Dave Airlie [Tue, 18 Feb 2014 22:20:14 +0000 (08:20 +1000)]
Merge branch 'drm-fixes-3.14' of git://people.freedesktop.org/~agd5f/linux into drm-fixes
Fix for 128x128 cursors, along with some misc fixes.
* 'drm-fixes-3.14' of git://people.freedesktop.org/~agd5f/linux:
drm/radeon/ni: fix typo in dpm sq ramping setup
drm/radeon/si: fix typo in dpm sq ramping setup
drm/radeon: fix CP semaphores on CIK
drm/radeon: delete a stray tab
drm/radeon: fix display tiling setup on SI
drm/radeon/dpm: reduce r7xx vblank mclk threshold to 200
drm/radeon: fill in DRM_CAPs for cursor size
drm: add DRM_CAPs for cursor size
drm/radeon: unify bpc handling
David S. Miller [Tue, 18 Feb 2014 21:57:42 +0000 (16:57 -0500)]
Merge branch 'for-davem' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless
John W. Linville says:
====================
Please pull this batch of fixes intended for the 3.14 stream...
For the iwlwifi one, Emmanuel says:
"As explicitly written in the commit message, we prefer to disable Tx
AMPDU on NICs supported by iwldvm. This feature gives a big boost in
Tx performance, but the firmware is buggy and we can't rely on it.
Our hope is that most of the users out there want wifi to surf on
the web which means that they care more for Rx traffic than for Tx.
People who want to enable it can do so with the help of a module
parameter."
On top of that...
Dan Carpenter fixes a typo/thinko in ath5k.
Olivier Langlois fixes a couple of rtlwifi issues, one which leaves
IRQs disabled too long (causing a variety of problems elsewhere),
and one which fixes an incorrect return code when failing to enable
the NIC.
Russell King fixes a NULL pointer dereference in hostap.
Stanislaw Gruszka fixes a DMA coherence issue in the rtl8187 driver.
Please let me know if there are problems!
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Lai Jiangshan [Sat, 15 Feb 2014 14:02:28 +0000 (22:02 +0800)]
workqueue: ensure @task is valid across kthread_stop()
When a kworker should die, the kworkre is notified through WORKER_DIE
flag instead of kthread_should_stop(). This, IIRC, is primarily to
keep the test synchronized inside worker_pool lock. WORKER_DIE is
first set while holding pool->lock, the lock is dropped and
kthread_stop() is called.
Unfortunately, this means that there's a slight chance that the target
kworker may see WORKER_DIE before kthread_stop() finishes and exits
and frees the target task before or during kthread_stop().
Fix it by pinning the target task before setting WORKER_DIE and
putting it after kthread_stop() is done.
tj: Improved patch description and comment. Moved pinning above
WORKER_DIE for better signify what it's protecting.
CC: stable@vger.kernel.org Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Tejun Heo <tj@kernel.org>
Daniel Borkmann [Mon, 17 Feb 2014 11:11:11 +0000 (12:11 +0100)]
net: sctp: fix sctp_connectx abi for ia32 emulation/compat mode
SCTP's sctp_connectx() abi breaks for 64bit kernels compiled with 32bit
emulation (e.g. ia32 emulation or x86_x32). Due to internal usage of
'struct sctp_getaddrs_old' which includes a struct sockaddr pointer,
sizeof(param) check will always fail in kernel as the structure in
64bit kernel space is 4bytes larger than for user binaries compiled
in 32bit mode. Thus, applications making use of sctp_connectx() won't
be able to run under such circumstances.
Introduce a compat interface in the kernel to deal with such
situations by using a 'struct compat_sctp_getaddrs_old' structure
where user data is copied into it, and then sucessively transformed
into a 'struct sctp_getaddrs_old' structure with the help of
compat_ptr(). That fixes sctp_connectx() abi without any changes
needed in user space, and lets the SCTP test suite pass when compiled
in 32bit and run on 64bit kernels.
Fixes: f9c67811ebc0 ("sctp: Fix regression introduced by new sctp_connectx api") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 18 Feb 2014 20:40:50 +0000 (15:40 -0500)]
Merge tag 'batman-adv-fix-for-davem' of git://git.open-mesh.org/linux-merge
Included changes:
- fix soft-interface MTU computation
- fix bogus pointer mangling when parsing the TT-TVLV
container. This bug led to a wrong memory access.
- fix memory leak by properly releasing the VLAN object
after CRC check
- properly check pskb_may_pull() return value
- avoid potential race condition while adding new neighbour
- fix potential memory leak by removing all the references
to the orig_node object in case of initialization failure
- fix the TT CRC computation by ensuring that every node uses
the same byte order when hosts with different endianess are
part of the same network
- fix severe memory leak by freeing skb after a successful
TVLV parsing
- avoid potential double free when orig_node initialization
fails
- fix potential kernel paging error caused by the usage of
the old value of skb->data after skb reallocation
Signed-off-by: David S. Miller <davem@davemloft.net>
phy: let phy_provider_register be the last step in registering PHY
Registering phy_provider before creating the PHY can result in PHY
callbacks being invoked which will lead to aborts. In order to avoid this
invoke phy_provider_register after phy_create and phy_set_drvdata.
Reported-by: Felipe Balbi <balbi@ti.com> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Acked-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Hans de Goede [Mon, 17 Feb 2014 08:59:24 +0000 (14:29 +0530)]
phy-core: Don't allow building phy-core as a module
include/phy/phy.h has stub code in there for when building without the
phy-core enabled. This is useful for generic drivers such as ahci-platform,
ehci-platoform and ohci-platform which have support for driving an optional
phy passed to them through the devicetree.
Since on some boards this phy functionality is not needed, being able to
disable the phy subsystem without needing a lot of #ifdef magic in the
driver using it is quite useful.
However this breaks when the module using the phy subsystem is build-in and
the phy-core is not, which leads to the build failing with missing symbol
errors in the linking stage of the zImage.
Which leads to gems such as this being added to the Kconfig for achi_platform:
depends on GENERIC_PHY || !GENERIC_PHY
Rather then duplicating this code in a lot of places using the phy-core,
I believe it is better to simply not allow the phy-core to be built as a
module. The phy core is quite small and has no external dependencies, so
always building it in when enabling it should not be an issue.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Roger Quadros <rogerq@ti.com> Acked-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Hans de Goede [Mon, 17 Feb 2014 08:59:23 +0000 (14:29 +0530)]
phy-core: Don't propagate -ENOSUPP from phy_pm_runtime_get_sync to caller
The phy-core allows phy_init and phy_power_on to be called multiple times,
but before this patch -ENOSUPP from phy_pm_runtime_get_sync would be
propagated to the caller for the 2nd and later calls.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Hans de Goede [Mon, 17 Feb 2014 08:59:22 +0000 (14:29 +0530)]
phy-core: phy_get: Leave error logging to the caller
In various cases errors may be expected, ie probe-deferral or a call to
phy_get from a driver where the use of a phy is optional.
