~shefty/rdma-dev.git
9 years agos3fb: Compute VGA base iomem pointer explicitly.
David Miller [Tue, 11 Jan 2011 23:54:21 +0000 (23:54 +0000)]
s3fb: Compute VGA base iomem pointer explicitly.

This allows the driver to work in multi-domain PCI
configurations.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agovt8623fb: Don't clobber par->state.vgabase during open method.
David Miller [Tue, 11 Jan 2011 23:54:07 +0000 (23:54 +0000)]
vt8623fb: Don't clobber par->state.vgabase during open method.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agos3fb: Don't clobber par->state.vgabase during open method.
David Miller [Tue, 11 Jan 2011 23:53:53 +0000 (23:53 +0000)]
s3fb: Don't clobber par->state.vgabase during open method.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agoarkfb: Don't clobber par->state.vgabase during open method.
David Miller [Tue, 11 Jan 2011 23:53:38 +0000 (23:53 +0000)]
arkfb: Don't clobber par->state.vgabase during open method.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agovt8623fb: Pass par->state.vgabase to vga_*() calls.
David Miller [Tue, 11 Jan 2011 23:53:11 +0000 (23:53 +0000)]
vt8623fb: Pass par->state.vgabase to vga_*() calls.

Instead of just plain NULL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agos3fb: Pass par->state.vgabase to vga_*() calls.
David Miller [Tue, 11 Jan 2011 23:52:57 +0000 (23:52 +0000)]
s3fb: Pass par->state.vgabase to vga_*() calls.

Instead of just plain NULL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agoarkfb: Pass par->state.vgabase to vga_*() calls.
David Miller [Tue, 11 Jan 2011 23:52:40 +0000 (23:52 +0000)]
arkfb: Pass par->state.vgabase to vga_*() calls.

Instead of just plain NULL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_set_timings() take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:52:25 +0000 (23:52 +0000)]
svga: Make svga_set_timings() take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_tilecursor() take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:52:11 +0000 (23:52 +0000)]
svga: Make svga_tilecursor() take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_set_textmode_vga_regs() take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:51:56 +0000 (23:51 +0000)]
svga: Make svga_set_textmode_vga_regs() take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_set_default_crt_regs() take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:51:41 +0000 (23:51 +0000)]
svga: Make svga_set_default_crt_regs() take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_wcrt_mask() take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:51:26 +0000 (23:51 +0000)]
svga: Make svga_wcrt_mask() take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_wseq_mask() take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:51:08 +0000 (23:51 +0000)]
svga: Make svga_wseq_mask() take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_set_default_seq_regs take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:50:54 +0000 (23:50 +0000)]
svga: Make svga_set_default_seq_regs take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_set_default_atc_regs take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:50:36 +0000 (23:50 +0000)]
svga: Make svga_set_default_atc_regs take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_set_default_gfx_regs take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:50:04 +0000 (23:50 +0000)]
svga: Make svga_set_default_gfx_regs take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_wseq_multi take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:49:49 +0000 (23:49 +0000)]
svga: Make svga_wseq_multi take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_wcrt_multi take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:49:34 +0000 (23:49 +0000)]
svga: Make svga_wcrt_multi take an iomem regbase pointer.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Make svga_wattr take an iomem regbase pointer.
David Miller [Tue, 11 Jan 2011 23:49:18 +0000 (23:49 +0000)]
svga: Make svga_wattr take an iomem regbase pointer.

And use vga_{r,w}().

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosvga: Use proper VGA register name macros in svga_wattr.
David Miller [Tue, 11 Jan 2011 23:49:03 +0000 (23:49 +0000)]
svga: Use proper VGA register name macros in svga_wattr.

Instead of magic register location constants.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosisfb: add support for XGI Z9 DDR2 POST
Aaro Koskinen [Sun, 13 Feb 2011 22:11:28 +0000 (22:11 +0000)]
sisfb: add support for XGI Z9 DDR2 POST

Add support for ZGI Z9 DDR2 POST. The init sequence is from XGI's
xgifb driver.

Tested with ARM board using a PCI card with XGI Z9s and 32 MB DDR2
memory. After a cold reset the POST succeeds.

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: Thomas Winischhofer <thomas@winischhofer.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosisfb: move XGI POST DDR2 bootup code into subroutines
Aaro Koskinen [Sun, 13 Feb 2011 22:11:27 +0000 (22:11 +0000)]
sisfb: move XGI POST DDR2 bootup code into subroutines

Move DDR2 register setting code into separate subroutines. No changes
in functionality.

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: Thomas Winischhofer <thomas@winischhofer.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosisfb: add RAM type detection for XGI Z9
Aaro Koskinen [Sun, 13 Feb 2011 22:11:26 +0000 (22:11 +0000)]
sisfb: add RAM type detection for XGI Z9

Detect the XGI Z9 RAM type as "documented" by the XGI's xgifb driver.

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: Thomas Winischhofer <thomas@winischhofer.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosisfb: add subroutine for detecting XGI Z9
Aaro Koskinen [Sun, 13 Feb 2011 22:11:25 +0000 (22:11 +0000)]
sisfb: add subroutine for detecting XGI Z9

Z7 and Z9 have the same PCI ID, so additional checking is needed to
detect Z9. The method was "documented" in XGI's xgifb driver.

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: Thomas Winischhofer <thomas@winischhofer.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosisfb: move XGI POST RAM type detection into a subroutine
Aaro Koskinen [Sun, 13 Feb 2011 22:11:24 +0000 (22:11 +0000)]
sisfb: move XGI POST RAM type detection into a subroutine

Move XGI POST RAM type detection into a separate subroutine to make
further code changes easier. No changes in functionality

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: Thomas Winischhofer <thomas@winischhofer.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agosisfb: POST should fail if R/W test fails
Aaro Koskinen [Sun, 13 Feb 2011 22:11:23 +0000 (22:11 +0000)]
sisfb: POST should fail if R/W test fails

Currently there is no indication if R/W test fails during POST. This can
happen e.g. when user plugs in a card with unsupported type of memory.
Since the driver will be unusable, it's better to fail the whole POST
if the memory cannot be configured properly.

