as well
(avoiding the awkward ioctl interface and providing a consistent
interface for both kernel mount and ceph-fuse users)."
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (143 commits)
libceph: add support for HASHPSPOOL pool flag
libceph: update osd request/reply encoding
libceph: calculate placement based on the internal data types
ceph: update support for PGID64, PGPOOL3, OSDENC protocol features
ceph: update "ceph_features.h"
libceph: decode into cpu-native ceph_pg type
libceph: rename ceph_pg -> ceph_pg_v1
rbd: pass length, not op for osd completions
rbd: move rbd_osd_trivial_callback()
libceph: use a do..while loop in con_work()
libceph: use a flag to indicate a fault has occurred
libceph: separate non-locked fault handling
libceph: encapsulate connection backoff
libceph: eliminate sparse warnings
ceph: eliminate sparse warnings in fs code
rbd: eliminate sparse warnings
libceph: define connection flag helpers
rbd: normalize dout() calls
rbd: barriers are hard
rbd: ignore zero-length requests
...
NFSv4.1: LAYOUTGET EDELAY loops timeout to the MDS
The client will currently try LAYOUTGETs forever if a server is returning
NFS4ERR_LAYOUTTRYLATER or NFS4ERR_RECALLCONFLICT - even if the client no
longer needs the layout (ie process killed, unmounted).
This patch uses the DS timeout value (module parameter 'dataserver_timeo'
via rpc layer) to set an upper limit of how long the client tries LATOUTGETs
in this situation. Once the timeout is reached, IO is redirected to the MDS.
This also changes how the client checks if a layout is on the clp list
to avoid a double list_add.
The client should have 60 second default timeouts for DS operations, not 6
seconds.
NFS4_DEF_DS_TIMEO is used as "timeout in tenths of a second" in
nfs_init_timeout_values (and is not used anywhere else).
This matches up with the description of the module param dataserver_timeo.
Trond Myklebust [Fri, 1 Mar 2013 00:19:59 +0000 (16:19 -0800)]
NFSv4: Fix another open/open_recovery deadlock
If we don't release the open seqid before we wait for state recovery,
then we may end up deadlocking the state recovery thread.
This patch addresses a new deadlock that was introduced by
commit c21443c2c792cd9b463646d982b0fe48aa6feb0f (NFSv4: Fix a reboot
recovery race when opening a file)
Linus Torvalds [Thu, 28 Feb 2013 21:21:44 +0000 (13:21 -0800)]
Merge tag 'writeback-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/wfg/linux
Pull writeback fixes from Wu Fengguang:
"Two writeback fixes
- fix negative (setpoint - dirty) in 32bit archs
- use down_read_trylock() in writeback_inodes_sb(_nr)_if_idle()"
* tag 'writeback-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/wfg/linux:
Negative (setpoint-dirty) in bdi_position_ratio()
vfs: re-implement writeback_inodes_sb(_nr)_if_idle() and rename them
Linus Torvalds [Thu, 28 Feb 2013 21:16:07 +0000 (13:16 -0800)]
Merge branch 'for-3.9/drivers' of git://git.kernel.dk/linux-block
Pull block driver bits from Jens Axboe:
"After the block IO core bits are in, please grab the driver updates
from below as well. It contains:
- Fix ancient regression in dac960. Nobody must be using that
anymore...
- Some good fixes from Guo Ghao for loop, fixing both potential
oopses and deadlocks.
- Improve mtip32xx for NUMA systems, by being a bit more clever in
distributing work.
- Add IBM RamSan 70/80 driver. A second round of fixes for that is
pending, that will come in through for-linus during the 3.9 cycle
as per usual.
- A few xen-blk{back,front} fixes from Konrad and Roger.
- Other minor fixes and improvements."
* 'for-3.9/drivers' of git://git.kernel.dk/linux-block:
loopdev: ignore negative offset when calculate loop device size
loopdev: remove an user triggerable oops
loopdev: move common code into loop_figure_size()
loopdev: update block device size in loop_set_status()
loopdev: fix a deadlock
xen-blkback: use balloon pages for persistent grants
xen-blkfront: drop the use of llist_for_each_entry_safe
xen/blkback: Don't trust the handle from the frontend.
xen-blkback: do not leak mode property
block: IBM RamSan 70/80 driver fixes
rsxx: add slab.h include to dma.c
drivers/block/mtip32xx: add missing GENERIC_HARDIRQS dependency
block: remove new __devinit/exit annotations on ramsam driver
block: IBM RamSan 70/80 device driver
drivers/block/mtip32xx/mtip32xx.c:1726:5: sparse: symbol 'mtip_send_trim' was not declared. Should it be static?