Rather then adding all sort of complicated checks for this, and/or adding
special functions like devm_phy_get_optional, simply don't log an error,
and let deciding if get_phy returning an error really should result in a
dev_err up to the caller.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
phy,phy-bcm-kona-usb2.c: Add dependency on HAS_IOMEM
On archs like S390 or um this driver cannot build nor work.
Make it depend on HAS_IOMEM to bypass build failures.
drivers/phy/phy-bcm-kona-usb2.c:114: undefined reference to `devm_ioremap_resource'
Signed-off-by: Richard Weinberger <richard@nod.at> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Daniel Mack [Fri, 14 Feb 2014 19:49:47 +0000 (20:49 +0100)]
usb: musb: correct use of schedule_delayed_work()
schedule_delayed_work() takes the delay in jiffies, not msecs. Fix the
caller sites in musb. This bug caused regressions with the cleanups
that went in for 3.14 (8ed1fb790ea: "usb: musb: finish suspend/reset
work independently from musb_hub_control()").
Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Daniel Mack [Fri, 14 Feb 2014 19:49:46 +0000 (20:49 +0100)]
usb: musb: do not sleep in atomic context
musb_port_reset() is called from musb_hub_control() which in turn holds
a spinlock, so musb_port_reset() is not allowed to call msleep().
With the asynchronous work helpers in place, this is fortunately easy to
fix by rescheduling the reset deassertion function to after the time
when the wait period is finished.
Note, however, that the MUSB_POWER_RESUME bit is only set on AM33xx
processors under rare conditions such as when to another driver
reporting an error during suspend. Hence, this didn't hit me yet in
normal operation.
Signed-off-by: Daniel Mack <zonque@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Alan Stern [Thu, 13 Feb 2014 20:49:17 +0000 (15:49 -0500)]
USB: EHCI: add delay during suspend to prevent erroneous wakeups
High-speed USB connections revert back to full-speed signalling when
the device goes into suspend. This takes several milliseconds, and
during that time it's not possible to tell reliably whether the device
has been disconnected.
On some platforms, the Wake-On-Disconnect circuitry gets confused
during this intermediate state. It generates a false wakeup signal,
which can prevent the controller from going to sleep.
To avoid this problem, this patch adds a 5-ms delay to the
ehci_bus_suspend() routine if any ports have to switch over to
full-speed signalling. (Actually, the delay was already present for
devices using a particular kind of PHY power management; the patch
merely causes the delay to be used more widely.)
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Reviewed-by: Peter Chen <Peter.Chen@freescale.com> CC: <stable@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Christian König [Tue, 18 Feb 2014 10:37:20 +0000 (11:37 +0100)]
drm/radeon: fix CP semaphores on CIK
The CP semaphore queue on CIK has a bug that triggers if uncompleted
waits use the same address while a signal is still pending. Work around
this by using different addresses for each sync.
Signed-off-by: Christian König <christian.koenig@amd.com> Cc: stable@vger.kernel.org Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Wed, 12 Feb 2014 17:48:23 +0000 (12:48 -0500)]
drm: add DRM_CAPs for cursor size
Some hardware may not support standard 64x64 cursors. Add
a drm cap to query the cursor size from the kernel. Some examples
include radeon CIK parts (128x128 cursors) and armada (32x64 or 64x32).
This allows things like device specific ddxes to remove asics specific
logic and also allows xf86-video-modesetting to work properly with hw
cursors on this hardware. Default to 64 if the driver doesn't specify
a size.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Rob Clark <robdclark@gmail.com>
Alex Deucher [Mon, 3 Feb 2014 20:53:25 +0000 (15:53 -0500)]
drm/radeon: unify bpc handling
We were already storing the bpc (bits per color) information
in radeon_crtc, so just use that everywhere rather than
calculating it everywhere we use it. This also allows us
to change it in one place if we ever want to override it.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Chao Bi [Wed, 12 Feb 2014 19:27:25 +0000 (21:27 +0200)]
mei: set client's read_cb to NULL when flow control fails
In mei_cl_read_start(), if it fails to send flow control request, it
will release "cl->read_cb" but forget to set pointer to NULL, leaving
"cl->read_cb" still pointing to random memory, next time this client is
operated like mei_release(), it has chance to refer to this wrong pointer.
Linus Torvalds [Tue, 18 Feb 2014 18:04:09 +0000 (10:04 -0800)]
Merge tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
Pull ext4 fixes from Ted Ts'o:
"Miscellaneous ext4 bug fixes for v3.14"
* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
jbd2: fix use after free in jbd2_journal_start_reserved()
ext4: don't leave i_crtime.tv_sec uninitialized
ext4: fix online resize with a non-standard blocks per group setting
ext4: fix online resize with very large inode tables
ext4: don't try to modify s_flags if the the file system is read-only
ext4: fix error paths in swap_inode_boot_loader()
ext4: fix xfstest generic/299 block validity failures
Florian Fainelli [Tue, 14 Jan 2014 23:36:29 +0000 (15:36 -0800)]
usb: gadget: bcm63xx_udc: fix build failure on DMA channel code
Commit 3dc6475 ("bcm63xx_enet: add support Broadcom BCM6345 Ethernet")
changed the ENETDMA[CS] macros such that they are no longer macros, but
actual register offset definitions. The bcm63xx_udc driver was not
updated, and as a result, causes the following build error to pop up:
CC drivers/usb/gadget/u_ether.o
drivers/usb/gadget/bcm63xx_udc.c: In function 'iudma_write':
drivers/usb/gadget/bcm63xx_udc.c:642:24: error: called object '0' is not
a function
drivers/usb/gadget/bcm63xx_udc.c: In function 'iudma_reset_channel':
drivers/usb/gadget/bcm63xx_udc.c:698:46: error: called object '0' is not
a function
drivers/usb/gadget/bcm63xx_udc.c:700:49: error: called object '0' is not
a function
Fix this by updating usb_dmac_{read,write}l and usb_dmas_{read,write}l to
take an extra channel argument, and use the channel width
(ENETDMA_CHAN_WIDTH) to offset the register we want to access, hence
doing again what the macro implicitely did for us.
Cc: Kevin Cernekee <cernekee@gmail.com> Cc: Jonas Gorski <jogo@openwrt.org> Cc: <stable@vger.kernel.org> Signed-off-by: Florian Fainelli <florian@openwrt.org> Signed-off-by: Felipe Balbi <balbi@ti.com>
Daniel Mack [Wed, 15 Jan 2014 13:09:49 +0000 (14:09 +0100)]
usb: musb: do not sleep in atomic context
musb_port_reset() is called from musb_hub_control() which in turn holds
a spinlock, so musb_port_reset() is not allowed to call msleep().
