Commit Graph

74727 Commits

Author SHA1 Message Date
Srivatsa Vaddagiri d842de871c sched: cpu accounting controller (V2)
Commit cfb5285660 removed a useful feature for
us, which provided a cpu accounting resource controller.  This feature would be
useful if someone wants to group tasks only for accounting purpose and doesnt
really want to exercise any control over their cpu consumption.

The patch below reintroduces the feature. It is based on Paul Menage's
original patch (Commit 62d0df6406), with
these differences:

        - Removed load average information. I felt it needs more thought (esp
	  to deal with SMP and virtualized platforms) and can be added for
	  2.6.25 after more discussions.
        - Convert group cpu usage to be nanosecond accurate (as rest of the cfs
	  stats are) and invoke cpuacct_charge() from the respective scheduler
	  classes
	- Make accounting scalable on SMP systems by splitting the usage
	  counter to be per-cpu
	- Move the code from kernel/cpu_acct.c to kernel/sched.c (since the
	  code is not big enough to warrant a new file and also this rightly
	  needs to live inside the scheduler. Also things like accessing
	  rq->lock while reading cpu usage becomes easier if the code lived in
	  kernel/sched.c)

The patch also modifies the cpu controller not to provide the same accounting
information.

Tested-by: Balbir Singh <balbir@linux.vnet.ibm.com>

 Tested the patches on top of 2.6.24-rc3. The patches work fine. Ran
 some simple tests like cpuspin (spin on the cpu), ran several tasks in
 the same group and timed them. Compared their time stamps with
 cpuacct.usage.

Signed-off-by: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2007-12-02 20:04:49 +01:00
Zhao Yakui 357dc4c3f1 ACPI: Delete the IRQ operation in throttling controll via PTC
The IRQ operation(enable/disable) should be avoided when throttling is
controlled via PTC method. It is replaced by the migration of task.

This fixes an oops on T61 -- a regression due to
f79f06ab9f b/c FixedHW support tried to read remote MSR with interrupts disabled.

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2007-12-01 23:27:15 -05:00
Mark Lord 02a121da5a sata_mv: Fix broken Marvell 7042 support.
sata_mv:  Fix broken Marvell 7042 support.

The Marvell 7042 chip is more or less the same as the 6042 internally,
but sports a PCIe bus.  Despite having identical SATA cores, the 7042
does differ from its PCI bus counterparts in placment and layout of
certain bus related registers.

This patch fixes sata_mv to distinguish between the PCI bus registers
of earlier chips, and the PCIe bus registers of the 7042.

Specifically, move the offsets and bit patterns for the
PCI/PCIe interrupt cause/mask registers into the struct mv_host_priv,
as these values differ between the 6xxx and 7xxx series chips.

This fixes the driver to not access reserved PCI addresses,
and prevents the lockups reported in linux-2.6.24 with 7042 boards.

Also add a new PCI ID for the Highpoint 2300 7042-based board
that I'm using for testing this stuff here.

Tested with Marvell 6081 + 7042 chips, on x86 & x86_64.

Signed-off-by: Mark Lord <mlord@pobox.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:44:10 -05:00
Alan Cox 0f9fe9b714 libata: Fix early use of port printk. (Was Re: ata4294967295: failed to start port (errno=-19))
On Fri, 30 Nov 2007 14:34:11 +0200 (EET)
Meelis Roos <mroos@linux.ee> wrote:

> > Can you stick a stack trace in at that point ? That would help diagnose
> > it a great deal quicker.
>
> Finally done - found out hard way that BUG() is too bad and
> dump_st5ack() suits me better.

Thanks. This should fix the real cause, and also allow for port start to
fail politely with -ENODEV.

Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:44:10 -05:00
Peter Schwenke 04d86d6fc0 ata_piix: add more toshiba laptops to broken suspend list
Add more toshiba laptops to broken suspend list.  This is from OSDL
bugzilla bug 7780.

tj: re-formatted patch and added description and SOB.