drivers/block/mtip32xx/mtip32xx.c:4029:1: sparse: symbol 'mtip_workq_sdbf0' was not declared. Should it be static?
dac960: return success instead of -ENOTTY
mtip32xx: add trim support
mtip32xx: Add workqueue and NUMA support
block: delete super ancient PC-XT driver for 1980's hardware
Linus Torvalds [Thu, 28 Feb 2013 20:52:24 +0000 (12:52 -0800)]
Merge branch 'for-3.9/core' of git://git.kernel.dk/linux-block
Pull block IO core bits from Jens Axboe:
"Below are the core block IO bits for 3.9. It was delayed a few days
since my workstation kept crashing every 2-8h after pulling it into
current -git, but turns out it is a bug in the new pstate code (divide
by zero, will report separately). In any case, it contains:
- The big cfq/blkcg update from Tejun and and Vivek.
- Additional block and writeback tracepoints from Tejun.
- Improvement of the should sort (based on queues) logic in the plug
flushing.
- _io() variants of the wait_for_completion() interface, using
io_schedule() instead of schedule() to contribute to io wait
properly.
- Various little fixes.
You'll get two trivial merge conflicts, which should be easy enough to
fix up"
Fix up the trivial conflicts due to hlist traversal cleanups (commit b67bfe0d42ca: "hlist: drop the node parameter from iterators").
* 'for-3.9/core' of git://git.kernel.dk/linux-block: (39 commits)
block: remove redundant check to bd_openers()
block: use i_size_write() in bd_set_size()
cfq: fix lock imbalance with failed allocations
drivers/block/swim3.c: fix null pointer dereference
block: don't select PERCPU_RWSEM
block: account iowait time when waiting for completion of IO request
sched: add wait_for_completion_io[_timeout]
writeback: add more tracepoints
block: add block_{touch|dirty}_buffer tracepoint
buffer: make touch_buffer() an exported function
block: add @req to bio_{front|back}_merge tracepoints
block: add missing block_bio_complete() tracepoint
block: Remove should_sort judgement when flush blk_plug
block,elevator: use new hashtable implementation
cfq-iosched: add hierarchical cfq_group statistics
cfq-iosched: collect stats from dead cfqgs
cfq-iosched: separate out cfqg_stats_reset() from cfq_pd_reset_stats()
blkcg: make blkcg_print_blkgs() grab q locks instead of blkcg lock
block: RCU free request_queue
blkcg: implement blkg_[rw]stat_recursive_sum() and blkg_[rw]stat_merge()
...
Linus Torvalds [Thu, 28 Feb 2013 20:43:43 +0000 (12:43 -0800)]
Merge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull first round of SCSI updates from James Bottomley:
"The patch set is mostly driver updates (bnx2fc, ipr, lpfc, qla4) and a
few bug fixes"
Pull delayed because google hates James, and sneakily considers his pull
requests spam. Why, google, why?
* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (60 commits)
[SCSI] aacraid: 1024 max outstanding command support for Series 7 and above
[SCSI] bnx2fc: adjust duplicate test
[SCSI] qla4xxx: Update driver version to 5.03.00-k4
[SCSI] qla4xxx: Fix return code for qla4xxx_session_get_param.
[SCSI] qla4xxx: wait for boot target login response during probe.
[SCSI] qla4xxx: Added support for force firmware dump
[SCSI] qla4xxx: Re-register IRQ handler while retrying initialize of adapter
[SCSI] qla4xxx: Throttle active IOCBs to firmware limits
[SCSI] qla4xxx: Remove unnecessary code from qla4xxx_init_local_data
[SCSI] qla4xxx: Quiesce driver activities while loopback
[SCSI] qla4xxx: Rename MBOX_ASTS_IDC_NOTIFY to MBOX_ASTS_IDC_REQUEST_NOTIFICATION
[SCSI] qla4xxx: Add spurious interrupt messages under debug level 2
[SCSI] cxgb4i: Remove the scsi host device when removing device
[SCSI] bfa: fix strncpy() limiter in bfad_start_ops()
[SCSI] qla4xxx: Update driver version to 5.03.00-k3
[SCSI] qla4xxx: Correct the validation to check in get_sys_info mailbox
[SCSI] qla4xxx: Pass correct function param to qla4_8xxx_rd_direct
[SCSI] lpfc 8.3.37: Update lpfc version for 8.3.37 driver release
[SCSI] lpfc 8.3.37: Fixed infinite loop in lpfc_sli4_fcf_rr_next_index_get.
