This change makes it so that igb_update_dca is broken into two halves, one
for Rx and one for Tx. The advantage to this is primarily readability.
In addition I am enabling relaxed ordering for reads from hardware since
this is supported on all of the igb parts.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change helps to address locking issues seen with
netif_set_real_num_tx_queues and netif_set_real_num_rx_queues when used in
the igb_set_interrupt_capability function. To resolve these locking issues
I have moved the two function calls into __igb_open so that they can be
called while the RTNL lock is held.
An added advantage to this is that the number of queues is not updated
until the last possible moment so if there are any issues in allocating
MSI-X interrupts or resources for the rings we have time to change the
values prior to updating the netdev.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change combines the the allocation of q_vectors and rings into a single
function. The advantage of this is that we are guaranteed we will avoid
overlap in the L1 cache sets.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change locks us in at 2K buffers even on a system that supports larger
frames. The reason for this change is to make better use of pages and to
reduce the overall truesize of frames generated by igb.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
In order to try and isolate things a bit further I am moving the code
related to retrieving data from the rx_buffer_info structure into a
separate function.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change makes it so that we map the entire page and just sync half of
it for the device at a time. The advantage to this approach is that we can
avoid the locking on map/unmap seen in many IOMMU implementations.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change is meant to just clean-up a number of function calls that were
made at the end of the Rx clean-up path by combining them into a single
function call.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change makes it so that we no longer use header split. The idea is to
reduce partial cache line writes by hardware when handling frames larger
then header size. We can compensate for the extra overhead of having to
memcpy the header buffer by avoiding the cache misses seen by leaving an
full skb allocated and sitting on the ring.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
In order to support page based receive we will need to split up the two
different types of timestamping into two separate functions. The first one
will handle legacy timestamps with the value in the register, and the new
one will handle timestamps in the Rx buffer itself.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Acked-by: Matthew Vick <matthew.vick@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Current implementation mess up the tail pointer. This patch sets skb->tail
correctly.
Also, the small packet check and padding is optimized by using unlikely and
calling skb_pad directly.
Signed-off-by: Tushar Dave <tushar.n.dave@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
These SoC's have an extended bank of GPIO's seperate to the main
GPIO block. This patch adds the additional 5 GPIO's located in this
block which control I2C and PWMOUT.
Signed-off-by: Tony Prisk <linux@prisktech.co.nz>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This change allows us to add a mailbox versioning API. This will allow us
to determine the features supported by the VFs from the PF. For example we
will be implementing a version 1.1 API for the VF that will indicate that
it can support us enabling Jumbo frames as the VF will support buffer
chaining.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Tested-by: Robert Garrett <RobertX.Garrett@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Instead of trying to maintain one large monolithic function that handles
most of the different messages from the VF it makes sense to break the
message handling function up so that we can just go through one switch
statement and call the correct routine for a given message.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This change makes it so that we can have limited support for jumbo frames
when SR-IOV is enabled. In order to accomplish this it is necessary to
disable all VFs when the PF has jumbo frames enabled. If the VFs then
request the same maximum frame size as the PF they will be re-enabled. A
follow on patch will add a means of identifying when a VF can support
spanning buffers and does not need to be worried about the actual supported
max frame size.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Tested-by: Robert Garrett <robertx.e.garrett@intel.com>
Tested-by: Sibai Li <Sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
When enabling DCB the rings belonging to a q_vector on CPU 0 were not
reinitializing their DCA registers. Upon closer inspection the issue was
that the q_vector CPU variable was left at 0 resulting in the driver not
updating the DCA registers.
In order to guarantee the DCA registers will be updated I am adding a
couple line change so that we initialize the CPU variable to -1 which will
force a DCA update the first time an interrupt fires on that q_vector.
In addition we were setting the CPU affinity hint to all CPUs when we were
not specifying a CPU. Instead we should leave it as all zeros to avoid any
possible confusion about the fact that we shouldn't be giving a hint.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
The only file left there is include/linux/dvb/video.h. The
only function for it is to include linux/compiler.h, but this
is already indirectly included. So, get rid of it.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The only reason for this header is to make sure that include
linux/time.h were added before uapi/*/dmx.h. Just push down the
time.h header on the few places where this is used, and drop
this new header.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
It turns out that S390 shouldn't use uapi/asm-generic/kvm_para.h since it
defines the functions in its non-UAPI kvm_para.h file.
