After certain errors, some devices report complete garbage on
IDENTIFY.  This can cause ata_dev_read_id() to fail with -EINVAL
resulting in immediate disabling of the device.  Give the device one
last chance after -EINVAL to allow recovery from such situations.  As
-EINVAL is triggered very rarely, this shouldn't cause any noticeable
affect on more common error paths.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Harald Dunkel <harald.dunkel@t-online.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

It seems the world isn't as frank as we thought and some devices lie
about who they are.  Fallback to the other IDENTIFY if IDENTIFY is
aborted by the device.  As this is the strategy used by IDE for a long
time, it shouldn't cause too much problem.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: William Thompson <wt@electro-mechanical.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

libata enables SCSI host during ATA host activation which happens
after IRQ handler is registered and IRQ is enabled.  All ATA ports are
in frozen state when IRQ is enabled but frozen ports may raise limited
number of IRQs after being frozen - IOW, ->freeze() is not responsible
for clearing pending IRQs.  During normal operation, the IRQ handler
is responsible for clearing spurious IRQs on frozen ports and it
usually doesn't require any extra code.

Unfortunately, during host initialization, the IRQ handler can end up
scheduling EH for a port whose SCSI host isn't initialized yet.  This
results in OOPS in the SCSI midlayer.  This is relatively short window
and scheduling EH for probing is the first thing libata does after
initialization, so ignoring EH scheduling until initialization is
complete solves the problem nicely.

This problem was spotted by Berck E. Nash in the following thread.

  http://thread.gmane.org/gmane.linux.kernel/519412Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Berck E. Nash <flyboy@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

The intention of using port_mask in SFF init helpers was to eventually
support exoctic configurations such as combination of legacy and
native port on the same controller.  This never became actually
necessary and the related code always has been subtly broken one way
or the other.  Now that new init model is in place, there is no reason
to make common helpers capable of handling all corner cases.  Exotic
cases can simply dealt within LLDs as necessary.

This patch removes port_mask handling in SFF init helpers.  SFF init
helpers don't take n_ports argument and interpret it into port_mask
anymore.  All information is carried via port_info.  n_ports argument
is dropped and always two ports are allocated.  LLD can tell SFF to
skip certain port by marking it dummy.  Note that SFF code has been
treating unuvailable ports this way for a long time until recent
breakage fix from Linus and is consistent with how other drivers
handle with unavailable ports.

This fixes 1-port legacy host handling still broken after the recent
native mode fix and simplifies SFF init logic.  The following changes
are made...

* ata_pci_init_native_host() and ata_init_legacy_host() both now try
  to initialized whatever they can and mark failed ports dummy.  They
  return 0 if any port is successfully initialized.

* ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't
  take n_ports argument.  All info should be specified via port_info
  array.  Always two ports are allocated.

* ata_pci_init_bmdma() exported to be used by LLDs in exotic cases.

* port_info handling in all LLDs are standardized - all port_info
  arrays are const stack variable named ppi.  Unless the second port
  is different from the first, its port_info is specified as NULL
  (tells libata that it's identical to the last non-NULL port_info).

* pata_hpt37x/hpt3x2n: don't modify static variable directly.  Make an
  on-stack copy instead as ata_piix does.

* pata_uli: It has 4 ports instead of 2.  Don't use
  ata_pci_prepare_native_host().  Allocate the host explicitly and use
  init helpers.  It's simple enough.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

Now that libata uses sd->manage_start_stop, libata spins down disk on
shutdown.  In an attempt to compensate libata's previous shortcoming,
some distros sync and spin down disks attached via libata in their
shutdown(8).  Some disks spin back up just to spin down again on
STANDBYNOW1 if the command is issued when the disk is spun down, so
this double spinning down causes problem.

This patch implements module parameter libata.spindown_compat which,
when set to one (default value), prevents libata from spinning down
disks on shutdown thus avoiding double spinning down.  Note that
libata spins down disks for suspend to mem and disk, so with
libata.spindown_compat set to one, disks should be properly spun down
in all cases without modifying shutdown(8).

shutdown(8) should be fixed eventually.  Some drive do spin up on
SYNCHRONZE_CACHE even when their cache is clean.  Those disks
currently spin up briefly when sd tries to shutdown the device and
then the machine powers off immediately, which can't be good for the
head.  We can't skip SYNCHRONIZE_CACHE during shudown as it can be
dangerous data integrity-wise.

So, this spindown_compat parameter is already scheduled for removal by
the end of the next year and here's what shutdown(8) should do.

  * Check whether /sys/modules/libata/parameters/spindown_compat
    exists.  If it does, write 0 to it.

