safe/jmp/linux-2.6
13 years agoof/dma: mpc512x_dma.c: Fix build failures
Anatolij Gustschin [Mon, 31 May 2010 16:39:13 +0000 (18:39 +0200)]
of/dma: mpc512x_dma.c: Fix build failures

Fixes build errors caused by the:
 - OF device_node pointer being moved into struct device
 - removal of the match_table field from struct of_platform_driver

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoof/mtd/nand: mpc5121_nfc.c: Fix build failures
Anatolij Gustschin [Mon, 31 May 2010 16:37:15 +0000 (18:37 +0200)]
of/mtd/nand: mpc5121_nfc.c: Fix build failures

Fixes build errors caused by the:
 - OF device_node pointer being moved into struct device
 - removal of the match_table field from struct of_platform_driver

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoof/spi: mpc512x_psc_spi.c: Fix build failures
Anatolij Gustschin [Mon, 31 May 2010 16:34:54 +0000 (18:34 +0200)]
of/spi: mpc512x_psc_spi.c: Fix build failures

Fixes build errors caused by the:
 - OF device_node pointer being moved into struct device
 - removal of the match_table field from struct of_platform_driver

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agowatchdog: Fix build failure with OF changes
Michael Guntsche [Wed, 2 Jun 2010 08:25:52 +0000 (02:25 -0600)]
watchdog: Fix build failure with OF changes

commit 61c7a080a5a061c976988fd4b844dfb468dda255 ( of: Always use
'struct device.of_node' to get device node pointer.) missed
drivers/watchdog/mpc8xxx_wdt.c. This patch fixes it

Signed-off-by: Michael Guntsche <mike@it-loops.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoof/spi: Fix build failure on spi_ppc4xx.c
Grant Likely [Wed, 26 May 2010 05:24:02 +0000 (23:24 -0600)]
of/spi: Fix build failure on spi_ppc4xx.c

This patch fixes a build error caused by the OF device_node pointer
being moved into struct device.

Fixes bug introduced by commit 61c7a080a5a061c976988fd4b844dfb468dda255
(of: Always use 'struct device.of_node' to get device node pointer)

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
CC: Sean MacLennan <smaclennan@pikatech.com>
CC: spi-devel-general@lists.sourceforge.net
CC: devicetree-discuss@lists.ozlabs.org
13 years agoof/usb: fix build error due to of_node pointer move
Grant Likely [Wed, 2 Jun 2010 19:35:02 +0000 (13:35 -0600)]
of/usb: fix build error due to of_node pointer move

Fix driver to use new location of of_node pointer (introduced by commit
use new location of of_node pointer (introduced by commit
61c7a080a5a061c976988fd4b844dfb468dda255; of: Always use 'struct
device.of_node' to get device node pointer)

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Reported-by: John Linn <john.linn@xilinx.com>
CC: Greg Kroah-Hartman <gregkh@suse.de>
CC: Alan Stern <stern@rowland.harvard.edu>
CC: linux-usb@vger.kernel.org
CC: devicetree-discuss@lists.ozlabs.org
13 years agolibata: implement on-demand HPA unlocking
Tejun Heo [Sat, 15 May 2010 18:09:34 +0000 (20:09 +0200)]
libata: implement on-demand HPA unlocking

Implement ata_scsi_unlock_native_capacity() which will be called
through SCSI layer when block layer notices that partitions on a
device extend beyond the end of the device.  It requests EH to unlock
HPA, waits for completion and returns the current device capacity.

This allows libata to unlock HPA on demand instead of having to decide
whether to unlock upfront.  Unlocking on demand is safer than
unlocking by upfront because some BIOSes write private data to the
area beyond HPA limit.  This was suggested by Ben Hutchings.

Signed-off-by: Tejun Heo <tj@kernel.org>
Suggested-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
13 years agolibata: use the enlarged capacity after late HPA unlock
Tejun Heo [Sat, 15 May 2010 18:09:33 +0000 (20:09 +0200)]
libata: use the enlarged capacity after late HPA unlock

After late HPA unlock, libata kept using the original capacity
ignoring the new larger native capacity.  Enlarging device on the fly
doesn't cause any harm.  Use the larger native capacity instead.  This
will enable on-demand HPA unlocking.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
13 years agoSCSI: implement sd_unlock_native_capacity()
Tejun Heo [Sat, 15 May 2010 18:09:32 +0000 (20:09 +0200)]
SCSI: implement sd_unlock_native_capacity()

Implement sd_unlock_native_capacity() method which calls into
hostt->unlock_native_capacity() if implemented.  This will be invoked
by block layer if partitions extend beyond the end of the device and
can be used to implement, for example, on-demand ATA host protected
area unlocking.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
13 years agolibata-sff: trivial corrections to Kconfig help text
Stefan Richter [Sat, 29 May 2010 10:47:45 +0000 (12:47 +0200)]
libata-sff: trivial corrections to Kconfig help text

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
13 years agosata_nv: don't diddle with nIEN on mcp55
Tejun Heo [Tue, 1 Jun 2010 15:29:21 +0000 (17:29 +0200)]
sata_nv: don't diddle with nIEN on mcp55

On mcp55, nIEN gets stuck once set and liteon blueray rom iHOS104-08
violates ATA specification and fails to set I on D2H Reg FIS if nIEN
is set when the command was issued.  When the other party is following
the spec, both devices can work fine but when the two flaws are put
together, they can't talk to each other.

mcp55 has its own IRQ masking mechanism and there's no reason to mess
with nIEN in the first place.  Fix it by dropping nIEN diddling from
nv_mcp55_freeze/thaw().

This was originally reported by Cengiz.  Although Cengiz hasn't
verified the fix yet, I could reproduce this problem and verfiy the
fix.  Even if Cengiz is experiencing different or additional problems,
this patch is needed.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Cengiz Günay <cgunay@emory.edu>
Cc: stable@kernel.org
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
13 years agosata_via: magic vt6421 fix for transmission problems w/ WD drives
Tejun Heo [Mon, 31 May 2010 14:26:48 +0000 (16:26 +0200)]
sata_via: magic vt6421 fix for transmission problems w/ WD drives

vt6421 has problems talking to recent WD drives.  It causes a lot of
transmission errors while high bandwidth transfer as reported in the
following bugzilla entry.

  https://bugzilla.kernel.org/show_bug.cgi?id=15173

Joseph Chan provided the following fix.  I don't have any idea what it
does but I can verify the issue is gone with the patch applied.

