Commit graph

287,753 commits

Author SHA1 Message Date
Linus Torvalds
1f0772fe27 Merge git://github.com/davem330/net
* git://github.com/davem330/net:
  tcp: fix validation of D-SACK
  tcp: fix build error if !CONFIG_SYN_COOKIES
2011-09-19 07:48:04 -07:00
Pekka Enberg
d20bbfab01 Merge branch 'slab/urgent' into slab/next 2011-09-19 17:46:07 +03:00
J. Bruce Fields
f7a4d87207 nfsd4: hash closed stateid's like any other
Look up closed stateid's in the stateid hash like any other stateid
rather than searching the close lru.

This is simpler, and fixes a bug: currently we handle only the case of a
close that is the last close for a given stateowner, but not the case of
a close for a stateowner that still has active opens on other files.
Thus in a case like:

	open(owner, file1)
	open(owner, file2)
	close(owner, file2)
	close(owner, file2)

the final close won't be recognized as a retransmission.

Signed-off-by: J. Bruce Fields <bfields@redhat.com>
2011-09-19 08:39:34 -04:00
J. Bruce Fields
d3b313a463 nfsd4: construct stateid from clientid and counter
Including the full clientid in the on-the-wire stateid allows more
reliable detection of bad vs. expired stateid's, simplifies code, and
ensures we won't reuse the opaque part of the stateid (as we currently
do when the same openowner closes and reopens the same file).

Signed-off-by: J. Bruce Fields <bfields@redhat.com>
2011-09-19 06:33:57 -04:00
Mark Brown
53daf20893 ASoC: Display the error code when we fail to add a DAPM control
Useful for diagnostics.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>
2011-09-19 11:27:22 +01:00
Fabio Estevam
d093d5bbdb ARM: mx31pdk: Fix build by passing IMX_HAVE_PLATFORM_MXC_MMC
mx31pdk target fails to build:

  CC      init/version.o
  LD      init/built-in.o
  LD      .tmp_vmlinux1
arch/arm/mach-imx/built-in.o: In function `mx31_3ds_init':
mach-mx31_3ds.c:(.init.text+0x548): undefined reference to `imx_add_mxc_mmc'
mach-mx31_3ds.c:(.init.text+0x69c): undefined reference to `imx31_mxc_mmc_data'
make: *** [.tmp_vmlinux1] Error

Fix this by passsing IMX_HAVE_PLATFORM_MXC_MMC option.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
2011-09-19 12:27:13 +02:00
Viresh Kumar
c12056466d dmaengine/amba-pl08x: Check txd->llis_va before freeing dma_pool
In pl08x_free_txd(), check if pool is allocated successfully before freeing it.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
2011-09-19 15:13:16 +05:30
Viresh Kumar
b7f69d9d42 dmaengine/amba-pl08x: Add support for sg len greater than one for slave transfers
Untill now, sg_len greater than one is not supported. This patch adds support to
do that.

Note: Still, if peripheral is flow controller, sg_len can't be greater that one.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
2011-09-19 15:13:06 +05:30
Takashi Iwai
8974bd51a7 ALSA: hda/realtek - Fix auto-mute with HP+LO configuration
When the system has only the headphone and the line-out jacks without
speakers, the current auto-mute code doesn't work.  It's because the
spec->automute_lines flag is wrongly referred in update_speakers().
This flag must be meaningless when spec->automute_hp_lo isn't set, thus
they should be always coupled.

The patch fixes the problem and add a comment to indicate the
relationship briefly.

BugLink: http://bugs.launchpad.net/bugs/851697

Reported-by: David Henningsson <david.henningsson@canonical.com>
Tested-By: Jayne Han <jayne.han@canonical.com>
Cc: stable@kernel.org (3.0)
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2011-09-19 11:31:34 +02:00
Tomasz Figa
fb5d375d35 ARM: S3C64XX: Add support for synchronous clock operation
Some boards based on S3C6410 use synchronous clocking, which means that HCLKx2
and other system clocks are generated from APLL instead of MPLL.