#including <asm-generic/kvm_para.h> will first pick up
include/asm-generic/kvm_para.h, which defines conflicting functions.
Instead, partially revert commit 0420c87e64,
ungenericising this file and just inserting a comment to prevent the patch
program from deleting it.
cc: Yuanhan Liu <yuanhan.liu@intel.com>
cc: fengguang.wu@intel.com
Reported-by: Yuanhan Liu <yuanhan.liu@intel.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The APF28 is a small SOM built around an i.MX28 processor with 128MBytes DDR2,
256MBytes NAND Flash and an Ethernet PHY.
Signed-off-by: Julien Boibessot <julien.boibessot@armadeus.com>
Signed-off-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@armadeus.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
commit 93abe8e (clk: add non CONFIG_HAVE_CLK routines)
added clk API stubs when !defined(CONFIG_HAVE_CLK).
This allows us to remove the HAVE_CLK dependency from
Exynos' glue layer and let it compile always.
Signed-off-by: Felipe Balbi <balbi@ti.com>
When we try to free object, there is some of case that we need
to take a node lock. This is the necessary step for preventing a race.
After taking a lock, then we try to cmpxchg_double_slab().
But, there is a possible scenario that cmpxchg_double_slab() is failed
with taking a lock. Following example explains it.
CPU A CPU B
need lock
... need lock
... lock!!
lock..but spin free success
spin... unlock
lock!!
free fail
In this case, retry with taking a lock is occured in CPU A.
I think that in this case for CPU A,
"release a lock first, and re-take a lock if necessary" is preferable way.
There are two reasons for this.
First, this makes __slab_free()'s logic somehow simple.
With this patch, 'was_frozen = 1' is "always" handled without taking a lock.
So we can remove one code path.
Second, it may reduce lock contention.
When we do retrying, status of slab is already changed,
so we don't need a lock anymore in almost every case.
"release a lock first, and re-take a lock if necessary" policy is
helpful to this.
Signed-off-by: Joonsoo Kim <js1304@gmail.com>
Acked-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
a2a47ca366
(ARM: __io abuse cleanup) cleanuped __io() -> IOMEM(),
but setup-r8a7779.c was out of target,
since it directly used (void __iomem __force *) instead of __io().
This patch cleanup it.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
This patch adds device tree entry for MFC v6 in the Exynos5
SoC. Makes the required changes in the clock files and adds
MFC to the DT device list.
Signed-off-by: Naveen Krishna Chatradhi <ch.naveen@samsung.com>
Signed-off-by: Arun Kumar K <arun.kk@samsung.com>
[kgene.kim@samsung.com: fixed section mismatches Seung-Woo Kim reported]
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
c->rate is rarely set, call clk_get_rate on the clock to print
the value in /d/clock/.../rate.
Signed-off-by: Colin Cross <ccross@android.com>
Signed-off-by: Jonghwan Choi <jhbird.choi@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Add a clock/clock_tree debugfs file that shows the entire clock
hierarchy including usage counts and rates.
Signed-off-by: Colin Cross <ccross@android.com>
Signed-off-by: Jonghwan Choi <jhbird.choi@samsung.com>
[kgene.kim@samsung.com: silence checkpatch complaining]
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Make arch/ia64/include/asm/kvm_para.h generic since it's the same as the
asm-generic version. This deals with the problem when ia64 is compiled with
allmodconfig you get:
In file included from include/uapi/linux/kvm_para.h:26:0,
from include/linux/kvm_para.h:4,
from kernel/watchdog.c:28:
arch/ia64/include/asm/kvm_para.h:24:28: error: redefinition of 'kvm_arch_para_features'
include/asm-generic/kvm_para.h:16:28: note: previous definition of 'kvm_arch_para_features' was here
arch/ia64/include/asm/kvm_para.h:29:20: error: redefinition of 'kvm_check_and_clear_guest_paused'
include/asm-generic/kvm_para.h:11:20: note: previous definition of 'kvm_check_and_clear_guest_paused' was here
This does not happen for ia64 defconfig.
