A user space program using the transactional execution facility
should be allowed to do program interrupt filtering. Do not set the
transactional-execution program-interruption-filtering override (PIFO)
bit in CR0.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
For a single-threaded KVM guest ptep_modify_prot_start will not use
IPTE, the invalid bit will therefore not be set. If DEBUG_VM is set
pgste_set_key called by ptep_modify_prot_commit will complain about
the missing invalid bit. ptep_modify_prot_start should set the
invalid bit in all cases.
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Modify the psw_idle waiting logic in entry[64].S to return with
interrupts disabled. This avoids potential issues with udelay
and interrupt loops as interrupts are not reenabled after
clock comparator interrupts.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Some requests - particularly 'discard' and 'read' are handled
differently depending on whether a reshape is active or not.
It is harmless to assume reshape is active if it isn't but wrong
to act as though reshape is not active when it is.
So when we start reshape - after making clear to all requests that
reshape has started - use mddev_suspend/mddev_resume to flush out all
requests. This will ensure that no requests will be assuming the
absence of reshape once it really starts.
Signed-off-by: NeilBrown <neilb@suse.de>
make_request() access various shape parameters (raid_disks, chunk_size
etc) which might be changed by raid5_start_reshape().
If the later is called at and awkward time during the form, the wrong
stripe_head might be used.
So introduce a 'seqcount' and after finding a stripe_head make sure
there is no reason to expect that we got the wrong one.
Signed-off-by: NeilBrown <neilb@suse.de>
Add a sysfs entry to control running workqueue thread number. If
group_thread_cnt is set to 0, we will disable workqueue offload handling of
stripes.
Signed-off-by: Shaohua Li <shli@fusionio.com>
Signed-off-by: NeilBrown <neilb@suse.de>
The net_device might be not set on the skb when we try refcounting.
This leads to a null pointer dereference in xdst_queue_output().
It turned out that the refcount to the net_device is not needed
after all. The dst_entry has a refcount to the net_device before
we queue the skb, so it can't go away. Therefore we can remove the
refcount on queueing to fix the null pointer dereference.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
This is another attempt to create multiple threads to handle raid5 stripes.
This time I use workqueue.
raid5 handles request (especially write) in stripe unit. A stripe is page size
aligned/long and acrosses all disks. Writing to any disk sector, raid5 runs a
state machine for the corresponding stripe, which includes reading some disks
of the stripe, calculating parity, and writing some disks of the stripe. The
state machine is running in raid5d thread currently. Since there is only one
thread, it doesn't scale well for high speed storage. An obvious solution is
multi-threading.
To get better performance, we have some requirements:
a. locality. stripe corresponding to request submitted from one cpu is better
handled in thread in local cpu or local node. local cpu is preferred but some
times could be a bottleneck, for example, parity calculation is too heavy.
local node running has wide adaptability.
b. configurablity. Different setup of raid5 array might need diffent
configuration. Especially the thread number. More threads don't always mean
better performance because of lock contentions.
My original implementation is creating some kernel threads. There are
interfaces to control which cpu's stripe each thread should handle. And
userspace can set affinity of the threads. This provides biggest flexibility
and configurability. But it's hard to use and apparently a new thread pool
implementation is disfavor.
Recent workqueue improvement is quite promising. unbound workqueue will be
bound to numa node. If WQ_SYSFS is set in workqueue, there are sysfs option to
do affinity setting. For example, we can only include one HT sibling in
affinity. Since work is non-reentrant by default, and we can control running
thread number by limiting dispatched work_struct number.
In this patch, I created several stripe worker group. A group is a numa node.
stripes from cpus of one node will be added to a group list. Workqueue thread
of one node will only handle stripes of worker group of the node. In this way,
stripe handling has numa node locality. And as I said, we can control thread
number by limiting dispatched work_struct number.