[SCSI] lpfc 8.3.37: Fixed crash due to SLI Port invalid resource count
...
Vlastimil Kosar [Thu, 28 Feb 2013 08:45:22 +0000 (08:45 +0000)]
net/phy: micrel: Disable asymmetric pause for KSZ9021
Phyter KSZ9021 has hardware bug. If asymmetric pause is enabled,
then it is necessary to disconnect and then reconnect the ethernet
cable to get the phyter working. The solution is to disable the
asymmetric pause.
Hauke Mehrtens [Thu, 28 Feb 2013 07:16:54 +0000 (07:16 +0000)]
bgmac: omit the fcs
Do not include the frame check sequence when adding the skb to
netif_receive_skb(). This causes problems when this interface was
bridged to a wifi ap and a big package should be forwarded from this
Ethernet driver through a bride to the wifi client.
Petr Malat [Thu, 28 Feb 2013 01:01:52 +0000 (01:01 +0000)]
phy: Fix phy_device_free memory leak
Fix memory leak in phy_device_free() for the case when phy_device*
returned by phy_device_create() is not registered in the system.
Bug description:
phy_device_create() sets name of kobject using dev_set_name(), which
allocates memory using kvasprintf(), but this memory isn't freed if
the underlying device isn't registered properly, because kobject_cleanup()
is not called in that case. This can happen (and actually is happening on
our machines) if phy_device_register(), called by mdiobus_scan(), fails.
Patch description:
Embedded struct device is initialized in phy_device_create() and it
counterpart phy_device_free() just drops one reference to the device,
which leads to proper deinitialization including releasing the kobject
name memory.
Yaniv Rosner [Wed, 27 Feb 2013 13:06:45 +0000 (13:06 +0000)]
bnx2x: Fix KR2 link
Fix KR2 link down problem after reboot when link speed is reconfigured via ethtool.
Since 1G/10G support link speed were missing by default, 1G/10G link speed were
not advertised.
Yaniv Rosner [Wed, 27 Feb 2013 13:06:44 +0000 (13:06 +0000)]
bnx2x: Fix port identification for the 84834
Fix the "ethtool -p" for boards with BCM84834, by using LED4 of the PHY
to toggle the link LED while keeping interrupt disabled to avoid NIG attentions,
and at the end restore NIG to previous state.
françois romieu [Wed, 27 Feb 2013 13:01:57 +0000 (13:01 +0000)]
r8169: honor jumbo settings when chipset is requested to start.
Some hardware start settings implicitely assume an usual 1500 bytes mtu
that can't be guaranteed because changes of mtu may be requested both
before and after the hardware is started.
Eric Dumazet [Wed, 27 Feb 2013 07:05:03 +0000 (07:05 +0000)]
tcp: avoid wakeups for pure ACK
TCP prequeue mechanism purpose is to let incoming packets
being processed by the thread currently blocked in tcp_recvmsg(),
instead of behalf of the softirq handler, to better adapt flow
control on receiver host capacity to schedule the consumer.
But in typical request/answer workloads, we send request, then
block to receive the answer. And before the actual answer, TCP
stack receives the ACK packets acknowledging the request.
Processing pure ACK on behalf of the thread blocked in tcp_recvmsg()
is a waste of resources, as thread has to immediately sleep again
because it got no payload.
This patch avoids the extra context switches and scheduler overhead.
David S. Miller [Thu, 28 Feb 2013 20:34:36 +0000 (15:34 -0500)]
Merge branch 'sctp'
Lee A. Roberts says:
====================
This series of patches resolves several SCTP association hangs observed during
SCTP stress testing. Observable symptoms include communications hangs with
data being held in the association reassembly and/or lobby (ordering) queues.
Close examination of reassembly/ordering queues may show either duplicated
or missing packets.
In version #2, corrected build failure in initial version of patch series
due to wrong calling sequence for sctp_ulpq_partial_delivery() being inserted
in sctp_ulpq_renege().
In version #3, adjusted patch documentation to be less repetitive.
====================
Lee A. Roberts [Thu, 28 Feb 2013 04:37:30 +0000 (04:37 +0000)]
sctp: fix association hangs due to partial delivery errors
In sctp_ulpq_tail_data(), use return values 0,1 to indicate whether
a complete event (with MSG_EOR set) was delivered. A return value
of -ENOMEM continues to indicate an out-of-memory condition was
encountered.
In sctp_ulpq_retrieve_partial() and sctp_ulpq_retrieve_first(),
correct message reassembly logic for SCTP partial delivery.