With the asynchronous work helpers in place, this is fortunately easy to
fix by rescheduling the reset deassertion function to after the time
when the wait period is finished.
Note, however, that the MUSB_POWER_RESUME bit is only set on AM33xx
processors under rare conditions such as when to another driver
reporting an error during suspend. Hence, this didn't hit me yet in
normal operation.
Signed-off-by: Daniel Mack <zonque@gmail.com> Signed-off-by: Felipe Balbi <balbi@ti.com>
Sachin Kamat [Mon, 3 Feb 2014 08:29:38 +0000 (13:59 +0530)]
usb: gadget: s3c2410_udc: Fix build error
Pass value instead of address as expected by 'usb_ep_set_maxpacket_limit'.
Fixes the following compilation error introduced by commit e117e742d310
("usb: gadget: add "maxpacket_limit" field to struct usb_ep"):
drivers/usb/gadget/s3c2410_udc.c: In function ‘s3c2410_udc_reinit’:
drivers/usb/gadget/s3c2410_udc.c:1632:3: error:
cannot take address of bit-field ‘maxpacket’
usb_ep_set_maxpacket_limit(&ep->ep, &ep->ep.maxpacket);
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org> Reviewed-by: Robert Baldyga <r.baldyga@samsung.com> Signed-off-by: Felipe Balbi <balbi@ti.com>
Roger Quadros [Tue, 4 Feb 2014 13:29:33 +0000 (15:29 +0200)]
usb: musb: core: Fix remote-wakeup resume
During resume don't touch SUSPENDM/RESUME bits of POWER register
while restoring controller context. These bits might be changed
by the controller during resume operation and so will be different
than what they were during suspend.
e.g. SUSPENDM bit is set by software during USB global suspend but
automatically cleared by the controller during remote wakeup or
during resume. Setting this bit back while restoring context
causes undesired behaviour. i.e. Babble interrupt is generated
and USB is broken.
Signed-off-by: Roger Quadros <rogerq@ti.com> Signed-off-by: Felipe Balbi <balbi@ti.com>
Disables PING on status phase of control transfer.
PING token is not mandatory in status phase of control transfer
and so some high speed USB devices don't support it. If such devices
are connected to MUSB then they would not respond to PING token
causing delayed or failed enumeration.
[Roger Q] Fixes enumeration issues with some Super-Speed USB hubs
e.g. Dlink DUB-1340
Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com> Signed-off-by: Roger Quadros <rogerq@ti.com> Signed-off-by: Felipe Balbi <balbi@ti.com>
Levente Kurusa [Tue, 18 Feb 2014 15:22:17 +0000 (10:22 -0500)]
ahci: disable NCQ on Samsung pci-e SSDs on macbooks
Samsung's pci-e SSDs with device ID 0x1600 which are found on some
macbooks time out on NCQ commands. Blacklist NCQ on the device so
that the affected machines can at least boot.
Tomasz Nowicki [Mon, 10 Feb 2014 13:00:11 +0000 (14:00 +0100)]
ACPI / PCI: Fix memory leak in acpi_pci_irq_enable()
acpi_pci_link_allocate_irq() can return negative gsi even if
entry != NULL. For that case we have a memory leak, so free
entry before returning from acpi_pci_irq_enable() for gsi < 0.
Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org> Cc: All applicable <stable@vger.kernel.org>
[rjw: Subject and changelog] Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Hui Wang [Tue, 18 Feb 2014 02:56:46 +0000 (10:56 +0800)]
ALSA: hda - add headset mic detect quirks for two Dell laptops
When we plug a 3-ring headset on the Dell machines (Vendor ID:
0x10ec0255, Subsystem ID: 0x10280657; Vendor ID: 0x10ec0255,
Subsystem ID: 0x1028065f), the headset mic can't be
detected, after apply this patch, the headset mic can work well.
BugLink: https://bugs.launchpad.net/bugs/1260303 Cc: David Henningsson <david.henningsson@canonical.com> Tested-by: Cyrus Lien <cyrus.lien@canonical.com> Cc: stable@vger.kernel.org Signed-off-by: Hui Wang <hui.wang@canonical.com> Signed-off-by: Takashi Iwai <tiwai@suse.de>
Dave Airlie [Tue, 18 Feb 2014 06:22:40 +0000 (16:22 +1000)]
Merge branch 'drm-nouveau-next' of git://anongit.freedesktop.org/git/nouveau/linux-2.6 into drm-fixes
Nothing too exciting, mostly fixes for ancient boards, but a pretty important fix for DP on some systems.
Thanks,
* 'drm-nouveau-next' of git://anongit.freedesktop.org/git/nouveau/linux-2.6:
drm/nouveau: fix TTM_PL_TT memtype on pre-nv50
drm/nv50/disp: use correct register to determine DP display bpp
drm/nouveau/fb: use correct ram oclass for nv1a hardware
drm/nv50/gr: add missing nv_error parameter priv
drm/nouveau: fix ENG_RUNLIST register address
drm/nv4c/bios: disallow retrieving from prom on nv4x igp's
drm/nv4c/vga: decode register is in a different place on nv4x igp's
drm/nv4c/mc: nv4x igp's have a different msi rearm register
drm/nouveau: set irq_enabled manually
Dave Airlie [Tue, 18 Feb 2014 06:21:49 +0000 (16:21 +1000)]
Merge tag 'drm-intel-fixes-2014-02-14' of ssh://git.freedesktop.org/git/drm-intel into drm-fixes
3 fixes plus 1 prep patch, all four cc: stable. Jani will take over from
here and the plan is that he'll do 3.14-fixes for the entire release just
to work things out a bit.
* tag 'drm-intel-fixes-2014-02-14' of ssh://git.freedesktop.org/git/drm-intel:
drm/i915/dp: add native aux defer retry limit
drm/i915/dp: increase native aux defer retry timeout
drm/i915: Prevent MI_DISPLAY_FLIP straddling two cachelines on IVB
drm/i915: Add intel_ring_cachline_align()
Ilia Mirkin [Sun, 16 Feb 2014 04:27:01 +0000 (23:27 -0500)]
drm/nouveau: fix TTM_PL_TT memtype on pre-nv50
Commit a55409066 ("drm/nv50-: map TTM_PL_SYSTEM through a BAR for CPU
access") made it possible to work with tiled memory. However
mem->mm_node is not a nouveau_mem for AGP-using pre-NV50 cards, but a
drm_mm_node, as created by the ttm_bo_manager_func. As such, extend the
untiled check to explicitly include all pre-nv50 cards.
Reported-by: Ronald <ronald645@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74613 Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Tested-by: Ronald Uitermark <ronald645@gmail.com> Acked-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Fri, 14 Feb 2014 02:57:15 +0000 (21:57 -0500)]
drm/nv50/disp: use correct register to determine DP display bpp
Commit 0a0afd282f ("drm/nv50-/disp: move DP link training to core and
train from supervisor") added code that uses the wrong register for
computing the display bpp, used for bandwidth calculation. Adjust to use
the same register as used by exec_clkcmp and nv50_disp_intr_unk20_2_dp.