The work_struct callback function handles several stripes in one run. A typical
work queue usage is to run one unit in each work_struct. In raid5 case, the
unit is a stripe. But we can't do that:
a. Though handling a stripe doesn't need lock because of reference accounting
and stripe isn't in any list, queuing a work_struct for each stripe will make
workqueue lock contended very heavily.
b. blk_start_plug()/blk_finish_plug() should surround stripe handle, as we
might dispatch request. If each work_struct only handles one stripe, such block
plug is meaningless.
This implementation can't do very fine grained configuration. But the numa
binding is most popular usage model, should be enough for most workloads.
Note: since we have only one stripe queue, switching to multi-thread might
decrease request size dispatching down to low level layer. The impact depends
on thread number, raid configuration and workload. So multi-thread raid5 might
not be proper for all setups.
Changes V1 -> V2:
1. remove WQ_NON_REENTRANT
2. disabling multi-threading by default
3. Add more descriptions in changelog
Signed-off-by: Shaohua Li <shli@fusionio.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This is the type-safe comparison function, so the double-underscore is
not related.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Gleb Natapov <gleb@redhat.com>
patch "make release_stripe lockless" changes the order stripes are released.
Originally I thought block layer can take care of request merge, but it appears
there are still some requests not merged. It's easy to fix the order.
Signed-off-by: Shaohua Li <shli@fusionio.com>
Signed-off-by: NeilBrown <neilb@suse.de>
The recent commit to delay the release of kobject triggered NULL
dereferences of opti9xx drivers. The cause is that all
snd-opti92x-ad1848, snd-opti92x-cs4231 and snd-opti93x drivers
register the PnP card driver with the very same name, and also
snd-opti92x-ad1848 and -cs4231 drivers register the ISA driver with
the same name, too. When these drivers are built in, quick
"register-release-and-re-register" actions occur, and this results in
Oops because of the same name is assigned to the kobject.
The fix is simply to assign individual names. As a bonus, by using
KBUILD_MODNAME, the patch reduces more lines than it adds.
The fix is based on the suggestion by Russell King.
Reported-and-tested-by: Fengguang Wu <fengguang.wu@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The pl330 DMA driver is broken in regard to handling a terminate all request
while it is processing the list of completed descriptors. This is most visible
when calling dmaengine_terminate_all() from within the descriptors callback for
cyclic transfers. In this case the TERMINATE_ALL transfer will clear the
work_list and stop the transfer. But after all callbacks for all completed
descriptors have been handled the descriptors will be re-enqueued into the (now
empty) work_list. So the next time dma_async_issue_pending() is called for the
channel these descriptors will be transferred again which will cause data
corruption. Similar issues can occur if dmaengine_terminate_all() is not called
from within the descriptor callback but runs on a different CPU at the same time
as the completed descriptor list is processed.
This patch introduces a new per channel list which will hold the completed
descriptors. While processing the list the channel's lock will be held to avoid
racing against dmaengine_terminate_all(). The lock will be released when calling
the descriptors callback though. Since the list of completed descriptors might
be modified (e.g. by calling dmaengine_terminate_all() from the callback) we can
not use the normal list iterator macros. Instead we'll need to check for each
loop iteration again if there are still items in the list. The drivers
TERMINATE_ALL implementation is updated to move descriptors from both the
work_list as well the new completed_list back to the descriptor pool. This makes
sure that none of the descripts finds its way back into the work list and also
that we do not call any futher complete callbacks after
dmaengine_terminate_all() has been called.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
commit 15b0beaa33 ("Add x64 support to debugfs") added
debugfs_create_x64(), but forgot to provide it when debugfs is
disabled, causing problems when code tries to use it even then.
Provide the appropriate static inline.
Cc: Huang Ying <ying.huang@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The dcdbas code was "hand rolling" a binary attribute group, which the
driver core now supports automatically. So remove the "create the files
by hand" logic, and just set the proper field in the attribute group
structure, saving lots of code and headache.