Signed-off-by: Peter Schwenke <peter@bluetoad.com.au>
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:44:10 -05:00
Peter Missel e9f3340673 libata: More IVB horkage from TSST
libata: Add more TSST (Samsung/Toshiba) IDE drives with broken
cable detection validation bits.

signed-off-by: Peter Missel (peter.missel@onlinehome.de)
Acked-by: Alan Cox <alan@redhat.com>

Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:35:59 -05:00
Tejun Heo abb6a88974 libata: report protocol and full CDB on error
Protocol and CDB allocation size field are important in determining
what went wrong with ATAPI commands.  Report them on failure.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:35:58 -05:00
Kristoffer Nyborg Gregertsen 1c20a493ca Several fixes for the AVR32 PATA driver
Several fixes for the AVR32 PATA driver:

* Updated to use new AVR32 SMC timing API. This removes the need for "magic"
constants in signal timing.

* Removed the ATA_FLAG_PIO_POLLING, the driver should use interrupts.

* Removed .port_disable and .irq_ack as these are no longer needed.

* Improved some comments.

Signed-off-by: Kristoffer Nyborg Gregertsen <kngregertsen@norway.atmel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:35:58 -05:00
Saeed Bishara 2d79ab8fd7 sata_mv: fix compilation error when enabling DEBUG
use sstatus instead status.

Signed-off-by: Saeed Bishara <saeed@marvell.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:35:58 -05:00
sonic zhang f88c480dac Set proper ATA UDMA mode for bf548 according to system clock.
UDMA Mode - Frequency compatibility

UDMA5 - 100 MB/s   - SCLK  = 133 MHz
UDMA4 - 66 MB/s    - SCLK >=  80 MHz
UDMA3 - 44.4 MB/s  - SCLK >=  50 MHz
UDMA2 - 33 MB/s    - SCLK >=  40 MHz

Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 17:35:58 -05:00
Thomas Bogendoerfer 9ec46c6dae LIB82596: correct data types for hardware addresses
dma_addr_t is 64bit wide on some architectures (for example 64bit MIPS),
so it's not a good idea to use it for 32bit wide addresses in descriptors.

Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:38:28 -05:00
Stephen Hemminger 48f6b05361 via-velocity: don't oops on MTU change (resend)
The VIA veloicty driver needs the following to allow changing MTU when down.
The buffer size needs to be computed when device is brought up, not when
device is initialized.  This also fixes a bug where the buffer size was
computed differently on change_mtu versus initial setting.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:35:51 -05:00
David Woodhouse dda93b486a Stop phy code from returning success to unknown ioctls.
This kind of sucks, and prevents the Fedora installer from using the
device for network installs...

[root@efika phy]# iwconfig eth0
Warning: Driver for device eth0 has been compiled with an ancient version
of Wireless Extension, while this program support version 11 and later.
Some things may be broken...

eth0        ESSID:off/any  Nickname:""
          NWID:0  Channel:0  Access Point: 00:00:BF:81:14:E0
          Bit Rate:-1.08206e+06 kb/s   Sensitivity=0/0
          RTS thr:off   Fragment thr:off
          Encryption key:<too big>
          Power Management:off

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:35:51 -05:00
David Woodhouse 3d26e69533 SET_NETDEV_DEV() in fec_mpc52xx.c
This helps to allow the Fedora installer to use the built-in Ethernet on
the Efika for a network install.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:35:51 -05:00
Paul Mundt 0313d9884f net: smc911x: only enable for mpr2 on sh.
The smc911x.h is a bit of a mess, not supporting any sort of generic
configuration. For the moment only ARCH_PXA and SH_MAGIC_PANEL_R2 have
suitable definitions, so we reflect this in the Kconfig also.

While there are other SH boards that will likely turn this on in the
2.6.25 time frame, it's not worth trying to stub around at the moment.

Fixes up the allmodconfig build, as noted by akpm.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:32 -05:00
Auke Kok f7bbb90983 e1000: Fix NAPI state bug when Rx complete
Don't exit polling when we have not yet used our budget, this causes
the NAPI system to end up with a messed up poll list.

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:32 -05:00
Stephen Hemminger e970d1f810 sky2: turn of dynamic Tx watermark workaround (FE+ only)
Add workaround for issues FE+ (A0) transmit watermark.
This is copied verbatim from vendor driver sk98lin (10.22.4.3).
Don't have that chip version and no more information seems to be available.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:31 -05:00
Stephen Hemminger 7782c8c4d7 sky2: don't use AER routines
Using PCIE advanced error recovery stuff creates more user problems than it's worth.
The AER stuff depends on MMCONFIG and in many configurations it just doesn't work.
Plus it doesn't add any real functionality to the driver. The sky2
driver handles its own errors fine as is.