Reported-by: Torsten Wagner <torsten.wagner@gmail.com> Reported-by: Michael Gulick <mgulick@mathworks.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=67628 Cc: stable@vger.kernel.org # 3.9+ Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
drm/nouveau/fb: remove ram oclass argument from base fb constructor
Introduced a unfortunate regression by using nv10 ram oclass for nv1a
hardware, causing corruption and eventually system lockup.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74866 Reported-by: John F. Godfrey <jfgodfrey@gmail.com> Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Cc: stable@vger.kernel.org # 3.13+ Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Sun, 9 Feb 2014 03:35:13 +0000 (22:35 -0500)]
drm/nv50/gr: add missing nv_error parameter priv
Commit ea7dce901 ("drm/nv50/gr: print mpc trap name when it's not an mp
trap") added an nv_error call that was missing the priv parameter. This
causes GPFs if the error is ever hit.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Wed, 5 Feb 2014 19:33:02 +0000 (14:33 -0500)]
drm/nv4c/mc: nv4x igp's have a different msi rearm register
See https://bugs.freedesktop.org/show_bug.cgi?id=74492
Reported-by: Ronald <ronald645@gmail.com> Suggested-by: Marcin Kościelnicki <koriakin@0x04.net> Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ilia Mirkin [Thu, 30 Jan 2014 00:53:00 +0000 (19:53 -0500)]
drm/nouveau: set irq_enabled manually
Since commit 0fa9061ae8c ("drm/nouveau/mc: handle irq-related setup
ourselves"), drm_device->irq_enabled remained unset. This is needed in
order to properly wait for a vblank event in the generic drm code.
See https://bugs.freedesktop.org/show_bug.cgi?id=74195
Reported-by: Jan Janecek <janjanjanx@gmail.com> Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Cc: stable@vger.kernel.org # 3.10+ Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Vinayak Kale [Wed, 12 Feb 2014 06:30:01 +0000 (07:30 +0100)]
ARM: 7957/1: add DSB after icache flush in __flush_icache_all()
Add DSB after icache flush to complete the cache maintenance operation.
Signed-off-by: Vinayak Kale <vkale@apm.com> Acked-by: Catalin Marinas <catalin.marinas@arm.com> Cc: <stable@vger.kernel.org> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Duan Jiong [Mon, 17 Feb 2014 07:23:43 +0000 (15:23 +0800)]
ipv4: fix counter in_slow_tot
since commit 89aef8921bf("ipv4: Delete routing cache."), the counter
in_slow_tot can't work correctly.
The counter in_slow_tot increase by one when fib_lookup() return successfully
in ip_route_input_slow(), but actually the dst struct maybe not be created and
cached, so we can increase in_slow_tot after the dst struct is created.
Signed-off-by: Duan Jiong <duanj.fnst@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Mon, 17 Feb 2014 21:51:00 +0000 (13:51 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client
Pull Ceph fixes from Sage Weil:
"We have some patches fixing up ACL support issues from Zheng and
Guangliang and a mount option to enable/disable this support. (These
fixes were somewhat delayed by the Chinese holiday.)
There is also a small fix for cached readdir handling when directories
are fragmented"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
ceph: fix __dcache_readdir()
ceph: add acl, noacl options for cephfs mount
ceph: make ceph_forget_all_cached_acls() static inline
ceph: add missing init_acl() for mkdir() and atomic_open()
ceph: fix ceph_set_acl()
ceph: fix ceph_removexattr()
ceph: remove xattr when null value is given to setxattr()
ceph: properly handle XATTR_CREATE and XATTR_REPLACE
Linus Torvalds [Mon, 17 Feb 2014 21:50:11 +0000 (13:50 -0800)]
Merge branch 'for-linus' of git://git.samba.org/sfrench/cifs-2.6
Pull CIFS fixes from Steve French:
"Three cifs fixes, the most important fixing the problem with passing
bogus pointers with writev (CVE-2014-0069).
Two additional cifs fixes are still in review (including the fix for
an append problem which Al also discovered)"
* 'for-linus' of git://git.samba.org/sfrench/cifs-2.6:
CIFS: Fix too big maxBuf size for SMB3 mounts
cifs: ensure that uncached writes handle unmapped areas correctly
[CIFS] Fix cifsacl mounts over smb2 to not call cifs
David Howells [Mon, 17 Feb 2014 15:01:47 +0000 (15:01 +0000)]
FS-Cache: Handle removal of unadded object to the fscache_object_list rb tree
When FS-Cache allocates an object, the following sequence of events can
occur:
-->fscache_alloc_object()
-->cachefiles_alloc_object() [via cache->ops->alloc_object]
<--[returns new object]
-->fscache_attach_object()
<--[failed]
-->cachefiles_put_object() [via cache->ops->put_object]
-->fscache_object_destroy()
-->fscache_objlist_remove()
-->rb_erase() to remove the object from fscache_object_list.
resulting in a crash in the rbtree code.
The problem is that the object is only added to fscache_object_list on
the success path of fscache_attach_object() where it calls
fscache_objlist_add().
So if fscache_attach_object() fails, the object won't have been added to
the objlist rbtree. We do, however, unconditionally try to remove the
object from the tree.
Thanks to NeilBrown for finding this and suggesting this solution.
Reported-by: NeilBrown <neilb@suse.de> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: (a customer of) NeilBrown <neilb@suse.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Dave Jones [Mon, 17 Feb 2014 21:21:24 +0000 (16:21 -0500)]
reiserfs: fix utterly brain-damaged indentation.
This has been this way for years, and every time I stumble across it I
lose my lunch. After coming across it for the nth time in the Coverity
results, I had to overcome the bystander effect and do something about
it.
This ignores the 79 column limit in favor of making it look like C
instead of gibberish.
The correct thing to do here would be to lose some of the indentation by
breaking this function up into several smaller ones. I might do that at
some point if I have the stomach to look at this again.
(Also some of those overlong ternary operations would likely be more
readable as regular if's)
Signed-off-by: Dave Jones <davej@fedoraproject.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Tommie Gannert [Mon, 17 Feb 2014 20:46:04 +0000 (20:46 +0000)]
irtty-sir.c: Do not set_termios() on irtty_close()
Issuing set_termios() from irtty_close() causes kernel Oops for
unplugged usb-serial devices.
Since no other tty_ldisc calls set_termios() on close and no tty driver
seem to check if tty->device_data is NULL or not on entry to set_termios(),
the only solution I can come up with is to remove the irtty_stop_receiver()
call, which only updates termios.