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: Thomas Winischhofer <thomas@winischhofer.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agovmlfb: use list_move() instead of list_del()/list_add() combination
Kirill A. Shutemov [Tue, 15 Mar 2011 22:53:19 +0000 (22:53 +0000)]
vmlfb: use list_move() instead of list_del()/list_add() combination

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agoomap: use list_move() instead of list_del()/list_add() combination
Kirill A. Shutemov [Tue, 15 Mar 2011 22:53:18 +0000 (22:53 +0000)]
omap: use list_move() instead of list_del()/list_add() combination

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Acked-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agomailmap: Update for OMAP DSS developers.
Paul Mundt [Tue, 22 Mar 2011 05:31:35 +0000 (14:31 +0900)]
mailmap: Update for OMAP DSS developers.

TI folks seem to have a rather schizophrenic relationship with author
naming conventions, stub in mailmap entries to match with the sign-off
convention.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agoMerge branch 'for-paul' of git://gitorious.org/linux-omap-dss2/linux
Paul Mundt [Tue, 22 Mar 2011 05:27:36 +0000 (14:27 +0900)]
Merge branch 'for-paul' of git://gitorious.org/linux-omap-dss2/linux

Conflicts:
arch/arm/mach-omap2/board-overo.c

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
9 years agoMerge branch 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6
Linus Torvalds [Mon, 21 Mar 2011 22:55:26 +0000 (15:55 -0700)]
Merge branch 'kbuild' of git://git./linux/kernel/git/mmarek/kbuild-2.6

* 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6:
  kbuild: Make DEBUG_SECTION_MISMATCH selectable, but not on by default
  genksyms: Regenerate lexer and parser
  genksyms: Track changes to enum constants
  genksyms: simplify usage of find_symbol()
  genksyms: Add helpers for building string lists
  genksyms: Simplify printing of symbol types
  genksyms: Simplify lexer
  genksyms: Do not paste the bison header file to lex.c
  modpost: fix trailing comma
  KBuild: silence "'scripts/unifdef' is up to date."
  kbuild: Add extra gcc checks
  kbuild: reenable section mismatch analysis
  unifdef: update to upstream version 2.5

9 years agoReduce sequential pointer derefs in scsi_error.c and reduce size as well
Jesper Juhl [Mon, 21 Mar 2011 19:47:31 +0000 (20:47 +0100)]
Reduce sequential pointer derefs in scsi_error.c and reduce size as well

This patch reduces the number of sequential pointer derefs in
drivers/scsi/scsi_error.c

This has been submitted a number of times over a couple of years.  I
believe this version adresses all comments it has gathered over time.
Please apply or reject with a reason.

The benefits are:

 - makes the code easier to read.  Lots of sequential derefs of the same
   pointers is not easy on the eye.

 - theoretically at least, just dereferencing the pointers once can
   allow the compiler to generally slightly faster code, so in theory
   this could also be a micro speed optimization.

 - reduces size of object file (tiny effect: on x86-64, in at least one
   configuration, the text size decreased from 9439 bytes to 9400)

 - removes some pointless (mostly trailing) whitespace.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agomatroxfb: remove incorrect Matrox G200eV support
Gary Hade [Mon, 21 Mar 2011 17:28:59 +0000 (10:28 -0700)]
matroxfb: remove incorrect Matrox G200eV support

Remove incorrect Matrox G200eV support that was previously added by
commit e3a1938805d2e81b27d3d348788644f3bad004f2

A serious issue with the incorrect G200eV support that reproduces on the
Matrox G200eV equipped IBM x3650 M2 is the total lack of text (login
banner, login prompt, etc) on the console when X is not running and
total lack of text on all of the virtual consoles after X is started.

Any concerns that the incorrect code (upstream since October 2008) has
been successfully used on non-IBM G200eV equipped system(s) appear to be
unwarranted.  In addition to the serious/non-intermittent nature of
issues that have been spotted on IBM systems, complete removal of the
incorrect code is clearly supported by the following Matrox (Yannick
Heneault) provided input:
 "It impossible that this patch should have work on a system.
 The patch only declare the G200eV as a regular G200 which is
 not case. Many registers are different, including at least the
 PLL programming sequence. If the G200eV is programmed like a
 regular G200, it will not display anything."

v1 - Initial patch that removed the incorrect code for _all_
     G200eV equipped systems.
v2 - Darrick Wong provided patch that blacklisted the incorrect
     code on G200eV equipped IBM systems leaving it enabled on
     all G200eV equipped non-IBM systems.
v3 - Same code changes included with v1 plus additional
     justification for complete removal of the incorrect code.

Signed-off-by: Gary Hade <garyhade@us.ibm.com>
Cc: Darrick J. Wong <djwong@us.ibm.com>
Cc: Krzysztof Helt <krzysztof.h1@wp.pl>
Cc: Petr Vandrovec <vandrove@vc.cvut.cz>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Yannick Heneault <yannick_heneault@matrox.com>
Cc: Christian Toutant <ctoutant@matrox.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoMerge branch 'for-linus' of git://oss.sgi.com/xfs/xfs
Linus Torvalds [Mon, 21 Mar 2011 21:24:56 +0000 (14:24 -0700)]
Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs

* 'for-linus' of git://oss.sgi.com/xfs/xfs: (23 commits)
  xfs: don't name variables "panic"
  xfs: factor agf counter updates into a helper
  xfs: clean up the xfs_alloc_compute_aligned calling convention
  xfs: kill support/debug.[ch]
  xfs: Convert remaining cmn_err() callers to new API
  xfs: convert the quota debug prints to new API
  xfs: rename xfs_cmn_err_fsblock_zero()
  xfs: convert xfs_fs_cmn_err to new error logging API
  xfs: kill xfs_fs_mount_cmn_err() macro
  xfs: kill xfs_fs_repair_cmn_err() macro
  xfs: convert xfs_cmn_err to xfs_alert_tag
  xfs: Convert xlog_warn to new logging interface
  xfs: Convert linux-2.6/ files to new logging interface
  xfs: introduce new logging API.
  xfs: zero proper structure size for geometry calls
  xfs: enable delaylog by default
  xfs: more sensible inode refcounting for ialloc
  xfs: stop using xfs_trans_iget in the RT allocator
  xfs: check if device support discard in xfs_ioc_trim()
  xfs: prevent leaking uninitialized stack memory in FSGEOMETRY_V1
  ...