Change logic to ensure that as much data as possible is sent
with the initial partial delivery and that following partial
deliveries contain all available data.
In sctp_ulpq_partial_delivery(), attempt partial delivery only
if the data on the head of the reassembly queue is at or before
the cumulative TSN ACK point.
In sctp_ulpq_renege(), use the modified return values from
sctp_ulpq_tail_data() to choose whether to attempt partial
delivery or to attempt to drain the reassembly queue as a
means to reduce memory pressure. Remove call to
sctp_tsnmap_mark(), as this is handled correctly in call to
sctp_ulpq_tail_data().
Lee A. Roberts [Thu, 28 Feb 2013 04:37:29 +0000 (04:37 +0000)]
sctp: fix association hangs due to errors when reneging events from the ordering queue
In sctp_ulpq_renege_list(), events being reneged from the
ordering queue may correspond to multiple TSNs. Identify
all affected packets; sum freed space and renege from the
tsnmap.
Lee A. Roberts [Thu, 28 Feb 2013 04:37:27 +0000 (04:37 +0000)]
sctp: fix association hangs due to off-by-one errors in sctp_tsnmap_grow()
In sctp_tsnmap_mark(), correct off-by-one error when calculating
size value for sctp_tsnmap_grow().
In sctp_tsnmap_grow(), correct off-by-one error when copying
and resizing the tsnmap. If max_tsn_seen is in the LSB of the
word, this bit can be lost, causing the corresponding packet
to be transmitted again and to be entered as a duplicate into
the SCTP reassembly/ordering queues. Change parameter name
from "gap" (zero-based index) to "size" (one-based) to enhance
code readability.
irq: Don't re-enable interrupts at the end of irq_exit
Commit 74eed0163d0def3fce27228d9ccf3d36e207b286
"irq: Ensure irq_exit() code runs with interrupts disabled"
restore interrupts flags in the end of irq_exit() for archs
that don't define __ARCH_IRQ_EXIT_IRQS_DISABLED.
However always returning from irq_exit() with interrupts
disabled should not be a problem for these archs. Prior to
this commit this was already happening anytime we processed
pending softirqs anyway.
David Sterba [Thu, 28 Feb 2013 14:54:18 +0000 (14:54 +0000)]
btrfs: use only inline_pages from extent buffer
The nodesize is capped at 64k and there are enough pages preallocated in
extent_buffer::inline_pages. The fallback to kmalloc never happened
because even on the smallest page size considered (4k) inline_pages
covered the needs.
Miao Xie [Thu, 28 Feb 2013 10:05:36 +0000 (10:05 +0000)]
Btrfs: fix wrong reserved space when deleting a snapshot/subvolume
When deleting a snapshot/subvolume, we need remove root ref/backref,
dir entries and update the dir inode, so we must reserve free space
for those operations.
Miao Xie [Thu, 28 Feb 2013 10:04:33 +0000 (10:04 +0000)]
Btrfs: fix wrong reserved space in qgroup during snap/subv creation
There are two problems in the space reservation of the snapshot/
subvolume creation.
- don't reserve the space for the root item insertion
- the space which is reserved in the qgroup is different with
the free space reservation. we need reserve free space for
7 items, but in qgroup reservation, we need reserve space only
for 3 items.
So we implement new metadata reservation functions for the
snapshot/subvolume creation.
Miao Xie [Thu, 28 Feb 2013 10:01:15 +0000 (10:01 +0000)]
Btrfs: remove unnecessary dget_parent/dput when creating the pending snapshot
Since we have grabbed the parent inode at the beginning of the
snapshot creation, and both sync and async snapshot creation
release it after the pending snapshots are actually created,
it is safe to access the parent inode directly during the snapshot
creation, we needn't use dget_parent/dput to fix the parent dentry
and get the dir inode.
David Sterba [Wed, 27 Feb 2013 23:13:55 +0000 (23:13 +0000)]
btrfs: remove a printk from scan_one_device
Dave pointed out that he saw messages from btrfs although there was no
such filesystem on his computers. The automatic device scan is called on
every new blockdevice if the usual distro udev rule set is used. The
printk introduced in 6f60cbd3ae442c was a remainder from copying
portions of code from btrfs_get_bdev_and_sb which is used under
different conditions and the warning makes sense there.
Liu Bo [Wed, 27 Feb 2013 13:28:25 +0000 (13:28 +0000)]
Btrfs: fix NULL pointer after aborting a transaction
While doing cleanup work on an aborted transaction, we've set
the global running transaction pointer to NULL _before_ waiting all
other transaction handles to finish, so others'd hit NULL pointer
crash when referencing the global running transaction pointer.