Signed-off-by: Tommie Gannert <tommie@gannert.se> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Mon, 17 Feb 2014 20:40:36 +0000 (12:40 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/egtvedt/linux-avr32
Pull AVR32 fixes from Hans-Christian Egtvedt.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/egtvedt/linux-avr32:
avr32: add generic vga.h to Kbuild
avr32: add generic ioremap_wc() definition in io.h
avr32: Makefile: add '-D__linux__' flag for gcc-4.4.7 use
avr32: fix missing module.h causing build failure in mimc200/fram.c
Yan, Zheng [Thu, 13 Feb 2014 11:40:26 +0000 (19:40 +0800)]
ceph: fix __dcache_readdir()
If directory is fragmented, readdir() read its dirfrags one by one.
After reading all dirfrags, the corresponding dentries are sorted in
(frag_t, off) order in the dcache. If dentries of a directory are all
cached, __dcache_readdir() can use the cached dentries to satisfy
readdir syscall. But when checking if a given dentry is after the
position of readdir, __dcache_readdir() compares numerical value of
frag_t directly. This is wrong, it should use ceph_frag_compare().
Sage Weil [Sun, 16 Feb 2014 18:05:29 +0000 (10:05 -0800)]
ceph: add acl, noacl options for cephfs mount
Make the 'acl' option dependent on having ACL support compiled in. Make
the 'noacl' option work even without it so that one can always ask it to
be off and not error out on mount when it is not supported.
Signed-off-by: Guangliang Zhao <lucienchao@gmail.com> Signed-off-by: Sage Weil <sage@inktank.com>
Yan, Zheng [Tue, 11 Feb 2014 05:08:51 +0000 (13:08 +0800)]
ceph: fix ceph_set_acl()
If acl is equivalent to file mode permission bits, ceph_set_acl()
needs to remove any existing acl xattr. Use __ceph_setxattr() to
handle both setting and removing acl xattr cases, it doesn't return
-ENODATA when there is no acl xattr.
Linus Torvalds [Mon, 17 Feb 2014 20:36:49 +0000 (12:36 -0800)]
Merge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Pull powerpc fixes from Ben Herrenschmidt:
"Here are some more powerpc fixes for 3.14
The main one is a nasty issue with the NUMA balancing support which
requires a small generic change and the addition of a new accessor to
set _PAGE_NUMA. Both have been reviewed and acked by Mel and Rik.
The changelog should have plenty of details but basically, without
this fix, we get random user segfaults and/or corruptions due to
missing TLB/hash flushes. Aneesh series of 3 patches fixes it.
We have some vDSO vs. perf fixes from Anton, some small EEH fixes
from Gavin, a ppc32 regression vs the stack overflow detector, and a
fix for the way we handle PCIe host bridge speed settings on pseries
(which is needed for proper operations of AMD graphics cards on
Power8)"
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
powerpc/eeh: Disable EEH on reboot
powerpc/eeh: Cleanup on eeh_subsystem_enabled
powerpc/powernv: Rework EEH reset
powerpc: Use unstripped VDSO image for more accurate profiling data
powerpc: Link VDSOs at 0x0
mm: Use ptep/pmdp_set_numa() for updating _PAGE_NUMA bit
mm: Dirty accountable change only apply to non prot numa case
powerpc/mm: Add new "set" flag argument to pte/pmd update function
powerpc/pseries: Add Gen3 definitions for PCIE link speed
powerpc/pseries: Fix regression on PCI link speed
powerpc: Set the correct ksp_limit on ppc32 when switching to irq stack
Linus Torvalds [Mon, 17 Feb 2014 20:24:45 +0000 (12:24 -0800)]
printk: fix syslog() overflowing user buffer
This is not a buffer overflow in the traditional sense: we don't
overflow any *kernel* buffers, but we do mis-count the amount of data we
copy back to user space for the SYSLOG_ACTION_READ_ALL case.
In particular, if the user buffer is too small to hold everything, and
*if* there is a continuation line at just the right place, we can end up
giving the user more data than he asked for.
The reason is that we first count up the number of bytes all the log
records contains, then we walk the records again until we've skipped the
records at the beginning that won't fit, and then we walk the rest of
the records and copy them to the user space buffer.
And in between that "skip the initial records that won't fit" and the
"copy the records that *will* fit to user space", we reset the 'prev'
variable that contained the record information for the last record not
copied. That meant that when we started copying to user space, we now
had a different character count than what we had originally calculated
in the first record walk-through.
The fix is to simply not clear the 'prev' flags value (in both cases
where we had the same logic: syslog_print_all and kmsg_dump_get_buffer:
the latter is used for pstore-like dumping)
Reported-and-tested-by: Debabrata Banerjee <dbanerje@akamai.com> Acked-by: Kay Sievers <kay@vrfy.org> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Jeff Mahoney <jeffm@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
David Herrmann [Thu, 19 Dec 2013 11:32:24 +0000 (12:32 +0100)]
HID: hyperv: make sure input buffer is big enough
We need at least HID_MAX_BUFFER_SIZE (4096) bytes as input buffer. HID
core depends on this as it requires every input report to be at least as
big as advertised.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
David Herrmann [Thu, 19 Dec 2013 11:09:32 +0000 (12:09 +0100)]
HID: Bluetooth: hidp: make sure input buffers are big enough
HID core expects the input buffers to be at least of size 4096
(HID_MAX_BUFFER_SIZE). Other sizes will result in buffer-overflows if an
input-report is smaller than advertised. We could, like i2c, compute the
biggest report-size instead of using HID_MAX_BUFFER_SIZE, but this will
blow up if report-descriptors are changed after ->start() has been called.
So lets be safe and just use the biggest buffer we have.
Note that this adds an additional copy to the HIDP input path. If there is
a way to make sure the skb-buf is big enough, we should use that instead.
The best way would be to make hid-core honor the @size argument, though,
that sounds easier than it is. So lets just fix the buffer-overflows for
now and afterwards look for a faster way for all transport drivers.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Jiri Bohac [Fri, 14 Feb 2014 17:13:50 +0000 (18:13 +0100)]
bonding: 802.3ad: make aggregator_identifier bond-private
aggregator_identifier is used to assign unique aggregator identifiers
to aggregators of a bond during device enslaving.
aggregator_identifier is currently a global variable that is zeroed in
bond_3ad_initialize().
This sequence will lead to duplicate aggregator identifiers for eth1 and eth3:
create bond0
change bond0 mode to 802.3ad
enslave eth0 to bond0 //eth0 gets agg id 1
enslave eth1 to bond0 //eth1 gets agg id 2
create bond1
change bond1 mode to 802.3ad
enslave eth2 to bond1 //aggregator_identifier is reset to 0
//eth2 gets agg id 1
enslave eth3 to bond0 //eth3 gets agg id 2
Fix this by making aggregator_identifier private to the bond.