This patch adds support for such boards, by calculating hclk2 depending on
the status of S3C_OTHERS_SYNCMUXSEL bit in S3C64XX_OTHERS regist

Signed-off-by: Tomasz Figa <tomasz.figa@gmail.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:38:11 +09:00
Tomasz Figa
d9018df007 ARM: S3C64XX: Save/restore S3C64XX_MODEM_MIFPCON on suspend/resume
This patch makes sure that the LCD bypass state is preserved during sleep mode.
It achieves this by saving the S3C64XX_MODEM_MIFPCON register, which is not
handled by any other code currently.

Signed-off-by: Tomasz Figa <tomasz.figa@gmail.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:38:05 +09:00
Tomasz Figa
348276f4a3 ARM: S3C64XX: Save/restore S3C64XX_SDMA_SEL on suspend/resume
This patch makes sure that S3C64XX_SDMA_SEL register is preserved during
sleep mode, as it is critical for DMA operation and the DMA driver alone
does not provide any power management facilities.

Signed-off-by: Tomasz Figa <tomasz.figa@gmail.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:38:00 +09:00
Tomasz Figa
6d8d492ef7 ARM: S3C64XX: Use S3C64XX_SDMA_SEL register name instead of numeric address
This patch adds a definition of S3C64XX_SDMA_SEL register and modifies
the s3c64xx DMA driver to use it instead of a magic register address.

Signed-off-by: Tomasz Figa <tomasz.figa@gmail.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:37:55 +09:00
Mark Brown
8c051ab47b ARM: S3C64XX: Register Lowland audio device on Cragganmore
The Lowland system is a combination of the Cragganmore CPU module and the
Tomatin and Kilchomin audio modules on Glenfarclas.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:33:04 +09:00
Mark Brown
3aa7779c74 ARM: S3C64XX: Add devices for Kilchoman audio module on Cragganmore
The Kilchoman audio module carries a WM5100 audio CODEC and WM9081 speaker
amplifier.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:33:03 +09:00
Mark Brown
4ed12b5070 ARM: S3C64XX: Hook up some additional supplies on Cragganmore
For further modules.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:33:02 +09:00
Mark Brown
d0f0b43f78 ARM: S3C64XX: Use module identification for Cragganmore system builds
The Cragganmore system is modular with I2C based identification chips on
the system allowing identification of the system build. Provide a stub I2C
driver which parses the module IDs and uses them to select the appropriate
audio subsystem components to register. To avoid confusion due to having
the mini-driver in the system the driver is placed in a separate file.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 16:32:36 +09:00
Vivek Natarajan
9df337a104 ath6kl: deinitialise wiphy on error
This fixes the following panic observed on card removal.

BUG: unable to handle kernel paging request at f86e22ac
EIP is at wiphy_update_regulatory+0x252/0x590 [cfg80211]
Call Trace:
 [<f92fb9e5>] set_regdom+0x165/0x600 [cfg80211]
 [<c02057ba>] ? __kmalloc+0x10a/0x190
 [<c0366557>] ? nla_parse+0xb7/0xd0
 [<f9300b92>] ? T.1400+0x12/0x20 [cfg80211]
 [<f9300c84>] nl80211_set_reg+0xe4/0x270 [cfg80211]
 [<f92fe560>] ? nl80211_pre_doit+0x0/0x160 [cfg80211]
 [<c050bdeb>] genl_rcv_msg+0x23b/0x280
 [<c050bbb0>] ? genl_rcv_msg+0x0/0x280
 [<c050ab76>] netlink_rcv_skb+0x86/0xb0
 [<c050bb80>] ? genl_rcv+0x0/0x30
 [<c050bb9c>] genl_rcv+0x1c/0x30

Signed-off-by: Vivek Natarajan <nataraja@qca.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
2011-09-19 10:01:55 +03:00
Jouni Malinen
151411e88f ath6kl: Fix static WEP configuration in AP mode
Configuration of the WEP keys needs to be delayed until the AP mode has
been properly started at the target. Partial support for delaying the
WEP key configuration was already in place in the driver, but the actual
part of deciding when to do this was missing.

Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
2011-09-19 09:57:22 +03:00
Guennadi Liakhovetski
937bb6e4c6 serial: sh-sci: don't filter on DMA device, use only channel ID
On some sh-mobile systems there are more than one DMA controllers, that
can be used for serial ports. Specifying a DMA device in sh-sci platform
data unnecessarily restricts the driver to only use one DMA controller.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
[Fixed the trivial conflict in include/linux/serial_sci.h]
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
2011-09-19 08:39:17 +05:30
Zheng Yan
f779b2d60a tcp: fix validation of D-SACK
D-SACK is allowed to reside below snd_una. But the corresponding check
in tcp_is_sackblock_valid() is the exact opposite. It looks like a typo.

Signed-off-by: Zheng Yan <zheng.z.yan@intel.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2011-09-18 22:37:34 -04:00
Eric Dumazet
e05c82d366 tcp: fix build error if !CONFIG_SYN_COOKIES
commit 946cedccbd (tcp: Change possible SYN flooding messages)
added a build error if CONFIG_SYN_COOKIES=n

Reported-by: Markus Trippelsdorf <markus@trippelsdorf.de>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2011-09-18 21:48:01 -04:00
Kukjin Kim
c9312209aa Merge branch 'next/topic-dma-pl330-devname-fix' into next/topic-dma-samsung
Conflicts:
	arch/arm/mach-exynos4/clock.c
	arch/arm/mach-s5p64x0/clock-s5p6440.c
	arch/arm/mach-s5p64x0/clock-s5p6450.c
	arch/arm/mach-s5pc100/clock.c
	arch/arm/mach-s5pv210/clock.c
2011-09-19 10:30:49 +09:00
Alim Akhtar
d5663e7a40 ARM: S5PC100: register the second instance of pl330 DMAC
Platform data is provided for two instance of pl330 DMAC,
but only one DMAC is register with amba_device.
This patch register the second instance.

Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 10:27:26 +09:00
Alim Akhtar
f86147cc8f ARM: S5PV210: register the second instance of pl330 DMAC
Platform data is provided for two instance of pl330 DMAC,
but only one DMAC is register with amba_device.
This patch register the second instance.

Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 10:27:23 +09:00
Alim Akhtar
fad23ddac3 ARM: EXYNOS4: register the second instance of pl330 DMAC
Platform data is provided for two instance of pl330 DMAC,
but only one DMAC is register with amba_device.
This patch register the second instance.

Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
2011-09-19 10:27:19 +09:00
Kukjin Kim
219c8c866c Merge branch 'samsung_dma' of git://git.infradead.org/users/vkoul/slave-dma into next/topic-dma-samsung 2011-09-19 10:26:54 +09:00
Linus Torvalds
6bf3b0dc32 Merge branch 'for-linus' of git://git.infradead.org/users/sameo/mfd-2.6
* 'for-linus' of git://git.infradead.org/users/sameo/mfd-2.6:
  mfd: Fix omap-usb-host build failure
  mfd: Make omap-usb-host TLL mode work again
  mfd: Set MAX8997 irq pointer
  mfd: Fix initialisation of tps65910 interrupts
  mfd: Check for twl4030-madc NULL pointer
  mfd: Copy the device pointer to the twl4030-madc structure
  mfd: Rename wm8350 static gpio_set_debounce()
  mfd: Fix value of WM8994_CONFIGURE_GPIO
2011-09-18 18:18:55 -07:00
Tetsuo Handa
6bce98edc3 TOMOYO: Allow specifying domain transition preference.
I got an opinion that it is difficult to use exception policy's domain
transition control directives because they need to match the pathname specified
to "file execute" directives. For example, if "file execute /bin/\*\-ls\-cat"
is given, corresponding domain transition control directive needs to be like
"no_keep_domain /bin/\*\-ls\-cat from any".

If we can specify like below, it will become more convenient.

  file execute /bin/ls keep exec.realpath="/bin/ls" exec.argv[0]="ls"
  file execute /bin/cat keep exec.realpath="/bin/cat" exec.argv[0]="cat"
  file execute /bin/\*\-ls\-cat child
  file execute /usr/sbin/httpd <apache> exec.realpath="/usr/sbin/httpd" exec.argv[0]="/usr/sbin/httpd"

In above examples, "keep" works as if keep_domain is specified, "child" works
as if "no_reset_domain" and "no_initialize_domain" and "no_keep_domain" are
specified, "<apache>" causes domain transition to <apache> domain upon
successful execve() operation.