Signed-off-by: Tejun Heo <tj@kernel.org>
Originally-from: Joseph Chan <JosephChan@via.com.tw>
Reported-by: Jorrit Tijben <sjorrit@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
13 years agocls_u32: use skb_header_pointer() to dereference data safely
Changli Gao [Wed, 2 Jun 2010 14:32:42 +0000 (07:32 -0700)]
cls_u32: use skb_header_pointer() to dereference data safely

use skb_header_pointer() to dereference data safely

the original skb->data dereference isn't safe, as there isn't any skb->len or
skb_is_nonlinear() check. skb_header_pointer() is used instead in this patch.
And when the skb isn't long enough, we terminate the function u32_classify()
immediately with -1.

Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoTCP: tcp_hybla: Fix integer overflow in slow start increment
Daniele Lacamera [Wed, 2 Jun 2010 02:02:04 +0000 (02:02 +0000)]
TCP: tcp_hybla: Fix integer overflow in slow start increment

For large values of rtt, 2^rho operation may overflow u32. Clamp down the increment to 2^16.

Signed-off-by: Daniele Lacamera <root@danielinux.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoact_nat: fix the wrong checksum when addr isn't in old_addr/mask
Changli Gao [Sat, 29 May 2010 14:26:59 +0000 (14:26 +0000)]
act_nat: fix the wrong checksum when addr isn't in old_addr/mask

fix the wrong checksum when addr isn't in old_addr/mask

For TCP and UDP packets, when addr isn't in old_addr/mask we don't do SNAT or
DNAT, and we should not update layer 4 checksum.

Signed-off-by: Changli Gao <xiaosuo@gmail.com>
----
 net/sched/act_nat.c |    4 ++++
 1 file changed, 4 insertions(+)
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agonet/fec: fix pm to survive to suspend/resume
Eric Bénard [Wed, 2 Jun 2010 13:13:34 +0000 (06:13 -0700)]
net/fec: fix pm to survive to suspend/resume

* in the actual driver, calling fec_stop and fec_enet_init doesn't
allow to have a working network interface at resume (where a
ifconfig down and up is required to recover the interface)
* by using fec_enet_close and fec_enet_open, this patch solves this
problem and handle the case where the link changed between suspend
and resume
* this patch also disable clock at suspend and reenable it at resume

Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agokorina: count RX DMA OVR as rx_fifo_error
Phil Sutter [Sat, 29 May 2010 13:23:36 +0000 (13:23 +0000)]
korina: count RX DMA OVR as rx_fifo_error

This way, RX DMA overruns (actually being caused by overrun of the
512byte input FIFO) show up in ifconfig output. The rx_fifo_errors
counter is unused otherwise.

Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agokorina: use netdev_alloc_skb_ip_align() here, too
Phil Sutter [Sat, 29 May 2010 13:23:35 +0000 (13:23 +0000)]
korina: use netdev_alloc_skb_ip_align() here, too

This patch completes commit 89d71a66c40d629e3b1285def543ab1425558cd5
which missed this spot, as it seems.

Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agokorina: fix deadlock on RX FIFO overrun
Phil Sutter [Sat, 29 May 2010 13:23:34 +0000 (13:23 +0000)]
korina: fix deadlock on RX FIFO overrun

By calling korina_restart(), the IRQ handler tries to disable the
interrupt it's currently serving. This leads to a deadlock since
disable_irq() waits for any running IRQ handlers to finish before
returning. This patch addresses the issue by turning korina_restart()
into a workqueue task, which is then scheduled when needed.

Reproducing the deadlock is easily done using e.g. GNU netcat to send
large amounts of UDP data to the host running this driver.

Note that the same problem (and fix) applies to TX FIFO underruns, but
apparently these are less easy to trigger.

Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoMerge branch 'fix/asoc' into for-linus
Takashi Iwai [Wed, 2 Jun 2010 12:18:13 +0000 (14:18 +0200)]
Merge branch 'fix/asoc' into for-linus

13 years agoMerge branch 'fix/hda' into for-linus
Takashi Iwai [Wed, 2 Jun 2010 12:18:06 +0000 (14:18 +0200)]
Merge branch 'fix/hda' into for-linus

13 years agoMerge branch 'for-2.6.35' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Takashi Iwai [Wed, 2 Jun 2010 12:17:44 +0000 (14:17 +0200)]
Merge branch 'for-2.6.35' of git://git./linux/kernel/git/broonie/sound-2.6 into fix/asoc

13 years agoMerge branch 'master' of git.alsa-project.org:alsa-kernel into fix/hda
Takashi Iwai [Wed, 2 Jun 2010 12:15:10 +0000 (14:15 +0200)]
Merge branch 'master' of git.alsa-project.org:alsa-kernel into fix/hda

13 years agoALSA: hda-intel - fix wallclk variable update and condition
Jaroslav Kysela [Wed, 2 Jun 2010 11:29:17 +0000 (13:29 +0200)]
ALSA: hda-intel - fix wallclk variable update and condition

This patch fixes thinko introduced in "last minutes" before commiting
of the last wallclk patch.

It also fixes the condition checking if the first period after last
wallclk update is processed. There is a little rounding error in
period_wallclk.

Signed-off-by: Jaroslav Kysela <perex@perex.cz>
13 years agonet: fix conflict between null_or_orig and null_or_bond
John Fastabend [Wed, 12 May 2010 21:31:11 +0000 (21:31 +0000)]
net: fix conflict between null_or_orig and null_or_bond

If a skb is received on an inactive bond that does not meet
the special cases checked for by skb_bond_should_drop it should
only be delivered to exact matches as the comment in
netif_receive_skb() says.