  * For each libata harddisk {
	* Check whether /sys/class/scsi_disk/h:c:i:l/manage_start_stop
	  exists.  Iff it doesn't, synchronize cache and spin the disk
	  down as before.
  }

The above procedure will make shutdown(8) work properly with kernels
before this change, ones with this workaround and later ones without
it.

To accelerate shutdown(8) updates, if the compat mode is in use, this
patch prints BIG FAT warning for five seconds during shutdown (the
optimal interval to annoy the user just the right amount discovered by
hours of tireless usability testing).
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

Reimplement suspend/resume support using sdev->manage_start_stop.

* Device suspend/resume is now SCSI layer's responsibility and the
  code is simplified a lot.

* DPM is dropped.  This also simplifies code a lot.  Suspend/resume
  status is port-wide now.

* ata_scsi_device_suspend/resume() and ata_dev_ready() removed.

* Resume now has to wait for disk to spin up before proceeding.  I
  couldn't find easy way out as libata is in EH waiting for the
  disk to be ready and sd is waiting for EH to complete to issue
  START_STOP.

* sdev->manage_start_stop is set to 1 in ata_scsi_slave_config().
  This fixes spindown on shutdown and suspend-to-disk.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

* 'linus' of master.kernel.org:/pub/scm/linux/kernel/git/perex/alsa: (122 commits)
  [ALSA] version 1.0.14rc4
  [ALSA] Add speaker pin sequencing to hda_codec.c:snd_hda_parse_pin_def_config()
  [ALSA] hda-codec - Add ALC861VD Lenovo support
  [ALSA] hda-codec - Fix connection list in generic parser
  [ALSA] usb-audio: work around wrong wMaxPacketSize on ESI M4U
  [ALSA] usb-audio: work around broken M-Audio MidiSport Uno firmware
  [ALSA] usb-audio: explicitly match Logitech QuickCam
  [ALSA] hda-codec - Fix a typo
  [ALSA] hda-codec - Fix ALC880 uniwill auto-mutes
  [ALSA] hda-codec - Fix AD1988 SPDIF playback route control
  [ALSA] wm8750 typo fix
  [ALSA] wavefront: only declare isapnp on CONFIG_PNP
  [ALSA] hda-codec - bug fixes for stac92xx HDA codecs.
  [ALSA] add MODULE_FIRMWARE entries
  [ALSA] do not depend on FW_LOADER when internal firmware images are used
  [ALSA] hda-codec - Fix resume of STAC92xx codecs
  [ALSA] usbaudio - Revert the minimal period size fix patch
  [ALSA] hda-codec - Add support for new HP DV series laptops
  [ALSA] usb-audio - Fix the minimum period size per transfer mode
  [ALSA] sound/pcmcia/vx/vxpocket.c: fix an if() condition
  ...

* 'master' of ssh://master.kernel.org/pub/scm/linux/kernel/git/mchehab/v4l-dvb: (44 commits)
  V4L/DVB (5571): V4l1-compat:  Make VIDIOCSPICT return errors in a useful way
  V4L/DVB (5624): Radio-maestro.c cleanup
  V4L/DVB (5623): Dsbr100.c Replace usb_dsbr100_do_ioctl to use video_ioctl2
  V4L/DVB (5622): Radio-zoltrix.c cleanup
  V4L/DVB (5621): Radio-cadet.c Replace cadet_do_ioctl to use video_ioctl2
  V4L/DVB (5619): Dvb-usb: fix typo
  V4L/DVB (5618): Cx88: Drop the generic i2c client from cx88-vp3054-i2c
  V4L/DVB (5617): V4L2: videodev, allow debugging
  V4L/DVB (5614): M920x: Disable second adapter on LifeView TV Walker Twin
  V4L/DVB (5613): M920x: loosen up 80-col limit
  V4L/DVB (5612): M920x: rename function prefixes from m9206_foo to m920x_foo
  V4L/DVB (5611): M920x: replace deb_rc with deb
  V4L/DVB (5610): M920x: remove duplicated code
  V4L/DVB (5609): M920x: group like functions together
  V4L/DVB (5608): M920x: various whitespace cleanups
  V4L/DVB (5607): M920x: Initial support for devices likely manufactured by Dposh
  V4L/DVB (5606): M920x: add "c-basic-offset: 8" to help emacs to enforce tabbing
  V4L/DVB (5605): M920x: Add support for LifeView TV Walker Twin
  V4L/DVB (5603): V4L: Prevent queueing queued buffers.
  V4L/DVB (5602): Enable DiSEqC in Starbox II (vp7021a)
  ...