Moreover, we can also allow transition to different domains based on conditions
like below example.

  <kernel> /usr/sbin/sshd
  file execute /bin/bash <kernel> /usr/sbin/sshd //batch-session exec.argc=2 exec.argv[1]="-c"
  file execute /bin/bash <kernel> /usr/sbin/sshd //root-session task.uid=0
  file execute /bin/bash <kernel> /usr/sbin/sshd //nonroot-session task.uid!=0

Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: James Morris <jmorris@namei.org>
2011-09-19 10:09:59 +10:00
Timur Tabi
2b7a905dd0 drivers/video: fsl-diu-fb: remove unusued MEM_ALLOC_THRESHOLD
If there was ever any code that used MEM_ALLOC_THRESHOLD, it was removed a
long time ago.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:58 +00:00
Timur Tabi
251b9b0d40 drivers/video: fsl-diu-fb: the video buffer is not I/O memory
The video buffer is not uncached memory-mapped I/O, so don't tag the virtual
address as __iomem.  It's also not a u8*.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:58 +00:00
Timur Tabi
3f78bbd1a2 drivers/video: fsl-diu-fb: use a normal for-loop to uninstall framebuffers
Uninstalling the framebuffers in reverse order is unnecessary and makes
the for-loop awkward.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:57 +00:00
Timur Tabi
589c797193 drivers/video: fsl-diu-fb: fix memory leak on error
We were forgetting to unmap the video memory if fsl_diu_check_var() fails.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:57 +00:00
Timur Tabi
ec02dd23dc drivers/video: fsl-diu-fb: fix potential memcpy buffer overflow bug
It makes no sense to limit the size of a strncpy() to the length of
the source string.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:57 +00:00
Timur Tabi
f8c6bf6ae6 drivers/video: fsl-diu-fb: set the driver name to "fsl-diu-fb"
Use the name "fsl-diu-fb" in the Freescale DIU framebuffer driver
during registrations.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:57 +00:00
Timur Tabi
bada04fcda drivers/video: fsl-diu-fb: improve local variable usage in some functions
Clean up the local variable usage in request_irq_local() and allocate_buf().
This streamlines the code without affecting functionality.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:57 +00:00
Timur Tabi
63cf8df44e drivers/video: fsl-diu-fb: remove redundant default video mode
The framebuffer layer already uses the first video mode defined in the
fb_videomode array as a default, so there's no need to duplicate the
first entry into a stand-alone structure.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:57 +00:00
Timur Tabi
9e52ba612f drivers/video: fsl-diu-fb: improve device tree usage
Implement various improvements to the way the Freescale DIU framebuffer
driver access the device tree.

1) Use of_iomap() instead of of_address_to_resource() and ioremap()

2) Use be32_to_cpup() instead of directly dereferencing the device_node
   pointer.

3) Rename variable 'ofdev' to 'pdev' to avoid any confusion that it's
   a platform_device pointer, not an of_device pointer (of_device no
   longer exists).

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:56 +00:00
Timur Tabi
89f08e3e5d drivers/video: fsl-diu-fb: fix compilation warning
Fix this compilation warning in the Freescale DIU framebuffer driver:

warning: 'dummy_ad_addr' may be used uninitialized in this function

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:56 +00:00
Timur Tabi
4a64e49df2 drivers/video: fsl-diu-fb: remove unused ioctls
Remove some unused ioctl commands, and treat those commands as unsupported
instead of ignored.

Also remove struct mfb_alpha, which isn't used by any ioctl.  It may have
been once intended for MFB_SET_ALPHA, but that ioctl uses a different
data structure.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:56 +00:00
Timur Tabi
154152aeea drivers/video: fsl-diu-fb: clean up printk usage
Remove debug printk messages (they don't help in debugging), replace
printk(KERN_xxx with its pr_xxx or dev_xxx equivalent, and add a couple
missing error messages.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:56 +00:00
Timur Tabi
4a85dc8b7d drivers/video: fsl-diu-fb: clean up whitespace and formatting
Fix various indentation and line length problems in the Freescale
DIU framebuffer driver.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
2011-09-18 20:08:56 +00:00
Alex Deucher
003cefe0c2 drm/radeon/kms: Make GPU/CPU page size handling consistent in blit code (v2)
The BO blit code inconsistenly handled the page size.  This wasn't
an issue on system with 4k pages since the GPU's page size is 4k as
well.  Switch the driver blit callbacks to take num pages in GPU
page units.