However because null_or_bond could also be null this is not
always true.  This patch renames null_or_bond to orig_or_bond
and initializes it to orig_dev.  This keeps the intent of
null_or_bond to pass frames received on VLAN interfaces stacked
on bonding interfaces without invalidating the statement for
null_or_orig.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agonet: init_vlan should not copy slave or master flags
John Fastabend [Wed, 12 May 2010 21:31:06 +0000 (21:31 +0000)]
net: init_vlan should not copy slave or master flags

The vlan device should not copy the slave or master flags from
the real device. It is not in the bond until added nor is it
a master.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoMerge branch 'fix/hda' into for-linus
Takashi Iwai [Wed, 2 Jun 2010 10:09:29 +0000 (12:09 +0200)]
Merge branch 'fix/hda' into for-linus

13 years agoALSA: asihpi - Fix uninitialized variable
Takashi Iwai [Wed, 2 Jun 2010 10:08:37 +0000 (12:08 +0200)]
ALSA: asihpi - Fix uninitialized variable

Initialize prev_ctl properly before reference:
  sound/pci/asihpi/asihpi.c: In function ‘snd_card_asihpi_mixer_new’:
  sound/pci/asihpi/asihpi.c:2568:30: warning: ‘prev_ctl.dst_node_index’ may be used uninitialized in this function

Signed-off-by: Takashi Iwai <tiwai@suse.de>
13 years agoenic: bug fix: make the set/get netlink VF_PORT support symmetrical
Scott Feldman [Tue, 1 Jun 2010 08:59:33 +0000 (08:59 +0000)]
enic: bug fix: make the set/get netlink VF_PORT support symmetrical

To make get/set netlink VF_PORT truly symmetrical, we need to keep track
of what items are set and only return those items on get.  Previously, the
driver wasn't differentiating between a set of attr with a NULL string,
for example, and not setting the attr at all.  We only want to return
the NULL string if the attr was actually set with a NULL string.  Otherwise,
don't return the attr.

Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agobnx2: Fix hang during rmmod bnx2.
Michael Chan [Tue, 1 Jun 2010 15:05:36 +0000 (15:05 +0000)]
bnx2: Fix hang during rmmod bnx2.

The regression is caused by:

commit 4327ba435a56ada13eedf3eb332e583c7a0586a9
    bnx2: Fix netpoll crash.

If ->open() and ->close() are called multiple times, the same napi structs
will be added to dev->napi_list multiple times, corrupting the dev->napi_list.
This causes free_netdev() to hang during rmmod.

We fix this by calling netif_napi_del() during ->close().

Also, bnx2_init_napi() must not be in the __devinit section since it is
called by ->open().

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoxfrm: force a dst reference in __xfrm_route_forward()
Eric Dumazet [Tue, 1 Jun 2010 10:04:49 +0000 (10:04 +0000)]
xfrm: force a dst reference in __xfrm_route_forward()

Packets going through __xfrm_route_forward() have a not refcounted dst
entry, since we enabled a noref forwarding path.

xfrm_lookup() might incorrectly release this dst entry.

It's a bit late to make invasive changes in xfrm_lookup(), so lets force
a refcount in this path.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agosh: Make intc messages consistent via pr_fmt.
Paul Mundt [Wed, 2 Jun 2010 09:10:00 +0000 (18:10 +0900)]
sh: Make intc messages consistent via pr_fmt.

Wrapping pr_fmt to the KBUILD_MODNAME prefix seems to be the trendy
thing to do these days, so just do that instead of manually tidying
up the stragglers.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agosh: make sure static declaration on ms7724se
Kuninori Morimoto [Wed, 2 Jun 2010 00:27:38 +0000 (00:27 +0000)]
sh: make sure static declaration on ms7724se

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agosh: make sure static declaration on mach-migor
Kuninori Morimoto [Wed, 2 Jun 2010 00:27:46 +0000 (00:27 +0000)]
sh: make sure static declaration on mach-migor

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agosh: make sure static declaration on mach-ecovec24
Kuninori Morimoto [Wed, 2 Jun 2010 00:27:31 +0000 (00:27 +0000)]
sh: make sure static declaration on mach-ecovec24

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agosh: make sure static declaration on mach-ap325rxa
Kuninori Morimoto [Wed, 2 Jun 2010 00:27:26 +0000 (00:27 +0000)]
sh: make sure static declaration on mach-ap325rxa

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agoclocksource: sh_cmt: compute mult and shift before registration
Paul Mundt [Wed, 2 Jun 2010 08:10:44 +0000 (17:10 +0900)]
clocksource: sh_cmt: compute mult and shift before registration

Based on the sh_tmu change in 66f49121ffa41a19c59965b31b046d8368fec3c7
("clocksource: sh_tmu: compute mult and shift before registration").
The same issues impact the sh_cmt driver, so we take the same approach
here.

Cc: stable@kernel.org
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agoclocksource: sh_tmu: compute mult and shift before registration
Aurelien Jarno [Mon, 31 May 2010 21:45:48 +0000 (21:45 +0000)]
clocksource: sh_tmu: compute mult and shift before registration

Since commit 98962465ed9e6ea99c38e0af63fe1dcb5a79dc25 ("nohz: Prevent
clocksource wrapping during idle"), the CPU of an R2D board never goes
to idle. This commit assumes that mult and shift are assigned before
the clocksource is registered. As a consequence the safe maximum sleep
time is negative and the CPU never goes into idle.

This patch fixes the problem by moving mult and shift initialization
from sh_tmu_clocksource_enable() to sh_tmu_register_clocksource().

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Cc: stable@kernel.org
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agopowerpc/macio: Fix probing of macio devices by using the right of match table
Benjamin Herrenschmidt [Wed, 2 Jun 2010 07:09:18 +0000 (17:09 +1000)]
powerpc/macio: Fix probing of macio devices by using the right of match table

Grant patches added an of mach table to struct device_driver. However,
while he changed the macio device code to use that, he left the match
table pointer in struct macio_driver and didn't update drivers to use
the "new" one, thus breaking the probing.

This completes the change by moving all drivers to setup the "new"
one, removing all traces of the old one, and while at it (since it
changes the exact same locations), I also remove two other duplicates
from struct driver which are the name and owner fields.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agoagp/uninorth: Fix oops caused by flushing too much
Paul Mackerras [Tue, 1 Jun 2010 19:33:08 +0000 (19:33 +0000)]
agp/uninorth: Fix oops caused by flushing too much

This fixes a sporadic oops at boot on G5 Power Macs.  The table_end
variable has the address of the last byte of the table.  Adding on
PAGE_SIZE means we flush too much, and if the page after the table
is not mapped for any reason, the kernel will oops.  Instead we add
on 1 because flush_dcache_range() interprets its second argument as
the first byte past the range to be flushed.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agopowerpc/pasemi: Update MAINTAINERS file
Olof Johansson [Sat, 22 May 2010 05:17:38 +0000 (05:17 +0000)]
powerpc/pasemi: Update MAINTAINERS file

PWRficient platform work is definitely in maintenance mode these
days, update MAINTAINERS file to reflect reality.