Reported-by: Yuanhan Liu <yuanhan.liu@intel.com>
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Yuanhan Liu <yuanhan.liu@intel.com>
cc: fengguang.wu@intel.com
Signed-off-by: Tony Luck <tony.luck@intel.com>
We want to remove plat/cpu.h. To do this, let's first split
it to private soc.h to mach-omap1 and mach-omap2. We have to
keep plat/cpu.h around until the remaining drivers are fixed,
so let's include the local soc.h in plat/cpu.h and for drivers
still including plat/cpu.h.
Once the drivers are fixed not to include plat/cpu.h, we
can remove the file.
This is needed for the ARM common zImage support.
[tony@atomide.com: updated to not print a warning]
Signed-off-by: Tony Lindgren <tony@atomide.com>
To facilitate the ARM single image work, split
arch/arm/plat-omap/include/plat/clkdev_omap.h into the
arch/arm/mach-omap1/clock.h and arch/arm/mach-omap2/clock.h files.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Remove arch/arm/plat-omap/include/plat/clock.h by merging it into
arch/arm/mach-omap1/clock.h and arch/arm/mach-omap2/clock.h.
The goal here is to facilitate ARM single image kernels by removing
includes via the "plat/" symlink.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
[tony@atomide.com: fixed to remove duplicate clock.h includes]
Signed-off-by: Tony Lindgren <tony@atomide.com>
Duplicate arch/arm/plat-omap/clock.c into arch/arm/mach-omap1/clock.c
and arch/arm/mach-omap2/clock.c. This is to support people who are working
on the ARM single image kernel and the OMAP common clock framework
conversion.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Tony Lindgren <tony@atomide.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Moving plat/omap-secure.h locally to mach-omap2/
as part of single zImage work
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
omap_reserve() callback is defned only for mach-omap2.
So, moving definition of omap_reserve() to mach-omap2.
This helps is moving plat/omap_secure.h local to
mach-omap2
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
omap_reserve() is a stub for omap1. So dropping
omap_reserve callback from all OMAP1 board files.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
We must move this for ARM common zImage support.
Note that neither drivers/media/rc/ir-rx51.c or
drivers/media/platform/omap3isp/ispvideo.c need
to include omap-pm.h, so this patch removes the
include for those files.
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: Timo Kokkonen <timo.t.kokkonen@iki.fi>
Cc: linux-media@vger.kernel.org
Signed-off-by: Tony Lindgren <tony@atomide.com>
This is private to cpu.h and no other places should
need to include it and we can drop the include
in mach-omap2/io.c.
Signed-off-by: Tony Lindgren <tony@atomide.com>
Pull media fixes from Mauro Carvalho Chehab:
"Media fixes for:
- one Kconfig fix patch;
- one patch fixing DocBook breakage due to the drivers/media UAPI
changes;
- the remaining UAPI media changes (DVB API).
I'm aware that is is a little late for the UAPI renames for the DVB
API, but IMHO, it is better to merge it for 3.7, due to two reasons:
1) There is a major rename at 3.7 (not only uapi changes, but also
the entire media drivers were reorganized on 3.7, in order to
simplify the Kconfig logic, and easy drivers selection, especially
for hybrid devices). By confining all those renames there at 3.7
it will cause all the harm at for media developers on just one
shot. Stable backports upstream and at distros will likely
welcome it as well, as they won't need to check what changed on
3.7 and what was postponed for on 3.8.
2) The V4L2 DocBook Makefile creates a cross-reference between the
media API headers and the specs. This helps us _a_lot_ to be sure
that all API improvements are properly documented. Every time a
header changes from one place to another, DocBook/media/Makefile
needs to be patched. Currently, the DocBook breakage patch
depends on the DVB UAPI."
* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
[media] Kconfig: Fix dependencies for driver autoselect options
DocBook/media/Makefile: Fix build due to uapi breakage
UAPI: (Scripted) Disintegrate include/linux/dvb
Switching stacks are xfs_alloc_vextent can cause deadlocks when we
run out of worker threads on the allocation workqueue. This can
occur because xfs_bmap_btalloc can make multiple calls to
xfs_alloc_vextent() and even if xfs_alloc_vextent() fails it can
return with the AGF locked in the current allocation transaction.
