deliverable/linux.git
8 years agolibata-scsi: avoid repeated calculation of number of TRIM ranges
Tom Yan [Tue, 12 Jul 2016 20:31:23 +0000 (04:31 +0800)] 
libata-scsi: avoid repeated calculation of number of TRIM ranges

Currently libata statically allows only 1-block (512-byte) payload
for each TRIM command. Each payload can carry 64 TRIM ranges since
each range requires 8 bytes.

It is silly to keep doing the calculation (512 / 8) in different
places. Hence, define the new ATA_MAX_TRIM_RNUM for the result.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: reject WRITE SAME (16) with n_block that exceeds limit
Tom Yan [Tue, 12 Jul 2016 20:31:22 +0000 (04:31 +0800)] 
libata-scsi: reject WRITE SAME (16) with n_block that exceeds limit

Currently if a WRITE SAME (16) command is issued to the SATL with
"number of blocks" that is larger than the "Maximum write same length"
(which is the maximum number of blocks per TRIM command allowed in
libata, currently 65535 * 512 / 8 blocks), the SATL will accept the
command and translate it to a TRIM command with the upper limit.

However, according to SBC (as of sbc4r11.pdf), the "device server"
should terminate the command with "Invalid field in CDB" in that case.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: rename ata_msense_ctl_mode() to ata_msense_control()
Tom Yan [Tue, 12 Jul 2016 18:54:12 +0000 (02:54 +0800)] 
libata-scsi: rename ata_msense_ctl_mode() to ata_msense_control()

To make it consistent with the recently added ata_mselect_control().
We probably shouldn't have the word "mode" in its name anyway, since
that's not the case for other ata_msense_*() / ata_mselect_*() either.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: fix D_SENSE bit relection in control mode page
Tom Yan [Tue, 12 Jul 2016 18:54:11 +0000 (02:54 +0800)] 
libata-scsi: fix D_SENSE bit relection in control mode page

The bit should always be set to 1 when the requested version of
page is "changeable" because we've made it so in ata_mselect_control().
Also, it should always be set to 1 if ATA_DFLAG_D_SENSE is set (when
the requested version of page is "current" or "default").

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoata: hpt366: fix incorrect mask when checking at cmd_high_time
Colin Ian King [Tue, 12 Jul 2016 11:16:19 +0000 (12:16 +0100)] 
ata: hpt366: fix incorrect mask when checking at cmd_high_time

According to the HPT366 data sheet, PCI config space dword 0x40-0x43
bits 11:8 specify the primary drive cmd_high_time, however,
currently just 3 bits of the 4 are being used because the mask
is 0x700 and not 0x0f00.  Fix the mask, allowing for the 40MHz clock
to be detected.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: correct SPC version descriptor
Tom Yan [Tue, 12 Jul 2016 13:29:35 +0000 (21:29 +0800)] 
libata-scsi: correct SPC version descriptor

The comment suggests we should be having an SPC-3 version descriptor
but the 0260h is the code for "SPC-2 (no version claimed)". Correct
it to 0300h so that it has the "SPC-3 (no version claimed)" descriptor.

Note that we are claiming SPC-3 version compatibility in the VERSION
field of the standard INQUIRY data. Therefore, I assume the typo was
on the code but not on the comment.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: set CmdQue bit in standard INQUIRY data to 1
Tom Yan [Tue, 12 Jul 2016 13:29:34 +0000 (21:29 +0800)] 
libata-scsi: set CmdQue bit in standard INQUIRY data to 1

Avoid performance bottleneck when being SCSI pass-through'd to
virtual machines with other OSes (e.g. Windows) via virtio-scsi
and scsi-block in qemu.

Ref.: https://github.com/YanVugenfirer/kvm-guest-drivers-windows/issues/63

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: fix SET FEATURES "filtering" for ata_msense_caching()
Tom Yan [Tue, 12 Jul 2016 13:37:02 +0000 (21:37 +0800)] 
libata-scsi: fix SET FEATURES "filtering" for ata_msense_caching()

Without this fix, the DRA bit of the caching mode page would not
be updated when the read look-ahead feature is toggled (e.g. with
`smartctl --set`), but will only be until, for example, the write
cache feature is touched.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: set correct VERSION field for ZAC devices
Tom Yan [Tue, 12 Jul 2016 14:12:01 +0000 (22:12 +0800)] 
libata-scsi: set correct VERSION field for ZAC devices

Commit 856c46639309 ("libata: support device-managed ZAC devices")
had the line that "bumps" the VERSION field in standard INQUIRY data
removed. Add it back and claim SPC-5 version compatibility, which
matches with the current version descriptor "SPC-5 (no version claimed)"
that is used for ZAC devices.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: correct cbd to CDB in comment
Tom Yan [Wed, 6 Jul 2016 17:13:09 +0000 (01:13 +0800)] 
libata-scsi: correct cbd to CDB in comment

It's Command Descriptor Block. Also capitalized it.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agolibata-scsi: do not respond with "invalid field" for FORMAT UNIT
Tom Yan [Wed, 6 Jul 2016 17:13:08 +0000 (01:13 +0800)] 
libata-scsi: do not respond with "invalid field" for FORMAT UNIT

It does not make sense and is confusing to respond with "Invalid
field in CDB" while we have no support at all implemented for
FORMAT UNIT. It is decent to let it go to the default, which
will respond with "Invalid command operation code" instead.

Signed-off-by: Tom Yan <tom.ty89@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agodrivers: ata: pata_arasan_cf: use the same name for the same lock
Iago Abal [Mon, 27 Jun 2016 07:51:42 +0000 (09:51 +0200)] 
drivers: ata: pata_arasan_cf: use the same name for the same lock

Note that `&acdev->host->lock' and `qc->ap->lock' denote the same lock, and it's
particularly confusing to spin_lock on the former but spin_unlock on the latter.

Signed-off-by: Iago Abal <mail@iagoabal.eu>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoata: fix "ering" sysfs time printing
Arnd Bergmann [Fri, 17 Jun 2016 15:37:12 +0000 (17:37 +0200)] 
ata: fix "ering" sysfs time printing

The sysfs file for the libata error handling has multiple issues
in the way it prints time stamps:

 * it prints a 9-digit nanosecond value using a %06lu format string,
   which drops some leading zeroes
 * it converts a 64-bit jiffes value to a timespec using
   jiffies_to_timespec(), which takes a 'long' argument, so the
   result is wrong after a jiffies overflow (49 days).
 * we try to avoid using timespec because that generally overflows
   in 2038, although this particular usage is ok.

This replaces the jiffies_to_timespec call with an open-coded
implementation that gets it right.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoARM: dts: nsp: Add sata device tree entry
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:35 +0000 (09:53 -0400)] 
ARM: dts: nsp: Add sata device tree entry

Add sata support to the Northstar Plus SoC device tree.

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agophy: Add SATA3 PHY support for Broadcom NSP SoC
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:34 +0000 (09:53 -0400)] 
phy: Add SATA3 PHY support for Broadcom NSP SoC

This patch adds support for Broadcom NSP SATA3 PHY in existing
Broadcom SATA PHY driver.

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoata: ahci_brcm: Add support for Broadcom NSP SoC
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:33 +0000 (09:53 -0400)] 
ata: ahci_brcm: Add support for Broadcom NSP SoC

Add SATA3 support for Broadcom NSP SoC

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoata: ahci_brcmstb: rename to support across Broadcom SoC's
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:32 +0000 (09:53 -0400)] 
ata: ahci_brcmstb: rename to support across Broadcom SoC's

Rename the existing Broadcom STB ahci driver to common
Broadcom SATA3 driver to share this across Broadcom SoCs.

