netif_tx_queue_stopped(txq) is most of the time false.
Yet its cost is very expensive on SMP.
static inline int netif_tx_queue_stopped(const struct netdev_queue *dev_queue)
{
return test_bit(__QUEUE_STATE_XOFF, &dev_queue->state);
}
I saw this on oprofile hunting and bnx2 driver bnx2_tx_int().
We probably should split "struct netdev_queue" in two parts, one
being read mostly.
__netif_tx_lock() touches _xmit_lock & xmit_lock_owner, these
deserve a separate cache line.
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In 2.6.25 we added UDP mem accounting.
This unfortunatly added a penalty when a frame is transmitted, since
we have at TX completion time to call sock_wfree() to perform necessary
memory accounting. This calls sock_def_write_space() and utimately
scheduler if any thread is waiting on the socket.
Thread(s) waiting for an incoming frame was scheduled, then had to sleep
again as event was meaningless.
(All threads waiting on a socket are using same sk_sleep anchor)
This adds lot of extra wakeups and increases latencies, as noted
by Christoph Lameter, and slows down softirq handler.
Reference : http://marc.info/?l=linux-netdev&m=124060437012283&w=2
Fortunatly, Davide Libenzi recently added concept of keyed wakeups
into kernel, and particularly for sockets (see commit
37e5540b3c
epoll keyed wakeups: make sockets use keyed wakeups)
Davide goal was to optimize epoll, but this new wakeup infrastructure
can help non epoll users as well, if they care to setup an appropriate
handler.
This patch introduces new DEFINE_WAIT_FUNC() helper and uses it
in wait_for_packet(), so that only relevant event can wakeup a thread
blocked in this function.
Trace of function calls from bnx2 TX completion bnx2_poll_work() is :
__kfree_skb()
skb_release_head_state()
sock_wfree()
sock_def_write_space()
__wake_up_sync_key()
__wake_up_common()
receiver_wake_function() : Stops here since thread is waiting for an INPUT
Reported-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Previous rework to ucc_geth.c to add of_mdio support (net: Rework
ucc_geth driver to use of_mdio infrastructure) added a block of
code which broke older openfirmware device trees which this case.
This patch removes the offending blurb.
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
CONFIG_DUET doesn't exist anymore, remove all the code that exists to
support it.
[ Simplify fs_init() even further -DaveM ]
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ayyappan at VMware noticed that we're missing this check from ixgbe which
is in our other drivers. The difference with this implementation from our
other drivers is that this checks all the tx queues rather than just tx[0].
Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update the interrupt management to correctly handle greater
than 16 queue vectors.
Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch enables hardware receive side coalescing for 82599 hardware.
82599 can merge multiple frames from the same TCP/IP flow into a single
structure that can span one ore more descriptors. The accumulated data is
arranged similar to how jumbo frames are arranged with the exception that
other packets can be interlaced inbetween. To overcome this issue a next
pointer is included in the written back descriptor which indicates the next
descriptor in the writeback sequence.
This feature sets the NETIF_F_LRO flag and clearing it via the ethtool set
flags operation will also disable hardware RSC.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Inspired by: Vlad Yasevich <vladislav.yasevich@hp.com>
This is the code to enable ixgbe for hardware offload support
of CRC32c on both transmit and receive of SCTP traffic.
only 82599 supports this offload, not 82598.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Originally from: Vlad Yasevich <vladislav.yasevich@hp.com>
This patch, both the driver portion and the sctp code was
modified by Jesse Brandeburg and is
Copyright(c) 2009 Intel Corporation.
Thanks go to Vlad for starting this work.
Intel 82576 chipset supports SCTP checksum offloading. This
patch enables this functionality in the driver. A new NETIF
feature is introduced for SCTP checksum offload. If the driver
supports CRC32c checksum, it can set this feature flag. The
hardware can offload both transmit and receive.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
this is the sctp code to enable hardware crc32c offload for
adapters that support it.
Originally by: Vlad Yasevich <vladislav.yasevich@hp.com>
modified by Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both of these drivers do a check to verify ip_summed is set to
CHECKSUM_UNNECESSARY prior to passing the packet to GRO. GRO itself
already does such a check so it is redundant and can be removed as this
will likely cause out of order issues when receiving a packet that didn't
pass checksum validation.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The igb driver was being incorrectly setup to only allow disabling receive
checksum if multiqueue was disabled. This change corrects that so that
RXCSUM is configured regardless of queue configuration.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This change updates the timeout logic so that it is not possible to have a
sucessful check for message and still return an error if countdown = 0.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Reported-by: Juha Leppanen <juha_motorsportscom@luukku.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is V2 of the smsc911x fifo byteswap patch.
