virtio_net receives packets from its pre-allocated vring buffers, then it
delivers these packets to upper layer protocols as skb buffs. So it's not
necessary to pre-allocate skb for each mergable buffer, then frees extra
skbs when buffers are merged into a large packet. This patch has deferred
skb allocation in receiving packets for both big packets and mergeable buffers
to reduce skb pre-allocations and skb frees. It frees unused buffers by calling
detach_unused_buf in vring, so recv skb queue is not needed.
Signed-off-by: Shirley Ma <xma@us.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
There's currently no way for a virtio driver to ask for unused
buffers, so it has to keep a list itself to reclaim them at shutdown.
This is redundant, since virtio_ring stores that information. So
add a new hook to do this.
Signed-off-by: Shirley Ma <xma@us.ibm.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use atomic_inc_return() in get_acqseq() to avoid taking a spinlock
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Free memory allocated using kmem_cache_zalloc using kmem_cache_free rather
than kfree.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression x,E,c;
@@
x = \(kmem_cache_alloc\|kmem_cache_zalloc\|kmem_cache_alloc_node\)(c,...)
... when != x = E
when != &x
?-kfree(x)
+kmem_cache_free(c,x)
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Acked-by: David Howells <dhowells@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Steve Dickson <steved@redhat.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: James Morris <jmorris@namei.org>
platfrom -> platform
monotor -> monitor
Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Cc: Anton Vorontsov <cbouatmailru@gmail.com>
Acked-by: Mike Rapoport <mike@compulab.co.il>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
This device generates ABS_Z and ABS_RX events, while it should be
generating ABS_X and ABS_Y instead. Using the MULTI_INPUT quirk solves
this issue.
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@holoscopio.com>
Signed-off-by: Daniel Oliveira Nascimento <don@syst.com.br>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
While debugging a dma driver I noticed a memleak after
unloading the driver module.
Caught by kmemleak.
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Cc: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Fix locking. Use two queues in the driver, one for pending transacions, and
one for transactions which are actually running on the hardware. Call
dma_run_dependencies() on descriptor cleanup so that the async_tx API works
correctly.
There are a number of places throughout the code where lists of descriptors
are freed in a loop. Create functions to handle this, and use them instead
of open-coding the loop each time.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The name fsl_chan seems too long, so it has been shortened to chan. There
are only a few places where the higher level "struct dma_chan *chan" name
conflicts. These have been changed to "struct dma_chan *dchan" instead.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The IRQ probing is needlessly complex. All off the 83xx device trees in
arch/powerpc/boot/dts/ specify 5 interrupts per DMA controller: one for the
controller, and one for each channel. These interrupts are all attached to
the same IRQ line.
This causes an interesting situation if two channels interrupt at the same
time. The per-controller handler will handle the first channel, and the
per-channel handler will handle the remaining channels.
Instead of this mess, we fix the bug in the per-controller handler, and
make it handle all channels that generated an interrupt. When a
per-controller handler is specified in the device tree, we prefer to use
the shared handler instead of the per-channel handler.
The 85xx/86xx controllers do not have a per-controller interrupt, and
instead use a per-channel interrupt. This behavior has not been changed.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This fixes some errors in the cleanup paths of the OF subsystem, including
missing checks for ioremap failing. Also, some variables were renamed for
brevity.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Most functions in the standard library use "dst" as a parameter, rather
than "dest". This renames all use of "dest" to "dst" to match the usual
convention.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This is the beginning of a cleanup which will change all instances of
"fsl_dma" to "fsldma" to match the name of the driver itself.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Remove some unused members from the fsldma data structures. A few trivial
uses of struct resource were converted to use the stack rather than keeping
the memory allocated for the lifetime of the driver.
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Some of the functions are written in a way where they use multiple reads
and writes where a single read/write pair could suffice. This shrinks the
kernel text size measurably, while making the functions easier to
understand.
add/remove: 0/0 grow/shrink: 1/4 up/down: 4/-196 (-192)
function old new delta
fsl_chan_set_request_count 120 124 +4
dma_halt 300 272 -28
fsl_chan_set_src_loop_size 208 156 -52
fsl_chan_set_dest_loop_size 208 156 -52
fsl_chan_xfer_ld_queue 500 436 -64
Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Some single chip family devices are sold in the market with
802.11n bonded out, these have no hardware capability for
802.11n but ath9k can still support them. These are called
AR2427.