Signed-off-by: Jiri Bohac <jbohac@suse.cz> Acked-by: Veaceslav Falico <vfalico@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Emil Goode [Thu, 13 Feb 2014 16:50:19 +0000 (17:50 +0100)]
usbnet: remove generic hard_header_len check
This patch removes a generic hard_header_len check from the usbnet
module that is causing dropped packages under certain circumstances
for devices that send rx packets that cross urb boundaries.
One example is the AX88772B which occasionally send rx packets that
cross urb boundaries where the remaining partial packet is sent with
no hardware header. When the buffer with a partial packet is of less
number of octets than the value of hard_header_len the buffer is
discarded by the usbnet module.
With AX88772B this can be reproduced by using ping with a packet
size between 1965-1976.
The bug has been reported here:
https://bugzilla.kernel.org/show_bug.cgi?id=29082
This patch introduces the following changes:
- Removes the generic hard_header_len check in the rx_complete
function in the usbnet module.
- Introduces a ETH_HLEN check for skbs that are not cloned from
within a rx_fixup callback.
- For safety a hard_header_len check is added to each rx_fixup
callback function that could be affected by this change.
These extra checks could possibly be removed by someone
who has the hardware to test.
- Removes a call to dev_kfree_skb_any() and instead utilizes the
dev->done list to queue skbs for cleanup.
The changes place full responsibility on the rx_fixup callback
functions that clone skbs to only pass valid skbs to the
usbnet_skb_return function.
Signed-off-by: Emil Goode <emilgoode@gmail.com> Reported-by: Igor Gnatenko <i.gnatenko.brain@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Mon, 17 Feb 2014 13:22:21 +0000 (14:22 +0100)]
gre: add link local route when local addr is any
This bug was reported by Steinar H. Gunderson and was introduced by commit f7cb8886335d ("sit/gre6: don't try to add the same route two times").
root@morgental:~# ip tunnel add foo mode gre remote 1.2.3.4 ttl 64
root@morgental:~# ip link set foo up mtu 1468
root@morgental:~# ip -6 route show dev foo
fe80::/64 proto kernel metric 256
but after the above commit, no such route shows up.
There is no link local route because dev->dev_addr is 0 (because local ipv4
address is 0), hence no link local address is configured.
In this scenario, the link local address is added manually: 'ip -6 addr add
fe80::1 dev foo' and because prefix is /128, no link local route is added by the
kernel.
Even if the right things to do is to add the link local address with a /64
prefix, we need to restore the previous behavior to avoid breaking userpace.
Reported-by: Steinar H. Gunderson <sesse@samfundet.no> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
batman-adv: fix potential kernel paging error for unicast transmissions
batadv_send_skb_prepare_unicast(_4addr) might reallocate the
skb's data. If it does then our ethhdr pointer is not valid
anymore in batadv_send_skb_unicast(), resulting in a kernel
paging error.
Fixing this by refetching the ethhdr pointer after the
potential reallocation.
Signed-off-by: Linus Lüssing <linus.luessing@web.de> Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
batman-adv: avoid double free when orig_node initialization fails
In the failure path of the orig_node initialization routine
the orig_node->bat_iv.bcast_own field is free'd twice: first
in batadv_iv_ogm_orig_get() and then later in
batadv_orig_node_free_rcu().
Fix it by removing the kfree in batadv_iv_ogm_orig_get().
Signed-off-by: Antonio Quartulli <antonio@meshcoding.com> Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
batman-adv: fix TT CRC computation by ensuring byte order
When computing the CRC on a 2byte variable the order of
the bytes obviously alters the final result. This means
that computing the CRC over the same value on two archs
having different endianess leads to different numbers.
The global and local translation table CRC computation
routine makes this mistake while processing the clients
VIDs. The result is a continuous CRC mismatching between
nodes having different endianess.
Fix this by converting the VID to Network Order before
processing it. This guarantees that every node uses the same
byte order.
Since batadv_orig_node_new() sets the refcount to two, assuming that
the calling function will use a reference for putting the orig_node into
a hash or similar, both references must be freed if initialization of
the orig_node fails. Otherwise that object may be leaked in that error
case.
Reported-by: Antonio Quartulli <antonio@meshcoding.com> Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de> Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch> Signed-off-by: Antonio Quartulli <antonio@meshcoding.com>
batman-adv: avoid potential race condition when adding a new neighbour
When adding a new neighbour it is important to atomically
perform the following:
- check if the neighbour already exists
- append the neighbour to the proper list
If the two operations are not performed in an atomic context
it is possible that two concurrent insertions add the same
neighbour twice.
Signed-off-by: Antonio Quartulli <antonio@open-mesh.com> Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
batman-adv: release vlan object after checking the CRC
There is a refcounter unbalance in the CRC checking routine
invoked on OGM reception. A vlan object is retrieved (thus
its refcounter is increased by one) but it is never properly
released. This leads to a memleak because the vlan object
will never be free'd.
Fix this by releasing the vlan object after having read the
CRC.
Reported-by: Russell Senior <russell@personaltelco.net> Reported-by: Daniel <daniel@makrotopia.org> Reported-by: cmsv <cmsv@wirelesspt.net> Signed-off-by: Antonio Quartulli <antonio@meshcoding.com> Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
The current MTU computation always returns a value
smaller than 1500bytes even if the real interfaces
have an MTU large enough to compensate the batman-adv
overhead.
Fix the computation by properly returning the highest
admitted value.
Reported-by: Russell Senior <russell@personaltelco.net> Signed-off-by: Antonio Quartulli <antonio@meshcoding.com> Signed-off-by: Marek Lindner <mareklindner@neomailbox.ch>
Archana Patni [Mon, 3 Feb 2014 07:14:16 +0000 (12:44 +0530)]
HID: hid-sensor-hub: quirk for STM Sensor hub
Added STM sensor hub vendor id in HID_SENSOR_HUB_ENUM_QUIRK to
fix report descriptors. These devices uses old FW which uses
logical 0 as minimum. In these, HID reports are not using proper
collection classes. So we need to fix report descriptors,for
such devices. This will not have any impact, if the FW uses
logical 1 as minimum.
We look for usage id for "power and report state", and modify
logical minimum value to 1.