The smsc911x hardware supports both big and little and endian
hardware configurations, and the linux smsc911x driver currently
detects word order.
For correct operation on big endian platforms lacking swapped
byte lanes the following patch is needed. Only fifo data is
swapped, register data does not require any swapping.
Signed-off-by: Magnus Damm <damm@igel.co.jp>
Acked-by: Steve Glendinning <steve.glendinning@smsc.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Disable L1/L0s when link detected. We enable L1/L0s when link connected
before, but there is some hareware error on some platform. So just diable
this feature when link connected. This feature is about power saving.
Signed-off-by: Jie Yang <jie.yang@atheros.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A non-SMP version of smp_send_stop() is now included in smp.h.
Remove the unneeded definition in the PS3 smp.c.
Fixes build errors like these when CONFIG_SMP=n:
arch/powerpc/platforms/ps3/setup.c:49: error: redefinition of 'smp_send_stop'
include/linux/smp.h:125: error: previous definition of 'smp_send_stop' was here
Reported-by: Subrata Modak <subrata@linux.vnet.ibm.com>
Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Currently PPC_CELL_NATIVE selects PPC_OF_PLATFORM_PCI, but does not
select PCI. This can lead to a config with the former and the latter
disabled, which does not build.
To fix this PPC_CELL_NATIVE should select PCI. However, that would
force PCI on for QPACE, which also selects PPC_CELL_NATIVE. So
instead move the select of PPC_OF_PLATFORM_PCI and PCI under both
IBM_CELL_BLADE and CELLEB.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
On ppc64 we implemented elf_read_implies_exec() for 32-bit binaries
because old toolchains had bugs where they didn't mark program
segments executable that needed to be. For some reason we didn't do
this on ppc32 builds. This hadn't been an issue until commit 8d30c14c
("powerpc/mm: Rework I$/D$ coherency (v3)"), which had as a side
effect that we are now enforcing execute permissions to some extent on
32-bit 4xx and Book E processors.
This fixes it by defining elf_read_implies_exec on 32-bit to turn on
the read-implies-exec behaviour on programs that are sufficiently old
that they don't have a PT_GNU_STACK program header.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
The old refok sections
.text.init.refok
.data.init.refok
.exit.text.refok
have been deprecated since commit
312b1485fb. After the other patches in
this patch series nothing is put in these sections, so clean things up
by eliminating all the remaining references to them.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The section .text.init.refok is deprecated and __REF (.ref.text)
should be used in assembly files instead. This patch cleans up a few
uses of .text.init.refok in the sparc architecture.
Also fix a reference to .text.init in a comment that wasn't updated to
.init.text.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Cc: David S. Miller <davem@davemloft.net>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The sh architecture has some code in the .text.init section, but it
does not reference that section in its linker scripts.
This change moves this code from the .text.init section to the
.init.text section, which is presumably where it belongs.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The section .text.init.refok is deprecated and __REF (.ref.text)
should be used in assembly files instead. This patch cleans up a few
uses of .text.init.refok in the powerpc architecture.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Rather than adding .ref.text to the powerpc linker script so that we
can use __REF on the powerpc architecture, it seems simpler to switch
to using the generic TEXT_TEXT macro.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
arm is placing some code in the .text.init section, but it does not
reference that section in its linker scripts.
This change moves this code from the .text.init section to the
.init.text section, which is presumably where it belongs.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
FRV is placing some code in the .text.init section but does not reference that
section in its linker scripts.
This change moves this code from the .text.init section to the .init.text
section, which is presumably where it belongs.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It seems nothing has included the frv asm/init.h header for some time, and its
actual contents are out of date with include/linux/init.h. So just delete it.
Signed-off-by: Tim Abbott <tabbott@mit.edu>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If reg_phys_mem() fails, we need to free memory allocated for MPA
frame with private data before returning the error. Also move
nes_add_ref() after the reg_phys_mem() is successful.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Running large cluster setup, we are hanging after many hours of
testing. Fixing this required going over the code and making sure the
rexmit entry was properly removed based on the cm_node's state and
packet received. Also when receiving a FIN packet, check seq# and
make sure there were no errors before calling handle_fin().
Following are the changes done in nes_cm.c:
* handle_ack_pkt() needs to return error value, so in case of error,
handle_fin() is not called. Some cleanup done while going over the code.