9 years agoPrevent rt_sigqueueinfo and rt_tgsigqueueinfo from spoofing the signal code
Julien Tinnes [Fri, 18 Mar 2011 22:05:21 +0000 (15:05 -0700)]
Prevent rt_sigqueueinfo and rt_tgsigqueueinfo from spoofing the signal code

Userland should be able to trust the pid and uid of the sender of a
signal if the si_code is SI_TKILL.

Unfortunately, the kernel has historically allowed sigqueueinfo() to
send any si_code at all (as long as it was negative - to distinguish it
from kernel-generated signals like SIGILL etc), so it could spoof a
SI_TKILL with incorrect siginfo values.

Happily, it looks like glibc has always set si_code to the appropriate
SI_QUEUE, so there are probably no actual user code that ever uses
anything but the appropriate SI_QUEUE flag.

So just tighten the check for si_code (we used to allow any negative
value), and add a (one-time) warning in case there are binaries out
there that might depend on using other si_code values.

Signed-off-by: Julien Tinnes <jln@google.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt...
Linus Torvalds [Mon, 21 Mar 2011 21:13:48 +0000 (14:13 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/rostedt/linux-2.6-ktest

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-2.6-ktest:
  ktest: Add STOP_TEST_AFTER to stop the test after a period of time
  ktest: Monitor kernel while running of user tests
  ktest: Fix bug where the test would not end after failure
  ktest: Add BISECT_FILES to run git bisect on paths
  ktest: Add BISECT_SKIP
  ktest: Add manual bisect
  ktest: Handle kernels before make oldnoconfig
  ktest: Start failure timeout on panic too
  ktest: Print logfile name on failure

9 years agoMerge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelv...
Linus Torvalds [Mon, 21 Mar 2011 21:02:55 +0000 (14:02 -0700)]
Merge branch 'hwmon-for-linus' of git://git./linux/kernel/git/jdelvare/staging

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  hwmon: (ads1015) Make gain and datarate configurable
  hwmon: (ads1015) Drop dynamic attribute group
  hwmon: Add support for Texas Instruments ADS1015
  hwmon: New driver for SMSC SCH5627
  hwmon: (abituguru*) Update my email address
  hwmon: (lm75) Speed up detection
  hwmon: (lm75) Add detection of the National Semiconductor LM75A
  hp_accel: Fix driver name
  Move lis3lv02d drivers to drivers/misc
  Move hp_accel to drivers/platform/x86
  Let Kconfig handle lis3lv02d dependencies
  hwmon: (sht15) Fix integer overflow in humidity calculation
  hwmon: (sht15) Spelling fix
  hwmon: (w83795) Document pin mapping

9 years agopstore: use mount option instead sysfs to tweak kmsg_bytes
Luck, Tony [Fri, 18 Mar 2011 22:33:43 +0000 (15:33 -0700)]
pstore: use mount option instead sysfs to tweak kmsg_bytes

/sys/fs is a somewhat strange way to tweak what could more
obviously be tuned with a mount option.

Suggested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6
Linus Torvalds [Mon, 21 Mar 2011 17:06:51 +0000 (10:06 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/viro/vfs-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
  FS: lookup_mnt() is only used in the core fs routines now
  bfs: fix bitmap size argument to find_first_zero_bit()
  fs: Use BUG_ON(!mnt) at dentry_open().
  fs: devpts_pty_new() return -ENOMEM if dentry allocation failed
  nfs: lock() vs unlock() typo
  pstore: fix leaking ->i_private
  introduce sys_syncfs to sync a single file system
  Small typo fix...
  Filesystem: fifo: Fixed coding style issue.
  fs/inode: Fix kernel-doc format for inode_init_owner
  select: remove unused MAX_SELECT_SECONDS
  vfs: cleanup do_vfs_ioctl()

9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394...
Linus Torvalds [Mon, 21 Mar 2011 17:05:22 +0000 (10:05 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/ieee1394/linux1394-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
  firewire: core: ignore link-active bit of new nodes, fix device recognition
  firewire: sbp2: revert obsolete 'fix stall with "Unsolicited response"'
  firewire: core: increase default SPLIT_TIMEOUT value
  firewire: ohci: Misleading kfree in ohci.c::pci_probe/remove
  firewire: ohci: omit IntEvent.busReset check rom AT queueing
  firewire: ohci: prevent starting of iso contexts with empty queue
  firewire: ohci: prevent iso completion callbacks after context stop
  firewire: core: rename some variables
  firewire: nosy: should work on Power Mac G4 PCI too
  firewire: core: fix card->reset_jiffies overflow
  firewire: cdev: remove unneeded reference
  firewire: cdev: always wait for outbound transactions to complete
  firewire: cdev: remove unneeded idr_find() from complete_transaction()
  firewire: ohci: log dead DMA contexts

9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/parisc-2.6
Linus Torvalds [Mon, 21 Mar 2011 17:04:53 +0000 (10:04 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/jejb/parisc-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/parisc-2.6:
  [PARISC] Convert to new irq_chip functions
  [PARISC] fix per-cpu flag problem in the cpu affinity checkers
  [PARISC] fix vmap flush/invalidate
  eliminate special FLUSH flag from page table
  parisc: flush pages through tmpalias space

9 years agohwmon: (ads1015) Make gain and datarate configurable
Dirk Eibach [Mon, 21 Mar 2011 16:59:37 +0000 (17:59 +0100)]
hwmon: (ads1015) Make gain and datarate configurable

Configuration for ads1015 gain and datarate is possible via
devicetree or platform data.

This is a followup patch to previous ads1015 patches on Jean Delvares
tree.

Signed-off-by: Dirk Eibach <eibach@gdsys.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agohwmon: (ads1015) Drop dynamic attribute group
Jean Delvare [Mon, 21 Mar 2011 16:59:37 +0000 (17:59 +0100)]
hwmon: (ads1015) Drop dynamic attribute group

It is cheaper to handle attributes individually.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Dirk Eibach <eibach@gdsys.de>
9 years agohwmon: Add support for Texas Instruments ADS1015
Dirk Eibach [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
hwmon: Add support for Texas Instruments ADS1015

Signed-off-by: Dirk Eibach <eibach@gdsys.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agohwmon: New driver for SMSC SCH5627
Hans de Goede [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
hwmon: New driver for SMSC SCH5627

SMSC SCH5627 Super I/O chips include complete hardware monitoring
capabilities. They can monitor up to 5 voltages, 4 fans and 8
temperatures.

