In the case where a non-MPLS packet is received and an MPLS stack is
added it may well be the case that the original skb is GSO but the
NIC used for transmit does not support GSO of MPLS packets.
The aim of this code is to provide GSO in software for MPLS packets
whose skbs are GSO.
SKB Usage:
When an implementation adds an MPLS stack to a non-MPLS packet it should do
the following to skb metadata:
* Set skb->inner_protocol to the old non-MPLS ethertype of the packet.
skb->inner_protocol is added by this patch.
* Set skb->protocol to the new MPLS ethertype of the packet.
* Set skb->network_header to correspond to the
end of the L3 header, including the MPLS label stack.
I have posted a patch, "[PATCH v3.29] datapath: Add basic MPLS support to
kernel" which adds MPLS support to the kernel datapath of Open vSwtich.
That patch sets the above requirements in datapath/actions.c:push_mpls()
and was used to exercise this code. The datapath patch is against the Open
vSwtich tree but it is intended that it be added to the Open vSwtich code
present in the mainline Linux kernel at some point.
Features:
I believe that the approach that I have taken is at least partially
consistent with the handling of other protocols. Jesse, I understand that
you have some ideas here. I am more than happy to change my implementation.
This patch adds dev->mpls_features which may be used by devices
to advertise features supported for MPLS packets.
A new NETIF_F_MPLS_GSO feature is added for devices which support
hardware MPLS GSO offload. Currently no devices support this
and MPLS GSO always falls back to software.
Alternate Implementation:
One possible alternate implementation is to teach netif_skb_features()
and skb_network_protocol() about MPLS, in a similar way to their
understanding of VLANs. I believe this would avoid the need
for net/mpls/mpls_gso.c and in particular the calls to
__skb_push() and __skb_push() in mpls_gso_segment().
I have decided on the implementation in this patch as it should
not introduce any overhead in the case where mpls_gso is not compiled
into the kernel or inserted as a module.
MPLS GSO suggested by Jesse Gross.
Based in part on "v4 GRE: Add TCP segmentation offload for GRE"
by Pravin B Shelar.
Cc: Jesse Gross <jesse@nicira.com>
Cc: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to mitigate ongoing incresase in the size of struct skbuff
use 16 bit integer offsets rather than pointers for inner_*_headers.
This appears to reduce the size of struct skbuff from 0xd0 to 0xc0
bytes on x86_64 with the following all unset.
CONFIG_XFRM
CONFIG_NF_CONNTRACK
CONFIG_NF_CONNTRACK_MODULE
NET_SKBUFF_NF_DEFRAG_NEEDED
CONFIG_BRIDGE_NETFILTER
CONFIG_NET_SCHED
CONFIG_IPV6_NDISC_NODETYPE
CONFIG_NET_DMA
CONFIG_NETWORK_SECMARK
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
libphy currently always reports a PHY as an external transceiver from
the ethtool output. This is inaccurate, because some drivers should be
able to tell that a PHY device is an internal transceiver of an Ethernet
MAC. Add a new flag (PHY_IS_INTERNAL) which can be set by PHY drivers
just like other flags, and a corresponding helper: phy_is_internal()
which can be used by networking drivers to query if a given
PHY device is internal.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
From Linus Walleij:
This is a set of patches from Lee Jones to start converting
the ux500 to fetch DMA channels from the device tree:
- Full DT support and channel mapping in the DMA40 driver
- Dropping of platform data for migrated devices on the DT
boot path.
* tag 'ux500-dma40-for-arm-soc' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson: (36 commits)
ARM: ux500: Register Cryp and Hash platform drivers on Snowball
crypto: ux500/[cryp|hash] - Show successful start-up in the bootlog
ARM: ux500: Stop passing Cryp DMA channel config information though pdata
crypto: ux500/cryp - Set DMA configuration though dma_slave_config()
crypto: ux500/cryp - Prepare clock before enabling it
ARM: ux500: Stop passing Hash DMA channel config information though pdata
crypto: ux500/hash - Set DMA configuration though dma_slave_config()
crypto: ux500/hash - Prepare clock before enabling it
ARM: ux500: Remove unnecessary attributes from DMA channel request pdata
dmaengine: ste_dma40: Correct copy/paste error
ARM: ux500: Remove DMA address look-up table
dmaengine: ste_dma40: Remove redundant address fetching function
dmaengine: ste_dma40: Only use addresses passed as configuration information
ARM: ux500: Stop passing UART's platform data for Device Tree boots
dmaengine: ste_dma40: Don't configure runtime configurable setup during allocate
dmaengine: ste_dma40: Remove unnecessary call to d40_phy_cfg()
dmaengine: ste_dma40: Separate Logical Global Interrupt Mask (GIM) unmasking
ARM: ux500: Pass remnant platform data though to DMA40 driver
dmaengine: ste_dma40: Supply full Device Tree parsing support
dmaengine: ste_dma40: Allow driver to be probe()able when DT is enabled
...