Cc: stable@kernel.org
Reported-by: Rolf Leggewie <bugzilla.kernel.org@rolf.leggewie.biz>
Tested-by: Bernhard Reiter <ockham@raz.or.at>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Patch prevents call set_wep_key() with zero key length. That fix long
standing regression since commit c038069352
"airo: clean up WEP key operations". Additionally print call trace when
someone will try to use improper parameters, and remove key.len = 0
assignment, because it is in not possible code path.
Reported-by: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu>
Bisected-by: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu>
Tested-by: Chris Siebenmann <cks@cs.toronto.edu>
Cc: Dan Williams <dcbw@redhat.com>
Cc: <stable@kernel.org>
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Introduce TKIP_HDR_LEN define for code clarity (in the same way as
CCMP_HDR_LEN).
Also odd len variable (not used) dropped from lib80211_tkip_hdr().
Signed-off-by: Andriy V. Tkachuk <andrit@ukr.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
cfq-iosched: Do not idle on async queues
blk-cgroup: Fix potential deadlock in blk-cgroup
block: fix bugs in bio-integrity mempool usage
block: fix bio_add_page for non trivial merge_bvec_fn case
drbd: null dereference bug
drbd: fix max_segment_size initialization
Improve handling of fragmented per-CPU vmaps. We previously don't free
up per-CPU maps until all its addresses have been used and freed. So
fragmented blocks could fill up vmalloc space even if they actually had
no active vmap regions within them.
Add some logic to allow all CPUs to have these blocks purged in the case
of failure to allocate a new vm area, and also put some logic to trim
such blocks of a current CPU if we hit them in the allocation path (so
as to avoid a large build up of them).
Christoph reported some vmap allocation failures when using the per CPU
vmap APIs in XFS, which cannot be reproduced after this patch and the
previous bug fix.
Cc: linux-mm@kvack.org
Cc: stable@kernel.org
Tested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Nick Piggin <npiggin@suse.de>
--
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
RCU list walking of the per-cpu vmap cache was broken. It did not use
RCU primitives, and also the union of free_list and rcu_head is
obviously wrong (because free_list is indeed the list we are RCU
walking).
While we are there, remove a couple of unused fields from an earlier
iteration.
These APIs aren't actually used anywhere, because of problems with the
XFS conversion. Christoph has now verified that the problems are solved
with these patches. Also it is an exported interface, so I think it
will be good to be merged now (and Christoph wants to get the XFS
changes into their local tree).
Cc: stable@kernel.org
Cc: linux-mm@kvack.org
Tested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Nick Piggin <npiggin@suse.de>
--
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'sh/for-2.6.33' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6:
sh: Fix access to released memory in clk_debugfs_register_one()
sh: Fix access to released memory in dwarf_unwinder_cleanup()
usb: r8a66597-hdc disable interrupts fix
spi: spi_sh_msiof: Fixed data sampling on the correct edge
Commit 221af7f87b ("Split 'flush_old_exec' into two functions") split
the function at the point of no return - ie right where there were no
more error cases to check. That made sense from a technical standpoint,
but when we then also combined it with the actual personality setting
going in between flush_old_exec() and setup_new_exec(), it needs to be a
bit more careful.
In particular, we need to make sure that we really flush the old
personality bits in the 'flush' stage, rather than later in the 'setup'
stage, since otherwise we might be flushing the _new_ personality state
that we're just setting up.
So this moves the flags and personality flushing (and 'flush_thread()',
which is the arch-specific function that generally resets lazy FP state
etc) of the old process into flush_old_exec(), so that it doesn't affect
any state that execve() is setting up for the new process environment.
This was reported by Michal Simek as breaking his Microblaze qemu
environment.
Reported-and-tested-by: Michal Simek <michal.simek@petalogix.com>
Cc: Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
I happen to own a keyboard identified as 05af:3062 which is labeled as
"FlatX Coldless Combo" by "Prodige", which exhibits input problems without
NOGET quirk. For some reason, lsusb reports this device as "Jing-Mold
Enterprise Co., Ltd", which is not mentioned anywhere on the package.