The hardware monitoring part of the SMSC SCH5627 is accessed by talking
through an embedded microcontroller. An application note describing the
protocol for communicating with the microcontroller is available upon
request. Please mail me if you want a copy.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agohwmon: (abituguru*) Update my email address
Hans de Goede [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
hwmon: (abituguru*) Update my email address

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agohwmon: (lm75) Speed up detection
Jean Delvare [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
hwmon: (lm75) Speed up detection

Make the LM75/LM75A device detection faster:

* Don't read the current temperature value when we don't use it.
* Check for unused bits in the configuration register as soon as we
  have read its value.
* Don't use word reads, not all devices support this, and some which
  don't misbehave when you try.
* Check for cycling register values every 40 register addresses
  instead of every 8, it's 5 times faster and just as efficient.

Some of these improvements come straight from the user-space
sensors-detect script, so both detection routines are in line now.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Len Sorensen <lsorense@csclub.uwaterloo.ca>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
9 years agohwmon: (lm75) Add detection of the National Semiconductor LM75A
Len Sorensen [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
hwmon: (lm75) Add detection of the National Semiconductor LM75A

Add support for detection of the National Semiconductor LM75A using the ID
register value.

Signed-off-by: Len Sorensen <lsorense@csclub.uwaterloo.ca>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agohp_accel: Fix driver name
Jean Delvare [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
hp_accel: Fix driver name

I suspect that the "lis3lv02d" driver name is a legacy from before
the split into several modules. Use a specific name for the hp_accel
driver, for better error messages and easier investigation of issues.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Acked-by: Eric Piel <eric.piel@tremplin-utc.net>
Acked-by: Jonathan Cameron <jic23@cam.ac.uk>
Tested-by: Eric Piel <eric.piel@tremplin-utc.net>
Tested-by: Takashi Iwai <tiwai@suse.de>
9 years agoMove lis3lv02d drivers to drivers/misc
Jean Delvare [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
Move lis3lv02d drivers to drivers/misc

The lis3lv02d drivers aren't hardware monitoring drivers, so the don't
belong to drivers/hwmon. Move them to drivers/misc, short of a better
home.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Acked-by: Eric Piel <eric.piel@tremplin-utc.net>
Acked-by: Jonathan Cameron <jic23@cam.ac.uk>
Tested-by: Eric Piel <eric.piel@tremplin-utc.net>
Tested-by: Takashi Iwai <tiwai@suse.de>
9 years agoMove hp_accel to drivers/platform/x86
Jean Delvare [Mon, 21 Mar 2011 16:59:36 +0000 (17:59 +0100)]
Move hp_accel to drivers/platform/x86

The hp_accel driver isn't a hardware monitoring driver, so it doesn't
belong to drivers/hwmon. Move it to drivers/platform/x86, assuming HP
doesn't ship non-x86 laptops.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Acked-by: Eric Piel <eric.piel@tremplin-utc.net>
Acked-by: Jonathan Cameron <jic23@cam.ac.uk>
Tested-by: Eric Piel <eric.piel@tremplin-utc.net>
Tested-by: Takashi Iwai <tiwai@suse.de>
9 years agoLet Kconfig handle lis3lv02d dependencies
Jean Delvare [Mon, 21 Mar 2011 16:59:35 +0000 (17:59 +0100)]
Let Kconfig handle lis3lv02d dependencies

The dependencies between the various lis3lv02d drivers make it
impossible to split them to different directories, while we really
want to do this. Move handling of dependencies from Makefile to
Kconfig, to make the move possible at all.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Acked-by: Eric Piel <eric.piel@tremplin-utc.net>
Acked-by: Jonathan Cameron <jic23@cam.ac.uk>
Tested-by: Eric Piel <eric.piel@tremplin-utc.net>
Tested-by: Takashi Iwai <tiwai@suse.de>
9 years agohwmon: (sht15) Fix integer overflow in humidity calculation
Vivien Didelot [Mon, 21 Mar 2011 16:59:35 +0000 (17:59 +0100)]
hwmon: (sht15) Fix integer overflow in humidity calculation

An integer overflow occurs in the calculation of RHlinear when the
relative humidity is greater than around 30%. The consequence is a subtle
(but noticeable) error in the resulting humidity measurement.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: stable@kernel.org
Cc: Jonathan Cameron <jic23@cam.ac.uk>
9 years agohwmon: (sht15) Spelling fix
Justin P. Mattock [Mon, 21 Mar 2011 16:59:35 +0000 (17:59 +0100)]
hwmon: (sht15) Spelling fix

Remove one too many "n" in a word.

Signed-off-by: Justin P. Mattock <justinmattock@gmail.com>
Acked-by: Jean Delvare <khali@linux-fr.org>
9 years agohwmon: (w83795) Document pin mapping
Jean Delvare [Mon, 21 Mar 2011 16:59:35 +0000 (17:59 +0100)]
hwmon: (w83795) Document pin mapping

Apparently users are interested in this information, so let's provide
it.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
9 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-2.6
Linus Torvalds [Mon, 21 Mar 2011 16:53:04 +0000 (09:53 -0700)]
Merge git://git./linux/kernel/git/davem/sparc-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-2.6:
  sparc: Add {open_by,name_to}_handle_at and clock_adjtime syscalls.
  sparc: Implement of_iomap().
  sparc: Implement of_address_to_resource().
  sparc: Provide NO_IRQ definition.

9 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Mon, 21 Mar 2011 16:52:27 +0000 (09:52 -0700)]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  powerpc/ptrace: Remove BUG_ON when full register set not available
  powerpc: Factoring mpic cpu id fetching into a function
  powerpc: Make MPIC honor the "pic-no-reset" device tree property
  powerpc: Document the Open PIC device tree binding
  powerpc/pci: Fix crash in PCI code on ppc64 when matching device nodes

9 years agoFS: lookup_mnt() is only used in the core fs routines now
David Howells [Mon, 21 Mar 2011 14:28:58 +0000 (14:28 +0000)]
FS: lookup_mnt() is only used in the core fs routines now

lookup_mnt() is only used in the core fs routines now, so it doesn't need to
be globally declared anymore.  It isn't exported to modules at the moment, so
nothing that can be modularised seems to be using it.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agobfs: fix bitmap size argument to find_first_zero_bit()
Akinobu Mita [Mon, 21 Mar 2011 12:32:53 +0000 (08:32 -0400)]
bfs: fix bitmap size argument to find_first_zero_bit()

The usage of find_first_zero_bit() in bfs_create() is wrong for two
reasons.