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agodt-bindings: phy: Add documentation for NSP SATA PHY
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:31 +0000 (09:53 -0400)] 
dt-bindings: phy: Add documentation for NSP SATA PHY

Add documentation for SATA PHY available in Broadcom's
NSP SoC.

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agodt-bindings: ata: add support for Broadcom NSP SoC ahci
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:30 +0000 (09:53 -0400)] 
dt-bindings: ata: add support for Broadcom NSP SoC ahci

Add support for Broadcom NSP SoC ahci

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agodt-bindings: ata: rename brcm, sata-brcmstb.txt to brcm, sata-brcm.txt
Yendapally Reddy Dhananjaya Reddy [Thu, 16 Jun 2016 13:53:29 +0000 (09:53 -0400)] 
dt-bindings: ata: rename brcm, sata-brcmstb.txt to brcm, sata-brcm.txt

Renames the dt bindings document of Broadcom STB ahci driver to
common Broadcom ahci driver bindings document

Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoata: Use IS_ENABLED() instead of checking for built-in or module
Javier Martinez Canillas [Wed, 18 May 2016 20:11:28 +0000 (16:11 -0400)] 
ata: Use IS_ENABLED() instead of checking for built-in or module

The IS_ENABLED() macro checks if a Kconfig symbol has been enabled either
built-in or as a module, use that macro instead of open coding the same.

Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoMAINTAINERS: Add file patterns for ata device tree bindings
Geert Uytterhoeven [Sun, 22 May 2016 09:06:01 +0000 (11:06 +0200)] 
MAINTAINERS: Add file patterns for ata device tree bindings

Submitters of device tree binding documentation may forget to CC
the subsystem maintainer if this is missing.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: linux-ide@vger.kernel.org
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agopata_atiixp: fix trivial indentation misalignment on if statement
Colin Ian King [Mon, 16 May 2016 11:35:03 +0000 (12:35 +0100)] 
pata_atiixp: fix trivial indentation misalignment on if statement

Remove extraneous space on if statement and on the following line,
trivial fix, no functional change

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
8 years agoMerge branch 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild
Linus Torvalds [Fri, 27 May 2016 05:32:05 +0000 (22:32 -0700)] 
Merge branch 'misc' of git://git./linux/kernel/git/mmarek/kbuild

Pull misc kbuild updates from Michal Marek:
 "This is the non-critical part of kbuild:

   - Coccinelle fixes, one semantic patch less in this round [Vaishali
     Thakkar, Wolfram Sang, Kees Cook]

   - rpm-pkg support for (open)SUSE's update-bootloader [Jiří Kosian]

   - rpm-pkg restored support for $RPMOPTS [Srinivas Pandruvada]

   - deb-pkg fixes for the linux-headers package [Bjørn Mork, Azriel
     Samson]"

* 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
  coccicheck: Fix missing 0 index in kill loop
  scripts/package/Makefile: rpmbuild add support of RPMOPTS
  builddeb: fix missing headers in linux-headers package
  builddeb: include objtool binary in headers package
  kbuild/mkspec: support 'update-bootloader'-based systems
  scripts: coccinelle: remove check to move constants to right
  Coccinelle: setup_timer: Add space in front of parentheses

8 years agoMerge branch 'kconfig' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild
Linus Torvalds [Fri, 27 May 2016 05:27:09 +0000 (22:27 -0700)] 
Merge branch 'kconfig' of git://git./linux/kernel/git/mmarek/kbuild

Pull kconfig update from Michal Marek:

 - fix for behavior of tristate choice items and fix for documentation
   of existing kconfig behavior [Dirk Gouders]

 - more helpful "unexpected data" kconfig warning [Paul Bolle]

* 'kconfig' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
  kconfig/symbol.c: handle choice_values that depend on 'm' symbols
  kconfig-language: elaborate on the type of a choice
  kconfig-language: fix comment on dependency-generated menu structures.
  kconfig: add unexpected data itself to warning

8 years agoMerge branch 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild
Linus Torvalds [Fri, 27 May 2016 05:01:22 +0000 (22:01 -0700)] 
Merge branch 'kbuild' of git://git./linux/kernel/git/mmarek/kbuild

Pull kbuild updates from Michal Marek:

 - new option CONFIG_TRIM_UNUSED_KSYMS which does a two-pass build and
   unexports symbols which are not used in the current config [Nicolas
   Pitre]

 - several kbuild rule cleanups [Masahiro Yamada]

 - warning option adjustments for gcov etc [Arnd Bergmann]

 - a few more small fixes

* 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild: (31 commits)
  kbuild: move -Wunused-const-variable to W=1 warning level
  kbuild: fix if_change and friends to consider argument order
  kbuild: fix adjust_autoksyms.sh for modules that need only one symbol
  kbuild: fix ksym_dep_filter when multiple EXPORT_SYMBOL() on the same line
  gcov: disable -Wmaybe-uninitialized warning
  gcov: disable tree-loop-im to reduce stack usage
  gcov: disable for COMPILE_TEST
  Kbuild: disable 'maybe-uninitialized' warning for CONFIG_PROFILE_ALL_BRANCHES
  Kbuild: change CC_OPTIMIZE_FOR_SIZE definition
  kbuild: forbid kernel directory to contain spaces and colons
  kbuild: adjust ksym_dep_filter for some cmd_* renames
  kbuild: Fix dependencies for final vmlinux link
  kbuild: better abstract vmlinux sequential prerequisites
  kbuild: fix call to adjust_autoksyms.sh when output directory specified
  kbuild: Get rid of KBUILD_STR
  kbuild: rename cmd_as_s_S to cmd_cpp_s_S
  kbuild: rename cmd_cc_i_c to cmd_cpp_i_c
  kbuild: drop redundant "PHONY += FORCE"
  kbuild: delete unnecessary "@:"
  kbuild: mark help target as PHONY
  ...

8 years agoMerge branch 'akpm' (patches from Andrew)
Linus Torvalds [Fri, 27 May 2016 04:32:40 +0000 (21:32 -0700)] 
Merge branch 'akpm' (patches from Andrew)

Merge fixes from Andrew Morton:
 "10 fixes"

* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
  drivers/pinctrl/intel/pinctrl-baytrail.c: fix build with gcc-4.4
  update "mm/zsmalloc: don't fail if can't create debugfs info"
  dma-debug: avoid spinlock recursion when disabling dma-debug
  mm: oom_reaper: remove some bloat
  memcg: fix mem_cgroup_out_of_memory() return value.
  ocfs2: fix improper handling of return errno
  mm: slub: remove unused virt_to_obj()
  mm: kasan: remove unused 'reserved' field from struct kasan_alloc_meta
  mm: make CONFIG_DEFERRED_STRUCT_PAGE_INIT depends on !FLATMEM explicitly
  seqlock: fix raw_read_seqcount_latch()

8 years agoMerge tag 'dax-locking-for-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 27 May 2016 03:00:28 +0000 (20:00 -0700)] 
Merge tag 'dax-locking-for-4.7' of git://git./linux/kernel/git/nvdimm/nvdimm

Pull DAX locking updates from Ross Zwisler:
 "Filesystem DAX locking for 4.7

   - We use a bit in an exceptional radix tree entry as a lock bit and
     use it similarly to how page lock is used for normal faults.  This
     fixes races between hole instantiation and read faults of the same
     index.