If we then need to make another allocation, and all the allocation
worker contexts are exhausted because the are blocked waiting for
the AGF lock, holder of the AGF cannot get it's xfs-alloc_vextent
work completed to release the AGF. Hence allocation effectively
deadlocks.
To avoid this, move the stack switch one layer up to
xfs_bmapi_allocate() so that all of the allocation attempts in a
single switched stack transaction occur in a single worker context.
This avoids the problem of an allocation being blocked waiting for
a worker thread whilst holding the AGF.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Certain allocation paths through xfs_bmapi_write() are in situations
where we have limited stack available. These are almost always in
the buffered IO writeback path when convertion delayed allocation
extents to real extents.
The current stack switch occurs for userdata allocations, which
means we also do stack switches for preallocation, direct IO and
unwritten extent conversion, even those these call chains have never
been implicated in a stack overrun.
Hence, let's target just the single stack overun offended for stack
switches. To do that, introduce a XFS_BMAPI_STACK_SWITCH flag that
the caller can pass xfs_bmapi_write() to indicate it should switch
stacks if it needs to do allocation.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Zero the kernel stack space that makes up the xfs_alloc_arg structures.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Michael Kerrisk <mtk.manpages@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Dave Jones <davej@redhat.com>
Signed-off-by: Richard Kuo <rkuo@codeaurora.org>
Calling convention for internal functions and 'asmlinkage' functions is
different on x86-32. Therefore do not directly cast aesni_enc as XTS tweak
function, but use wrapper function in between. Fixes crash with "XTS +
aesni_intel + x86-32" combination.
Cc: stable@vger.kernel.org
Reported-by: Krzysztof Kolasa <kkolasa@winsoft.pl>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
- Fix broadcast packet CRC calculation which can lead to ~80% broadcast packet
loss
- Fix a race condition in duplicate broadcast packet check
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
iEYEABECAAYFAlCANjIACgkQpGgxIkP9cwfeUwCgkGfv2Y8evZYX6d1bW2kFg7di
LJ8AoISKwbTCJUvKrUuQHAvTss7bcX4k
=/dxF
-----END PGP SIGNATURE-----
Merge tag 'batman-adv-fix-for-davem' of git://git.open-mesh.org/linux-merge
Included fixes:
- Fix broadcast packet CRC calculation which can lead to ~80% broadcast packet
loss
- Fix a race condition in duplicate broadcast packet check
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 38f3865744 ("xattr: extract simple_xattr code from tmpfs") moved
some code from tmpfs but introduced a subtle bug along the way.
If the name passed to simple_xattr_remove() does not exist in the list of
xattrs, then it is possible to call kfree(new_xattr) when new_xattr is
actually initialized to itself on the stack via uninitialized_var().
This causes a BUG() since the memory was not allocated via the slab
allocator and was not bypassed through to the page allocator because it
was too large.
Initialize the local variable to NULL so the kfree() never takes place.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Acked-by: Hugh Dickins <hughd@google.com>
Acked-by: Aristeu Rozanski <aris@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
tcp_ioctl() tries to take into account if tcp socket received a FIN
to report correct number bytes in receive queue.
But its flaky because if the application ate the last skb,
we return 1 instead of 0.
Correct way to detect that FIN was received is to test SOCK_DONE.
Reported-by: Elliot Hughes <enh@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On some suspend/resume operations involving wimax device, we have
noticed some intermittent memory corruptions in netlink code.
Stéphane Marchesin tracked this corruption in netlink_update_listeners()
and suggested a patch.
It appears netlink_release() should use kfree_rcu() instead of kfree()
for the listeners structure as it may be used by other cpus using RCU
protection.
netlink_release() must set to NULL the listeners pointer when
it is about to be freed.
Also have to protect netlink_update_listeners() and
netlink_has_listeners() if listeners is NULL.
Add a nl_deref_protected() lockdep helper to properly document which
locks protects us.
Reported-by: Jonathan Kliegman <kliegs@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Stéphane Marchesin <marcheu@google.com>
Cc: Sam Leffler <sleffler@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>