This first sets a hint to avoid new transaction handle joining, then
waits other existing handles to abort or finish so that we can safely
set the above global pointer to NULL.
Josef Bacik [Thu, 28 Feb 2013 18:23:38 +0000 (13:23 -0500)]
Btrfs: copy everything if we've created an inline extent
I noticed while looking into a tree logging bug that we aren't logging inline
extents properly. Since this requires copying and it shouldn't happen too often
just force us to copy everything for the inode into the tree log when we have an
inline extent. With this patch we have valid data after a crash when we write
an inline extent. Thanks,
arm: mvebu: enable the SD card slot on Armada 370 Reference Design board
The Armada 370 Reference Design board has one SD card slot, directly
connected to the SDIO IP of the SoC, so we enable this IP. there are no
GPIOs for card-detect and write-protect so we do not specify any.
ARM: dove: convert serial DT nodes to clocks property
of_serial now has support for using clocks property and we have
a DT clock provider. This patch replaces the hard coded clock-frequency
property with a clocks phandle to tclk.
Gregory CLEMENT [Tue, 5 Feb 2013 20:54:55 +0000 (21:54 +0100)]
arm: mvebu: Add SPI flash on Armada 370 DB board
This patch add support for the SPI flash MX25l25635E which is present
on the Armada 370 DB board. This flash stores the bootloader and its
environment.
Ezequiel Garcia [Wed, 6 Feb 2013 13:06:21 +0000 (10:06 -0300)]
arm: mvebu: Add support for SPI controller in Armada 370/XP
The Armada 370 and Armada XP SoC has an SPI controller.
This patch adds support for this controller in Armada 370
and Armada XP SoC common device tree files.
Note that the Armada XP SPI register length is 0x50 bytes,
while Armada 370 SPI register length is 0x28 bytes,
so we choose the smaller of the two.
Gregory CLEMENT [Fri, 25 Jan 2013 17:32:45 +0000 (18:32 +0100)]
clocksource: update and move armada-370-xp-timer documentation to timer directory
Timer driver for Armada 370 and Armada XP have gained local timers
support. So it needs new resources information regarding the IRQs
and the registers.
Also move the documentation in the new and more accurate directory
Ezequiel Garcia [Wed, 23 Jan 2013 15:26:31 +0000 (12:26 -0300)]
arm: mvebu: Enable USB controllers on Armada 370/XP boards
This patch activates every USB port provided by each SoC.
Except for Armada XP Openblocks AX3-4 board,
where we enable only the first two USB ports
until we have more information on the third one usage.
Ezequiel Garcia [Wed, 23 Jan 2013 15:26:30 +0000 (12:26 -0300)]
arm: mvebu: Add support for USB host controllers in Armada 370/XP
The Armada 370 and Armada XP SoC has an Orion EHCI USB controller.
This patch adds support for this controller in Armada 370
and Armada XP SoC common device tree files.
The OpenBlocks AX3-4 board has one software-controlled button on the
front side, labeled "INIT", so we add minimal support for this button
in the kernel.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:12 +0000 (15:49 +0100)]
arm: kirkwood: mplcec4: use Device Tree to probe SDIO
Now that the mvsdio driver has a Device Tree binding, and the SDIO
controller is declared in kirkwood.dtsi, migrate the mplcec4 board to
use the Device Tree to probe the SDIO controller and to mux the pins
of the SDIO interface correctly.
This patch has not been tested, it remains to be tested by a person
having access to the hardware.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:11 +0000 (15:49 +0100)]
arm: kirkwood: dreamplug: use Device Tree to probe SDIO
Now that the mvsdio driver has a Device Tree binding, and the SDIO
controller is declared in kirkwood.dtsi, migrate the dreamplug board
to use the Device Tree to probe the SDIO controller and to mux this
interface properly.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:09 +0000 (15:49 +0100)]
arm: mvebu: enable the SDIO interface on the Globalscale Mirabox
The Globalscale Mirabox uses the SDIO interface of the Armada 370 to
connect to a Wifi/Bluetooth SD8787 chip, so we enable the SDIO
interface of this board in its Device Tree file.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:08 +0000 (15:49 +0100)]
arm: mvebu: enable the SD card slot on Armada 370 DB board
The Armada XP DB evaluation board has one SD card slot, directly
connected to the SDIO IP of the SoC, so we add a device tree
description for it.