   - Filesystem DAX PMD faults are disabled, and will be re-enabled when
     PMD locking is implemented"

* tag 'dax-locking-for-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
  dax: Remove i_mmap_lock protection
  dax: Use radix tree entry lock to protect cow faults
  dax: New fault locking
  dax: Allow DAX code to replace exceptional entries
  dax: Define DAX lock bit for radix tree exceptional entry
  dax: Make huge page handling depend of CONFIG_BROKEN
  dax: Fix condition for filling of PMD holes

8 years agoMerge tag 'dax-misc-for-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm...
Linus Torvalds [Fri, 27 May 2016 02:34:26 +0000 (19:34 -0700)] 
Merge tag 'dax-misc-for-4.7' of git://git./linux/kernel/git/nvdimm/nvdimm

Pull misc DAX updates from Vishal Verma:
 "DAX error handling for 4.7

   - Until now, dax has been disabled if media errors were found on any
     device.  This enables the use of DAX in the presence of these
     errors by making all sector-aligned zeroing go through the driver.

   - The driver (already) has the ability to clear errors on writes that
     are sent through the block layer using 'DSMs' defined in ACPI 6.1.

  Other misc changes:

   - When mounting DAX filesystems, check to make sure the partition is
     page aligned.  This is a requirement for DAX, and previously, we
     allowed such unaligned mounts to succeed, but subsequent
     reads/writes would fail.

   - Misc/cleanup fixes from Jan that remove unused code from DAX
     related to zeroing, writeback, and some size checks"

* tag 'dax-misc-for-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
  dax: fix a comment in dax_zero_page_range and dax_truncate_page
  dax: for truncate/hole-punch, do zeroing through the driver if possible
  dax: export a low-level __dax_zero_page_range helper
  dax: use sb_issue_zerout instead of calling dax_clear_sectors
  dax: enable dax in the presence of known media errors (badblocks)
  dax: fallback from pmd to pte on error
  block: Update blkdev_dax_capable() for consistency
  xfs: Add alignment check for DAX mount
  ext2: Add alignment check for DAX mount
  ext4: Add alignment check for DAX mount
  block: Add bdev_dax_supported() for dax mount checks
  block: Add vfs_msg() interface
  dax: Remove redundant inode size checks
  dax: Remove pointless writeback from dax_do_io()
  dax: Remove zeroing from dax_io()
  dax: Remove dead zeroing code from fault handlers
  ext2: Avoid DAX zeroing to corrupt data
  ext2: Fix block zeroing in ext2_get_blocks() for DAX
  dax: Remove complete_unwritten argument
  DAX: move RADIX_DAX_ definitions to dax.c

8 years agodrivers/pinctrl/intel/pinctrl-baytrail.c: fix build with gcc-4.4
Andrew Morton [Thu, 26 May 2016 22:16:30 +0000 (15:16 -0700)] 
drivers/pinctrl/intel/pinctrl-baytrail.c: fix build with gcc-4.4

gcc-4.4 and thereabouts has issues with initializers of anonymous
unions, and it generates the following warnings:

  drivers/pinctrl/intel/pinctrl-baytrail.c:413: error: unknown field 'simple_funcs' specified in initializer
  drivers/pinctrl/intel/pinctrl-baytrail.c:413: warning: missing braces around initializer
  drivers/pinctrl/intel/pinctrl-baytrail.c:413: warning: (near initialization for 'byt_score_groups[0].<anonymous>')
  drivers/pinctrl/intel/pinctrl-baytrail.c:415: error: unknown field 'simple_funcs' specified in initializer
  drivers/pinctrl/intel/pinctrl-baytrail.c:417: error: unknown field 'simple_funcs' specified in initializer
  ...

Work around this.

Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agoupdate "mm/zsmalloc: don't fail if can't create debugfs info"
Dan Streetman [Thu, 26 May 2016 22:16:27 +0000 (15:16 -0700)] 
update "mm/zsmalloc: don't fail if can't create debugfs info"

Some updates to commit d34f615720d1 ("mm/zsmalloc: don't fail if can't
create debugfs info"):

 - add pr_warn to all stat failure cases
 - do not prevent module loading on stat failure

Link: http://lkml.kernel.org/r/1463671123-5479-1-git-send-email-ddstreet@ieee.org
Signed-off-by: Dan Streetman <ddstreet@ieee.org>
Reviewed-by: Ganesh Mahendran <opensource.ganesh@gmail.com>
Acked-by: Minchan Kim <minchan@kernel.org>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Dan Streetman <dan.streetman@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agodma-debug: avoid spinlock recursion when disabling dma-debug
Ville Syrjälä [Thu, 26 May 2016 22:16:25 +0000 (15:16 -0700)] 
dma-debug: avoid spinlock recursion when disabling dma-debug

With netconsole (at least) the pr_err("...  disablingn") call can
recurse back into the dma-debug code, where it'll try to grab
free_entries_lock again.  Avoid the problem by doing the printk after
dropping the lock.

Link: http://lkml.kernel.org/r/1463678421-18683-1-git-send-email-ville.syrjala@linux.intel.com
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agomm: oom_reaper: remove some bloat
Michal Hocko [Thu, 26 May 2016 22:16:22 +0000 (15:16 -0700)] 
mm: oom_reaper: remove some bloat

mmput_async is currently used only from the oom_reaper which is defined
only for CONFIG_MMU.  We can save work_struct in mm_struct for
!CONFIG_MMU.

[akpm@linux-foundation.org: fix typo, per Minchan]
Link: http://lkml.kernel.org/r/20160520061658.GB19172@dhcp22.suse.cz
Reported-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Minchan Kim <minchan@kernel.org>
Cc: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agomemcg: fix mem_cgroup_out_of_memory() return value.
Tetsuo Handa [Thu, 26 May 2016 22:16:19 +0000 (15:16 -0700)] 
memcg: fix mem_cgroup_out_of_memory() return value.

mem_cgroup_out_of_memory() is returning "true" if it finds a TIF_MEMDIE
task after an eligible task was found, "false" if it found a TIF_MEMDIE
task before an eligible task is found.

This difference confuses memory_max_write() which checks the return
value of mem_cgroup_out_of_memory().  Since memory_max_write() wants to
continue looping, mem_cgroup_out_of_memory() should return "true" in
this case.

This patch sets a dummy pointer in order to return "true".

Link: http://lkml.kernel.org/r/1463753327-5170-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Vladimir Davydov <vdavydov@virtuozzo.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agoocfs2: fix improper handling of return errno
Eric Ren [Thu, 26 May 2016 22:16:16 +0000 (15:16 -0700)] 
ocfs2: fix improper handling of return errno

Previously, if a bad inode was found in ocfs2_iget(), -ESTALE was
returned back to the caller anyway.  Since commit d2b9d71a2da7 ("ocfs2:
check/fix inode block for online file check") can handle with return
value from ocfs2_read_locked_inode() now, we know the exact errno
returned for us.

Link: http://lkml.kernel.org/r/1463970656-18413-1-git-send-email-zren@suse.com
Signed-off-by: Eric Ren <zren@suse.com>
Reviewed-by: Joseph Qi <joseph.qi@huawei.com>
Cc: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agomm: slub: remove unused virt_to_obj()
Andrey Ryabinin [Thu, 26 May 2016 22:16:14 +0000 (15:16 -0700)] 
mm: slub: remove unused virt_to_obj()

It's unused since commit 7ed2f9e66385 ("mm, kasan: SLAB support")

Link: http://lkml.kernel.org/r/1464020961-2242-1-git-send-email-aryabinin@virtuozzo.com
Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agomm: kasan: remove unused 'reserved' field from struct kasan_alloc_meta
Andrey Ryabinin [Thu, 26 May 2016 22:16:11 +0000 (15:16 -0700)] 
mm: kasan: remove unused 'reserved' field from struct kasan_alloc_meta

Commit cd11016e5f52 ("mm, kasan: stackdepot implementation.  Enable
stackdepot for SLAB") added 'reserved' field, but never used it.