Signed-off-by: Olof Johansson <olof@lixom.net>
Here we introduce a new interface to replace alloc_pci_dev():
struct pci_dev *pci_alloc_dev(struct pci_bus *bus)
It takes a "struct pci_bus *" argument, so we can alloc a PCI device
on a target PCI bus, and it acquires a reference on the pci_bus.
We use pci_alloc_dev(NULL) to simplify the old alloc_pci_dev(),
and keep it for a while but mark it as __deprecated.
Holding a reference to the pci_bus ensures that referencing
pci_dev->bus is valid as long as the pci_dev is valid.
[bhelgaas: keep existing "return error early" structure in pci_alloc_dev()]
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jiang Liu <jiang.liu@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
When we don't have any file in cpu/cpufreq directory we shouldn't
create it. Specially with the introduction of per-policy governor
instance patchset, even governors are moved to
cpu/cpu*/cpufreq/governor-name directory and so this directory is
just not required.
Lets have it only when required.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Governors other than ondemand and conservative can also use
get_cpu_idle_time() and they aren't required to compile
cpufreq_governor.c. So, move these independent routines to
cpufreq.c instead.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
get_governor_parent_kobj() can be used by any governor, generic
cpufreq governors or platform specific ones and so must be present in
cpufreq.c instead of cpufreq_governor.c.
This patch moves it to cpufreq.c. This also adds
EXPORT_SYMBOL_GPL(get_governor_parent_kobj) so that modules can use
this function too.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
If card power is dependent on SD bus power then the host controller
must not be runtime suspended while the card is powered up. Add
the ability to stay runtime-resumed in that case and enable it with a new
quirk SDHCI_QUIRK2_CARD_ON_NEEDS_BUS_ON.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
This removes the deprecated use of the .private member of struct dma_chan
and switches the sdhi / tmio mmc driver to using the
dmaengine_slave_config() channel configuration method.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
So far only the SDHI implementation uses TMIO MMC with DMA. That way a DMA
channel filter function, defined in the TMIO driver wasn't a problem.
However, such a filter function is DMA controller specific. Since the SDHI
glue is only running on systems with the SHDMA DMA controller, the filter
function can safely be provided by it. Move it into SDHI.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
Use the saved values in card->ext_csd when selecting power class.
By doing this the power class will be selected even if mmc_init_card
is called with oldcard != NULL, which is the case after a suspend/resume.
Today ext_csd is NULL if mmc_init_card is called with oldcard != NULL
and power class will not be selected.
According to the eMMC specification the POWER_CLASS value is reset after
power failure, H/W reset assertion and any CMD0 reset.
Signed-off-by: Fredrik Soderstedt <fredrik.soderstedt@stericsson.com>
Reviewed-by: Johan Rudholm <jrudholm@gmail.com>
Acked By: Girish K S <girish.shivananjappa@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
The mmc_card_sleep|awake APIs are not being used since the support is
already properly encapsulated within the suspend sequence. Sleep|awake
command is also specific for eMMC.
We remove the sleep|awake bus_ops, the mmc_card_sleep|awake APIs and
move the code into the mmc specific core instead. This also includes
the mmc ops function, mmc_sleepawake. All releated functions have then
become static and we have got far less code to maintain.
Additionally this patch also simplifies the code from mmc_sleepawake,
since it is only used to put the card to sleep and not awake.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
Aggressive power management is suitable when saving power is
essential. At request inactivity timeout, aka pm runtime
autosuspend timeout, the card will be suspended.
Once a new request arrives, the card will be re-initalized and
thus the first request will suffer from a latency. This latency
is card-specific, experiments has shown in general that SD-cards
has quite poor initialization time, around 300ms-1100ms. eMMC is
not surprisingly far better but still a couple of hundreds of ms
has been observed.
