Oliver Neukum [Tue, 27 Mar 2007 14:02:34 +0000 (16:02 +0200)]
USB: fix omninet write vs. close race
omninet kills all URBs in close. However write() returns as soon as
the URB has been submitted. Killing the last URB means a race that
can lose that date written in the last call to write().
As a fix this is moved to shutdown().
Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Oliver Neukum [Fri, 30 Mar 2007 11:11:00 +0000 (13:11 +0200)]
USB: fix catc error handling
this driver ignores errors while starting the transmit queue. It will
never be reported stopped as the completion handler won't run
and it will never be started again as it will be considered started.
This patch adds error handling.
Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Jelle Foks [Mon, 26 Mar 2007 01:08:35 +0000 (21:08 -0400)]
USB: ftdi_sio: Add USB ID of ADSTech USBX-707
This patch adds the USB ID of the ADS Tech USBX-707 USB IR blaster (that
comes with the ADS Tech PTV-305 grabber card), which has a ftdi232bm
inside hooked up to a pic.
With this it should be fairly straightforward to make at least lirc
receiving work with this device. I will submit a patch to lirc for that
as soon as I have one ready, I'm getting data with minicom with this
patch, but need to figure out some more details such as best/correct
baudrate.
Kay Sievers [Tue, 13 Mar 2007 14:59:31 +0000 (15:59 +0100)]
USB: make usbdevices export their device nodes instead of using a separate class
o The "real" usb-devices export now a device node which can
populate /dev/bus/usb.
o The usb_device class is optional now and can be disabled in the
kernel config. Major/minor of the "real" devices and class devices
are the same.
o The environment of the usb-device event contains DEVNUM and BUSNUM to
help udev and get rid of the ugly udev rule we need for the class
devices.
o The usb-devices and usb-interfaces share the same bus, so I used
the new "struct device_type" to let these devices identify
themselves. This also removes the current logic of using a magic
platform-pointer.
The name of the device_type is also added to the environment
which makes it easier to distinguish the different kinds of devices
on the same subsystem.
It looks like this:
add@/devices/pci0000:00/0000:00:1d.1/usb2/2-1
ACTION=add
DEVPATH=/devices/pci0000:00/0000:00:1d.1/usb2/2-1
SUBSYSTEM=usb
SEQNUM=1533
MAJOR=189
MINOR=131
DEVTYPE=usb_device
PRODUCT=46d/c03e/2000
TYPE=0/0/0
BUSNUM=002
DEVNUM=004
This udev rule works as a replacement for usb_device class devices:
SUBSYSTEM=="usb", ACTION=="add", ENV{DEVTYPE}=="usb_device", \
NAME="bus/usb/$env{BUSNUM}/$env{DEVNUM}", MODE="0644"
Updated patch, which needs the device_type patches in Greg's tree.
I also got a bugzilla assigned for this. :)
https://bugzilla.novell.com/show_bug.cgi?id=250659
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Tue, 20 Mar 2007 18:59:39 +0000 (14:59 -0400)]
USB: add power/level sysfs attribute
This patch (as874) adds another piece to the user-visible part of the
USB autosuspend interface. The new power/level sysfs attribute allows
users to force the device on (with autosuspend off), force the device
to sleep (with autoresume off), or return to normal automatic operation.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Peter Stokes [Sat, 17 Mar 2007 14:14:12 +0000 (16:14 +0200)]
USB: ati_remote2: Add channel support
Add logical channel support for ATI Remote Wonder II
The ATI Remote Wonder II can be configured with one of 16 unique logical
channels. Allowing up to 16 remotes to be used independently within
range of each other. This change adds functionality to configure the
receiver and filter the input data to respond or exclude remotes
configured with different logical channels.
Signed-off-by: Peter Stokes <linux@dadeos.freeserve.co.uk> Acked-by: Ville Syrjala <syrjala@sci.fi> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Oliver Neukum [Tue, 20 Mar 2007 12:54:05 +0000 (13:54 +0100)]
USB: sierra close race
the sierra driver does not directly use usb_kill_urb(). It uses a wrapper.
This wrapper means that callbacks which are running are not killed during
close, resubmitting and illicitly pushing data into the tty layer.
The whole purpose of usb_kill_urb() is subverted. The wrapper must be removed.
The same problem as the option driver.
Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Oliver Neukum [Tue, 20 Mar 2007 12:41:21 +0000 (13:41 +0100)]
USB: option close race
the option driver does not directly use usb_kill_urb(). It uses a wrapper.
This wrapper means that callbacks which are running are not killed during
close, resubmitting and illicitly pushing data into the tty layer.
The whole purpose of usb_kill_urb() is subverted. The wrapper must be removed.