Cc: Doug Warzecha <Douglas_Warzecha@dell.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
--
Doug, I can take this through my driver-core tree if you don't object.
drivers/firmware/dcdbas.c | 19 ++-----------------
1 file changed, 2 insertions(+), 17 deletions(-)
cpm_uart serial driver uses GPIO for control signals. In order to be used
properly, GPIOs have to be reserved. Comment in gpiolib.c considers illegal
the use of GPIOs without requesting them. In addition, the direction of the
GPIO has to be set properly.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A lot of files contain reference to my old e-mail address.
Now I'm going not to read mail from it anymore, so update it
with my current address everywhere.
Signed-off-by: Andrea Merello <andrea.merello@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch removes the dgnc_driver_kzmalloc function from
driver.c and driver.h. A previous patch replaced all
dgnc_driver_kzmalloc function calls with kzalloc.
Signed-off-by: Lidza Louina <lidza.louina@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch replaces dgnc_driver_kzmalloc with kzalloc.
A patch that follows removes the dgnc_driver_kzmalloc
function.
Signed-off-by: Lidza Louina <lidza.louina@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch edits the type casts neo_uart_struct and
cls_uart_struct. A previous patch added the marker __iomem
to these structs. This patch ensures that the change to
the marker is consistent. This also removes these
sparse warnings:
warning: incorrect type in assignment (different address spaces)
expected struct neo_uart_struct [noderef] <asn:2>*ch_neo_uart
got struct neo_uart_struct *<noident>
warning: incorrect type in assignment (different address spaces)
expected struct cls_uart_struct [noderef] <asn:2>*ch_cls_uart
got struct cls_uart_struct *<noident>
Signed-off-by: Lidza Louina <lidza.louina@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The channel state should be correctly set before registering the device. In the current
code the driver probe would fail for channels that have been rescinded and subsequently
re-offered. Fix the bug.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Cc: stable <stable@vger.kernel.org> # 3.11
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
free_irq() expects the same device identity that was passed to
corresponding request_irq(), otherwise the IRQ is not freed.
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds support for virtual IOMMU to the vmci module. We switch
to DMA consistent mappings for guest queuepair and doorbell pages that
are passed to the device. We still allocate each page individually,
since there's no guarantee that we'll get a contiguous block of physical
for an entire queuepair (especially since we allow up to 128 MiB!).
Also made the split between guest and host in the kernelIf struct much
clearer. Now it's obvious which fields are which.
Acked-by: George Zhang <georgezhang@vmware.com>
Acked-by: Aditya Sarwade <asarwade@vmware.com>
Signed-off-by: Andy King <acking@vmware.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We added this for a special case that doesn't exist on Linux. Remove
the non-blocking/pinned queuepair code and simplify the driver in
preparation for adding virtual IOMMU support.
Acked-by: Aditya Sarwade <asarwade@vmware.com>
Signed-off-by: Andy King <acking@vmware.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver core clears the driver data to NULL after device_release
or on probe failure. Thus, it is not needed to manually clear the
device driver data to NULL.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver core clears the driver data to NULL after device_release
or on probe failure. Thus, it is not needed to manually clear the
device driver data to NULL.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since ohci-hcd supports runtime PM, the .pm field in its pci_driver
structure should be protected by CONFIG_PM rather than
CONFIG_PM_SLEEP.
Without this change, OHCI controllers won't do runtime suspend if
system suspend or hibernation isn't enabled.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
cc: <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Unnecessary dev_set_drvdata() is removed, because the driver core
clears the driver data to NULL after device_release or on probe failure.
Signed-off-by: Libo Chen <libo.chen@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the wrapper functions for getting and setting the driver data using
platform_device instead of using dev_{get,set}_drvdata() with &pdev->dev,
so we can directly pass a struct platform_device.
Signed-off-by: Libo Chen <libo.chen@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the wrapper functions for getting and setting the driver data using
platform_device instead of using dev_{get,set}_drvdata() with &pdev->dev,
so we can directly pass a struct platform_device.