Website is long gone as well.

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agopowerpc/cell: Fix integer constant warning
Denis Kirjanov [Thu, 27 May 2010 04:19:52 +0000 (04:19 +0000)]
powerpc/cell: Fix integer constant warning

Fix smatch warning:  warning: constant 0x800000000 is so big it is long

Signed-off-by: Denis Kirjanov <dkirjanov@kernel.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agopowerpc/kprobes: Remove resume_execution() in kprobes
Ananth N Mavinakayanahalli [Thu, 27 May 2010 19:19:20 +0000 (19:19 +0000)]
powerpc/kprobes: Remove resume_execution() in kprobes

emulate_step() in kprobe_handler() would've already determined if the
probed instruction can be emulated. We single-step in hardware only if
the instruction couldn't be emulated. resume_execution() therefore is
superfluous -- all we need is to fix up the instruction pointer after
single-stepping.

Thanks to Paul Mackerras for catching this.

Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agopowerpc/macio: Don't dereference pointer before null check
Andreas Schwab [Tue, 1 Jun 2010 05:58:40 +0000 (05:58 +0000)]
powerpc/macio: Don't dereference pointer before null check

Signed-off-by: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agoof/powerpc: fix 85xx RapidIO device node pointer
Alexandre Bounine [Fri, 28 May 2010 17:56:17 +0000 (13:56 -0400)]
of/powerpc: fix 85xx RapidIO device node pointer

Fixes bug introduced by commit 61c7a080a5a061c976988fd4b844dfb468dda255
(of: Always use 'struct device.of_node' to get device node pointer)

Signed-off-by: Alexandre Bounine <alexandre.bounine@idt.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoInput: ads7846 - fix compiler warning in ads7846_probe()
Dmitry Torokhov [Wed, 2 Jun 2010 07:40:06 +0000 (00:40 -0700)]
Input: ads7846 - fix compiler warning in ads7846_probe()

This patch fixes the follwing warning introduced by commit
067fb2f648543894ce775082c5636f4c32b99e4f ("Input: ads7846 - return error on
regulator_get() failure"):

drivers/input/touchscreen/ads7846.c: In function 'ads7846_probe':
drivers/input/touchscreen/ads7846.c:1167: warning: format '%ld' expects
type 'long int', but argument 4 has type 'int'

Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
13 years agox86, smpboot: Fix cores per node printing on boot
Borislav Petkov [Tue, 1 Jun 2010 19:04:55 +0000 (21:04 +0200)]
x86, smpboot: Fix cores per node printing on boot

Percpu initialization happens now after booting the cores on the
machine and this causes them all to be displayed as belonging to
node 0:

Jun  8 05:57:21 kepek kernel: [    0.106999] Booting Node   0,
Processors  #1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 #13 #14 #15 #16 #17 #18 #19 #20 #21 #22 #23 Ok.

Use early_cpu_to_node() to get the correct node of each core
instead.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Cc: Mike Travis <travis@sgi.com>
LKML-Reference: <20100601190455.GA14237@aftab>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
13 years agoMerge branch 'sh/iomap'
Paul Mundt [Wed, 2 Jun 2010 07:32:12 +0000 (16:32 +0900)]
Merge branch 'sh/iomap'

13 years agosh: PIO disabling for x3proto and urquell.
Paul Mundt [Wed, 26 May 2010 06:58:45 +0000 (15:58 +0900)]
sh: PIO disabling for x3proto and urquell.

urquell only provides PIO in the PCI case, while the x3proto board never
had a working PCIe controller, so it can simply disable it outright.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agosh: mach-sdk7786: conditionally disable PIO support.
Paul Mundt [Tue, 25 May 2010 11:10:24 +0000 (20:10 +0900)]
sh: mach-sdk7786: conditionally disable PIO support.

SDK7786 only supports PIO via the PCI I/O space, so we disable PIO
completely for the non-PCI case.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agosh: support for platforms without PIO.
Paul Mundt [Tue, 25 May 2010 11:06:13 +0000 (20:06 +0900)]
sh: support for platforms without PIO.

This extends some of the existing special casing for HAS_IOPORT
platforms and gets it to the point where platforms can begin to
conditionally select it.

The major changes here are that the PIO routines themselves go away
completely, including all of the machvec port mapping wrappers. With this
in place it's possible for any non-machvec abusing platform to disable
PIO completely. At present this is left as an opt-in until the abusers
are the odd ones out instead of the majority.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agousb: r8a66597-hcd pio to mmio accessor conversion.
Paul Mundt [Wed, 2 Jun 2010 07:27:54 +0000 (16:27 +0900)]
usb: r8a66597-hcd pio to mmio accessor conversion.

r8a66597-hcd is erroneously using PIO routines on MMIO registers, which
presently blows up for any platform that elects to either override or do
away with PIO routines. This managed to work for the common cases since
the PIO routines were simply wrapped to their MMIO counterparts. This
switches over to using the MMIO routines directly, and enables us to kill
off a lot of superfluous casting in the process.

Acked-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agousb: gadget: r8a66597-udc pio to mmio accessor conversion.
Paul Mundt [Wed, 2 Jun 2010 07:27:12 +0000 (16:27 +0900)]
usb: gadget: r8a66597-udc pio to mmio accessor conversion.

r8a66597-udc is erroneously using PIO routines on MMIO registers, which
presently blows up for any platform that elects to either override or do
away with PIO routines. This managed to work for the common cases since
the PIO routines were simply wrapped to their MMIO counterparts. This
switches over to using the MMIO routines directly, and enables us to kill
off a lot of superfluous casting in the process.

Acked-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agousb: gadget: m66592-udc pio to mmio accessor conversion.
Paul Mundt [Wed, 2 Jun 2010 07:26:13 +0000 (16:26 +0900)]
usb: gadget: m66592-udc pio to mmio accessor conversion.

m66592-udc is erroneously using PIO routines on MMIO registers, which
presently blows up for any platform that elects to either override or do
away with PIO routines. This managed to work for the common cases since
the PIO routines were simply wrapped to their MMIO counterparts. This
switches over to using the MMIO routines directly, and enables us to kill
off a lot of superfluous casting in the process.