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] Quicklist support for IA64
  [IA64] fix Kprobes reentrancy
  [IA64] SN: validate smp_affinity mask on intr redirect
  [IA64] drivers/char/snsc_event.c:206: warning: unused variable `p'
  [IA64] mca.c:121: warning: 'cpe_poll_timer' defined but not used
  [IA64] Fix - Section mismatch: reference to .init.data:mvec_name
  [IA64] more warning cleanups
  [IA64] Wire up epoll_pwait and utimensat
  [IA64] Fix warnings resulting from type-checking in dev_dbg()
  [IA64] typo s/kenrel/kernel/

The code was ok, but triggered warnings for calling __init from
__cpuinit. Instead call it from check_bugs instead.
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

I'm using a custom BIOS to configure the northbridge GART at address
0x80000000, size 2G.  Linux complains:

"Aperture from northbridge cpu 0 beyond 4GB. Ignoring."

I think there's an off-by-two error in arch/x86_64/kernel/aperture.c:

AK: use correct types for i386
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* git://git.kernel.org/pub/scm/linux/kernel/git/bunk/trivial:
  further UTF-8 fixes and name correction
  Fix wrong identifier name in Documentation/kref.txt

> -** Copyright 1994 by Bj<94>rn Brauel
> +** Copyright 1994 by Bj”rn Brauel

I think these were cp437, and it should read 'Björn'.
(asm-m68k/atari*.h)

Also note that Arnaldo just put more legacy noise into CREDITS...
Signed-off-by: Adrian Bunk <bunk@stusta.de>

There's a typo / wrong identifier name in Documentation/kref.txt. Fix it.
Signed-off-by: Satyam Sharma <ssatyam@cse.iitk.ac.in>
Acked-by: Corey Minyard <minyard@acm.org>
Signed-off-by: Adrian Bunk <bunk@stusta.de>

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-ip22:
  Convert SGI IP22 and specific drivers to platform_device.

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus: (28 commits)
  [MIPS] Rework cobalt_board_id
  [MIPS] Use RTC_CMOS for Cobalt
  [MIPS] Use platform_device for Cobalt UART
  [MIPS] Separate Alchemy processor based boards config
  [MIPS] Fix build error in atomic64_cmpxchg
  [MIPS] Run checksyscalls for N32 and O32 ABI
  [MIPS] tlbex: use __maybe_unused
  [MIPS] excite: use __maybe_unused
  [MIPS] Add extern cobalt_board_id
  [MIPS] Remove unused CONFIG_TOSHIBA_BOARDS
  [MIPS] Rename tb0229_defconfig to tb0219_defconfig
  [MIPS] Update tb0229_defconfig; add CONFIG_GPIO_TB0219.
  [MIPS] Add minimum defconfig for RBHMA4200
  [MIPS] SB1: Build fix.
  [MIPS] Drop __devinit tag from allocate_irqno() and free_irqno()
  [MIPS] clocksource: use CLOCKSOURCE_MASK() macro
  [MIPS] Remove LIMITED_DMA support
  [MIPS] Remove Momenco Jaguar ATX support
  [MIPS] Remove Momenco Ocelot G support
  [MIPS] FPU hazard handling
  ...

* 'for-linus' of git://git.kernel.dk/data/git/linux-2.6-block:
  Fix compile/link of init/do_mounts.c with !CONFIG_BLOCK
  When stacked block devices are in-use (e.g. md or dm), the recursive calls

* 'audit.b38' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/audit-current:
  [PATCH] Abnormal End of Processes
  [PATCH] match audit name data
  [PATCH] complete message queue auditing
  [PATCH] audit inode for all xattr syscalls
  [PATCH] initialize name osid
  [PATCH] audit signal recipients
  [PATCH] add SIGNAL syscall class (v3)
  [PATCH] auditing ptrace

* 'upstream-fixes' of master.kernel.org:/pub/scm/linux/kernel/git/jikos/hid:
  USB HID: hiddev - fix race between hiddev_send_event() and hiddev_release()
  HID: add hooks for getkeycode() and setkeycode() methods
  HID: switch to using input_dev->dev.parent
  USB HID: Logitech wheel 0x046d/0xc294 needs HID_QUIRK_NOGET quirk
  USB HID: usb_buffer_free() cleanup
  USB HID: report descriptor of Cypress USB barcode readers needs fixup
  Bluetooth HID: HIDP - don't initialize force feedback
  USB HID: update CONFIG_USB_HIDINPUT_POWERBOOK description
  HID: add input mappings for non-working keys on Logitech S510 remote

IA64 is the origin of the quicklist implementation.  So cut out the pieces
that are now in core code and modify the functions called.
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>

In case of reentrance i.e when a probe handler calls a functions which
inturn has a probe, we save a previous kprobe information and just single
step the reentrant probe without calling the actual probe handler.  During
this reentracy period, if an interrupt occurs and if probe happens to
trigger in the inturrupt path, then we were corrupting the previous kprobe(
as we were overriding the previous kprobe info) info their by crashing the
system.  This patch fixes this issues by having a an array of previous
kprobe info struct(with the array size of 2).