This reverts 555382cbfc

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:31 -05:00
Stephen Hemminger b32f40c485 sky2: revert to access PCI config via device space
Using the hardware window into PCI config space is more reliable
and smaller/faster than using the pci_config routines. It avoids issues
with MMCONFIG etc.

Reverts: 167f53d05f

Please apply for 2.6.24

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:31 -05:00
Divy Le Ray e0348b9ae5 cxgb - fix stats
Fix MAC stats accounting.
Fix get_stats.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:31 -05:00
Divy Le Ray 445cf803ca cxgb - fix NAPI
netif_rx_complete() should be called only
when work_done < budget.

Signed-off-by: Divy Le ray <divy@chelsio.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:31 -05:00
Divy Le Ray 7832ee034b cxgb - fix T2 GSO
The patch ensures that a GSO skb has enough headroom
to push an encapsulating cpl_tx_pkt_lso header.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:31 -05:00
Kim Phillips bd0ceaab86 ucc_geth: handle passing of RX-only and TX-only internal delay PHY connection type parameters
Extend the RGMII-Internal Delay specification case to include
TX-only and RX-only variants.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Tested-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Acked-by: Li Yang <leoli@freescale.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:30 -05:00
Kim Phillips 9daf5a7695 phylib: marvell: add support for TX-only and RX-only Internal Delay
Previously, Internal Delay specification implied the delay be
applied to both TX and RX.  This patch allows for separate TX/RX-only
internal delay specification.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Tested-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Acked-by: Li Yang <leoli@freescale.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:30 -05:00
Kim Phillips 7d400a4c58 phylib: add PHY interface modes for internal delay for tx and rx only
Allow phylib specification of cases where hardware needs to configure
PHYs for Internal Delay only on either RX or TX (not both).

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Tested-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Acked-by: Li Yang <leoli@freescale.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:30 -05:00
Stephen Hemminger 1a8098be98 skge: MTU changing fix
The code to change MTU doesn't correctly handle all the chip variations
and requirements for restarting.  On Genesis chips changing MTU would just
cause receiver to hang.
Use a simpler approach of just taking link down/up if needed.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:30 -05:00
Stephen Hemminger 44c7fccec4 skge: serial mode register values
For compatiablity with sk98lin, make sure and set same values
in serial mode register.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:30 -05:00
Stephen Hemminger bf9f56d585 skge version 1.13
Version for 2.6.24

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:29 -05:00
Stephen Hemminger 485982a99a skge: increase TX threshold for Jumbo
Need to increase TX threshold when doing Jumbo frames on dual port board
to avoid underruns. (Code from sk98lin).

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:29 -05:00
Stephen Hemminger 21d7f67700 skge: fiber link up/down fix
The driver would not work over fibre if other end when down then
came back up (would require reloading driver). The correct way
to manage the link the same way for both TP and fibre.

Resloves problem described in: 	 http://lkml.org/lkml/2007/11/6/395

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:29 -05:00
Stephen Hemminger 799b21d2bd skge: retry on MAC shutdown
Make sure and retry when shutting down the MAC. This code is copied
from sk98lin driver.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:29 -05:00
Stephen Hemminger d08b9bdf02 skge: receive flush logic
Receive FIFO overrun is not catastrophic condition, so don't flush when
it happens.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:29 -05:00
Stephen Hemminger 29816d9aa5 skge: FIFO Ram calculation error
The calculation of usable FIFO RAM is wrong in the skge driver.
First, is doesn't take into account the reserved area on the original
SysKonnect Genesis boards. Second it has an off-by-one error because
hw->ports is either 1 or 2.

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:29 -05:00
Peter Tiedemann 1c14788590 ctc: make use of alloc_netdev()
Currently ctc-device initialization is broken (kernel bug in
ctc_new_device).
The new network namespace code reveals a deficiency of the
ctc driver. It should make use of alloc_netdev() as described
in Documentation/networking/netdevices.txt.