Link: http://lkml.kernel.org/r/1464021054-2307-1-git-send-email-aryabinin@virtuozzo.com
Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agomm: make CONFIG_DEFERRED_STRUCT_PAGE_INIT depends on !FLATMEM explicitly
Yang Shi [Thu, 26 May 2016 22:16:08 +0000 (15:16 -0700)] 
mm: make CONFIG_DEFERRED_STRUCT_PAGE_INIT depends on !FLATMEM explicitly

Per the suggestion from Michal Hocko [1], DEFERRED_STRUCT_PAGE_INIT
requires some ordering wrt other initialization operations, e.g.
page_ext_init has to happen after the whole memmap is initialized
properly.

For SPARSEMEM this requires to wait for page_alloc_init_late.  Other
memory models (e.g.  flatmem) might have different initialization
layouts (page_ext_init_flatmem).  Currently DEFERRED_STRUCT_PAGE_INIT
depends on MEMORY_HOTPLUG which in turn

depends on SPARSEMEM || X86_64_ACPI_NUMA
depends on ARCH_ENABLE_MEMORY_HOTPLUG

and X86_64_ACPI_NUMA depends on NUMA which in turn disable FLATMEM
memory model:

config ARCH_FLATMEM_ENABLE
def_bool y
depends on X86_32 && !NUMA

so FLATMEM is ruled out via dependency maze.  Be explicit and disable
FLATMEM for DEFERRED_STRUCT_PAGE_INIT so that we do not reintroduce
subtle initialization bugs

[1] http://lkml.kernel.org/r/20160523073157.GD2278@dhcp22.suse.cz

Link: http://lkml.kernel.org/r/1464027356-32282-1-git-send-email-yang.shi@linaro.org
Signed-off-by: Yang Shi <yang.shi@linaro.org>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agoseqlock: fix raw_read_seqcount_latch()
Alexey Dobriyan [Thu, 26 May 2016 22:16:06 +0000 (15:16 -0700)] 
seqlock: fix raw_read_seqcount_latch()

lockless_dereference() is supposed to take pointer not integer.

Link: http://lkml.kernel.org/r/20160521201448.GA7429@p183.telecom.by
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph...
Linus Torvalds [Thu, 26 May 2016 21:10:32 +0000 (14:10 -0700)] 
Merge branch 'for-linus' of git://git./linux/kernel/git/sage/ceph-client

Pull Ceph updates from Sage Weil:
 "This changeset has a few main parts:

   - Ilya has finished a huge refactoring effort to sync up the
     client-side logic in libceph with the user-space client code, which
     has evolved significantly over the last couple years, with lots of
     additional behaviors (e.g., how requests are handled when cluster
     is full and transitions from full to non-full).

     This structure of the code is more closely aligned with userspace
     now such that it will be much easier to maintain going forward when
     behavior changes take place.  There are some locking improvements
     bundled in as well.

   - Zheng adds multi-filesystem support (multiple namespaces within the
     same Ceph cluster)

   - Zheng has changed the readdir offsets and directory enumeration so
     that dentry offsets are hash-based and therefore stable across
     directory fragmentation events on the MDS.

   - Zheng has a smorgasbord of bug fixes across fs/ceph"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (71 commits)
  ceph: fix wake_up_session_cb()
  ceph: don't use truncate_pagecache() to invalidate read cache
  ceph: SetPageError() for writeback pages if writepages fails
  ceph: handle interrupted ceph_writepage()
  ceph: make ceph_update_writeable_page() uninterruptible
  libceph: make ceph_osdc_wait_request() uninterruptible
  ceph: handle -EAGAIN returned by ceph_update_writeable_page()
  ceph: make fault/page_mkwrite return VM_FAULT_OOM for -ENOMEM
  ceph: block non-fatal signals for fault/page_mkwrite
  ceph: make logical calculation functions return bool
  ceph: tolerate bad i_size for symlink inode
  ceph: improve fragtree change detection
  ceph: keep leaf frag when updating fragtree
  ceph: fix dir_auth check in ceph_fill_dirfrag()
  ceph: don't assume frag tree splits in mds reply are sorted
  ceph: fix inode reference leak
  ceph: using hash value to compose dentry offset
  ceph: don't forbid marking directory complete after forward seek
  ceph: record 'offset' for each entry of readdir result
  ceph: define 'end/complete' in readdir reply as bit flags
  ...

8 years agoMerge tag 'nfs-for-4.7-1' of git://git.linux-nfs.org/projects/anna/linux-nfs
Linus Torvalds [Thu, 26 May 2016 17:33:33 +0000 (10:33 -0700)] 
Merge tag 'nfs-for-4.7-1' of git://git.linux-nfs.org/projects/anna/linux-nfs

Pull NFS client updates from Anna Schumaker:
 "Highlights include:

  Features:
   - Add support for the NFS v4.2 COPY operation
   - Add support for NFS/RDMA over IPv6

  Bugfixes and cleanups:
   - Avoid race that crashes nfs_init_commit()
   - Fix oops in callback path
   - Fix LOCK/OPEN race when unlinking an open file
   - Choose correct stateids when using delegations in setattr, read and
     write
   - Don't send empty SETATTR after OPEN_CREATE
   - xprtrdma: Prevent server from writing a reply into memory client
     has released
   - xprtrdma: Support using Read list and Reply chunk in one RPC call"

* tag 'nfs-for-4.7-1' of git://git.linux-nfs.org/projects/anna/linux-nfs: (61 commits)
  pnfs: pnfs_update_layout needs to consider if strict iomode checking is on
  nfs/flexfiles: Use the layout segment for reading unless it a IOMODE_RW and reading is disabled
  nfs/flexfiles: Helper function to detect FF_FLAGS_NO_READ_IO
  nfs: avoid race that crashes nfs_init_commit
  NFS: checking for NULL instead of IS_ERR() in nfs_commit_file()
  pnfs: make pnfs_layout_process more robust
  pnfs: rework LAYOUTGET retry handling
  pnfs: lift retry logic from send_layoutget to pnfs_update_layout
  pnfs: fix bad error handling in send_layoutget
  flexfiles: add kerneldoc header to nfs4_ff_layout_prepare_ds
  flexfiles: remove pointless setting of NFS_LAYOUT_RETURN_REQUESTED
  pnfs: only tear down lsegs that precede seqid in LAYOUTRETURN args
  pnfs: keep track of the return sequence number in pnfs_layout_hdr
  pnfs: record sequence in pnfs_layout_segment when it's created
  pnfs: don't merge new ff lsegs with ones that have LAYOUTRETURN bit set
  pNFS/flexfiles: When initing reads or writes, we might have to retry connecting to DSes
  pNFS/flexfiles: When checking for available DSes, conditionally check for MDS io
  pNFS/flexfile: Fix erroneous fall back to read/write through the MDS
  NFS: Reclaim writes via writepage are opportunistic
  NFSv4: Use the right stateid for delegations in setattr, read and write
  ...

8 years agoMerge tag 'xfs-for-linus-4.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 26 May 2016 17:13:40 +0000 (10:13 -0700)] 
Merge tag 'xfs-for-linus-4.7-rc1' of git://git./linux/kernel/git/dgc/linux-xfs

Pull xfs updates from Dave Chinner:
 "A pretty average collection of fixes, cleanups and improvements in
  this request.