However, in the default configuration of the board, the SD card slot
is not usable: the connector plugged into CON40 must be changed
against a different one, provided with the board by the
manufacturer. Since such a manual modification of the hardware is
needed, we did not enable the SDIO interface by default, and left it
to the board user to modify the Device Tree if needed. Since this
board is really only an evaluation board for developers and not a
final product, it is not too bad.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:07 +0000 (15:49 +0100)]
arm: mvebu: enable the SD card slot on Armada XP DB board
The Armada XP DB evaluation board has one SD card slot, directly
connected to the SDIO IP of the SoC, so we enable this
IP. Unfortunately, there are no GPIOs for card-detect and
write-protect.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:06 +0000 (15:49 +0100)]
arm: mvebu: add pin muxing options for the SDIO interface on Armada XP
The SDIO interface is only available on pins MPP30/31/32/33/34/35 on
the various Armada XP variants, so we provide a pin muxing option for
this in the Armada XP .dtsi files.
Even though those muxing options are the same for MV78230, MV78260 and
MV78460, we keep them in each .dtsi file, because the number of pins,
and therefore the declaration of the pinctrl node, is different for
each SoC variant.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:05 +0000 (15:49 +0100)]
arm: mvebu: add pin muxing options for the SDIO interface on Armada 370
The SDIO interface is available either on pins MPP9/11/12/13/14/15 or
MPP47/48/49/50/51/52 on the Armada 370. Even though all combinations
are potentially possible, those two muxing options are the most
probable ones, so we provide those at the SoC level .dtsi file.
In practice, in turns out the Armada 370 DB board uses the former,
while the Armada 370 Mirabox uses the latter.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:04 +0000 (15:49 +0100)]
arm: mvebu: add DT information for the SDIO interface of Armada 370/XP
Now that the mvsdio MMC driver has a Device Tree binding, we add the
Device Tree informations to describe the SDIO interface available in
the Armada 370/XP SoCs.
Gregory CLEMENT [Wed, 12 Dec 2012 09:06:24 +0000 (10:06 +0100)]
arm: mvebu: Add RTC support for Armada 370 and Armada XP
The Armada 370 and Armada XP Socs have the same controller that the
one used in the orion platforms. This patch updates the device tree
for these SoCs.
Gregory CLEMENT [Mon, 4 Feb 2013 17:21:07 +0000 (18:21 +0100)]
arm: mvebu: update defconfig with ATAG support when using DT
Some of the mvebu boards (mainly the development board) come with
plug-in RAM modules. This patch allows to let the bootloaders which
have no support for DTS to give the real amount of memory available on
the board.
Gregory CLEMENT [Fri, 25 Jan 2013 17:32:42 +0000 (18:32 +0100)]
clocksource: time-armada-370-xp: add local timer support
On the SOCs Armada 370 and Armada XP, each CPU comes with two private
timers. This patch use the timer 0 of each CPU as local timer for the
clockevent if CONFIG_LOCAL_TIMER is selected. In the other case, use
only the private Timer 0 of CPU 0.
arm: mvebu: add DTS file for Marvell RD-A370-A1 board
This patch adds the DTS file to support the Marvell RD-A370-A1
(Reference Design board) also known as RD-88F6710 board. It is almost
entirely similar to the DB-A370 board except that the first Ethernet PHY
is SGMII-wired and the second is a switch which is RGMII-wired.
Gregory CLEMENT [Wed, 5 Dec 2012 20:43:23 +0000 (21:43 +0100)]
arm: mvebu: Improve the SMP support of the interrupt controller
This patch makes the interrupt controller driver more SMP aware for
the Armada XP SoCs. It adds the support for the per-CPU irq. It also
adds the implementation for the set_affinity hook.
Patch initialy wrote by Yehuda Yitschak and reworked by Gregory
CLEMENT.
Arnd Bergmann [Thu, 28 Feb 2013 17:19:16 +0000 (18:19 +0100)]
ARM: Dove: split legacy and DT setup
In the beginning of DT for Dove it was reasonable to have it close to
non-DT code. With improved DT support, it became more and more difficult
to not break non-DT while changing DT code.
This patch splits up DT board setup and introduces a DOVE_LEGACY config
to allow to remove legacy code for DT-only kernels.
The OpenBlocks AX3-4 platform has several LEDs, so it sounds wise to
enable LED support in mvebu_defconfig. We anticipate that more
platforms using Marvell EBU SoCs will have LEDs in the future.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:19 +0000 (15:49 +0100)]
arm: mvebu: enable btmrvl driver in mvebu_defconfig
The Globalscale Mirabox platform, based on the Armada 370 from
Marvell, has a SD8787 Wireless/Bluetooth chip connected on the SDIO
interface. Now that the mvsdio has a Device Tree binding, and the
necessary Device Tree informations have been added at the SoC and
board level, let's enable the btmrvl driver for the Bluetooth part of
the SD8787 chip.