Except for the request latency, it is important to know that
suspending the card will also prevent the card from executing
internal house-keeping operations in idle mode. This could mean
degradation in performance.
To use this feature make sure the request inactivity timeout is
chosen carefully. This has not been done as a part of this patch.
Enable this feature by using host cap MMC_CAP_AGGRESSIVE_PM and
by setting CONFIG_MMC_UNSAFE_RESUME.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
Once the mmc blkdevice is being probed, runtime pm will be enabled.
By using runtime autosuspend, the power save operations can be done
when request inactivity occurs for a certain time. Right now the
selected timeout value is set to 3 s. Obviously this value will likely
need to be configurable somehow since it needs to be trimmed depending
on the power save algorithm.
For SD-combo cards, we are still leaving the enablement of runtime PM
to the SDIO init sequence since it depends on the capabilities of the
SDIO func driver.
Moreover, when the blk device is being suspended, we make sure the device
will be runtime resumed. The reason for doing this is that we want the
host suspend sequence to be unaware of any runtime power save operations
done for the card in this phase. Thus it can just handle the suspend as
the card is fully powered from a runtime perspective.
Finally, this patch prepares to make it possible to move BKOPS handling
into the runtime callbacks for the mmc bus_ops. Thus IDLE BKOPS can be
accomplished.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
The sanitize support is added as a user-app ioctl call, and
was removed from the block-device request, since its purpose is
to be invoked not via File-System but by a user.
This feature deletes the unmap memory region of the eMMC card,
by writing to a specific register in the EXT_CSD.
unmap region is the memory region that was previously deleted
(by erase, trim or discard operation).
In order to avoid timeout when sanitizing large-scale cards,
the timeout for sanitize operation is 240 seconds.
Signed-off-by: Yaniv Gardi <ygardi@codeaurora.org>
Signed-off-by: Maya Erez <merez@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
Previously the MMC_CAP2_DETECT_ON_ERR was invented for detecting
slow card removal. In was never a realy good solution and a proper
fix has been merged using gpio debouncing instead. We remove this
cap in this patch.
Although when using polling card detect mode, the code invented for
MMC_CAP2_DETECT_ON_ERR is re-used to complete card removal in an
earlier phase. There are no need waiting for the polling timeout to
elapse in this case.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Kevin Liu <kliu5@marvell.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
Now when upper device is changed, event is not propagated via RT Netlink
to userspace. Userspace might never now about the change. Fix this by
adding upper-device-change notifier event.
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull slave-dma fixes from Vinod Koul:
"We have two patches from Andy & Rafael fixing the Lynxpoint dma"
* 'fixes' of git://git.infradead.org/users/vkoul/slave-dma:
ACPI / LPSS: register clock device for Lynxpoint DMA properly
dma: acpi-dma: parse CSRT to extract additional resources
This patch adds a ASoC CODEC driver for the SSM2516. The SSM2516 is a stereo
Class-D audio amplifier with an I2S interface for audio in and a built-in
dynamic range control processor.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Merge fixes from Andrew Morton:
"A bunch of fixes and one simple fbdev driver which missed the merge
window because people will still talking about it (to no great
effect)."
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (30 commits)
aio: fix kioctx not being freed after cancellation at exit time
mm/pagewalk.c: walk_page_range should avoid VM_PFNMAP areas
drivers/rtc/rtc-max8998.c: check for pdata presence before dereferencing
ocfs2: goto out_unlock if ocfs2_get_clusters_nocache() failed in ocfs2_fiemap()
random: fix accounting race condition with lockless irq entropy_count update
drivers/char/random.c: fix priming of last_data
mm/memory_hotplug.c: fix printk format warnings
nilfs2: fix issue of nilfs_set_page_dirty() for page at EOF boundary
drivers/block/brd.c: fix brd_lookup_page() race
fbdev: FB_GOLDFISH should depend on HAS_DMA
drivers/rtc/rtc-pl031.c: pass correct pointer to free_irq()
auditfilter.c: fix kernel-doc warnings
aio: fix io_getevents documentation
revert "selftest: add simple test for soft-dirty bit"
drivers/leds/leds-ot200.c: fix error caused by shifted mask
mm/THP: use pmd_populate() to update the pmd with pgtable_t pointer
linux/kernel.h: fix kernel-doc warning
mm compaction: fix of improper cache flush in migration code
rapidio/tsi721: fix bug in MSI interrupt handling
hfs: avoid crash in hfs_bnode_create
...