  Summary:
   - fixes for mount line parsing, sparse warnings, read-only compat
     feature remount behaviour
   - allow fast path symlink lookups for inline symlinks.
   - attribute listing cleanups
   - writeback goes direct to bios rather than indirecting through
     bufferheads
   - transaction allocation cleanup
   - optimised kmem_realloc
   - added configurable error handling for metadata write errors,
     changed default error handling behaviour from "retry forever" to
     "retry until unmount then fail"
   - fixed several inode cluster writeback lookup vs reclaim race
     conditions
   - fixed inode cluster writeback checking wrong inode after lookup
   - fixed bugs where struct xfs_inode freeing wasn't actually RCU safe
   - cleaned up inode reclaim tagging"

* tag 'xfs-for-linus-4.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs: (39 commits)
  xfs: fix warning in xfs_finish_page_writeback for non-debug builds
  xfs: move reclaim tagging functions
  xfs: simplify inode reclaim tagging interfaces
  xfs: rename variables in xfs_iflush_cluster for clarity
  xfs: xfs_iflush_cluster has range issues
  xfs: mark reclaimed inodes invalid earlier
  xfs: xfs_inode_free() isn't RCU safe
  xfs: optimise xfs_iext_destroy
  xfs: skip stale inodes in xfs_iflush_cluster
  xfs: fix inode validity check in xfs_iflush_cluster
  xfs: xfs_iflush_cluster fails to abort on error
  xfs: remove xfs_fs_evict_inode()
  xfs: add "fail at unmount" error handling configuration
  xfs: add configuration handlers for specific errors
  xfs: add configuration of error failure speed
  xfs: introduce table-based init for error behaviors
  xfs: add configurable error support to metadata buffers
  xfs: introduce metadata IO error class
  xfs: configurable error behavior via sysfs
  xfs: buffer ->bi_end_io function requires irq-safe lock
  ...

8 years agoMerge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelv...
Linus Torvalds [Thu, 26 May 2016 16:48:23 +0000 (09:48 -0700)] 
Merge branch 'hwmon-for-linus' of git://git./linux/kernel/git/jdelvare/staging

Pull hwmon fixlets from Jean Delvare.

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  Documentation/hwmon: Update links in max34440
  hwmon: (emc2103) Fix typo in MODULE_PARM_DESC

8 years agoMerge tag 'mmc-v4.7-rc1' of git://git.linaro.org/people/ulf.hansson/mmc
Linus Torvalds [Thu, 26 May 2016 16:36:10 +0000 (09:36 -0700)] 
Merge tag 'mmc-v4.7-rc1' of git://git.linaro.org/people/ulf.hansson/mmc

Pull MMC fixes from Ulf Hansson:
 "Here are some mmc fixes intended for v4.7 rc1.  They are based on a
  commit earlier in the merge window and have been tested in linux-next
  for a while.

  MMC core:
   - Prevent re-tuning while serving requests for RPMB partitions
   - Extend timeout for long read time quirk to support more eMMCs

  MMC host:
   - sdhci-acpi: Ensure connected devices are powered when probing
   - sdhci-pci|acpi: Remove unreliable MMC_CAP_BUS_WIDTH_TEST for Intel HWs
   - dw_mmc: Correct the assigning of max_blk_size
   - dw_mmc-rockchip: Allow RPMB partitions to be created
   - dw_mmc-rockchip: Set the drive phase properly"

* tag 'mmc-v4.7-rc1' of git://git.linaro.org/people/ulf.hansson/mmc:
  mmc: sdhci-acpi: Remove MMC_CAP_BUS_WIDTH_TEST for Intel controllers
  mmc: sdhci-pci: Remove MMC_CAP_BUS_WIDTH_TEST for Intel controllers
  mmc: longer timeout for long read time quirk
  mmc: dw_mmc: rockchip: Set the drive phase properly
  mmc: dw_mmc: fix the wrong max_blk_size
  mmc: dw_mmc-rockchip: add MMC_CAP_CMD23 capabilities
  mmc: sdhci-acpi: Ensure connected devices are powered when probing
  ACPI / PM: Export acpi_device_fix_up_power()
  mmc: block: Pause re-tuning while switched to the RPMB partition
  mmc: block: Always switch back to main area after RPMB access
  mmc: core: Add a facility to "pause" re-tuning

8 years agoMerge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux
Linus Torvalds [Thu, 26 May 2016 16:23:43 +0000 (09:23 -0700)] 
Merge branch 'next' of git://git./linux/kernel/git/rzhang/linux

Pull thermal management updates from Zhang Rui:

 - Introduce generic ADC thermal driver, based on OF thermal (Laxman
   Dewangan)

 - Introduce new thermal driver for Tango chips (Marc Gonzalez)

 - Rockchip driver support for RK3399, RK3366, and some fixes (Caesar
   Wang, Elaine Zhang and Shawn Lin)

 - Add CPU power cooling model to Mediatek thermal driver (Dawei Chien)

 - Wider usage of dev_thermal_zone_of_sensor_register (Eduardo Valentin)

 - TI thermal driver gained a new maintainer (Keerthy).

 - Enabled powerclamp driver by checking CPU feature and package cstate
   counter instead of CPU whitelist (Jacob Pan)

 - Various fixes on thermal governor, OF thermal, Tegra, and RCAR

* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux: (50 commits)
  thermal: tango: initialize TEMPSI_CFG
  thermal: rockchip: use the usleep_range instead of udelay
  thermal: rockchip: add the notes for better reading
  thermal: rockchip: Support RK3366 SoCs in the thermal driver
  thermal: rockchip: handle the power sequence for tsadc controller
  thermal: rockchip: update the tsadc table for rk3399
  thermal: rockchip: fixes the code_to_temp for tsadc driver
  thermal: rockchip: disable thermal->clk in err case
  thermal: tegra: add Tegra132 specific SOC_THERM driver
  thermal: fix ptr_ret.cocci warnings
  thermal: mediatek: Add cpu dynamic power cooling model.
  thermal: generic-adc: Add ADC based thermal sensor driver
  thermal: generic-adc: Add DT binding for ADC based thermal sensor
  thermal: tegra: fix static checker warning
  thermal: tegra: mark PM functions __maybe_unused
  thermal: add temperature sensor support for tango SoC
  thermal: hisilicon: fix IRQ imbalance enabling
  thermal: hisilicon: support to use any sensor
  thermal: rcar: Remove binding docs for r8a7794
  thermal: tegra: add PM support
  ...

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris...
Linus Torvalds [Thu, 26 May 2016 16:15:19 +0000 (09:15 -0700)] 
Merge branch 'for-linus' of git://git./linux/kernel/git/jmorris/linux-security

Pull Yama locking fix from James Morris:
 "Fix for the Yama LSM"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
  Yama: fix double-spinlock and user access in atomic context

8 years agopnfs: pnfs_update_layout needs to consider if strict iomode checking is on
Tom Haynes [Wed, 25 May 2016 14:31:14 +0000 (07:31 -0700)] 
pnfs: pnfs_update_layout needs to consider if strict iomode checking is on

As flexfiles has FF_FLAGS_NO_READ_IO, there is a need to generically
support enforcing that a IOMODE_RW segment will not allow READ I/O.

Signed-off-by: Tom Haynes <loghyr@primarydata.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
8 years agonfs/flexfiles: Use the layout segment for reading unless it a IOMODE_RW and reading...
Tom Haynes [Wed, 25 May 2016 14:31:13 +0000 (07:31 -0700)] 
nfs/flexfiles: Use the layout segment for reading unless it a IOMODE_RW and reading is disabled

Signed-off-by: Tom Haynes <loghyr@primarydata.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
8 years agoDocumentation/hwmon: Update links in max34440
Glenn Dayton [Thu, 26 May 2016 09:06:53 +0000 (11:06 +0200)] 
Documentation/hwmon: Update links in max34440

It appears the website for maxim-ic.com changed to
maximintegrated.com.

Signed-off-by: Glenn Dayton <glenn.dayton24@gmail.com>
Signed-off-by: Jean Delvare <jdelvare@suse.de>
8 years agohwmon: (emc2103) Fix typo in MODULE_PARM_DESC
Dan Carpenter [Thu, 26 May 2016 09:06:53 +0000 (11:06 +0200)] 
hwmon: (emc2103) Fix typo in MODULE_PARM_DESC

"apd" was intended here instead of "init".