Chen Gang [Sun, 16 Feb 2014 11:36:06 +0000 (19:36 +0800)]
avr32: add generic vga.h to Kbuild
Need add generic "vga.h", or can not pass building for allmodconfig,
the related error:
CC [M] drivers/gpu/drm/drm_irq.o
In file included from include/linux/vgaarb.h:34,
from drivers/gpu/drm/drm_irq.c:42:
include/video/vga.h:22:21: error: asm/vga.h: No such file or directory
Signed-off-by: Chen Gang <gang.chen.5i5j@gmail.com> Acked-by: Hans-Christian Egtvedt <hegtvedt@cisco.com>
Chen Gang [Sun, 16 Feb 2014 11:39:30 +0000 (19:39 +0800)]
avr32: add generic ioremap_wc() definition in io.h
Need generic ioremap_wc(), or can not pass compiling with allmodconfig,
the related error:
CC [M] drivers/gpu/drm/drm_bufs.o
drivers/gpu/drm/drm_bufs.c: In function 'drm_addmap_core':
drivers/gpu/drm/drm_bufs.c:217: error: implicit declaration of function 'ioremap_wc'
drivers/gpu/drm/drm_bufs.c:218: warning: assignment makes pointer from integer without a cast
Signed-off-by: Chen Gang <gang.chen.5i5j@gmail.com> Acked-by: Hans-Christian Egtvedt <hegtvedt@cisco.com>
Chen Gang [Sat, 1 Feb 2014 12:35:54 +0000 (20:35 +0800)]
avr32: Makefile: add '-D__linux__' flag for gcc-4.4.7 use
For avr32 cross compiler, do not define '__linux__' internally, so it
will cause issue with allmodconfig.
The related error:
CC [M] fs/coda/psdev.o
In file included from include/linux/coda.h:64,
from fs/coda/psdev.c:45:
include/uapi/linux/coda.h:221: error: expected specifier-qualifier-list before 'u_quad_t'
The related toolchain version (which only download, not re-compile):
Paul Gortmaker [Fri, 10 Jan 2014 14:29:39 +0000 (09:29 -0500)]
avr32: fix missing module.h causing build failure in mimc200/fram.c
Causing this:
In file included from arch/avr32/boards/mimc200/fram.c:13:
include/linux/miscdevice.h:51: error: field 'list' has incomplete type
include/linux/miscdevice.h:55: error: expected specifier-qualifier-list before 'mode_t'
arch/avr32/boards/mimc200/fram.c:42: error: 'THIS_MODULE' undeclared here (not in a function)
Clemens Ladisch [Sun, 16 Feb 2014 16:11:10 +0000 (17:11 +0100)]
ALSA: usb-audio: work around KEF X300A firmware bug
When the driver tries to access Function Unit 10, the KEF X300A
speakers' firmware apparently locks up, making even PCM streaming
impossible. Work around this by ignoring this FU.
Cc: <stable@vger.kernel.org> Signed-off-by: Clemens Ladisch <clemens@ladisch.de> Signed-off-by: Takashi Iwai <tiwai@suse.de>
Daniel Borkmann [Sun, 16 Feb 2014 14:55:22 +0000 (15:55 +0100)]
packet: check for ndo_select_queue during queue selection
Mathias reported that on an AMD Geode LX embedded board (ALiX)
with ath9k driver PACKET_QDISC_BYPASS, introduced in commit d346a3fae3ff ("packet: introduce PACKET_QDISC_BYPASS socket
option"), triggers a WARN_ON() coming from the driver itself
via 066dae93bdf ("ath9k: rework tx queue selection and fix
queue stopping/waking").
The reason why this happened is that ndo_select_queue() call
is not invoked from direct xmit path i.e. for ieee80211 subsystem
that sets queue and TID (similar to 802.1d tag) which is being
put into the frame through 802.11e (WMM, QoS). If that is not
set, pending frame counter for e.g. ath9k can get messed up.
So the WARN_ON() in ath9k is absolutely legitimate. Generally,
the hw queue selection in ieee80211 depends on the type of
traffic, and priorities are set according to ieee80211_ac_numbers
mapping; working in a similar way as DiffServ only on a lower
layer, so that the AP can favour frames that have "real-time"
requirements like voice or video data frames.
Therefore, check for presence of ndo_select_queue() in netdev
ops and, if available, invoke it with a fallback handler to
__packet_pick_tx_queue(), so that driver such as bnx2x, ixgbe,
or mlx4 can still select a hw queue for transmission in
relation to the current CPU while e.g. ieee80211 subsystem
can make their own choices.
Reported-by: Mathias Kretschmer <mathias.kretschmer@fokus.fraunhofer.de> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Sun, 16 Feb 2014 14:55:21 +0000 (15:55 +0100)]
netdevice: move netdev_cap_txqueue for shared usage to header
In order to allow users to invoke netdev_cap_txqueue, it needs to
be moved into netdevice.h header file. While at it, also add kernel
doc header to document the API.
Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Sun, 16 Feb 2014 14:55:20 +0000 (15:55 +0100)]
netdevice: add queue selection fallback handler for ndo_select_queue
Add a new argument for ndo_select_queue() callback that passes a
fallback handler. This gets invoked through netdev_pick_tx();
fallback handler is currently __netdev_pick_tx() as most drivers
invoke this function within their customized implementation in
case for skbs that don't need any special handling. This fallback
handler can then be replaced on other call-sites with different
queue selection methods (e.g. in packet sockets, pktgen etc).
This also has the nice side-effect that __netdev_pick_tx() is
then only invoked from netdev_pick_tx() and export of that
function to modules can be undone.
Suggested-by: David S. Miller <davem@davemloft.net> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ingo Molnar [Fri, 14 Feb 2014 14:32:20 +0000 (15:32 +0100)]
drivers/net: tulip_remove_one needs to call pci_disable_device()
Otherwise the device is not completely shut down.
Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: David S. Miller <davem@davemloft.net>
net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer
Implementation of (a)rwnd calculation might lead to severe performance issues
and associations completely stalling. These problems are described and solution
is proposed which improves lksctp's robustness in congestion state.
1) Sudden drop of a_rwnd and incomplete window recovery afterwards
Data accounted in sctp_assoc_rwnd_decrease takes only payload size (sctp data),
but size of sk_buff, which is blamed against receiver buffer, is not accounted
in rwnd. Theoretically, this should not be the problem as actual size of buffer
is double the amount requested on the socket (SO_RECVBUF). Problem here is
that this will have bad scaling for data which is less then sizeof sk_buff.
E.g. in 4G (LTE) networks, link interfacing radio side will have a large portion
of traffic of this size (less then 100B).
An example of sudden drop and incomplete window recovery is given below. Node B
exhibits problematic behavior. Node A initiates association and B is configured
to advertise rwnd of 10000. A sends messages of size 43B (size of typical sctp
message in 4G (LTE) network). On B data is left in buffer by not reading socket
in userspace.