Merge net into net-next because some upcoming net-next changes
build on top of bug fixes that went into net.
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull ARM SoC fixes from Olof Johansson:
"We didn't have any fixes sent up for -rc2, so this is a slightly
larger batch. A bit all over the place platform-wise; OMAP, at91,
marvell, renesas, sunxi, ux500, etc.
I tried to summarize highlights but there isn't a whole lot to point
out. Lots of little things fixed all over. A couple of defconfig
updates due to new/changing options."
* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (44 commits)
ARM: at91/sama5: fix incorrect PMC pcr div definition
ARM: at91/dt: fix macb pinctrl_macb_rmii_mii_alt definition
ARM: at91: at91sam9n12: move external irq declatation to DT
ARM: shmobile: marzen: Use error values in usb_power_*
ARM: tegra: defconfig fixes
ARM: nomadik: fix IRQ assignment for SMC ethernet
ARM: vt8500: Add missing NULL terminator in dt_compat
clk: tegra: add ac97 controller clock
clk: tegra: remove USB from clk init table
ARM: dts: mvebu: Fix wrong the address reg value for the L2-cache node
ARM: plat-orion: Fix num_resources and id for ge10 and ge11
ARM: OMAP2+: hwmod: Remove sysc slave idle and auto idle apis
SERIAL: OMAP: Remove the slave idle handling from the driver
ARM: OMAP2+: serial: Remove the un-used slave idle hooks
ARM: OMAP2+: hwmod-data: UART IP needs software control to manage sidle modes
ARM: OMAP2+: hwmod: Add a new flag to handle SIDLE in SWSUP only in active
ARM: OMAP2+: hwmod: Fix sidle programming in _enable_sysc()/_idle_sysc()
arm: mvebu: fix the 'ranges' property to handle PCIe
ARM: mvebu: select ARCH_REQUIRE_GPIOLIB for mvebu platform
ARM: AM33XX: Add missing .clkdm_name to clkdiv32k_ick clock
...
Fix kernel-doc warning in <linux/kernel.h>:
Warning(include/linux/kernel.h:590): No description found for parameter 'ip'
scripts/kernel-doc cannot handle macros, functions, or function
prototypes between the function or macro that is being documented and
its definition, so move these prototypes above the function that is
being documented.
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Many callers of the wait_event_timeout() and
wait_event_interruptible_timeout() expect that the return value will be
positive if the specified condition becomes true before the timeout
elapses. However, at the moment this isn't guaranteed. If the wake-up
handler is delayed enough, the time remaining until timeout will be
calculated as 0 - and passed back as a return value - even if the
condition became true before the timeout has passed.
Fix this by returning at least 1 if the condition becomes true. This
semantic is in line with what wait_for_condition_timeout() does; see
commit bb10ed09 ("sched: fix wait_for_completion_timeout() spurious
failure under heavy load").
Daniel said "We have 3 instances of this bug in drm/i915. One case even
where we switch between the interruptible and not interruptible
wait_event_timeout variants, foolishly presuming they have the same
semantics. I very much like this."
One such bug is reported at
https://bugs.freedesktop.org/show_bug.cgi?id=64133
Signed-off-by: Imre Deak <imre.deak@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: David Howells <dhowells@redhat.com>
Acked-by: Jens Axboe <axboe@kernel.dk>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Lukas Czerner <lczerner@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Systems that use RapidIO fabric may need to implement their own
enumeration and discovery methods which are better suitable for needs of
a target application.
The following set of patches is intended to simplify process of
introduction of new RapidIO fabric enumeration/discovery methods.
The first patch offers ability to add new RapidIO enumeration/discovery
methods using kernel configuration options. This new configuration
option mechanism allows to select statically linked or modular
enumeration/discovery method(s) from the list of existing methods or use
external module(s).
This patch also updates the currently existing enumeration/discovery
code to be used as a statically linked or modular method.
The corresponding configuration option is named "Basic
enumeration/discovery" method. This is the only one configuration
option available today but new methods are expected to be introduced
after adoption of provided patches.
The second patch address a long time complaint of RapidIO subsystem
users regarding fabric enumeration/discovery start sequence. Existing
implementation offers only a boot-time enumeration/discovery start which
requires synchronized boot of all endpoints in RapidIO network. While
it works for small closed configurations with limited number of
endpoints, using this approach in systems with large number of endpoints
is quite challenging.