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jean Delvare <jdelvare@suse.de>
8 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 26 May 2016 00:37:33 +0000 (17:37 -0700)] 
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Ingo Molnar:
 "Misc fixes: EFI, entry code, pkeys and MPX fixes, TASK_SIZE cleanups
  and a tsc frequency table fix"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mm: Switch from TASK_SIZE to TASK_SIZE_MAX in the page fault code
  x86/fsgsbase/64: Use TASK_SIZE_MAX for FSBASE/GSBASE upper limits
  x86/mm/mpx: Work around MPX erratum SKD046
  x86/entry/64: Fix stack return address retrieval in thunk
  x86/efi: Fix 7-parameter efi_call()s
  x86/cpufeature, x86/mm/pkeys: Fix broken compile-time disabling of pkeys
  x86/tsc: Add missing Cherrytrail frequency to the table

8 years agoMerge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 26 May 2016 00:11:43 +0000 (17:11 -0700)] 
Merge branch 'sched-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull scheduler fixes from Ingo Molnar:
 "Two fixes: one for a lost wakeup, the other to fix the compiler
  optimizing out preempt operations on ARM64 (and possibly other non-x86
  architectures)"

* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/core: Fix remote wakeups
  sched/preempt: Fix preempt_count manipulations

8 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 26 May 2016 00:05:40 +0000 (17:05 -0700)] 
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf updates from Ingo Molnar:
 "Mostly tooling and PMU driver fixes, but also a number of late updates
  such as the reworking of the call-chain size limiting logic to make
  call-graph recording more robust, plus tooling side changes for the
  new 'backwards ring-buffer' extension to the perf ring-buffer"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (34 commits)
  perf record: Read from backward ring buffer
  perf record: Rename variable to make code clear
  perf record: Prevent reading invalid data in record__mmap_read
  perf evlist: Add API to pause/resume
  perf trace: Use the ptr->name beautifier as default for "filename" args
  perf trace: Use the fd->name beautifier as default for "fd" args
  perf report: Add srcline_from/to branch sort keys
  perf evsel: Record fd into perf_mmap
  perf evsel: Add overwrite attribute and check write_backward
  perf tools: Set buildid dir under symfs when --symfs is provided
  perf trace: Only auto set call-graph to "dwarf" when syscalls are being traced
  perf annotate: Sort list of recognised instructions
  perf annotate: Fix identification of ARM blt and bls instructions
  perf tools: Fix usage of max_stack sysctl
  perf callchain: Stop validating callchains by the max_stack sysctl
  perf trace: Fix exit_group() formatting
  perf top: Use machine->kptr_restrict_warned
  perf trace: Warn when trying to resolve kernel addresses with kptr_restrict=1
  perf machine: Do not bail out if not managing to read ref reloc symbol
  perf/x86/intel/p4: Trival indentation fix, remove space
  ...

8 years agoYama: fix double-spinlock and user access in atomic context
Jann Horn [Sun, 22 May 2016 04:01:34 +0000 (06:01 +0200)] 
Yama: fix double-spinlock and user access in atomic context

Commit 8a56038c2aef ("Yama: consolidate error reporting") causes lockups
when someone hits a Yama denial. Call chain:

process_vm_readv -> process_vm_rw -> process_vm_rw_core -> mm_access
-> ptrace_may_access
task_lock(...) is taken
__ptrace_may_access -> security_ptrace_access_check
-> yama_ptrace_access_check -> report_access -> kstrdup_quotable_cmdline
-> get_cmdline -> access_process_vm -> get_task_mm
task_lock(...) is taken again

task_lock(p) just calls spin_lock(&p->alloc_lock), so at this point,
spin_lock() is called on a lock that is already held by the current
process.

Also: Since the alloc_lock is a spinlock, sleeping inside
security_ptrace_access_check hooks is probably not allowed at all? So it's
not even possible to print the cmdline from in there because that might
involve paging in userspace memory.

It would be tempting to rewrite ptrace_may_access() to drop the alloc_lock
before calling the LSM, but even then, ptrace_may_access() itself might be
called from various contexts in which you're not allowed to sleep; for
example, as far as I understand, to be able to hold a reference to another
task, usually an RCU read lock will be taken (see e.g. kcmp() and
get_robust_list()), so that also prohibits sleeping. (And using e.g. FUSE,
a user can cause pagefault handling to take arbitrary amounts of time -
see https://bugs.chromium.org/p/project-zero/issues/detail?id=808.)

Therefore, AFAIK, in order to print the name of a process below
security_ptrace_access_check(), you'd have to either grab a reference to
the mm_struct and defer the access violation reporting or just use the
"comm" value that's stored in kernelspace and accessible without big
complications. (Or you could try to use some kind of atomic remote VM
access that fails if the memory isn't paged in, similar to
copy_from_user_inatomic(), and if necessary fall back to comm, but
that'd be kind of ugly because the comm/cmdline choice would look
pretty random to the user.)

Fix it by deferring reporting of the access violation until current
exits kernelspace the next time.

v2: Don't oops on PTRACE_TRACEME, call report_access under
task_lock(current). Also fix nonsensical comment. And don't use
GPF_ATOMIC for memory allocation with no locks held.
This patch is tested both for ptrace attach and ptrace traceme.

Fixes: 8a56038c2aef ("Yama: consolidate error reporting")
Signed-off-by: Jann Horn <jann@thejh.net>
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: James Morris <james.l.morris@oracle.com>
8 years agoMerge branch 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Wed, 25 May 2016 23:52:19 +0000 (16:52 -0700)] 
Merge branch 'core-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull objtool build fix from Ingo Molnar:
 "An libtool fix for older libelf versions"

* 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  objtool: Allow building with older libelf

8 years agoceph: fix wake_up_session_cb()
Yan, Zheng [Thu, 19 May 2016 11:15:19 +0000 (19:15 +0800)] 
ceph: fix wake_up_session_cb()

We should reset i_requested_max_size before waking the waiters.
(zero i_requested_max_size make waiter re-request the max size)

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: don't use truncate_pagecache() to invalidate read cache
Yan, Zheng [Wed, 18 May 2016 12:58:26 +0000 (20:58 +0800)] 
ceph: don't use truncate_pagecache() to invalidate read cache

truncate_pagecache() drops dirty pages, it's dangerous to use it
to invalidate read cache. Besides, we shouldn't start invalidating
read cache while there are buffer writers. Because buffer writers
may add dirty pages later.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: SetPageError() for writeback pages if writepages fails
Yan, Zheng [Fri, 13 May 2016 09:54:17 +0000 (17:54 +0800)] 
ceph: SetPageError() for writeback pages if writepages fails

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: handle interrupted ceph_writepage()
Yan, Zheng [Fri, 13 May 2016 09:29:51 +0000 (17:29 +0800)] 
ceph: handle interrupted ceph_writepage()

writepage() can be interrupted when it's called by direct memory
reclaimer (the direct memory relaimer is killed). To avoid lossing
data, we redirty the page.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: make ceph_update_writeable_page() uninterruptible
Yan, Zheng [Fri, 13 May 2016 03:30:24 +0000 (11:30 +0800)] 
ceph: make ceph_update_writeable_page() uninterruptible

ceph_update_writeable_page() is used by ceph_write_begin(). It beaks
atomicity of write operation if it's interruptible.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agolibceph: make ceph_osdc_wait_request() uninterruptible
Yan, Zheng [Fri, 13 May 2016 03:04:33 +0000 (11:04 +0800)] 
libceph: make ceph_osdc_wait_request() uninterruptible

Ceph_osdc_wait_request() is used when cephfs issues sync IO. In most
cases, the sync IO should be uninterruptible. The fix is use killale
wait function in ceph_osdc_wait_request().