The bitmap size argument to find_first_zero_bit() is info->si_lasti but
the correct bitmap size is info->si_lasti + 1 as info->si_lasti is the
last valid index in info->si_imap bitmap.

Another problem is that it is impossible to detect that info->si_imap
bitmap is full because there is an off-by-one bug in the return value
check for find_first_zero_bit().  If no zero bits exist in info->si_imap,
find_first_zero_bit() returns info->si_lasti.  But the check can't catch
it due to the off-by-one.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: "Tigran A. Aivazian" <tigran@aivazian.fsnet.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agokbuild: Make DEBUG_SECTION_MISMATCH selectable, but not on by default
Michal Marek [Mon, 21 Mar 2011 09:44:30 +0000 (10:44 +0100)]
kbuild: Make DEBUG_SECTION_MISMATCH selectable, but not on by default

CONFIG_DEBUG_SECTION_MISMATCH has also runtime effects due to the
-fno-inline-functions-called-once compiler flag, so forcing it on
everyone is not a good idea.

Signed-off-by: Michal Marek <mmarek@suse.cz>
9 years agofs: Use BUG_ON(!mnt) at dentry_open().
Tetsuo Handa [Wed, 19 Jan 2011 12:08:41 +0000 (21:08 +0900)]
fs: Use BUG_ON(!mnt) at dentry_open().

dentry_open() requires callers to pass a valid vfsmount.

Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agofs: devpts_pty_new() return -ENOMEM if dentry allocation failed
Andrey Vagin [Mon, 7 Feb 2011 21:14:52 +0000 (00:14 +0300)]
fs: devpts_pty_new() return -ENOMEM if dentry allocation failed

In this case nobody can open a slave point, so will be better return
from devpts_pty_new()

Now we should not check error code from d_find_alias() in
devpts_pty_kill(), because the dentry exists all times.

Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agonfs: lock() vs unlock() typo
Dan Carpenter [Sun, 20 Mar 2011 11:22:07 +0000 (14:22 +0300)]
nfs: lock() vs unlock() typo

These should be spin_unlock() instead of spin_lock().  It's a typo.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agopstore: fix leaking ->i_private
Tony Luck [Fri, 18 Mar 2011 18:44:48 +0000 (11:44 -0700)]
pstore: fix leaking ->i_private

Move kfree() of i_private out of ->unlink() and into ->evict_inode()

Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agointroduce sys_syncfs to sync a single file system
Sage Weil [Thu, 10 Mar 2011 19:31:30 +0000 (11:31 -0800)]
introduce sys_syncfs to sync a single file system

It is frequently useful to sync a single file system, instead of all
mounted file systems via sync(2):

 - On machines with many mounts, it is not at all uncommon for some of
   them to hang (e.g. unresponsive NFS server).  sync(2) will get stuck on
   those and may never get to the one you do care about (e.g., /).
 - Some applications write lots of data to the file system and then
   want to make sure it is flushed to disk.  Calling fsync(2) on each
   file introduces unnecessary ordering constraints that result in a large
   amount of sub-optimal writeback/flush/commit behavior by the file
   system.

There are currently two ways (that I know of) to sync a single super_block:

 - BLKFLSBUF ioctl on the block device: That also invalidates the bdev
   mapping, which isn't usually desirable, and doesn't work for non-block
   file systems.
 - 'mount -o remount,rw' will call sync_filesystem as an artifact of the
   current implemention.  Relying on this little-known side effect for
   something like data safety sounds foolish.

Both of these approaches require root privileges, which some applications
do not have (nor should they need?) given that sync(2) is an unprivileged
operation.

This patch introduces a new system call syncfs(2) that takes an fd and
syncs only the file system it references.  Maybe someday we can

 $ sync /some/path

and not get

 sync: ignoring all arguments

The syscall is motivated by comments by Al and Christoph at the last LSF.
syncfs(2) seems like an appropriate name given statfs(2).

A similar ioctl was also proposed a while back, see
http://marc.info/?l=linux-fsdevel&m=127970513829285&w=2

Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agoSmall typo fix...
Holger Hans Peter Freyther [Thu, 24 Feb 2011 16:46:49 +0000 (17:46 +0100)]
Small typo fix...

Hi,

I was backporting the coredump over pipe feature and noticed this small typo,
I wish I would have something bigger to contribute...

>From 15d6080e0ed4267da103c706917a33b1015e8804 Mon Sep 17 00:00:00 2001
From: Holger Hans Peter Freyther <holger@moiji-mobile.com>
Date: Thu, 24 Feb 2011 17:42:50 +0100
Subject: [PATCH] fs: Fix a small typo in the comment

The function is called umh_pipe_setup not uhm_pipe_setup.

Signed-off-by: Holger Hans Peter Freyther <holger@moiji-mobile.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agoFilesystem: fifo: Fixed coding style issue.
David Jenni [Wed, 23 Feb 2011 15:51:05 +0000 (16:51 +0100)]
Filesystem: fifo: Fixed coding style issue.

Fixed coding style issue.

Signed-off-by: David Jenni <dave.j@gmx.ch>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agofs/inode: Fix kernel-doc format for inode_init_owner
Ben Hutchings [Tue, 15 Feb 2011 12:48:09 +0000 (12:48 +0000)]
fs/inode: Fix kernel-doc format for inode_init_owner

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agoselect: remove unused MAX_SELECT_SECONDS
Namhyung Kim [Fri, 21 Jan 2011 04:59:59 +0000 (13:59 +0900)]
select: remove unused MAX_SELECT_SECONDS

Remove the leftover from the commit 8ff3e8e85fa6 ("select:
switch select() and poll() over to hrtimers").

Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Acked-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agovfs: cleanup do_vfs_ioctl()
Namhyung Kim [Sun, 16 Jan 2011 15:48:17 +0000 (00:48 +0900)]
vfs: cleanup do_vfs_ioctl()

Move declaration of 'inode' to beginning of the function. Since it
is referenced directly or indirectly (in case of FIFREEZE/FITHAW/
FS_IOC_FIEMAP) it's not harmful IMHO. And remove unnecessary casts
using 'argp' instead.

Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
9 years agoMerge branch 'trivial' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6
Linus Torvalds [Mon, 21 Mar 2011 01:14:55 +0000 (18:14 -0700)]
Merge branch 'trivial' of git://git./linux/kernel/git/mmarek/kbuild-2.6

* 'trivial' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6: (25 commits)
  video: change to new flag variable
  scsi: change to new flag variable
  rtc: change to new flag variable
  rapidio: change to new flag variable
  pps: change to new flag variable
  net: change to new flag variable
  misc: change to new flag variable
  message: change to new flag variable
  memstick: change to new flag variable
  isdn: change to new flag variable
  ieee802154: change to new flag variable
  ide: change to new flag variable
  hwmon: change to new flag variable
  dma: change to new flag variable
  char: change to new flag variable
  fs: change to new flag variable
  xtensa: change to new flag variable
  um: change to new flag variables
  s390: change to new flag variable
  mips: change to new flag variable
  ...

Fix up trivial conflict in drivers/hwmon/Makefile

9 years agoMerge branch 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6
Linus Torvalds [Mon, 21 Mar 2011 01:13:09 +0000 (18:13 -0700)]
Merge branch 'misc' of git://git./linux/kernel/git/mmarek/kbuild-2.6

* 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6:
  scripts/extract-ikconfig: add xz compression support
  kbuild: add GNU GLOBAL tags generation
  setlocalversion: update mercurial tag parsing

9 years agoMerge branch 'packaging' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek...
Linus Torvalds [Mon, 21 Mar 2011 01:12:42 +0000 (18:12 -0700)]
Merge branch 'packaging' of git://git./linux/kernel/git/mmarek/kbuild-2.6

* 'packaging' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6:
  kbuild: Add make tarxz-pkg build option

9 years agoMerge branch 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvar...
Linus Torvalds [Mon, 21 Mar 2011 00:54:13 +0000 (17:54 -0700)]
Merge branch 'i2c-for-linus' of git://git./linux/kernel/git/jdelvare/staging

* 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  i2c-i801: SMBus patch for Intel DH89xxCC DeviceIDs
  i2c: Drop i2c_adapter.id
  i2c: Deprecate i2c_driver.attach_adapter and .detach_adapter
  i2c-dev: Use standard bus notification mechanism
  i2c: Export i2c_for_each_dev
  i2c: Get rid of <linux/i2c-id.h>
  i2c: Minor fixes to upgrading-clients document
  i2c: make i2c_get_adapter prototype clearer
  i2c: Fix typo in instantiating-devices document
  i2c-boardinfo: Fix typo in comment

9 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile
Linus Torvalds [Mon, 21 Mar 2011 00:53:50 +0000 (17:53 -0700)]
Merge git://git./linux/kernel/git/cmetcalf/linux-tile

* git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile:
  asm-generic: support clock_adjtime() in <asm-generic/unistd.h>
  arch/tile: fix futex sanitization definition/prototype mismatch

9 years agor8169: fix a bug in rtl8169_init_phy()
Eric Dumazet [Sat, 19 Mar 2011 15:39:11 +0000 (16:39 +0100)]
r8169: fix a bug in rtl8169_init_phy()

commit 54405cde7624 (r8169: support control of advertising.)
introduced a bug in rtl8169_init_phy()

Reported-and-tested-by: Piotr Hosowicz <piotr@hosowicz.com>
Reported-and-tested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by:: Oliver Neukum <oliver@neukum.org>
Cc: Francois Romieu <romieu@fr.zoreil.com>
Tested-by: Anca Emanuel <anca.emanuel@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agopowerpc/ptrace: Remove BUG_ON when full register set not available
Mike Wolf [Mon, 21 Mar 2011 00:14:53 +0000 (11:14 +1100)]
powerpc/ptrace: Remove BUG_ON when full register set not available

In some cases during a threaded core dump not all the threads will have
a full register set. This happens when the signal causing the core dump
races with a thread exiting.  The race happens when the exiting thread
has entered the kernel for the last time before the signal arrives, but
doesn't get far enough through the exit code to avoid being included
in the core dump.

So we get a thread included in the core dump which is never going to go
out to userspace again and only has a partial register set recorded

Normally we would catch each thread as it is about to go into userspace
and capture the full register set then.

However, this exiting thread is never going to go out to userspace
again, so we have no way to capture its full register set.  It doesn't
really matter, though, as this is a thread which is effectively
already dead.

So instead of hitting a BUG() in this case (a really bad choice of
action in the first place), we use a poison value for the register
values.

[BenH]: Some cosmetic/stylistic changes and fix build on ppc32

Signed-off-by: Mike Wolf <mjw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
9 years agopowerpc: Factoring mpic cpu id fetching into a function
Meador Inge [Mon, 14 Mar 2011 10:01:07 +0000 (10:01 +0000)]
powerpc: Factoring mpic cpu id fetching into a function

The following code snippet:

unsigned int cpu = 0;
if (mpic->flags & MPIC_PRIMARY)
cpu = hard_smp_processor_id();

is seen in several places in the 'mpic.c' code.  This changeset factors
that pattern out into a helper function called 'mpic_processor_id'.

Signed-off-by: Meador Inge <meador_inge@mentor.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
9 years agopowerpc: Make MPIC honor the "pic-no-reset" device tree property
Meador Inge [Mon, 14 Mar 2011 10:01:06 +0000 (10:01 +0000)]
powerpc: Make MPIC honor the "pic-no-reset" device tree property

This property, defined in the Open PIC binding, tells the kernel not to use
the reset bit in the global configuration register.  Additionally, its
presence mandates that only sources which are actually used (i.e. appear in
the device tree) should have their VECPRI bits initialized.

Although, "pic-no-reset" can be used for the same use cases that
"protected-sources" is covering, the "protected-sources" implementation was
left completely intact.  This is a more pragmatic approach as there are
already several existing systems which use protected sources.  If
"pic-no-reset" *and* "protected-sources" are both used, however, then
"pic-no-reset" takes precedence in terms of the init behavior and the
sanity checks done by protected sources will still take place.