Lets examine when we will hit pressure state and declare rwnd to be 0 for
scenario with above stated parameters (rwnd == 10000, chunk size == 43, each
chunk is sent in separate sctp packet)
Logic is implemented in sctp_assoc_rwnd_decrease:
socket_buffer (see below) is maximum size which can be held in socket buffer
(sk_rcvbuf). current_alloced is amount of data currently allocated (rx_count)
A simple expression is given for which it will be examined after how many
packets for above stated parameters we enter pressure state:
We start by condition which has to be met in order to enter pressure state:
socket_buffer < currently_alloced;
currently_alloced is represented as size of sctp packets received so far and not
yet delivered to userspace. x is the number of chunks/packets (since there is no
bundling, and each chunk is delivered in separate packet, we can observe each
chunk also as sctp packet, and what is important here, having its own sk_buff):
socket_buffer < x*each_sctp_packet;
each_sctp_packet is sctp chunk size + sizeof(struct sk_buff). socket_buffer is
twice the amount of initially requested size of socket buffer, which is in case
of sctp, twice the a_rwnd requested:
2*rwnd < x*(payload+sizeof(struc sk_buff));
sizeof(struct sk_buff) is 190 (3.13.0-rc4+). Above is stated that rwnd is 10000
and each payload size is 43
20000 < x(43+190);
x > 20000/233;
x ~> 84;
After ~84 messages, pressure state is entered and 0 rwnd is advertised while
received 84*43B ~= 3612B sctp data. This is why external observer notices sudden
drop from 6474 to 0, as it will be now shown in example:
At this point, rwnd_press stores current rwnd value so it can be later restored
in sctp_assoc_rwnd_increase. This however doesn't happen as condition to start
slowly increasing rwnd until rwnd_press is returned to rwnd is never met. This
condition is not met since rwnd, after it hit 0, must first reach rwnd_press by
adding amount which is read from userspace. Let us observe values in above
example. Initial a_rwnd is 10000, pressure was hit when rwnd was ~6500 and the
amount of actual sctp data currently waiting to be delivered to userspace
is ~3500. When userspace starts to read, sctp_assoc_rwnd_increase will be blamed
only for sctp data, which is ~3500. Condition is never met, and when userspace
reads all data, rwnd stays on 3569.
Reproduction is straight forward, it is enough for sender to send packets of
size less then sizeof(struct sk_buff) and receiver keeping them in its buffers.
2) Minute size window for associations sharing the same socket buffer
In case multiple associations share the same socket, and same socket buffer
(sctp.rcvbuf_policy == 0), different scenarios exist in which congestion on one
of the associations can permanently drop rwnd of other association(s).
Situation will be typically observed as one association suddenly having rwnd
dropped to size of last packet received and never recovering beyond that point.
Different scenarios will lead to it, but all have in common that one of the
associations (let it be association from 1)) nearly depleted socket buffer, and
the other association blames socket buffer just for the amount enough to start
the pressure. This association will enter pressure state, set rwnd_press and
announce 0 rwnd.
When data is read by userspace, similar situation as in 1) will occur, rwnd will
increase just for the size read by userspace but rwnd_press will be high enough
so that association doesn't have enough credit to reach rwnd_press and restore
to previous state. This case is special case of 1), being worse as there is, in
the worst case, only one packet in buffer for which size rwnd will be increased.
Consequence is association which has very low maximum rwnd ('minute size', in
our case down to 43B - size of packet which caused pressure) and as such
unusable.
Scenario happened in the field and labs frequently after congestion state (link
breaks, different probabilities of packet drop, packet reordering) and with
scenario 1) preceding. Here is given a deterministic scenario for reproduction:
>From node A establish two associations on the same socket, with rcvbuf_policy
being set to share one common buffer (sctp.rcvbuf_policy == 0). On association 1
repeat scenario from 1), that is, bring it down to 0 and restore up. Observe
scenario 1). Use small payload size (here we use 43). Once rwnd is 'recovered',
bring it down close to 0, as in just one more packet would close it. This has as
a consequence that association number 2 is able to receive (at least) one more
packet which will bring it in pressure state. E.g. if association 2 had rwnd of
10000, packet received was 43, and we enter at this point into pressure,
rwnd_press will have 9957. Once payload is delivered to userspace, rwnd will
increase for 43, but conditions to restore rwnd to original state, just as in
1), will never be satisfied.
--> Read everything from the socket, both association recover up to maximum rwnd
they are capable of reaching, note that association 1 recovered up to 3698,
and association 2 recovered only to 43
A careful reader might wonder why it is necessary to reproduce 1) prior
reproduction of 2). It is simply easier to observe when to send packet over
association 2 which will push association into the pressure state.
Proposed solution:
Both problems share the same root cause, and that is improper scaling of socket
buffer with rwnd. Solution in which sizeof(sk_buff) is taken into concern while
calculating rwnd is not possible due to fact that there is no linear
relationship between amount of data blamed in increase/decrease with IP packet
in which payload arrived. Even in case such solution would be followed,
complexity of the code would increase. Due to nature of current rwnd handling,
slow increase (in sctp_assoc_rwnd_increase) of rwnd after pressure state is
entered is rationale, but it gives false representation to the sender of current
buffer space. Furthermore, it implements additional congestion control mechanism
which is defined on implementation, and not on standard basis.
Proposed solution simplifies whole algorithm having on mind definition from rfc:
o Receiver Window (rwnd): This gives the sender an indication of the space
available in the receiver's inbound buffer.
Core of the proposed solution is given with these lines:
We advertise to sender (half of) actual space we have. Half is in the braces
depending whether you would like to observe size of socket buffer as SO_RECVBUF
or twice the amount, i.e. size is the one visible from userspace, that is,
from kernelspace.
In this way sender is given with good approximation of our buffer space,
regardless of the buffer policy - we always advertise what we have. Proposed
solution fixes described problems and removes necessity for rwnd restoration
algorithm. Finally, as proposed solution is simplification, some lines of code,
along with some bytes in struct sctp_association are saved.
Version 2 of the patch addressed comments from Vlad. Name of the function is set
to be more descriptive, and two parts of code are changed, in one removing the
superfluous call to sctp_assoc_rwnd_update since call would not result in update
of rwnd, and the other being reordering of the code in a way that call to
sctp_assoc_rwnd_update updates rwnd. Version 3 corrected change introduced in v2
in a way that existing function is not reordered/copied in line, but it is
correctly called. Thanks Vlad for suggesting.
Signed-off-by: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Reviewed-by: Alexander Sverdlin <alexander.sverdlin@nsn.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Duan Jiong [Fri, 14 Feb 2014 10:26:22 +0000 (18:26 +0800)]
ipv4: distinguish EHOSTUNREACH from the ENETUNREACH
since commit 251da413("ipv4: Cache ip_error() routes even when not forwarding."),
the counter IPSTATS_MIB_INADDRERRORS can't work correctly, because the value of
err was always set to ENETUNREACH.
Signed-off-by: Duan Jiong <duanj.fnst@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Haiyang Zhang [Thu, 13 Feb 2014 00:54:27 +0000 (16:54 -0800)]
hyperv: Fix the carrier status setting
Without this patch, the "cat /sys/class/net/ethN/operstate" shows
"unknown", and "ethtool ethN" shows "Link detected: yes", when VM
boots up with or without vNIC connected.
This patch fixed the problem.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Reviewed-by: K. Y. Srinivasan <kys@microsoft.com> Acked-by: Jason Wang <jasowang@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>