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: handle -EAGAIN returned by ceph_update_writeable_page()
Yan, Zheng [Tue, 10 May 2016 11:09:06 +0000 (19:09 +0800)] 
ceph: handle -EAGAIN returned by ceph_update_writeable_page()

when ceph_update_writeable_page() return -EAGAIN, caller should
lock the page and call ceph_update_writeable_page() again.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: make fault/page_mkwrite return VM_FAULT_OOM for -ENOMEM
Yan, Zheng [Tue, 10 May 2016 10:59:13 +0000 (18:59 +0800)] 
ceph: make fault/page_mkwrite return VM_FAULT_OOM for -ENOMEM

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: block non-fatal signals for fault/page_mkwrite
Yan, Zheng [Tue, 10 May 2016 10:40:28 +0000 (18:40 +0800)] 
ceph: block non-fatal signals for fault/page_mkwrite

Fault and page_mkwrite are supposed to be uninterruptable. But they
call ceph functions that are interruptible. So they should block
signals before calling functions that are interruptible

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: make logical calculation functions return bool
Zhang Zhuoyu [Fri, 25 Mar 2016 09:18:39 +0000 (05:18 -0400)] 
ceph: make logical calculation functions return bool

This patch makes serverl logical caculation functions return bool to
improve readability due to these particular functions only using 0/1
as their return value.

No functional change.

Signed-off-by: Zhang Zhuoyu <zhangzhuoyu@cmss.chinamobile.com>
8 years agoceph: tolerate bad i_size for symlink inode
Yan, Zheng [Thu, 5 May 2016 08:40:17 +0000 (16:40 +0800)] 
ceph: tolerate bad i_size for symlink inode

A mds bug can cause symlink's size to be truncated to zero.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: improve fragtree change detection
Yan, Zheng [Wed, 4 May 2016 03:40:30 +0000 (11:40 +0800)] 
ceph: improve fragtree change detection

check if number of splits in i_fragtree is equal to number of splits
in mds reply

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: keep leaf frag when updating fragtree
Yan, Zheng [Wed, 4 May 2016 03:05:10 +0000 (11:05 +0800)] 
ceph: keep leaf frag when updating fragtree

Nodes in i_fragtree are sorted according to ceph_compare_frag().
It means frag node in i_fragtree always follow its direct parent
node. To check if a leaf node is valid, we just need to check if
it's child of previous split node.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: fix dir_auth check in ceph_fill_dirfrag()
Yan, Zheng [Tue, 3 May 2016 14:33:20 +0000 (22:33 +0800)] 
ceph: fix dir_auth check in ceph_fill_dirfrag()

-1 is CDIR_AUTH_PARENT, it means dir's auth mds is the same as
inode's auth mds

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: don't assume frag tree splits in mds reply are sorted
Yan, Zheng [Tue, 3 May 2016 12:55:50 +0000 (20:55 +0800)] 
ceph: don't assume frag tree splits in mds reply are sorted

The algorithm that updates i_fragtree relies on that the frag tree
splits in mds reply are of the same order of i_fragtree. This is not
true because current MDS encodes frag tree splits in ascending order
of (unsigned)frag_t. But nodes in i_fragtree are sorted according to
ceph_frag_compare().

The fix is sort the frag tree splits first, then updates i_fragtree.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: fix inode reference leak
Yan, Zheng [Fri, 29 Apr 2016 15:40:23 +0000 (23:40 +0800)] 
ceph: fix inode reference leak

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: using hash value to compose dentry offset
Yan, Zheng [Fri, 29 Apr 2016 03:27:30 +0000 (11:27 +0800)] 
ceph: using hash value to compose dentry offset

If MDS sorts dentries in dirfrag in hash order, we use hash value to
compose dentry offset. dentry offset is:

  (0xff << 52) | ((24 bits hash) << 28) |
  (the nth entry hash hash collision)

This offset is stable across directory fragmentation. This alos means
there is no need to reset readdir offset if directory get fragmented
in the middle of readdir.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: don't forbid marking directory complete after forward seek
Yan, Zheng [Thu, 28 Apr 2016 14:56:44 +0000 (22:56 +0800)] 
ceph: don't forbid marking directory complete after forward seek

Forward seek within same frag does not update fi->last_name, it will
not affect contents of later readdir reply. So there is no need to
forbid marking directory complete

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: record 'offset' for each entry of readdir result
Yan, Zheng [Thu, 28 Apr 2016 07:17:40 +0000 (15:17 +0800)] 
ceph: record 'offset' for each entry of readdir result

This is preparation for using hash value as dentry 'offset'

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: define 'end/complete' in readdir reply as bit flags
Yan, Zheng [Wed, 27 Apr 2016 09:48:30 +0000 (17:48 +0800)] 
ceph: define 'end/complete' in readdir reply as bit flags

Set a flag in readdir request, which indicates that client interprets
'end/complete' as bit flags. So that mds can reply additional flags in
readdir reply.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: define struct for dir entry in readdir reply
Yan, Zheng [Thu, 28 Apr 2016 01:37:39 +0000 (09:37 +0800)] 
ceph: define struct for dir entry in readdir reply

This avoids defining multiple arrays for entries in readdir reply

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: simplify 'offset in frag'
Yan, Zheng [Wed, 27 Apr 2016 09:32:34 +0000 (17:32 +0800)] 
ceph: simplify 'offset in frag'

don't distinguish leftmost frag from other frags. always use 2 as
first entry's offset.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: remove unnecessary checks in __dcache_readdir
Yan, Zheng [Fri, 29 Apr 2016 07:58:32 +0000 (15:58 +0800)] 
ceph: remove unnecessary checks in __dcache_readdir

we never add snapdir and the hidden .ceph dir into readdir cache

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: search cache postion for dcache readdir
Yan, Zheng [Thu, 28 Apr 2016 09:43:35 +0000 (17:43 +0800)] 
ceph: search cache postion for dcache readdir

use binary search to find cache index that corresponds to readdir
postion.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: use CEPH_MDS_OP_RMXATTR request to remove xattr
Yan, Zheng [Thu, 21 Apr 2016 04:11:54 +0000 (12:11 +0800)] 
ceph: use CEPH_MDS_OP_RMXATTR request to remove xattr

Setxattr with NULL value and XATTR_REPLACE flag should be equivalent
to removexattr. But current MDS does not support deleting vxattrs through
MDS_OP_SETXATTR request. The workaround is sending MDS_OP_RMXATTR request
if setxattr actually removs xattr.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: report mount root in session metadata
Yan, Zheng [Thu, 21 Apr 2016 03:09:55 +0000 (11:09 +0800)] 
ceph: report mount root in session metadata

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: don't show symlink target in debugfs/mdsc
Yan, Zheng [Mon, 18 Apr 2016 08:51:37 +0000 (16:51 +0800)] 
ceph: don't show symlink target in debugfs/mdsc

symlink target is useless for debug and can be very long. It's annoying
to show it in debugfs/mdsc.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: don't call truncate_pagecache in ceph_writepages_start
Yan, Zheng [Fri, 15 Apr 2016 05:56:12 +0000 (13:56 +0800)] 
ceph: don't call truncate_pagecache in ceph_writepages_start

truncate_pagecache() may decrease inode's reference. This can cause
deadlock if inode's last reference is dropped and iput_final() wants
to evict the inode. (evict() calls inode_wait_for_writeback(), which
waits for ceph_writepages_start() to return).