Signed-off-by: Meador Inge <meador_inge@mentor.com>
Cc: Hollis Blanchard <hollis_blanchard@mentor.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
9 years agopowerpc: Document the Open PIC device tree binding
Meador Inge [Mon, 14 Mar 2011 10:01:05 +0000 (10:01 +0000)]
powerpc: Document the Open PIC device tree binding

This binding documents several properties that have been in use for quite
some time, and adds one new property 'pic-no-reset', which controls the
runtime initialization behavior of the PIC.  More specifically, the presence
of 'pic-no-reset' mandates that the PIC shall not be reset during runtime
initialization and that any initialization related to interrupt sources
shall be limited to sources explicitly referenced in the device tree.  This
functionality is useful in AMP systems where multiple OSes are sharing the
PIC and the reinitialization of the PIC can interfere with OSes that are
already up and running.

The interrupt specifier definition is based off of Stuart Yoder's FSL MPIC
binding.

Signed-off-by: Meador Inge <meador_inge@mentor.com>
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: Stuart Yoder <stuart.yoder@freescale.com>
Cc: Hollis Blanchard <hollis_blanchard@mentor.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
9 years agopowerpc/pci: Fix crash in PCI code on ppc64 when matching device nodes
Benjamin Herrenschmidt [Sun, 20 Mar 2011 23:57:57 +0000 (10:57 +1100)]
powerpc/pci: Fix crash in PCI code on ppc64 when matching device nodes

Commit b5d937de0367d26f65b9af1aef5f2c34c1939be0 has a bug which causes
basically a NULL dereference in the PCI code during boot on ppc64
machines.

fetch_dev_dn() is called when dev->dev.of_node is NULL, so using that
as the starting point for the search makes no sense. It should instead
start from the device node of the PHB.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
9 years agofirewire: core: ignore link-active bit of new nodes, fix device recognition
Stefan Richter [Mon, 14 Mar 2011 23:08:41 +0000 (00:08 +0100)]
firewire: core: ignore link-active bit of new nodes, fix device recognition

Like the older ieee1394 core driver, firewire-core skipped scanning of
any new node whose PHY sent a self ID without "link active" bit.  If a
device had this bit off mistakenly, it meant that it was inaccessible to
kernel drivers with the old IEEE 1394 driver stack but could still be
accessed by userspace drivers through the raw1394 interface.

But with firewire-core, userspace drivers don't get to see such buggy
devices anymore.  This is effectively a driver regression since this
device bug is otherwise harmless.

We now attempt to scan all devices, even repeaters that don't have a
link or powered-down devices that have everything but their PHY shut
down when plugged in.  This results in futile repeated scanning attempts
in case of such devices that really don't have an active link, but this
doesn't hurt since recent workqueue infrastructure lets us run more
concurrent scanning jobs than we can shake a stick at.

This should fix accessibility of Focusrite Saffire PRO 26 I/O:
http://sourceforge.net/mailarchive/forum.php?thread_name=20110314215622.5c751bb0%40stein&forum_name=ffado-user

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
9 years agofirewire: sbp2: revert obsolete 'fix stall with "Unsolicited response"'
Stefan Richter [Mon, 14 Mar 2011 23:04:42 +0000 (00:04 +0100)]
firewire: sbp2: revert obsolete 'fix stall with "Unsolicited response"'

Now that firewire-core sets the local node's SPLIT_TIMEOUT to 2 seconds
per default, commit a481e97d3cdc40b9d58271675bd4f0abb79d4872 is no
longer required.

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
9 years agofirewire: core: increase default SPLIT_TIMEOUT value
Clemens Ladisch [Mon, 7 Mar 2011 10:21:15 +0000 (11:21 +0100)]
firewire: core: increase default SPLIT_TIMEOUT value

The SPLIT_TIMEOUT mechanism is intended to detect requests that somehow
got lost.  However, when the timeout value is too low, transactions that
could have been completed successfully will be cancelled.  Furthermore,
there are chips whose firmwares ignore the configured split timeout and
send late split response; known examples are the DM1x00 (BeBoB), TCD22x0
(DICE), and some OXUF936QSE firmwares.

This patch changes the default timeout to two seconds, which happens to
be the default on other OSes, too.

Actual lost requests are extremely rare, so there should be no practical
downside to increasing the split timeout even on devices that work
correctly.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
9 years agoi2c-i801: SMBus patch for Intel DH89xxCC DeviceIDs
Seth Heasley [Sun, 20 Mar 2011 13:50:53 +0000 (14:50 +0100)]
i2c-i801: SMBus patch for Intel DH89xxCC DeviceIDs

Add the SMBus Controller DeviceIDs for the Intel DH89xxCC PCH.

Signed-off-by: Seth Heasley <seth.heasley@intel.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: Drop i2c_adapter.id
Jean Delvare [Sun, 20 Mar 2011 13:50:53 +0000 (14:50 +0100)]
i2c: Drop i2c_adapter.id

There is no user left of i2c_adapter.id, so we can get rid of it.
Finally! :)

Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: Deprecate i2c_driver.attach_adapter and .detach_adapter
Jean Delvare [Sun, 20 Mar 2011 13:50:53 +0000 (14:50 +0100)]
i2c: Deprecate i2c_driver.attach_adapter and .detach_adapter

The last legitimate user of i2c_driver.attach_adapter and
.detach_adapter is gone, so we can finally deprecate these callbacks.
The last few drivers which still use these will have to be updated to
make use of standard I2C device instantiation ways instead.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c-dev: Use standard bus notification mechanism
Jean Delvare [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c-dev: Use standard bus notification mechanism

Use the standard driver core mechanism to keep track of i2c adapters
present on the system: i2c_for_each_dev and a notifier. This will let
us deprecate and ultimately remove the legacy attach_adapter and
detach_adapter callbacks in i2c_driver.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: Export i2c_for_each_dev
Jean Delvare [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c: Export i2c_for_each_dev

Introduce i2c_for_each_dev(), an i2c device iterator with proper
locking for use by i2c-dev. This is needed so that we can get rid of
the attach_adapter and detach_adapter legacy callback functions.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: Get rid of <linux/i2c-id.h>
Jean Delvare [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c: Get rid of <linux/i2c-id.h>

The last remaining ID in <linux/i2c-id.h> is no longer used anywhere,
so we can finally get rid of it.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: Minor fixes to upgrading-clients document
Jean Delvare [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c: Minor fixes to upgrading-clients document

* Typical legacy drivers implemented method .detach_client, not
  .detach_adapter.
* Drop all references to __devexit, as i2c drivers shouldn't use it.