Fixes lemote mipsel systems using AMD rs780/rs880 chipsets.

v2: incorporate suggestions from Michel.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-09-18 19:44:36 +01:00
Alex Deucher
18b4fada27 drm/radeon/kms: fix typo in r100_blit_copy
cur_pages is the number of pages per loop iteration.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-09-18 19:43:42 +01:00
Linus Torvalds
b0e7031ac0 Merge git://github.com/davem330/net
* git://github.com/davem330/net: (62 commits)
  ipv6: don't use inetpeer to store metrics for routes.
  can: ti_hecc: include linux/io.h
  IRDA: Fix global type conflicts in net/irda/irsysctl.c v2
  net: Handle different key sizes between address families in flow cache
  net: Align AF-specific flowi structs to long
  ipv4: Fix fib_info->fib_metrics leak
  caif: fix a potential NULL dereference
  sctp: deal with multiple COOKIE_ECHO chunks
  ibmveth: Fix checksum offload failure handling
  ibmveth: Checksum offload is always disabled
  ibmveth: Fix issue with DMA mapping failure
  ibmveth: Fix DMA unmap error
  pch_gbe: support ML7831 IOH
  pch_gbe: added the process of FIFO over run error
  pch_gbe: fixed the issue which receives an unnecessary packet.
  sfc: Use 64-bit writes for TX push where possible
  Revert "sfc: Use write-combining to reduce TX latency" and follow-ups
  bnx2x: Fix ethtool advertisement
  bnx2x: Fix 578xx link LED
  bnx2x: Fix XMAC loopback test
  ...
2011-09-18 11:02:26 -07:00
Josef Bacik
a66e7cc626 Btrfs: only clear the need lookup flag after the dentry is setup
We can race with readdir and the RCU path walking stuff.  This is because we
clear the need lookup flag before actually instantiating the inode.  This will
lead the RCU path walk stuff to find a dentry it thinks is valid without a
d_inode attached.  So instead unhash the dentry when we first start the lookup,
and then clear the flag after we've instantiated the dentry so we're garunteed
to either try the slow lookup, or have the d_inode set properly.

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2011-09-18 10:34:03 -04:00
Jeff Liu
48802c8ae2 BTRFS: Fix lseek return value for error
The recent reworking of btrfs' lseek lead to incorrect
values being returned.  This adds checks for seeking
beyond EOF in SEEK_HOLE and makes sure the error
values come back correct.

Andi Kleen also sent in similar patches.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Reported-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2011-09-18 10:34:02 -04:00
Chris Mason
2cf4ce7c2a Merge branch 'btrfs-3.0' into for-linus 2011-09-18 10:31:44 -04:00
Li Zefan
dde820fbf7 Btrfs: don't change inode flag of the dest clone file
The dst file will have the same inode flags with dst file after
file clone, and I think it's unexpected.

For example, the dst file will suddenly become immutable after
getting some share of data with src file, if the src is immutable.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2011-09-18 10:20:46 -04:00
Li Zefan
0e7b824c4e Btrfs: don't make a file partly checksummed through file clone
To reproduce the bug:

  # mount /dev/sda7 /mnt
  # dd if=/dev/zero of=/mnt/src bs=4K count=1
  # umount /mnt

  # mount -o nodatasum /dev/sda7 /mnt
  # dd if=/dev/zero of=/mnt/dst bs=4K count=1
  # clone_range -s 4K -l 4K /mnt/src /mnt/dst

  # echo 3 > /proc/sys/vm/drop_caches
  # cat /mnt/dst
  # dmesg
  ...
  btrfs no csum found for inode 258 start 0
  btrfs csum failed ino 258 off 0 csum 2566472073 private 0

It's because part of the file is checksummed and the other part is not,
and then btrfs will complain checksum is not found when we read the file.

Disallow file clone if src and dst file have different checksum flag,
so we ensure a file is completely checksummed or unchecksummed.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2011-09-18 10:20:46 -04:00