Acked-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
13 years agoMerge branch 'perf/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/acme...
Ingo Molnar [Wed, 2 Jun 2010 07:13:12 +0000 (09:13 +0200)]
Merge branch 'perf/urgent' of git://git./linux/kernel/git/acme/linux-2.6 into perf/urgent

13 years agoInput: tps6507x-ts - a couple work queue cleanups
Dan Carpenter [Wed, 2 Jun 2010 04:29:40 +0000 (21:29 -0700)]
Input: tps6507x-ts - a couple work queue cleanups

1) Use msecs_to_jiffies() instead of calculating by hand.
2) Call cancel_delayed_work_sync() instead of cancel_delayed_work()
   followed by a separate flush_workqueue().
3) Remove the "tsc->wq = 0;"  Sparse complains about that because
   tsc->wq is a pointer, not an int.  It's not needed because we just
   free the pointer anyway.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
13 years agoMerge branches 'bugzilla-14668' and 'misc-2.6.35' into release
Len Brown [Wed, 2 Jun 2010 02:53:36 +0000 (22:53 -0400)]
Merge branches 'bugzilla-14668' and 'misc-2.6.35' into release

13 years agoMerge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Tue, 1 Jun 2010 21:13:14 +0000 (14:13 -0700)]
Merge branch 'next' of git://git./linux/kernel/git/benh/powerpc

* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  powerpc: Don't export cvt_fd & _df when CONFIG_PPC_FPU is not set
  powerpc/44x: icon: select SM502 and frame buffer console support
  powerpc/85xx: Add P1021MDS board support
  powerpc/85xx: Change MPC8572DS camp dtses for MSI sharing
  powerpc/fsl_msi: add removal path and probe failing path
  powerpc/fsl_msi: enable msi sharing through AMP OSes
  powerpc/fsl_msi: enable msi allocation in all banks
  powerpc/fsl_msi: fix the conflict of virt_msir's chip_data
  powerpc/fsl_msi: Add multiple MSI bank support
  powerpc/kexec: Add support for FSL-BookE
  powerpc/fsl-booke: Move the entry setup code into a seperate file
  powerpc/fsl-booke: fix the case where we are not in the first page
  powerpc/85xx: Enable support for ports 3 and 4 on 8548 CDS
  powerpc/fsl-booke: Add hibernation support for FSL BookE processors
  powerpc/e500mc: Implement machine check handler.
  powerpc/44x: Add basic ICON PPC440SPe board support
  powerpc/44x: Fix UART clocks on 440SPe
  powerpc/44x: Add reset-type to katmai.dts
  powerpc/44x: Adding PCI-E support for PowerPC 460SX based SOC.

13 years agoMerge branch 'drm-intel-next' of git://git.kernel.org/pub/scm/linux/kernel/git/anholt...
Linus Torvalds [Tue, 1 Jun 2010 21:12:27 +0000 (14:12 -0700)]
Merge branch 'drm-intel-next' of git://git./linux/kernel/git/anholt/drm-intel

* 'drm-intel-next' of git://git.kernel.org/pub/scm/linux/kernel/git/anholt/drm-intel: (41 commits)
  drm/i915: add HAS_BSD check to i915_getparam
  drm/i915: Honor sync polarity from VBT panel timing descriptors
  drm/i915: Unmask interrupt for render engine on Sandybridge
  drm/i915: Fix PIPE_CONTROL command on Sandybridge
  drm/i915: Fix up address spaces in slow_kernel_write()
  drm/i915: Use non-atomic kmap for slow copy paths
  drm/i915: Avoid moving from CPU domain during pwrite
  drm/i915: Cleanup after failed initialization of ringbuffers
  drm/i915: Reject bind_to_gtt() early if object > aperture
  drm/i915: Check error code whilst moving buffer to GTT domain.
  drm/i915: Remove spurious warning "Failure to install fence"
  drm/i915: Rebind bo if currently bound with incorrect alignment.
  drm/i915: Include pitch in set_base debug statement.
  drm/i915: Only print "nothing to do" debug message as required.
  drm/i915: Propagate error from unbinding an unfenceable object.
  drm/i915: Avoid nesting of domain changes when setting display plane
  drm/i915: Hold the spinlock whilst resetting unpin_work along error path
  drm/i915: Only print an message if there was an error
  drm/i915: Clean up leftover bits from hws move to ring structure.
  drm/i915: Add CxSR support on Pineview DDR3
  ...

13 years agoFS-Cache: Remove unneeded null checks
Dan Carpenter [Tue, 1 Jun 2010 19:58:22 +0000 (20:58 +0100)]
FS-Cache: Remove unneeded null checks

fscache_write_op() makes unnecessary checks of the page variable to see if it
is NULL.  It can't be NULL at those points as the kernel would already have
crashed a little higher up where we examined page->index.

Furthermore, unless radix_tree_gang_lookup_tag() can return 1 but no page, a
NULL pointer crash should not be encountered there as we can only get there if
r_t_g_l_t() returned 1.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoperf buildid-list: Fix --with-hits event processing
Arnaldo Carvalho de Melo [Tue, 1 Jun 2010 15:37:05 +0000 (12:37 -0300)]
perf buildid-list: Fix --with-hits event processing

When we use plain 'perf buildid-list' we use only what is in the buildid
table in the perf.data header. And those have absolute pathnames because
at 'perf record' time we used __perf_session__process_events and that
doesn't sets up the path shortening code in map__new() that happens if
symbol_conf.full_paths is false, the default.

On the other hand, when we use 'perf buildid-list --with-hits' we
process all the events using perf_session__process_events, adding
entries to the global DSO list _after_ removing the current directory
from the DSO name, for presentation purposes.

Because of that we end up having two entries in the DSO list when
recording events for binaries using relative pathnames.

Fix it minimally by setting symbol_conf.full_paths to true when marking
the DSOs with hits in 'perf buildid-list --with-hits', as used by 'perf
archive'

Right fix longer term is to shorten the path only at presentation time.
Will be done for 2.6.36.