* handle_rst_pkt(), handling of cm_node's NES_CM_STATE_LAST_ACK is missing.
* process_packet(), in case of FIN only packet is received, call
check_seq() before processing.
* in handle_fin_pkt(), we are calling cleanup_retrans_entry() for all
conditions, even if the packets need to be dropped.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Under heavy load with large cluster testing, it may take longer to
receive a response to MPA requests. Change the driver to wait longer
after each rexmit to max time value.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
check_seq() was not checking if the seq#s have wrapped. Fix it.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When a connect request comes, apbvt should only be set for
non-loopback connections.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Remove the NES_DEBUG that is causing the compile warning about an
unused variable when INFINIBAND_NES_DEBUG is not enabled.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
/sys/class/infiniband/nes?/fw_ver is not displaying firmware version
properly (it shows 0.0.0 with the current code). Fill in the correct
firmware version number.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
With updated PHY firmware for SFP_D, setting the trace length to 1
inch for SFP_D provides a more stable link.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Enable repause timer for port 1. Without this setting, under stress,
the chip may misbehave.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
In commit 1b949324 ("RDMA/nes: Fix SFP+ PHY initialization") there is
a mistake in the clean up code that removed port 1 CDR loop filter
settings for 10G cards other than CX4. Put the correct setting back
for appropriate PHY types.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Change thermo mitigation code to flip the SerDes1 reference clock to
internal, to match the change in commit a4849fc1 ("RDMA/nes: Add
wide_ppm_offset parm for switch compatibility").
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
We need to check only if the WM8350 is master and only when starting
the stream so if either is not true then we can skip the check.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
This reverts commit 8032b526d1.
Hey, it was only meant to be a single release. Now they can all die as
far as I'm concerned.
[ Just kidding. They're cute and cuddly.
Except when they have horrible nasty facial diseases. Oh, and I guess
they're not actually that cuddly even when disease-free. ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6:
PCI: only save/restore existent registers in the PCIe capability
x86/PCI: don't bother with root quirks if _CRS is used
docbooks: add/fix PCI kernel-doc
PCI: cleanup debug output resources
x86/PCI: set_pci_bus_resources_arch_default cleanups
x86/PCI: Move set_pci_bus_resources_arch_default into arch/x86
x86/PCI: don't call e820_all_mapped with -1 in the mmconfig case
PCI quirk: disable MSI on VIA VT3364 chipsets
This warning shows up on 64 bit builds:
fs/ecryptfs/inode.c:693: warning: comparison of distinct pointer types
lacks a cast
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
Btrfs: look for acls during btrfs_read_locked_inode
Btrfs: fix acl caching
Btrfs: Fix a bunch of printk() warnings.
Btrfs: Fix a trivial warning using max() of u64 vs ULL.
Btrfs: remove unused btrfs_bit_radix slab
Btrfs: ratelimit IO error printks
Btrfs: remove #if 0 code
Btrfs: When shrinking, only update disk size on success
Btrfs: fix deadlocks and stalls on dead root removal
Btrfs: fix fallocate deadlock on inode extent lock
Btrfs: kill btrfs_cache_create
Btrfs: don't export symbols
Btrfs: simplify makefile
Btrfs: try to keep a healthy ratio of metadata vs data block groups
fs/ecryptfs/inode.c:670: warning: format '%d' expects type 'int', but argument 3 has type 'size_t'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Cc: Dustin Kirkland <kirkland@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The 'device_type = "soc";' line *is* needed in the DTS for get_immrbase()
to return the correct address.
Signed-off-by: Martyn Welch <martyn.welch@gefanuc.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
This changes btrfs_read_locked_inode() to peek ahead in the btree for acl items.
If it is certain a given inode has no acls, it will set the in memory acl
fields to null to avoid acl lookups completely.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Linus noticed the btrfs code to cache acls wasn't properly caching
a NULL acl when the inode didn't have any acls. This meant the common
case of no acls resulted in expensive btree searches every time the
kernel checked permissions (which is quite often).
This is a modified version of Linus' original patch:
Properly set initial acl fields to BTRFS_ACL_NOT_CACHED in the inode.
This forces an acl lookup when permission checks are done.
Fix btrfs_get_acl to avoid lookups and locking when the inode acls fields
are set to null.
Fix btrfs_get_acl to use the right return value from __btrfs_getxattr
when deciding to cache a NULL acl. It was storing a NULL acl when
__btrfs_getxattr return -ENOENT, but __btrfs_getxattr was actually returning
-ENODATA for this case.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