This similar technique is not needed on i386 and x86_64 because by default
interrupts are turn off in the break/int3 exception handler.
Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>

On SN, only allow one bit to be set in the smp_affinty mask when
redirecting an interrupt.  Currently setting multiple bits is allowed, but
only the first bit is used in determining the CPU to redirect to.  This has
caused confusion among some customers.

[akpm@linux-foundation.org: fixes]
Signed-off-by: John Keller <jpk@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6: (31 commits)
  [NETFILTER]: xt_conntrack: add compat support
  [NETFILTER]: iptable_raw: ignore short packets sent by SOCK_RAW sockets
  [NETFILTER]: iptable_{filter,mangle}: more descriptive "happy cracking" message
  [NETFILTER]: nf_nat: Clears helper private area when NATing
  [NETFILTER]: ctnetlink: clear helper area and handle unchanged helper
  [NETFILTER]: nf_conntrack: Removes unused destroy operation of l3proto
  [NETFILTER]: nf_conntrack: Removes duplicated declarations
  [NETFILTER]: nf_nat: remove unused argument of function allocating binding
  [NETFILTER]: Clean up table initialization
  [NET_SCHED]: Avoid requeue warning on dev_deactivate
  [NET_SCHED]: Reread dev->qdisc for NETDEV_TX_OK
  [NET_SCHED]: Rationalise return value of qdisc_restart
  [NET]: Fix dev->qdisc race for NETDEV_TX_LOCKED case
  [UDP]: Fix AF-specific references in AF-agnostic code.
  [IrDA]: KingSun/DonShine USB IrDA dongle support.
  [IPV6] ROUTE: Assign rt6i_idev for ip6_{prohibit,blk_hole}_entry.
  [IPV6]: Do no rely on skb->dst before it is assigned.
  [IPV6]: Send ICMPv6 error on scope violations.
  [SCTP]: Do not include ABORT chunk header in the notification.
  [SCTP]: Correctly copy addresses in sctp_copy_laddrs
  ...

When exporting input device bitmaps via compat_ioctl on BIG_ENDIAN
platforms evdev calculates data size incorrectly. This causes buffer
overflow if user specifies buffer smaller than maxlen.
Signed-off-by: Kenichi Nagai <kenichi3.nagai@toshiba.co.jp>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

During a 'resync' or similar activity, md checks if the devices in the
array are otherwise active and winds back resync activity when they are.
This test in done in is_mddev_idle, and it is somewhat fragile - it
sometimes thinks there is non-sync io when there isn't.

The test compares the total sectors of io (disk_stat_read) with the sectors
of resync io (disk->sync_io).  This has problems because total sectors gets
updated when a request completes, while resync io gets updated when the
request is submitted.  The time difference can cause large differenced
between the two which do not actually imply non-resync activity.  The test
currently allows for some fuzz (+/- 4096) but there are some cases when it
is not enough.

The test currently looks for any (non-fuzz) difference, either positive or
negative.  This clearly is not needed.  Any non-sync activity will cause
the total sectors to grow faster than the sync_io count (never slower) so
we only need to look for a positive differences.

If we do this then the amount of in-flight sync io will never cause the
appearance of non-sync IO.  Once enough non-sync IO to worry about starts
happening, resync will be slowed down and the measurements will thus be
more precise (as there is less in-flight) and control of resync will still
be suitably responsive.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Remove the obsolete "if [ ]" construct from the video console Kconfig
file.
Signed-off-by: Robert P. J. Day <rpjday@mindspring.com>
Acked-by: James Simmons <jsimmons@infradead.org>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Convert internal wait_pm2() function to fb API fb_sync() method.
Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Let the user enable debugging messages in nvidiafb.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Fix rivafb's I2C getscl callback function, as was done in nvidiafb recently.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Adds a framebuffer driver to ATMEL AT91SAM9x and AT32 aka AVR32 platforms.
Those chips share quite the same IP and this code is suitable for both
architectures.
Signed-off-by: Nicolas Ferre <nicolas.ferre@rfo.atmel.com>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

This is a basic port from 2.4 kernel to 2.6.  Acceleration is lost and big
endian support probably too.  The driver works in 8, 16 and 32 bit mode.