The fix is use work thead to truncate dirty pages. Also add 'forced
umount' check to ceph_update_writeable_page(), which prevents new
pages getting dirty.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: renew caps for read/write if mds session got killed.
Yan, Zheng [Fri, 8 Apr 2016 07:27:16 +0000 (15:27 +0800)] 
ceph: renew caps for read/write if mds session got killed.

When mds session gets killed, read/write operation may hang.
Client waits for Frw caps, but mds does not know what caps client
wants. To recover this, client sends an open request to mds. The
request will tell mds what caps client wants.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: CEPH_FEATURE_MDSENC support
Yan, Zheng [Thu, 31 Mar 2016 07:53:01 +0000 (15:53 +0800)] 
ceph: CEPH_FEATURE_MDSENC support

Signed-off-by: Yan, Zheng <zyan@redhat.com>
8 years agoceph: multiple filesystem support
Yan, Zheng [Wed, 30 Mar 2016 09:18:34 +0000 (17:18 +0800)] 
ceph: multiple filesystem support

To access non-default filesystem, we just need to subscribe to
mdsmap.<MDS_NAMESPACE_ID> and add a new mount option for mds
namespace id.

Signed-off-by: Yan, Zheng <zyan@redhat.com>
[idryomov@gmail.com: switch to a new libceph API]
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: support for subscribing to "mdsmap.<id>" maps
Ilya Dryomov [Wed, 25 May 2016 22:05:01 +0000 (00:05 +0200)] 
libceph: support for subscribing to "mdsmap.<id>" maps

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: replace ceph_monc_request_next_osdmap()
Ilya Dryomov [Thu, 28 Apr 2016 14:07:28 +0000 (16:07 +0200)] 
libceph: replace ceph_monc_request_next_osdmap()

... with a wrapper around maybe_request_map() - no need for two
osdmap-specific functions.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: take osdc->lock in osdmap_show() and dump flags in hex
Ilya Dryomov [Thu, 28 Apr 2016 14:07:27 +0000 (16:07 +0200)] 
libceph: take osdc->lock in osdmap_show() and dump flags in hex

There is now about a dozen CEPH_OSDMAP_* flags.  This is a debugging
interface, so just dump in hex instead of spelling each flag out.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: pool deletion detection
Ilya Dryomov [Thu, 28 Apr 2016 14:07:27 +0000 (16:07 +0200)] 
libceph: pool deletion detection

This adds the "map check" infrastructure for sending osdmap version
checks on CALC_TARGET_POOL_DNE and completing in-flight requests with
-ENOENT if the target pool doesn't exist or has just been deleted.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: async MON client generic requests
Ilya Dryomov [Thu, 28 Apr 2016 14:07:27 +0000 (16:07 +0200)] 
libceph: async MON client generic requests

For map check, we are going to need to send CEPH_MSG_MON_GET_VERSION
messages asynchronously and get a callback on completion.  Refactor MON
client to allow firing off generic requests asynchronously and add an
async variant of ceph_monc_get_version().  ceph_monc_do_statfs() is
switched over and remains sync.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: support for checking on status of watch
Ilya Dryomov [Thu, 28 Apr 2016 14:07:27 +0000 (16:07 +0200)] 
libceph: support for checking on status of watch

Implement ceph_osdc_watch_check() to be able to check on status of
watch.  Note that the time it takes for a watch/notify event to get
delivered through the notify_wq is taken into account.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: support for sending notifies
Ilya Dryomov [Thu, 28 Apr 2016 14:07:27 +0000 (16:07 +0200)] 
libceph: support for sending notifies

Implement ceph_osdc_notify() for sending notifies.

Due to the fact that the current messenger can't do read-in into
pagelists (it can only do write-out from them), I had to go with a page
vector for a NOTIFY_COMPLETE payload, for now.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph, rbd: ceph_osd_linger_request, watch/notify v2
Ilya Dryomov [Wed, 25 May 2016 23:15:02 +0000 (01:15 +0200)] 
libceph, rbd: ceph_osd_linger_request, watch/notify v2

This adds support and switches rbd to a new, more reliable version of
watch/notify protocol.  As with the OSD client update, this is mostly
about getting the right structures linked into the right places so that
reconnects are properly sent when needed.  watch/notify v2 also
requires sending regular pings to the OSDs - send_linger_ping().

A major change from the old watch/notify implementation is the
introduction of ceph_osd_linger_request - linger requests no longer
piggy back on ceph_osd_request.  ceph_osd_event has been merged into
ceph_osd_linger_request.

All the details are now hidden within libceph, the interface consists
of a simple pair of watch/unwatch functions and ceph_osdc_notify_ack().
ceph_osdc_watch() does return ceph_osd_linger_request, but only to keep
the lifetime management simple.

ceph_osdc_notify_ack() accepts an optional data payload, which is
relayed back to the notifier.

Portions of this patch are loosely based on work by Douglas Fuller
<dfuller@redhat.com> and Mike Christie <michaelc@cs.wisc.edu>.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agorbd: rbd_dev_header_unwatch_sync() variant
Ilya Dryomov [Thu, 28 Apr 2016 14:07:26 +0000 (16:07 +0200)] 
rbd: rbd_dev_header_unwatch_sync() variant

Introduce __rbd_dev_header_unwatch_sync(), which doesn't flush notify
callbacks.  This is for the new rados_watcherrcb_t, which would be
called from a notify callback.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: wait_request_timeout()
Ilya Dryomov [Thu, 28 Apr 2016 14:07:26 +0000 (16:07 +0200)] 
libceph: wait_request_timeout()

The unwatch timeout is currently implemented in rbd.  With
watch/unwatch code moving into libceph, we are going to need
a ceph_osdc_wait_request() variant with a timeout.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: request_init() and request_release_checks()
Ilya Dryomov [Thu, 28 Apr 2016 14:07:26 +0000 (16:07 +0200)] 
libceph: request_init() and request_release_checks()

These are going to be used by request_reinit() code.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: a major OSD client update
Ilya Dryomov [Thu, 28 Apr 2016 14:07:26 +0000 (16:07 +0200)] 
libceph: a major OSD client update

This is a major sync up, up to ~Jewel.  The highlights are:

- per-session request trees (vs a global per-client tree)
- per-session locking (vs a global per-client rwlock)
- homeless OSD session
- no ad-hoc global per-client lists
- support for pool quotas
- foundation for watch/notify v2 support
- foundation for map check (pool deletion detection) support

The switchover is incomplete: lingering requests can be setup and
teared down but aren't ever reestablished.  This functionality is
restored with the introduction of the new lingering infrastructure
(ceph_osd_linger_request, linger_work, etc) in a later commit.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: protect osdc->osd_lru list with a spinlock
Ilya Dryomov [Thu, 28 Apr 2016 14:07:26 +0000 (16:07 +0200)] 
libceph: protect osdc->osd_lru list with a spinlock

OSD client is getting moved from the big per-client lock to a set of
per-session locks.  The big rwlock would only be held for read most of
the time, so a global osdc->osd_lru needs additional protection.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
8 years agolibceph: allocate ceph_osd with GFP_NOFAIL
Ilya Dryomov [Thu, 28 Apr 2016 14:07:25 +0000 (16:07 +0200)] 
libceph: allocate ceph_osd with GFP_NOFAIL

create_osd() is called way too deep in the stack to be able to error
out in a sane way; a failing create_osd() just messes everything up.
The current req_notarget list solution is broken - the list is never
traversed as it's not entirely clear when to do it, I guess.

If we were to start traversing it at regular intervals and retrying
each request, we wouldn't be far off from what __GFP_NOFAIL is doing,
so allocate OSD sessions with __GFP_NOFAIL, at least until we come up
with a better fix.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
This page took 0.048577 seconds and 5 git commands to generate.