A quick search on the intenet shows that there a other people who have
this in their lsusb output, but apparently they don't have the problem
I am seeing (or they are not such furious typists as myself).
Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
Cc: Jiri Kosina <trivial@kernel.org>
Cc: André Goddard Rosa <andre.goddard@gmail.com>
Cc: jfs-discussion@lists.sourceforge.net
Few weeks back, Shaohua Li had posted similar patch. I am reposting it
with more test results.
This patch does two things.
- Do not idle on async queues.
- It also changes the write queue depth CFQ drives (cfq_may_dispatch()).
Currently, we seem to driving queue depth of 1 always for WRITES. This is
true even if there is only one write queue in the system and all the logic
of infinite queue depth in case of single busy queue as well as slowly
increasing queue depth based on last delayed sync request does not seem to
be kicking in at all.
This patch will allow deeper WRITE queue depths (subjected to the other
WRITE queue depth contstraints like cfq_quantum and last delayed sync
request).
Shaohua Li had reported getting more out of his SSD. For me, I have got
one Lun exported from an HP EVA and when pure buffered writes are on, I
can get more out of the system. Following are test results of pure
buffered writes (with end_fsync=1) with vanilla and patched kernel. These
results are average of 3 sets of run with increasing number of threads.
AVERAGE[bufwfs][vanilla]
-------
job Set NR ReadBW(KB/s) MaxClat(us) WriteBW(KB/s) MaxClat(us)
--- --- -- ------------ ----------- ------------- -----------
bufwfs 3 1 0 0 95349 474141
bufwfs 3 2 0 0 100282 806926
bufwfs 3 4 0 0 109989 2.7301e+06
bufwfs 3 8 0 0 116642 3762231
bufwfs 3 16 0 0 118230 6902970
AVERAGE[bufwfs] [patched kernel]
-------
bufwfs 3 1 0 0 270722 404352
bufwfs 3 2 0 0 206770 1.06552e+06
bufwfs 3 4 0 0 195277 1.62283e+06
bufwfs 3 8 0 0 260960 2.62979e+06
bufwfs 3 16 0 0 299260 1.70731e+06
I also ran buffered writes along with some sequential reads and some
buffered reads going on in the system on a SATA disk because the potential
risk could be that we should not be driving queue depth higher in presence
of sync IO going to keep the max clat low.
With some random and sequential reads going on in the system on one SATA
disk I did not see any significant increase in max clat. So it looks like
other WRITE queue depth control logic is doing its job. Here are the
results.
AVERAGE[brr, bsr, bufw together] [vanilla]
-------
job Set NR ReadBW(KB/s) MaxClat(us) WriteBW(KB/s) MaxClat(us)
--- --- -- ------------ ----------- ------------- -----------
brr 3 1 850 546345 0 0
bsr 3 1 14650 729543 0 0
bufw 3 1 0 0 23908 8274517
brr 3 2 981.333 579395 0 0
bsr 3 2 14149.7 1175689 0 0
bufw 3 2 0 0 21921 1.28108e+07
brr 3 4 898.333 1.75527e+06 0 0
bsr 3 4 12230.7 1.40072e+06 0 0
bufw 3 4 0 0 19722.3 2.4901e+07
brr 3 8 900 3160594 0 0
bsr 3 8 9282.33 1.91314e+06 0 0
bufw 3 8 0 0 18789.3 23890622
AVERAGE[brr, bsr, bufw mixed] [patched kernel]
-------
job Set NR ReadBW(KB/s) MaxClat(us) WriteBW(KB/s) MaxClat(us)
--- --- -- ------------ ----------- ------------- -----------
brr 3 1 837 417973 0 0
bsr 3 1 14357.7 591275 0 0
bufw 3 1 0 0 24869.7 8910662
brr 3 2 1038.33 543434 0 0
bsr 3 2 13351.3 1205858 0 0
bufw 3 2 0 0 18626.3 13280370
brr 3 4 913 1.86861e+06 0 0
bsr 3 4 12652.3 1430974 0 0
bufw 3 4 0 0 15343.3 2.81305e+07
brr 3 8 890 2.92695e+06 0 0
bsr 3 8 9635.33 1.90244e+06 0 0
bufw 3 8 0 0 17200.3 24424392
So looks like it might make sense to include this patch.