Reported-by: Stephane Eranian <eranian@google.com>
Tested-by: Stephane Eranian <eranian@google.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <20100601183837.GC4093@ghostprotocols.net>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
13 years agomac80211: fix dialog token allocator
Johannes Berg [Mon, 31 May 2010 10:00:12 +0000 (12:00 +0200)]
mac80211: fix dialog token allocator

The dialog token allocator has apparently been broken
since b83f4e15 ("mac80211: fix deadlock in sta->lock")
because it got moved out under the spinlock. Fix it.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: fix blockack-req processing
Johannes Berg [Sun, 30 May 2010 12:52:58 +0000 (14:52 +0200)]
mac80211: fix blockack-req processing

Daniel reported that the paged RX changes had
broken blockack request frame processing due
to using data that wasn't really part of the
skb data.

Fix this using skb_copy_bits() for the needed
data. As a side effect, this adds a check on
processing too short frames, which previously
this code could do.

Reported-by: Daniel Halperin <dhalperi@cs.washington.edu>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Acked-by: Daniel Halperin <dhalperi@cs.washington.edu>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath5k: wake queues on reset
Bruno Randolf [Wed, 19 May 2010 01:30:49 +0000 (10:30 +0900)]
ath5k: wake queues on reset

We can wake all queues after a chip reset since everything should be set up and
we are ready to transmit. If we don't do that we might end up starting up with
stopped queues, not beeing able to transmit. (This started to happen after
"ath5k: clean up queue manipulation" but since periodic calibration also
stopped and started the queues this effect was hidden most of the time).

This way we can also get rid of the superfluous ath5k_reset_wake() function.

Signed-off-by: Bruno Randolf <br1@einfach.org>
Acked-by: Nick Kossifidis <mickflemm@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agodrm/i915: add HAS_BSD check to i915_getparam
Zou Nan hai [Mon, 31 May 2010 05:58:47 +0000 (13:58 +0800)]
drm/i915: add HAS_BSD check to i915_getparam

This will let userland only try to use the new media decode
functionality when the appropriate kernel is present.

Signed-off-by: Zou Nan hai <nanhai.zou@intel.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
13 years agodrm/i915: Honor sync polarity from VBT panel timing descriptors
Adam Jackson [Fri, 28 May 2010 21:17:37 +0000 (17:17 -0400)]
drm/i915: Honor sync polarity from VBT panel timing descriptors

I'm actually kind of shocked that it works at all otherwise.

Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
13 years agocifs: fix page refcount leak
Jeff Layton [Tue, 1 Jun 2010 14:54:45 +0000 (10:54 -0400)]
cifs: fix page refcount leak

Commit 315e995c63a15cb4d4efdbfd70fe2db191917f7a is causing OOM kills
when stress-testing a CIFS filesystem. The VFS readpages operation takes
a page reference. The older code just handed this reference off to the
page cache, but the new code takes an extra one. The simplest fix is to
put the new reference after add_to_page_cache_lru.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Steve French <sfrench@us.ibm.com>
13 years agoDocumentation/timers/hpet_example.c: only build on X86
Randy Dunlap [Tue, 1 Jun 2010 16:55:07 +0000 (09:55 -0700)]
Documentation/timers/hpet_example.c: only build on X86

We should only build hpet_example on x86[-64], where it is implemented.
It can cause build errors on other architectures.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoAFS: Fix possible null pointer dereference in afs_alloc_server()
Denis Kirjanov [Tue, 1 Jun 2010 16:15:39 +0000 (17:15 +0100)]
AFS: Fix possible null pointer dereference in afs_alloc_server()

Fix a possible null pointer dereference in afs_alloc_server(): the server
pointer is NULL if there was an allocation failure, and under such a
condition, we can't dereference it in the _leave() statement.

Signed-off-by: Denis Kirjanov <dkirjanov@kernel.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agofix cpu_chain section mismatch...
Daniel J Blueman [Tue, 1 Jun 2010 11:15:11 +0000 (12:15 +0100)]
fix cpu_chain section mismatch...

In commit e9fb7631ebcd ("cpu-hotplug: introduce cpu_notify(),
__cpu_notify(), cpu_notify_nofail()") the new helper functions access
cpu_chain.  As a result, it shouldn't be marked __cpuinitdata (via
section mismatch warning).

Alternatively, the helper functions should be forced inline, or marked
__ref or __cpuinit.  In the meantime, this patch silences the warning
the trivial way.

Signed-off-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoACPI: update feature-removal.txt to reflect deleted acpi=ht option
Len Brown [Tue, 1 Jun 2010 16:13:23 +0000 (12:13 -0400)]
ACPI: update feature-removal.txt to reflect deleted acpi=ht option

Per plan, acpi=ht was removed in 2.6.35-rc1.

Signed-off-by: Len Brown <len.brown@intel.com>
13 years agoSFI: do not return freed pointer
Dan Carpenter [Fri, 28 May 2010 10:33:15 +0000 (12:33 +0200)]
SFI: do not return freed pointer

We never actually use the return value of sfi_sysfs_install_table() but
it still seems wrong to return a freed pointer.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Len Brown <len.brown@intel.com>
13 years agoMerge branch 'for-35' of git://repo.or.cz/linux-kbuild
Linus Torvalds [Tue, 1 Jun 2010 15:55:52 +0000 (08:55 -0700)]
Merge branch 'for-35' of git://repo.or.cz/linux-kbuild

* 'for-35' of git://repo.or.cz/linux-kbuild: (81 commits)
  kbuild: Revert part of e8d400a to resolve a conflict
  kbuild: Fix checking of scm-identifier variable
  gconfig: add support to show hidden options that have prompts
  menuconfig: add support to show hidden options which have prompts
  gconfig: remove show_debug option
  gconfig: remove dbg_print_ptype() and dbg_print_stype()
  kconfig: fix zconfdump()
  kconfig: some small fixes
  add random binaries to .gitignore
  kbuild: Include gen_initramfs_list.sh and the file list in the .d file
  kconfig: recalc symbol value before showing search results
  .gitignore: ignore *.lzo files
  headerdep: perlcritic warning
  scripts/Makefile.lib: Align the output of LZO
  kbuild: Generate modules.builtin in make modules_install
  Revert "kbuild: specify absolute paths for cscope"
  kbuild: Do not unnecessarily regenerate modules.builtin
  headers_install: use local file handles
  headers_check: fix perl warnings
  export_report: fix perl warnings
  ...