For now, the driver gets probed correctly, detects the device but
apparently fails to push the firmware to the device:
Bluetooth: vendor=0x2df, device=0x911a, class=255, fn=2
Bluetooth: FW failed to be active in time!
Bluetooth: Downloading firmware failed!
Bluetooth: vendor=0x2df, device=0x911b, class=255, fn=3
Bluetooth: FW failed to be active in time!
Bluetooth: Downloading firmware failed!
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:18 +0000 (15:49 +0100)]
arm: mvebu: enable mwifiex driver in mvebu_defconfig
The Globalscale Mirabox platform, based on the Armada 370 from
Marvell, has a SD8787 Wireless chip connected on the SDIO
interface. Now that the mvsdio has a Device Tree binding, and the
necessary Device Tree informations have been added at the SoC and
board level, let's enable the mwifiex driver for the Wireless part of
the SD8787 chip.
For now, the driver gets probed correctly, detects a device and shows
the network interfaces. However, scanning Wifi networks doesn't work
for now, with a 'CMD_RESP: cmd 0x6 error, result=0x1' message. This
will have to be investigated separately.
Thomas Petazzoni [Fri, 21 Dec 2012 14:49:17 +0000 (15:49 +0100)]
arm: mvebu: enable SDIO support in mvebu_defconfig
Now that the mvsdio driver has gained Device Tree support and the
necessary Device Tree informations has been added for Armada 370 and
Armada XP platforms, we enable the MMC subsystem and the mvsdio driver
in mvebu_defconfig.
Gregory CLEMENT [Wed, 12 Dec 2012 09:06:26 +0000 (10:06 +0100)]
arm: mvebu: Update defconfig with Marvell RTC support
The RTC class driver is already part of the mvebu_defconfig but the
Marvell internal RTC not yet. Now that its support is added for mvebu
let's update the config file.
J. Bruce Fields [Wed, 20 Feb 2013 22:52:19 +0000 (17:52 -0500)]
SUNRPC: make AF_LOCAL connect synchronous
It doesn't appear that anyone actually needs to connect asynchronously.
Also, using a workqueue for the connect means we lose the namespace
information from the original process. This is a problem since there's
no way to explicitly pass in a filesystem namespace for resolution of an
AF_LOCAL address.
Ouyang Maochun [Mon, 18 Feb 2013 15:54:52 +0000 (09:54 -0600)]
cifs: bugfix for unreclaimed writeback pages in cifs_writev_requeue()
Pages get the PG_writeback flag set before cifs sends its
request to SMB server in cifs_writepages(), if the SMB service
goes down, cifs may try to recommit the writing requests in
cifs_writev_requeue(). However, it does not clean its PG_writeback
flag and relaimed the pages even if it fails again in
cifs_writev_requeue(), which may lead to the hanging of the
processes accessing the cifs directory. This patch just cleans
the PG_writeback flags and reclaims the pages under that circumstances.
Steps to reproduce the bug(trying serveral times may trigger the issue):
1.Write from cifs client continuously.(e.g dd if=/dev/zero of=<cifs file>)
2.Stop SMB service from server.(e.g service smb stop)
3.Wait for two minutes， and then start SMB service from
server.(e.g service smb start)
4.The processes which are accessing cifs directory may hang up.
The git commit 8eaffa67b43e99ae581622c5133e20b0f48bcef1
(xen/pat: Disable PAT support for now) explains in details why
we want to disable PAT for right now. However that
change was not enough and we should have also disabled
the pat_enabled value. Otherwise we end up with:
mmap-example:3481 map pfn expected mapping type write-back for
[mem 0x00010000-0x00010fff], got uncached-minus
------------[ cut here ]------------
WARNING: at /build/buildd/linux-3.8.0/arch/x86/mm/pat.c:774 untrack_pfn+0xb8/0xd0()
mem 0x00010000-0x00010fff], got uncached-minus
------------[ cut here ]------------
WARNING: at /build/buildd/linux-3.8.0/arch/x86/mm/pat.c:774
untrack_pfn+0xb8/0xd0()
...