Cc: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: make i2c_get_adapter prototype clearer
Jean Delvare [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c: make i2c_get_adapter prototype clearer

Rename the parameter of i2c_get_adapter() to "nr", to make it clear we
are passing an adapter number and not an adapter ID (which have gone
away by now.)

Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoi2c: Fix typo in instantiating-devices document
Roman Fietze [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c: Fix typo in instantiating-devices document

The struct i2c_board_info member holding the name is "type", not
"name".

Signed-off-by: Roman Fietze <roman.fietze@telemotive.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: stable@kernel.org
9 years agoi2c-boardinfo: Fix typo in comment
Wolfram Sang [Sun, 20 Mar 2011 13:50:52 +0000 (14:50 +0100)]
i2c-boardinfo: Fix typo in comment

Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
Linus Torvalds [Sun, 20 Mar 2011 05:27:06 +0000 (22:27 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/dtor/input

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: (64 commits)
  Input: tsc2005 - remove 'disable' sysfs attribute
  Input: tsc2005 - add open/close
  Input: tsc2005 - handle read errors from SPI layer
  Input: tsc2005 - do not rearm timer in hardirq handler
  Input: tsc2005 - don't use work for 'pen up' handling
  Input: tsc2005 - do not use 0 in place of NULL
  Input: tsc2005 - use true/false for boolean variables
  Input: tsc2005 - hide selftest attribute if we can't reset
  Input: tsc2005 - rework driver initialization code
  Input: tsc2005 - set up bus type in input device
  Input: tsc2005 - set up parent device
  Input: tsc2005 - clear driver data after unbinding
  Input: tsc2005 - add module description
  Input: tsc2005 - remove driver banner message
  Input: tsc2005 - remove incorrect module alias
  Input: tsc2005 - convert to using dev_pm_ops
  Input: tsc2005 - use spi_get/set_drvdata()
  Input: introduce tsc2005 driver
  Input: xen-kbdfront - move to drivers/input/misc
  Input: xen-kbdfront - add grant reference for shared page
  ...

9 years agoasm-generic: support clock_adjtime() in <asm-generic/unistd.h>
Chris Metcalf [Sat, 19 Mar 2011 15:47:53 +0000 (11:47 -0400)]
asm-generic: support clock_adjtime() in <asm-generic/unistd.h>

A syscall was added without being added to asm-generic, which
makes tile (and presumably score and unicore32) break.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: Chen Liqin <liqin.chen@sunplusct.com>
Cc: Lennox Wu <lennox.wu@gmail.com>
9 years agoarch/tile: fix futex sanitization definition/prototype mismatch
Chris Metcalf [Sat, 19 Mar 2011 15:45:34 +0000 (11:45 -0400)]
arch/tile: fix futex sanitization definition/prototype mismatch

Commit 8d7718aa082aaf30a0b4989e1f04858952f941bc changed "int"
to "u32" in the prototypes but not the definition.
I missed this when I saw the patch go by on LKML.

We cast "u32 *" to "int *" since we are tying into the underlying
atomics framework, and atomic_t uses int as its value type.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Reviewed-by: Michel Lespinasse <walken@google.com>
9 years agoMerge branch 'next' into for-linus
Dmitry Torokhov [Sat, 19 Mar 2011 06:38:50 +0000 (23:38 -0700)]
Merge branch 'next' into for-linus

9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke...
Linus Torvalds [Sat, 19 Mar 2011 05:33:38 +0000 (22:33 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/ryusuke/nilfs2

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
  nilfs2: move NILFS_SUPER_MAGIC to linux/magic.h
  nilfs2: get rid of nilfs_sb_info structure
  nilfs2: use sb instance instead of nilfs_sb_info struct
  nilfs2: get rid of sc_sbi back pointer
  nilfs2: move log writer onto nilfs object
  nilfs2: move next generation counter into nilfs object
  nilfs2: move s_inode_lock and s_dirty_files into nilfs object
  nilfs2: move parameters on nilfs_sb_info into nilfs object
  nilfs2: move mount options to nilfs object
  nilfs2: record used amount of each checkpoint in checkpoint list
  nilfs2: optimize rec_len functions
  nilfs2: append blocksize info to warnings during loading super blocks
  nilfs2: add compat ioctl
  nilfs2: implement FS_IOC_GETFLAGS/SETFLAGS/GETVERSION
  nilfs2: tighten restrictions on inode flags
  nilfs2: mark S_NOATIME on inodes only if NOATIME attribute is set
  nilfs2: use common file attribute macros
  nilfs2: add free entries count only if clear bit operation succeeded
  nilfs2: decrement inodes count only if raw inode was successfully deleted

9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cjb/mmc
Linus Torvalds [Sat, 19 Mar 2011 05:32:40 +0000 (22:32 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/cjb/mmc

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cjb/mmc: (53 commits)
  mmc: dw_mmc: support mmc power control with regulator
  mmc: dw_mmc: fix suspend/resume operation
  mmc: dw_mmc: add quirks for unreliable card detect, and capabilities
  mmc: tmio: fix address in kunmap_atomic() calls
  mmc: core: reset card voltage after power off
  mmc: core: export function mmc_do_release_host()
  mmc: sdio: remember new card RCA when redetecting card
  mmc: dw_mmc: Remove set-but-unused variable.
  mmc: sdhci-esdhc-imx: add card detect on custom GPIO for mx25/35
  mmc: sdhci-esdhc: broken card detection is not a default quirk
  mmc: sdhci-esdhc-imx: add write protect on custom GPIO on mx25/35
  mmc: msm_sdcc: remove needless cache flush after dma_unmap_sg()
  mmc: sh_mmcif: support aggressive clock gating
  mmc: check if mmc cards < 2GB do sector addressing
  mmc: core: comment on why sdio_reset is done at init time
  mmc: dw_mmc: support DDR mode
  mmc: via-sdmmc: Remove set-but-unused variable.
  mmc: cb710: Return err value in cb710_wait_while_busy()
  mmc: sdhci-pci: Remove set-but-unused variable.
  mmc: mxs-mmc: add mmc host driver for i.MX23/28
  ...