To eliminate requirement for synchronized start the second patch
introduces RapidIO enumeration/discovery start from user space.
For compatibility with the existing RapidIO subsystem implementation,
automatic boot time enumeration/discovery start can be configured in by
specifying "rio-scan.scan=1" command line parameter if statically linked
basic enumeration method is selected.
This patch:
Rework to implement RapidIO enumeration/discovery method selection
combined with ability to use enumeration/discovery as a kernel module.
This patch adds ability to introduce new RapidIO enumeration/discovery
methods using kernel configuration options. Configuration option
mechanism allows to select statically linked or modular
enumeration/discovery method from the list of existing methods or use
external modules. If a modular enumeration/discovery is selected each
RapidIO mport device can have its own method attached to it.
The existing enumeration/discovery code was updated to be used as
statically linked or modular method. This configuration option is named
"Basic enumeration/discovery" method.
Several common routines have been moved from rio-scan.c to make them
available to other enumeration methods and reduce number of exported
symbols.
Signed-off-by: Alexandre Bounine <alexandre.bounine@idt.com>
Cc: Matt Porter <mporter@kernel.crashing.org>
Cc: Li Yang <leoli@freescale.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: Andre van Herk <andre.van.herk@Prodrive.nl>
Cc: Micha Nelissen <micha.nelissen@Prodrive.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull networking fixes from David Miller:
"It's been a while since my last pull request so quite a few fixes have
piled up."
Indeed.
1) Fix nf_{log,queue} compilation with PROC_FS disabled, from Pablo
Neira Ayuso.
2) Fix data corruption on some tg3 chips with TSO enabled, from Michael
Chan.
3) Fix double insertion of VLAN tags in be2net driver, from Sarveshwar
Bandi.
4) Don't have TCP's MD5 support pass > PAGE_SIZE page offsets in
scatter-gather entries into the crypto layer, the crypto layer can't
handle that. From Eric Dumazet.
5) Fix lockdep splat in 802.1Q MRP code, also from Eric Dumazet.
6) Fix OOPS in netfilter log module when called from conntrack, from
Hans Schillstrom.
7) FEC driver needs to use netif_tx_{lock,unlock}_bh() rather than the
non-BH disabling variants. From Fabio Estevam.
8) TCP GSO can generate out-of-order packets, fix from Eric Dumazet.
9) vxlan driver doesn't update 'used' field of fdb entries when it
should, from Sridhar Samudrala.
10) ipv6 should use kzalloc() to allocate inet6 socket cork options,
otherwise we can OOPS in ip6_cork_release(). From Eric Dumazet.
11) Fix races in bonding set mode, from Nikolay Aleksandrov.
12) Fix checksum generation regression added by "r8169: fix 8168evl
frame padding.", from Francois Romieu.
13) ip_gre can look at stale SKB data pointer, fix from Eric Dumazet.
14) Fix checksum handling when GSO is enabled in bnx2x driver with
certain chips, from Yuval Mintz.
15) Fix double free in batman-adv, from Martin Hundebøll.
16) Fix device startup synchronization with firmware in tg3 driver, from
Nithin Sujit.
17) perf networking dropmonitor doesn't work at all due to mixed up
trace parameter ordering, from Ben Hutchings.
18) Fix proportional rate reduction handling in tcp_ack(), from Nandita
Dukkipati.
19) IPSEC layer doesn't return an error when a valid state is detected,
causing an OOPS. Fix from Timo Teräs.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (85 commits)
be2net: bug fix on returning an invalid nic descriptor
tcp: xps: fix reordering issues
net: Revert unused variable changes.
xfrm: properly handle invalid states as an error
virtio_net: enable napi for all possible queues during open
tcp: bug fix in proportional rate reduction.
net: ethernet: sun: drop unused variable
net: ethernet: korina: drop unused variable
net: ethernet: apple: drop unused variable
qmi_wwan: Added support for Cinterion's PLxx WWAN Interface
perf: net_dropmonitor: Remove progress indicator
perf: net_dropmonitor: Use bisection in symbol lookup
perf: net_dropmonitor: Do not assume ordering of dictionaries
perf: net_dropmonitor: Fix symbol-relative addresses
perf: net_dropmonitor: Fix trace parameter order
net: fec: use a more proper compatible string for MVF type device
qlcnic: Fix updating netdev->features
qlcnic: remove netdev->trans_start updates within the driver
qlcnic: Return proper error codes from probe failure paths
tg3: Update version to 3.132
...