Thanks
Vivek
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Linux kernel 2.6.32 and later allocate address space from the top of the
kernel virtual memory address space.
This patch implements virtual memory size detection for 64 bit MIPS CPUs
to avoid resulting crashes.
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
Cc: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/935/
Reviewed-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Distributions now have lsmod in /bin instead of /sbin. But to handle
both cases, we look for it in /sbin /bin /usr/bin and /usr/sbin.
If lsmod is not found in any of those paths, it defaults to use
just lsmod and hopes that it lies in the path of the user.
Tested-by: Xavier Chantry <shiningxc@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
This is a cleanup for the dummy driver. The model kernel module parameter
is introduced to select the soundcard emulation.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
cev-isr reg offset for each function is better calculated using (any) eq-id
alloted to that function instead of using pci-func number(which
does not work in some configurations...)
Signed-off-by: Sathya Perla <sathyap@serverengines.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is very simple IrDA SIR driver for SuperH.
This driver was tested by irdaping/ircp on SH7724 EcoVec24 board
Signed-off-by: Kuninori Morimoto <morimoto.kuninori@renesas.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Almost all igmp functions accessing inet->mc_list are protected by
rtnl_lock(), but there is one exception which is ip_mc_sf_allow(),
so there is a chance of either ip_mc_drop_socket or ip_mc_leave_group
remove an entry while ip_mc_sf_allow is running causing a crash.
Signed-off-by: Flavio Leitner <fleitner@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 20dd3850bc "can: Speed up CAN frame
receiption by using ml_priv" the formerly used hlist of receiver lists for
each CAN netdevice has been replaced.
The hlist content ensured only CAN netdevices to be accessed by the
can_rx_(un)register() functions which accidently dropped away together with
the hlist receiver implementation.
This patch re-introduces the check for CAN netdevices in can_rx_(un)register().
Signed-off-by: Oliver Hartkopp <oliver@hartkopp.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
When we cat <debugfs>/tracing/stack_trace, we may cause circular lock:
sys_read()
t_start()
arch_spin_lock(&max_stack_lock);
t_show()
seq_printf(), vsnprintf() .... /* they are all trace-able,
when they are traced, max_stack_lock may be required again. */
The following script can trigger this circular dead lock very easy:
#!/bin/bash
echo 1 > /proc/sys/kernel/stack_tracer_enabled
mount -t debugfs xxx /mnt > /dev/null 2>&1
(
# make check_stack() zealous to require max_stack_lock
for ((; ;))
{
echo 1 > /mnt/tracing/stack_max_size
}
) &
for ((; ;))
{
cat /mnt/tracing/stack_trace > /dev/null
}
To fix this bug, we increase the percpu trace_active before
require the lock.
Reported-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
LKML-Reference: <4B67D4F9.9080905@cn.fujitsu.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The TCPMSS target is dropping SYN packets where:
1) There is data, or
2) The data offset makes the TCP header larger than the packet.
Both of these result in an error level printk. This printk has been
removed.
This change avoids dropping SYN packets containing data. If there
is also no MSS option (as well as data), one will not be added
because of possible complications due to the increased packet size.
Signed-off-by: Simon Arlott <simon@fire.lp0.eu>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Unconditionally save the register states when suspending and restore
them again at resume time. Register contents were not preserved over
suspend, and hence the driver takes false assumptions about them.
The clock must be enabled to access the register block.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
String constants that are continued on subsequent lines with \
are not good.
Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
The dependency on MFD_WM8994 rather than I2C went awry.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
IRQ autoprobing hasn't actually worked for us at all since very early in
2.6, but no one seems to have noticed given that none of the drivers
that use it see much testing.
yenta_socket is the odd one out, and that depends on PCI IRQs which are
fixed on all SH platforms anyways. Consequently, turning off autoprobing
fixes up crashes triggered by yenta_socket and at least gets it working
again on r7785rp.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