Signed-off-by: Peter Tiedemann <ptiedem@de.ibm.com>
Signed-off-by: Ursula Braun <braunu@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:28 -05:00
Sreenivasa Honnur 9f74ffdebf S2io: Fixed the case when the card initialization fails on mtu change
Fix the case when the card initialization fails on a mtu change and then
close is called (due to ifdown), which frees non existent rx buffers.
- Returning appropriate error codes in init_nic function.
- In s2io_close function s2io_card_down is called only when device is up.
- In s2io_change_mtu function return value of s2io_card_up function
  is checked and returned if it failed.

Signed-off-by: Surjit Reang <surjit.reang@neterion.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-01 16:32:28 -05:00
Jeff Garzik 85b161a826 Merge branch 'fixes-jgarzik' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-2.6 into upstream-fixes 2007-12-01 16:20:02 -05:00
Jeff Garzik c99da91e7a Merge branch 'master' into upstream-fixes 2007-12-01 16:18:56 -05:00
Eric W. Biederman 2b1e300a9d [NETNS]: Fix /proc/net breakage
Well I clearly goofed when I added the initial network namespace support
for /proc/net.  Currently things work but there are odd details visible to
user space, even when we have a single network namespace.

Since we do not cache proc_dir_entry dentries at the moment we can just
modify ->lookup to return a different directory inode depending on the
network namespace of the process looking at /proc/net, replacing the
current technique of using a magic and fragile follow_link method.

To accomplish that this patch:
- introduces a shadow_proc method to allow different dentries to
  be returned from proc_lookup.
- Removes the old /proc/net follow_link magic
- Fixes a weakness in our not caching of proc generic dentries.

As shadow_proc uses a task struct to decided which dentry to return we can
go back later and fix the proc generic caching without modifying any code
that uses the shadow_proc method.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2007-12-02 00:33:17 +11:00
Linus Torvalds 92d499d991 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  IB/ehca: Fix static rate if path faster than link
  IPoIB: Fix oops if xmit is called when priv->broadcast is NULL
2007-11-30 20:07:54 -08:00
Linus Torvalds 1811534a80 Merge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] Fix build.
  [MIPS] Fix use of smp_processor_id() in preemptible code.
2007-11-30 17:07:38 -08:00
Ralf Baechle e6a1bb725e [MIPS] Fix build.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-12-01 00:39:37 +00:00
Pavel Kiryukhin 54fd6441e0 [MIPS] Fix use of smp_processor_id() in preemptible code.
Freeing prom memory: 956kb freed
Freeing firmware memory: 978944k freed
Freeing unused kernel memory: 180k freed
BUG: using smp_processor_id() in preemptible [00000000] code: swapper/1
caller is r4k_dma_cache_wback_inv+0x144/0x2a0
Call Trace:
 [<80117af8>] r4k_dma_cache_wback_inv+0x144/0x2a0
 [<802e4b84>] debug_smp_processor_id+0xd4/0xf0
 [<802e4b7c>] debug_smp_processor_id+0xcc/0xf0
...
CONFIG_DEBUG_PREEMPT is enabled.
--
Bug cause is blast_dcache_range() in preemptible code [in
r4k_dma_cache_wback_inv()].
blast_dcache_range() is constructed via __BUILD_BLAST_CACHE_RANGE that
uses cpu_dcache_line_size(). It uses current_cpu_data that use
smp_processor_id() in turn. In case of CONFIG_DEBUG_PREEMPT
smp_processor_id emits BUG if we are executing with preemption
enabled.

Cpu options of cpu0 are assumed to be the superset of all processors.

Can I make the same assumptions for cache line size  and fix this
issue the following way:

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-12-01 00:39:37 +00:00
Joachim Fenkes b1812582ba IB/ehca: Fix static rate if path faster than link
The formula would yield -1 if the path is faster than the link, which
is wrong in a bad way (max throttling).  Clamp to 0, which is the
correct value.

Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-11-30 16:19:41 -08:00
Linus Torvalds b62c855938 Merge branch 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm
* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] 4680/1: parentheses around NR_IRQS definition
  [ARM] 4679/1: AT91: Change maintainer email address
  [ARM] 4675/1: pxa: fix mfp address definition error for pxa320
  [ARM] 4674/1: pxa: increase LCD PCLK drive strength to fast 2mA for PXA300/PXA310
  [ARM] 4673/1: pxa: add missing IRQ_SSP4 definitions for PXA3xx
  [ARM] 4672/1: pxa: fix DRCMR(n) to support PXA27x and later processors
  [ARM] 4665/1: fix __und_usr wrt accessing the undefined insn in user space
  [ARM] 4659/1: remove possibilities for spurious false negative with __kuser_cmpxchg
  [ARM] 4661/1: fix do_undefinstr wrt the enabling of IRQs
  [ARM] uengine: fix memset size error
  [ARM] 4648/1: i.MX/MX1 ensure more complete AITC initialization
  [ARM] 4611/2: AT91: Fix GPIO buttons pins on SAM9261-EK.
  [ARM] 4650/1: AT91: New-style init of I2C, support for i2c-gpio
  [ARM] 4604/2: AT91: Master clock divistor on SAM9
  [ARM] 4662/1: Fix PXA serial driver compilation if SERIAL_PXA_CONSOLE is disabled
  [ARM] PXA ssp: unlock when ssp tries to close an invalid port
  [ARM] 4654/1: pxa: update default MFP register value
  [ARM] 4653/1: pxa: fix a gpio typo in mfp-pxa320.h
  [ARM] 4652/1: pxa: fix a typo of pxa27x usb host clk definition
  [ARM] 4651/1: pxa: add PXA3xx specific IRQ definitions
2007-11-30 08:11:18 -08:00
Matthew Wilcox 80cbd911ca Fix kmem_cache_free performance regression in slab
The database performance group have found that half the cycles spent
in kmem_cache_free are spent in this one call to BUG_ON.  Moving it
into the CONFIG_SLAB_DEBUG-only function cache_free_debugcheck() is a
performance win of almost 0.5% on their particular benchmark.

The call was added as part of commit ddc2e812d5
with the comment that "overhead should be minimal".  It may have been
minimal at the time, but it isn't now.

[ Quoth Pekka Enberg: "I don't think the BUG_ON per se caused the
  performance regression but rather the virt_to_head_page() changes to
  virt_to_cache() that were added later." ]

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
Acked-by: Pekka J Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-11-30 08:08:05 -08:00
Pablo Neira Ayuso e03ba84adb [TEXTSEARCH]: Do not allow zero length patterns in the textsearch infrastructure
If a zero length pattern is passed then return EINVAL.
Avoids infinite loops (bm) or invalid memory accesses (kmp).

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2007-12-01 00:03:52 +11:00
Jan Engelhardt 67b4af2970 [NETFILTER]: fix forgotten module release in xt_CONNMARK and xt_CONNSECMARK
Fix forgotten module release in xt_CONNMARK and xt_CONNSECMARK

When xt_CONNMARK is used outside the mangle table and the user specified
"--restore-mark", the connmark_tg_check() function will (correctly)
error out, but (incorrectly) forgets to release the L3 conntrack module.
Same for xt_CONNSECMARK.

Fix is to move the call to acquire the L3 module after the basic
constraint checks.

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2007-12-01 00:01:50 +11:00
Patrick McHardy 9dc0564e86 [NETFILTER]: xt_TCPMSS: remove network triggerable WARN_ON
ipv6_skip_exthdr() returns -1 for invalid packets. don't WARN_ON
that.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2007-11-30 23:58:03 +11:00
Pavel Emelyanov 3ccd86241b [DECNET]: dn_nl_deladdr() almost always returns no error
As far as I see from the err variable initialization
the dn_nl_deladdr() routine was designed to report errors
like "EADDRNOTAVAIL" and probaby "ENODEV".

But the code sets this err to 0 after the first nlmsg_parse
and goes on, returning this 0 in any case.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2007-11-30 23:43:31 +11:00
Evgeniy Polyakov d31c7b8fa3 [IPV6]: Restore IPv6 when MTU is big enough
Avaid provided test application, so bug got fixed.

IPv6 addrconf removes ipv6 inner device from netdev each time cmu
changes and new value is less than IPV6_MIN_MTU (1280 bytes).
When mtu is changed and new value is greater than IPV6_MIN_MTU,
it does not add ipv6 addresses and inner device bac.

This patch fixes that.

Tested with Avaid's application, which works ok now.

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2007-11-30 23:36:08 +11:00