This patch converts cgroup_for_each_child(),
cgroup_next_descendant_pre/post() and thus
cgroup_for_each_descendant_pre/post() to use cgroup_next_sibling()
instead of manually dereferencing ->sibling.next.
The only reason the iterators couldn't allow dropping RCU read lock
while iteration is in progress was because they couldn't determine the
next sibling safely once RCU read lock is dropped. Using
cgroup_next_sibling() removes that problem and enables all iterators
to allow dropping RCU read lock in the middle. Comments are updated
accordingly.
This makes the iterators easier to use and will simplify controllers.
Note that @cgroup argument is renamed to @cgrp in
cgroup_for_each_child() because it conflicts with "struct cgroup" used
in the new macro body.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Currently, there's no easy way to find out the next sibling cgroup
unless it's known that the current cgroup is accessed from the
parent's children list in a single RCU critical section. This in turn
forces all iterators to require whole iteration to be enclosed in a
single RCU critical section, which sometimes is too restrictive. This
patch implements cgroup_next_sibling() which can reliably determine
the next sibling regardless of the state of the current cgroup as long
as it's accessible.
It currently is impossible to determine the next sibling after
dropping RCU read lock because the cgroup being iterated could be
removed anytime and if RCU read lock is dropped, nothing guarantess
its ->sibling.next pointer is accessible. A removed cgroup would
continue to point to its next sibling for RCU accesses but stop
receiving updates from the sibling. IOW, the next sibling could be
removed and then complete its grace period while RCU read lock is
dropped, making it unsafe to dereference ->sibling.next after dropping
and re-acquiring RCU read lock.
This can be solved by adding a way to traverse to the next sibling
without dereferencing ->sibling.next. This patch adds a monotonically
increasing cgroup serial number, cgroup->serial_nr, which guarantees
that all cgroup->children lists are kept in increasing serial_nr
order. A new function, cgroup_next_sibling(), is implemented, which,
if CGRP_REMOVED is not set on the current cgroup, follows
->sibling.next; otherwise, traverses the parent's ->children list
until it sees a sibling with higher ->serial_nr.
This allows the function to always return the next sibling regardless
of the state of the current cgroup without adding overhead in the fast
path.
Further patches will update the iterators to use cgroup_next_sibling()
so that they allow dropping RCU read lock and blocking while iteration
is in progress which in turn will be used to simplify controllers.
v2: Typo fix as per Serge.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
cgroup_is_removed() no longer has external users and it shouldn't grow
any - controllers should deal with cgroup_subsys_state on/offline
state instead of cgroup removal state. Make it static.
While at it, make it return bool.
Signed-off-by: Tejun Heo <tj@kernel.org>
Merging to receive 7805d000db ("cgroup: fix a subtle bug in descendant
pre-order walk") so that further iterator updates can build upon it.
Signed-off-by: Tejun Heo <tj@kernel.org>
When cgroup_next_descendant_pre() initiates a walk, it checks whether
the subtree root doesn't have any children and if not returns NULL.
Later code assumes that the subtree isn't empty. This is broken
because the subtree may become empty inbetween, which can lead to the
traversal escaping the subtree by walking to the sibling of the
subtree root.
There's no reason to have the early exit path. Remove it along with
the later assumption that the subtree isn't empty. This simplifies
the code a bit and fixes the subtle bug.
While at it, fix the comment of cgroup_for_each_descendant_pre() which
was incorrectly referring to ->css_offline() instead of
->css_online().
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: stable@vger.kernel.org
Pull PCI updates from Bjorn Helgaas:
"Here are some more fixes for v3.10. The Moorestown update broke Intel
Medfield devices, so I reverted it. The acpiphp change fixes a
regression: we broke hotplug notifications to host bridges when we
split acpiphp into the host-bridge related part and the
endpoint-related part.
Moorestown
Revert "x86/pci/mrst: Use configuration mechanism 1 for 00:00.0, 00:02.0, 00:03.0"
Hotplug
PCI: acpiphp: Re-enumerate devices when host bridge receives Bus Check"
* tag 'pci-v3.10-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
Revert "x86/pci/mrst: Use configuration mechanism 1 for 00:00.0, 00:02.0, 00:03.0"
PCI: acpiphp: Re-enumerate devices when host bridge receives Bus Check
DMA addresses are now passed as part of the dmaengine API by invoking
dmaengine_slave_config(). So there's no requirement for the DMA40
driver to look them up in a table provided by platform data. This
method does not fit in well using Device Tree either.