Pid: 3481, comm: mmap-example Tainted: GF 3.8.0-6-generic #13-Ubuntu
Call Trace:
[<ffffffff8105879f>] warn_slowpath_common+0x7f/0xc0
[<ffffffff810587fa>] warn_slowpath_null+0x1a/0x20
[<ffffffff8104bcc8>] untrack_pfn+0xb8/0xd0
[<ffffffff81156c1c>] unmap_single_vma+0xac/0x100
[<ffffffff81157459>] unmap_vmas+0x49/0x90
[<ffffffff8115f808>] exit_mmap+0x98/0x170
[<ffffffff810559a4>] mmput+0x64/0x100
[<ffffffff810560f5>] dup_mm+0x445/0x660
[<ffffffff81056d9f>] copy_process.part.22+0xa5f/0x1510
[<ffffffff81057931>] do_fork+0x91/0x350
[<ffffffff81057c76>] sys_clone+0x16/0x20
[<ffffffff816ccbf9>] stub_clone+0x69/0x90
[<ffffffff816cc89d>] ? system_call_fastpath+0x1a/0x1f
---[ end trace 4918cdd0a4c9fea4 ]---
(a similar message shows up if you end up launching 'mcelog')
The call chain is (as analyzed by Liu, Jinsong):
do_fork
--> copy_process
--> dup_mm
--> dup_mmap
--> copy_page_range
--> track_pfn_copy
--> reserve_pfn_range
--> line 624: flags != want_flags
It comes from different memory types of page table (_PAGE_CACHE_WB) and MTRR
(_PAGE_CACHE_UC_MINUS).
Stefan Bader dug in this deep and found out that:
"That makes it clearer as this will do
And that can return -1/0xff in case of MTRR not being enabled/initialized. Which
is not the case (given there are no messages for it in dmesg). This is not equal
to MTRR_TYPE_WRBACK and thus becomes _PAGE_CACHE_UC_MINUS.
It looks like the problem starts early in reserve_memtype:
if (!pat_enabled) {
/* This is identical to page table setting without PAT */
if (new_type) {
if (req_type == _PAGE_CACHE_WC)
*new_type = _PAGE_CACHE_UC_MINUS;
else
*new_type = req_type & _PAGE_CACHE_MASK;
}
return 0;
}
This would be what we want, that is clearing the PWT and PCD flags from the
supported flags - if pat_enabled is disabled."
Heiko Carstens [Wed, 27 Feb 2013 15:28:20 +0000 (16:28 +0100)]
s390/module: fix compile warning
Get rid of this one (false positive):
arch/s390/kernel/module.c: In function ‘apply_relocate_add’:
arch/s390/kernel/module.c:404:5: warning: ‘rc’ may be used uninitialized
in this function [-Wmaybe-uninitialized]
arch/s390/kernel/module.c:225:6: note: ‘rc’ was declared here
Play safe and preinitialize rc with an error value, so we see an error
if new users indeed don't initialize it.
Heiko Carstens [Mon, 25 Feb 2013 08:10:23 +0000 (09:10 +0100)]
s390/uaccess: fix kernel ds access for page table walk
When the kernel resides in home space and the mvcos instruction is not
available uaccesses for kernel ds happen via simple strnlen() or memcpy()
calls.
This however can break badly, since uaccesses in kernel space may fail as
well, especially if CONFIG_DEBUG_PAGEALLOC is turned on.
To fix this implement strnlen_kernel() and copy_in_kernel() functions
which can only be used by the page table uaccess functions. These two
functions detect invalid memory accesses and return the correct length
of processed data.. Both functions are more or less a copy of the std
variants without sacf calls.
Fixes ipl crashes on 31 bit machines as well on 64 bit machines without
mvcos. Caused by changing the default address space of the kernel being
home space.
The "standard" and page table walk variants of strncpy_from_user() first
check the length of the to be copied string in userspace.
The string is then copied to kernel space and the length returned to the
caller.
However userspace can modify the string at any time while the kernel
checks for the length of the string or copies the string. In result the
returned length of the string is not necessarily correct.
Fix this by copying in a loop which mimics the mvcos variant of
strncpy_from_user(), which handles this correctly.
Heiko Carstens [Thu, 21 Feb 2013 12:30:42 +0000 (13:30 +0100)]
s390/uaccess: fix strncpy_from_user/strnlen_user zero maxlen case
If the maximum length specified for the to be accessed string for
strncpy_from_user() and strnlen_user() is zero the following incorrect
values would be returned or incorrect memory accesses would happen:
strnlen_user_std() and strnlen_user_pt() incorrectly return "1"
strncpy_from_user_pt() would incorrectly access "dst[maxlen - 1]"
strncpy_from_user_mvcos() would incorrectly return "-EFAULT"
projects / linux.git / log