[adaplas]
- change VESA_* to FB_BLANK_* constants
- removed unused function clear_memory
- fix uninitialized variable compiler warning
- some whitespace cleaning

[akpm@linux-foundation.org: Nuke pestiferous CVS string]
Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ups... coding style.
Signed-off-by: Michal Piotrowski <michal.k.k.piotrowski@gmail.com>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Re-arrange epoll code to avoid static functions pre-declarations, and apply
akpm-filter on it.
Signed-off-by: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Epoll is either compiled it, or not (if EMBEDDED). Remove the module code
and use fs_initcall().
Signed-off-by: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Cut out lots of code from epoll, by reusing the anonymous inode source
patch (fs/anon_inodes.c).
Signed-off-by: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

This is an example about how to add eventfd support to the current KAIO code,
in order to enable KAIO to post readiness events to a pollable fd (hence
compatible with POSIX select/poll).  The KAIO code simply signals the eventfd
fd when events are ready, and this triggers a POLLIN in the fd.  This patch
uses a reserved for future use member of the struct iocb to pass an eventfd
file descriptor, that KAIO will use to post events every time a request
completes.  At that point, an aio_getevents() will return the completed result
to a struct io_event.  I made a quick test program to verify the patch, and it
runs fine here:

http://www.xmailserver.org/eventfd-aio-test.c

The test program uses poll(2), but it'd, of course, work with select and epoll
too.

This can allow to schedule both block I/O and other poll-able devices
requests, and wait for results using select/poll/epoll.  In a typical
scenario, an application would submit KAIO request using aio_submit(), and
will also use epoll_ctl() on the whole other class of devices (that with the
addition of signals, timers and user events, now it's pretty much complete),
and then would:

	epoll_wait(...);
	for_each_event {
		if (curr_event_is_kaiofd) {
			aio_getevents();
			dispatch_aio_events();
		} else {
			dispatch_epoll_event();
		}
	}
Signed-off-by: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

This patch wires the eventfd system call to the x86 architectures.
Signed-off-by: Davide Libenzi <davidel@xmailserver.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

This is a very simple and light file descriptor, that can be used as event
wait/dispatch by userspace (both wait and dispatch) and by the kernel
(dispatch only).  It can be used instead of pipe(2) in all cases where those
would simply be used to signal events.  Their kernel overhead is much lower
than pipes, and they do not consume two fds.  When used in the kernel, it can
offer an fd-bridge to enable, for example, functionalities like KAIO or
syslets/threadlets to signal to an fd the completion of certain operations.
But more in general, an eventfd can be used by the kernel to signal readiness,
in a POSIX poll/select way, of interfaces that would otherwise be incompatible
with it.  The API is:

int eventfd(unsigned int count);

The eventfd API accepts an initial "count" parameter, and returns an eventfd
fd.  It supports poll(2) (POLLIN, POLLOUT, POLLERR), read(2) and write(2).

The POLLIN flag is raised when the internal counter is greater than zero.

The POLLOUT flag is raised when at least a value of "1" can be written to the
internal counter.

The POLLERR flag is raised when an overflow in the counter value is detected.

The write(2) operation can never overflow the counter, since it blocks (unless
O_NONBLOCK is set, in which case -EAGAIN is returned).

But the eventfd_signal() function can do it, since it's supposed to not sleep
during its operation.

The read(2) function reads the __u64 counter value, and reset the internal
value to zero.  If the value read is equal to (__u64) -1, an overflow happened
on the internal counter (due to 2^64 eventfd_signal() posts that has never
been retired - unlickely, but possible).

The write(2) call writes an __u64 count value, and adds it to the current
counter.  The eventfd fd supports O_NONBLOCK also.

On the kernel side, we have:

struct file *eventfd_fget(int fd);
int eventfd_signal(struct file *file, unsigned int n);

The eventfd_fget() should be called to get a struct file* from an eventfd fd
(this is an fget() + check of f_op being an eventfd fops pointer).

The kernel can then call eventfd_signal() every time it wants to post an event
to userspace.  The eventfd_signal() function can be called from any context.
An eventfd() simple test and bench is available here:

http://www.xmailserver.org/eventfd-bench.c

This is the eventfd-based version of pipetest-4 (pipe(2) based):

http://www.xmailserver.org/pipetest-4.c

Not that performance matters much in the eventfd case, but eventfd-bench
shows almost as double as performance than pipetest-4.

[akpm@linux-foundation.org: fix i386 build]
[akpm@linux-foundation.org: add sys_eventfd to sys_ni.c]
Signed-off-by: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>