Acked-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This change will cost ~25KB of memory, but it's worth the trade-off,
as it removes a great deal of overhead. It means that instead of only
allocating memory for the logical channels in use, it does so for all
available ones, which is 32 per physical channel. However, this now
means we can remove some platform data and we don't have to worry
about adding vendor specific variables to Device Tree.
Acked-by: Vinod Koul <vnod.koul@intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Devices which utilise DMA use the same device numbers for transmitting
and receiving. In this patch we encode the source and destination
information into one single attribute. We can subsequently exploit the
direction attribute to see which of the transfer directions are being
described. This also lessens the burden on platform data.
Cc: Dan Williams <djbw@fb.com>
Cc: Per Forlin <per.forlin@stericsson.com>
Cc: Rabin Vincent <rabin@rab.in>
Acked-by: Vinod Koul <vinod.koul@intel.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
There are only two default memcpy configurations used for the DMA40
driver; one for physical memcpy and one for logical memcpy. Instead
of invariably passing the same configurations though platform data,
we're moving them into the driver instead.
Acked-by: Vinod Koul <vnod.koul@intel.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The channels reserved for memcpy are the same for all currently
supported platforms. With this in mind, we can ease the platform
data passing requirement by moving these assignments out from
platform code and place them directly into the driver.
Acked-by: Vinod Koul <vnod.koul@intel.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Pull tty/serial fixes from Greg Kroah-Hartman:
"Here are some tty / serial driver fixes for 3.10-rc2.
Nothing huge, although the rocket driver fix looks large, it's just
moving the code around to fix the reported build issues in it. Other
than that, this has the fix for the of-reported lockdep warning from
the vt layer, as well as some other needed bugfixes.
All of these have been in linux-next for a while"
* tag 'tty-3.10-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
tty: mxser: Fix build warning introduced by dfc7b837c7 (Re: linux-next: build warning after merge of the tty.current tree)
tty: mxser: fix usage of opmode_ioaddr
serial: 8250_dw: add ACPI ID for Intel BayTrail
TTY: Fix tty miss restart after we turn off flow-control
tty/vt: Fix vc_deallocate() lock order
TTY: ehv_bytechan: add missing platform_driver_unregister() when module exit
TTY: rocket, fix more no-PCI warnings
serial: mcf: missing uart_unregister_driver() on error in mcf_init()
tty: serial: mpc5xxx: fix error handing in mpc52xx_uart_init()
serial: samsung: add missing platform_driver_unregister() when module exit
serial: pl011: protect attribute read from NULL platform data struct
tty: nwpserial: Pass correct pointer to free_irq()
serial: 8250_dw: Add valid clk pointer check
Pull USB fixes from Greg Kroah-Hartman:
"Here are a number of tiny USB bugfixes / new device ids for 3.10-rc2
The majority of these are USB gadget fixes, but they are all small.
Other than that, some USB host controller fixes, and USB serial driver
fixes for problems reported with them.
Also hopefully a fixed up USB_OTG Kconfig dependancy, that one seems
to be almost impossible to get right for all of the different
platforms these days."
* tag 'usb-3.10-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (56 commits)
USB: cxacru: potential underflow in cxacru_cm_get_array()
USB: ftdi_sio: Add support for Newport CONEX motor drivers
USB: option: add device IDs for Dell 5804 (Novatel E371) WWAN card
usb: ohci: fix goto wrong tag in err case
usb: isp1760-if: fix memleak when platform_get_resource fail
usb: ehci-s5p: fix memleak when fallback to pdata
USB: serial: clean up chars_in_buffer
USB: ti_usb_3410_5052: fix chars_in_buffer overhead
USB: io_ti: fix chars_in_buffer overhead
USB: ftdi_sio: fix chars_in_buffer overhead
USB: ftdi_sio: clean up get_modem_status
USB: serial: add generic wait_until_sent implementation
USB: serial: add wait_until_sent operation
USB: set device dma_mask without reference to global data
USB: Blacklisted Cinterion's PLxx WWAN Interface
usb: option: Add Telewell TW-LTE 4G
USB: EHCI: remove bogus #error
USB: reset resume quirk needed by a hub
USB: usb-stor: realtek_cr: Fix compile error
usb, chipidea: fix link error when USB_EHCI_HCD is a module
...