13 years agoMerge branch 'for-linus' of git://www.jni.nu/cris
Linus Torvalds [Tue, 1 Jun 2010 15:51:25 +0000 (08:51 -0700)]
Merge branch 'for-linus' of git://www.jni.nu/cris

* 'for-linus' of git://www.jni.nu/cris:
  CRIS: Don't use mask_irq as symbol name
  CRIS: Simplify param.h by simply including <asm-generic/param.h>
  CRISv10: Whitespace fixes for hw_settings.S
  CRISv10: Trivial fixes.
  CRISv32: Fix RS485 port 4 CD Kconfig item.
  CRISv32: Remove duplicated Kconfig items.
  cris: push down BKL into some device drivers

13 years agointel_scu_ipc: Length fix
Alan Cox [Tue, 1 Jun 2010 11:51:00 +0000 (12:51 +0100)]
intel_scu_ipc: Length fix

Commands with data must set the length in the message.

Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Tue, 1 Jun 2010 15:13:20 +0000 (08:13 -0700)]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (28 commits)
  drivers/isdn/hardware/mISDN: Use GFP_ATOMIC when a lock is held
  ksz884x: Add missing validate_addr hook
  ksz884x: convert to netdev_tx_t
  virtio-net: pass gfp to add_buf
  be2net: convert hdr.timeout in be_cmd_loopback_test() to le32
  can: mpc5xxx_can.c: Fix build failure
  net/ipv4/tcp_input.c: fix compilation breakage when FASTRETRANS_DEBUG > 1
  net: sock_queue_err_skb() dont mess with sk_forward_alloc
  netfilter: xtables: stackptr should be percpu
  netfilter: don't xt_jumpstack_alloc twice in xt_register_table
  greth: Fix build after OF device conversions.
  net: fix sk_forward_alloc corruptions
  Phonet: listening socket lock protects the connected socket list
  caif: unlock on error path in cfserl_receive()
  be2net: remove superfluous externs
  be2net: add unlock on error path
  net/rds: Add missing mutex_unlock
  drivers/isdn/hardware/mISDN: Add missing spin_unlock
  fs_enet: Adjust BDs after tx error
  skb: make skb_recycle_check() return a bool value
  ...

13 years agobinfmt_elf_fdpic: Fix clear_user() error handling
Takuya Yoshikawa [Tue, 1 Jun 2010 13:10:47 +0000 (14:10 +0100)]
binfmt_elf_fdpic: Fix clear_user() error handling

clear_user() returns the number of bytes that could not be copied rather than
an error code.  So we should return -EFAULT rather than directly returning the
results.

Without this patch, positive values may be returned to elf_fdpic_map_file()
and the following error handlings do not function as expected.

1.
ret = elf_fdpic_map_file_constdisp_on_uclinux(params, file, mm);
if (ret < 0)
return ret;
2.
ret = elf_fdpic_map_file_by_direct_mmap(params, file, mm);
if (ret < 0)
return ret;

Signed-off-by: Takuya Yoshikawa <yoshikawa.takuya@oss.ntt.co.jp>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Mike Frysinger <vapier@gentoo.org>
CC: Alexander Viro <viro@zeniv.linux.org.uk>
CC: Andrew Morton <akpm@linux-foundation.org>
CC: Daisuke HATAYAMA <d.hatayama@jp.fujitsu.com>
CC: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoMerge branch 'master' into for-linus
Jens Axboe [Tue, 1 Jun 2010 10:42:12 +0000 (12:42 +0200)]
Merge branch 'master' into for-linus

Conflicts:
fs/pipe.c

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoblock: disable preemption before using sched_clock()
Jens Axboe [Tue, 1 Jun 2010 10:23:18 +0000 (12:23 +0200)]
block: disable preemption before using sched_clock()

Commit 9195291e5f05e01d67f9a09c756b8aca8f009089 added calls to
sched_clock() from preemptible code. sched_clock() is both the
wrong interface AND cannot be called without preempt disabled.

Apply a temporary fix to get rid of the warnings, a real patch
is in the works.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agocciss: call BUG() earlier
Dan Carpenter [Tue, 1 Jun 2010 10:17:48 +0000 (12:17 +0200)]
cciss: call BUG() earlier

I moved the range check after the increment.  The current code would
write past the end of the array once before calling BUG().

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoMerge branch 'amd-iommu/2.6.35' of git://git.kernel.org/pub/scm/linux/kernel/git...
Ingo Molnar [Tue, 1 Jun 2010 09:45:45 +0000 (11:45 +0200)]
Merge branch 'amd-iommu/2.6.35' of git://git./linux/kernel/git/joro/linux-2.6-iommu into x86/urgent

13 years agoperf scripts python: Give field dict to unhandled callback
Pierre Tardy [Mon, 31 May 2010 21:12:09 +0000 (23:12 +0200)]
perf scripts python: Give field dict to unhandled callback

trace_unhandled() callback does not allow to access event fields, this patch
resolves the problem.

It can also been used as a more pythonic and flexible way for script writters
to demux event types

This will for example greatly simplify pytimechart event demux.

Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Tom Zanussi <tzanussi@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>,
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <1275340329-2397-1-git-send-email-tardyp@gmail.com>
Signed-off-by: Pierre Tardy <tardyp@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
13 years agoPreparing 8.3.8rc2
Philipp Reisner [Thu, 27 May 2010 11:46:35 +0000 (13:46 +0200)]
Preparing 8.3.8rc2

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: Reduce verbosity
Philipp Reisner [Wed, 26 May 2010 15:59:55 +0000 (17:59 +0200)]
drbd: Reduce verbosity

The "Local READ/WRITE failed" messages are too verbose.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: use drbd specific ratelimit instead of global printk_ratelimit
Lars Ellenberg [Thu, 27 May 2010 09:51:56 +0000 (11:51 +0200)]
drbd: use drbd specific ratelimit instead of global printk_ratelimit

using the global printk_ratelimit() may mask other messages.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: fix hang on local read errors while disconnected
Lars Ellenberg [Thu, 27 May 2010 07:45:45 +0000 (09:45 +0200)]
drbd: fix hang on local read errors while disconnected

"canceled" w_read_retry_remote never completed, if they have been
canceled after drbd_disconnect connection teardown cleanup has already
run (or we are currently not connected anyways).