Signed-off-by: Libo Chen <libo.chen@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Merge the MSM driver from Rob Clark
* 'drm-next' of git://people.freedesktop.org/~robclark/linux:
drm/msm: add basic hangcheck/recovery mechanism
drm/msm: add a3xx gpu support
drm/msm: add register definitions for gpu
drm/msm: basic KMS driver for snapdragon
drm/msm: add register definitions
Jesse Gross says:
====================
A number of significant new features and optimizations for net-next/3.12.
Highlights are:
* "Megaflows", an optimization that allows userspace to specify which
flow fields were used to compute the results of the flow lookup.
This allows for a major reduction in flow setups (the major
performance bottleneck in Open vSwitch) without reducing flexibility.
* Converting netlink dump operations to use RCU, allowing for
additional parallelism in userspace.
* Matching and modifying SCTP protocol fields.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
The following patchset contains Netfilter updates for your net-next tree,
they are:
* The new SYNPROXY target for iptables, including IPv4 and IPv6 support,
from Patrick McHardy.
* nf_defrag_ipv6.o should be only linked to nf_defrag_ipv6.ko, from
Nathan Hintz.
* Fix an old bug in REJECT, which replies with wrong MAC source address
from the bridge, by Phil Oester.
* Fix uninitialized helper variable in the expectation support over
nfnetlink_queue, from Florian Westphal.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PF gathers statistics for the VF, when the VF is about to unload
we must synchronize the release of its statistics buffer with the PF, so that
no DMA operation will be made to that address after the buffer release.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Due to incorrect VF/PF conditions, when unloading a VF it will not release
part of the memory it has previously allocated.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The check on return code of bnx2x_vfop_config_vlan0() would lead to error
handling flow as the return value indicating an existing pending ramrod would
be erroneously considered as an error.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If driver will fail to allocate all queues, it will shrink the number of
queues and move the storage queue to its correct place (i.e., the last
queue among the newly supported number).
When changing the pointers of the new location of the FCoE queue, we need
to pay special attention to the aggregations pointer - that memory is allocated
during probe and released upon driver removal. Current implementation has 2
pointers pointing to the same chunk of allocated memory, meaning upon removal
there will be two kfree() of the same chunk while the other won't be released.
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Solve issue where no stats were being collected for VF devices due to missing
configuration in the stats' atomic synchronization mechanism.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ben Hutchings says:
====================
More refactoring and cleanup, particularly around filter management.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
release_stripe still has big lock contention. We just add the stripe to a llist
without taking device_lock. We let the raid5d thread to do the real stripe
release, which must hold device_lock anyway. In this way, release_stripe
doesn't hold any locks.
The side effect is the released stripes order is changed. But sounds not a big
deal, stripes are never handled in order. And I thought block layer can already
do nice request merge, which means order isn't that important.
I kept the unplug release batch, which is unnecessary with this patch from lock
contention avoid point of view, and actually if we delete it, the stripe_head
release_list and lru can share storage. But the unplug release batch is also
helpful for request merge. We probably can delay wakeup raid5d till unplug, but
I'm still afraid of the case which raid5d is running.
Signed-off-by: Shaohua Li <shli@fusionio.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This patch, originally authored by Arve Hjonnevag and Todd Poynor,
prevents the system from entering suspend mode until the power supply
plug, unplug, or any other change of state event is fully processed. This
guarantees that the screen lights up and displays the battery charging
state. The implementation uses the power supply wakeup_source object.
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Arve Hjonnevag <arve@android.com>
Cc: Todd Poynor <toddpoynor@google.com>
Cc: John Stultz <john.stultz@linaro.org>
Signed-off-by: Zoran Markovic <zoran.markovic@linaro.org>
Signed-off-by: Anton Vorontsov <anton@enomsg.org>
__initdata should be placed between the variable name and equal
sign for the variable to be placed in the intended section.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
__initconst should be placed between the variable name and equal
sign for the variable to be placed in the intended section.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Emilio López <emilio@elopez.com.ar>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
[mturquette@linaro.org: refreshed patch based on sunxi changes]