This is intended for use in loops which read data protected by RCU and may
have a large number of iterations. Such an example is dumping the list of
connections known to IPVS: ip_vs_conn_array() and ip_vs_conn_seq_next().
The benefits are for CONFIG_PREEMPT_RCU=y where we save CPU cycles
by moving rcu_read_lock and rcu_read_unlock out of large loops
but still allowing the current task to be preempted after every
loop iteration for the CONFIG_PREEMPT_RCU=n case.
The call to cond_resched() is not needed when CONFIG_PREEMPT_RCU=y.
Thanks to Paul E. McKenney for explaining this and for the
final version that checks the context with CONFIG_DEBUG_ATOMIC_SLEEP=y
for all possible configurations.
The function can be empty in the CONFIG_PREEMPT_RCU case,
rcu_read_lock and rcu_read_unlock are not needed in this case
because the task can be preempted on indication from scheduler.
Thanks to Peter Zijlstra for catching this and for his help
in trying a solution that changes __might_sleep.
Initial cond_resched_rcu_lock() function suggested by Eric Dumazet.
Tested-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Simon Horman <horms@verge.net.au>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Don't panic if we hit an error while adding the nf_log or pernet
netfilter support, just bail out.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acked-by: Gao feng <gaofeng@cn.fujitsu.com>
Quoting https://bugzilla.netfilter.org/show_bug.cgi?id=812:
[ ip6tables -m addrtype ]
When I tried to use in the nat/PREROUTING it messes up the
routing cache even if the rule didn't matched at all.
[..]
If I remove the --limit-iface-in from the non-working scenario, so just
use the -m addrtype --dst-type LOCAL it works!
This happens when LOCAL type matching is requested with --limit-iface-in,
and the default ipv6 route is via the interface the packet we test
arrived on.
Because xt_addrtype uses ip6_route_output, the ipv6 routing implementation
creates an unwanted cached entry, and the packet won't make it to the
real/expected destination.
Silently ignoring --limit-iface-in makes the routing work but it breaks
rule matching (--dst-type LOCAL with limit-iface-in is supposed to only
match if the dst address is configured on the incoming interface;
without --limit-iface-in it will match if the address is reachable
via lo).
The test should call ipv6_chk_addr() instead. However, this would add
a link-time dependency on ipv6.
There are two possible solutions:
1) Revert the commit that moved ipt_addrtype to xt_addrtype,
and put ipv6 specific code into ip6t_addrtype.
2) add new "nf_ipv6_ops" struct to register pointers to ipv6 functions.
While the former might seem preferable, Pablo pointed out that there
are more xt modules with link-time dependeny issues regarding ipv6,
so lets go for 2).
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pull MIPS update from Ralf Baechle:
- Fix a build error if <linux/printk.h> is included without
<linux/linkage.h> having been included before.
- Cleanup and fix the damage done by the generic idle loop patch.
- A kprobes fix that brings the MIPS code in line with what other
architectures are for quite a while already.
- Wire up the native getdents64(2) syscall for 64 bit - for some reason
it was only for the compat ABIs. This has been reported to cause an
application issue. This turned out bigger than I meant but the wait
instruction support code was driving me nuts.
* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
MIPS: N64: Wire getdents64(2)
kprobes/mips: Fix to check double free of insn slot
MIPS: Idle: Break r4k_wait into two functions and fix it.
MIPS: Idle: Do address fiddlery in helper functions.
MIPS: Idle: Consolidate all declarations in <asm/idle.h>.
MIPS: Idle: Don't call local_irq_disable() in cpu_wait() implementations.
MIPS: Idle: Re-enable irqs at the end of r3081, au1k and loongson2 cpu_wait.
MIPS: Idle: Make call of function pointer readable.
MIPS: Idle: Consistently reformat inline assembler.
MIPS: Idle: cleaup SMTC idle hook as per Linux coding style.
MIPS: Consolidate idle loop / WAIT instruction support in a single file.
MIPS: clock.h: Remove declaration of cpu_wait.
Add include dependencies to <linux/printk.h>.
MIPS: Rewrite pfn_valid to work in modules, too.
Simon Horman