Fixed by not queueing a remote retry if we already know it won't work
(pdsk not uptodate), and cleanup ourselves on "cancel", in case we hit a
race with drbd_disconnect.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: Removed the now empty w_io_error() function
Philipp Reisner [Wed, 26 May 2010 15:13:18 +0000 (17:13 +0200)]
drbd: Removed the now empty w_io_error() function

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: removed duplicated #includes
Andrea Gelmini [Sun, 23 May 2010 19:48:13 +0000 (21:48 +0200)]
drbd: removed duplicated #includes

drbd/drbd_receiver.c: linux/mm.h is included more than once.

Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: improve usage of MSG_MORE
Lars Ellenberg [Tue, 25 May 2010 14:26:16 +0000 (16:26 +0200)]
drbd: improve usage of MSG_MORE

It seems to improve performance if we allow the "p_data" header in its
own frame (no MSG_MORE), but sendpage all but the last page with MSG_MORE.
This is also in preparation of a later zero copy receive implementation.

Suggested by Eduard.Guzovsky@stratus.com on drbd-dev.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: need to set socket bufsize early to take effect
Lars Ellenberg [Tue, 25 May 2010 14:18:01 +0000 (16:18 +0200)]
drbd: need to set socket bufsize early to take effect

quoting tcp(7):
    On individual connections, the socket buffer size must be set prior to the
    listen(2) or connect(2) calls in order to have it take effect.

This adds a wrapper to do so, and uses it appropriately.
Improves performance in certain situations.

Note that because we cannot easily determine which socket will be
"meta" and wich "data" (bulk) socket, we adjust both sockets.
Previously, DRBD only adjusted the bufsizes of the "data" socket.

Thanks again to Eduard.Guzovsky@stratus.com.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: improve network latency, TCP_QUICKACK
Lars Ellenberg [Tue, 25 May 2010 12:23:57 +0000 (14:23 +0200)]
drbd: improve network latency, TCP_QUICKACK

On Thu, Apr 29, 2010 at 04:00:50PM -0400, Eduard.Guzovsky@stratus.com
 wrote on drbd-dev@lists.linbit.com
 Subject: [Drbd-dev] DRBD small synchronous writes performance improvements

> 1. TCP_QUICKACK option is set incorrectly. The goal was force TCP to
> send and ACK as a  "one time" event.  Instead the code permanently sets
> connection in the QUICKACK mode.

He is right, we actually want to use an even val with TCP_QUICKACK.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agodrbd: Revert "drbd: Create new current UUID as late as possible"
Philipp Reisner [Tue, 25 May 2010 12:32:03 +0000 (14:32 +0200)]
drbd: Revert "drbd: Create new current UUID as late as possible"

The late-UUID writing is delayed until the next release.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agobrd: support discard
Nick Piggin [Wed, 26 May 2010 13:41:14 +0000 (15:41 +0200)]
brd: support discard

Support discard requests in brd by zeroing or deleting the underlying backing
pages. This is simply to help with testing and documentation nature of
brd code.

Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
13 years agoRevert "writeback: fix WB_SYNC_NONE writeback from umount"
Jens Axboe [Tue, 1 Jun 2010 09:08:43 +0000 (11:08 +0200)]
Revert "writeback: fix WB_SYNC_NONE writeback from umount"

This reverts commit e913fc825dc685a444cb4c1d0f9d32f372f59861.

We are investigating a hang associated with the WB_SYNC_NONE changes,
so revert them for now.

Conflicts:

fs/fs-writeback.c
mm/page-writeback.c

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoRevert "writeback: ensure that WB_SYNC_NONE writeback with sb pinned is sync"
Jens Axboe [Tue, 1 Jun 2010 09:05:22 +0000 (11:05 +0200)]
Revert "writeback: ensure that WB_SYNC_NONE writeback with sb pinned is sync"

This reverts commit 7c8a3554c683f512dbcee26faedb42e4c05f12fa.

We are investigating a hang associated with the WB_SYNC_NONE changes,
so revert them for now.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoperf hist: fix objdump output parsing
Konstantin Stepanyuk [Wed, 26 May 2010 22:10:11 +0000 (02:10 +0400)]
perf hist: fix objdump output parsing

hist_entry__annotate() runs objdump with -S option so the output may contain
lines of any format. If a line starts with a colon strtoull() returns 0 and
calculated offset will be negative. This causes perf annotate segfaults.

Make sure that strtoull() has parsed at least one digit.

Cc: David S. Miller <davem@davemloft.net>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Konstantin Stepanyuk <konstantin.stepanyuk@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
13 years agox86/amd-iommu: Fall back to GART if initialization fails
Joerg Roedel [Mon, 31 May 2010 13:05:20 +0000 (15:05 +0200)]
x86/amd-iommu: Fall back to GART if initialization fails

This patch implements a fallback to the GART IOMMU if this
is possible and the AMD IOMMU initialization failed.
Otherwise the fallback would be nommu which is very
problematic on machines with more than 4GB of memory or
swiotlb which hurts io-performance.

Cc: stable@kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
13 years agox86/amd-iommu: Fix crash when request_mem_region fails
Joerg Roedel [Fri, 28 May 2010 12:26:48 +0000 (14:26 +0200)]
x86/amd-iommu: Fix crash when request_mem_region fails

When request_mem_region fails the error path tries to
disable the IOMMUs. This accesses the mmio-region which was
not allocated leading to a kernel crash. This patch fixes
the issue.

Cc: stable@kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
13 years agoMerge commit 'v2.6.35-rc1' into amd-iommu/2.6.35
Joerg Roedel [Tue, 1 Jun 2010 07:57:49 +0000 (09:57 +0200)]
Merge commit 'v2.6.35-rc1' into amd-iommu/2.6.35

13 years agosched, trace: Fix sched_switch() prev_state argument
Peter Zijlstra [Mon, 31 May 2010 16:13:25 +0000 (18:13 +0200)]
sched, trace: Fix sched_switch() prev_state argument

For CONFIG_PREEMPT=y kernels the sched_switch(.prev_state) argument isn't
useful because we can get preempted with current->state != TASK_RUNNING
without actually getting removed from the runqueue.

Cure this by treating all preempted tasks as runnable from the tracer's
point of view.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cautiously-acked-by: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <1275322715.27810.23323.camel@twins>
Signed-off-by: Ingo Molnar <mingo@elte.hu>