Alan Stern [Mon, 19 Mar 2007 19:33:11 +0000 (15:33 -0400)]
UHCI: Add some WARN_ON()s
This patch (as872) adds some WARN_ON()s to various error checks which
are never supposed to fail. Unsettlingly, one of them has shown up in
a user's log! Maybe making the warning more visible and having the
call-stack information available will help pinpoint the source of the
problem.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Randy Dunlap [Sat, 17 Mar 2007 01:33:13 +0000 (18:33 -0700)]
USB: dm9601: fix sparse NULL warnings
Fix sparse NULL warnings:
drivers/usb/net/dm9601.c:88:23: warning: Using plain integer as NULL pointer
drivers/usb/net/dm9601.c:174:22: warning: Using plain integer as NULL pointer
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Acked-by: Peter Korsgaard <jacmet@sunsite.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Oliver Neukum [Fri, 16 Mar 2007 19:28:28 +0000 (20:28 +0100)]
USB: overhaul of mos7840 driver
This fixes:
- breaking DMA rules about buffers
- usage of _global_ variables to save a single device's attributes
- racy access to urb->status
- smp monotonity issue with statistics
- use of one buffer for many simultaneous URBs
- error handling introduced
- several instances of following NULL pointers
- use after free
- unnecessary GFP_ATOMIC
- GFP_KERNEL in interrupt
- various cleanups
- write room granularity issue that bit cdc-acm
- race in shutdown
Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Thu, 15 Mar 2007 19:54:30 +0000 (15:54 -0400)]
EHCI: add delay to bus_resume before accessing ports
This patch (as870) adds a delay to ehci-hcd's bus_resume routine.
Apparently there are controllers and/or BIOSes out there which need
such a delay to get the ports back into their correct state. This
fixes Bugzilla #8190.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Cc: David Brownell <david-b@pacbell.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Tue, 13 Mar 2007 20:39:15 +0000 (16:39 -0400)]
USB: Allow autosuspend delay to equal 0
This patch (as867) adds an entry for the new power/autosuspend
attribute in Documentation/ABI/testing, and it changes the behavior of
the delay value. Now a delay of 0 means to autosuspend as soon as
possible, and negative values will prevent autosuspend.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Tue, 13 Mar 2007 20:37:30 +0000 (16:37 -0400)]
USB: separate autosuspend from external suspend
This patch (as866) adds new entry points for external USB device
suspend and resume requests, as opposed to internally-generated
autosuspend or autoresume. It also changes the existing
remote-wakeup code paths to use the new routines, since remote wakeup
is not the same as autoresume.
As part of the change, it turns out to be necessary to do remote
wakeup of root hubs from a workqueue. We had been using khubd, but it
does autoresume rather than an external resume. Using the
ksuspend_usb_wq workqueue for this purpose seemed a logical choice.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
ehci-fsl: change SI_CTRL, PRI_CTRL register offsets according to errata
Correct the offsets of the SI_CTRL, PRI_CTRL registers according to
the Reference Manual errata sheet in order to prevent unwanted
settings regarding burst transactions and priority states.
Signed-off-by: Christian Engelmayer <Christian.Engelmayer@frequentis.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Tue, 13 Mar 2007 15:10:52 +0000 (11:10 -0400)]
USB: fix race in HCD removal
This patch (as865) fixes a race in the HCD removal code discovered by
Milan Plzik. Arrival of an interrupt after the root hub was
unregistered could cause the root-hub status timer to start up, even
after it was supposed to have been shut down. The problem is fixed by
moving the del_timer_sync() call to after the HCD's stop() method, at
which time IRQ generation should be disabled.
Cc: Milan Plzik <milan.plzik@gmail.com> Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Simon Arlott [Tue, 6 Mar 2007 10:47:46 +0000 (02:47 -0800)]
USB: MAINTAINERS: cxacru
I've acquired a second device for testing and plan to make some changes in
the near future to export all the device stats to sysfs (based on my
proposed patch to add them to the proc file ~2007-01-30).
Signed-off-by: Simon Arlott <simon@fire.lp0.eu> Acked-by: Duncan Sands <baldrick@free.fr> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Simon Arlott [Tue, 6 Mar 2007 10:47:45 +0000 (02:47 -0800)]
USB: cxacru: export detailed device info through sysfs
When the device is polled for status there is a lot of useful status
information available that is ignored. This patch stores the device info
array when the status is polled and adds sysfs files to the usb device to
allow userspace to query it. Since the device updates its status
internally once a second the poll time is changed to this, and
round_jiffies_relative is used to avoid waking the cpu unnecessarily.
Signed-off-by: Simon Arlott <simon@fire.lp0.eu> Cc: Duncan Sands <duncan.sands@free.fr> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Simon Arlott [Tue, 6 Mar 2007 10:47:43 +0000 (02:47 -0800)]
usbatm: create sysfs link "device" from atm class device to usb interface
There is currently no path from the ATM device in /sys to the USB device's
interface that the driver is using; this patch creates a "device" symlink. It
is then possible to get to the cxacru ADSL statistics
(http://lkml.org/lkml/2007/2/23/328):
If this link is not appropriate I'd have to create device files in
/sys/class/atm/cxacru0 instead - which seems less appropriate since the ADSL
statistics are for the USB device not ATM (which is running over the ADSL).
[akpm@linux-foundation.org: cleanups] Signed-off-by: Simon Arlott <simon@fire.lp0.eu> Cc: Duncan Sands <duncan.sands@math.u-psud.fr> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Fri, 9 Mar 2007 20:41:13 +0000 (15:41 -0500)]
usbcore: move usb_autosuspend_work
This patch (as864) moves the work routine for USB autosuspend from one
source file to another. This permits the removal of one whole global
symbol (!) and should smooth the way for more changes in the future.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Pete Zaitcev [Fri, 9 Mar 2007 04:02:26 +0000 (20:02 -0800)]
libusual: change block scope variable to function scope
Someone changed the code to kthread and used his style instead of mine.
The problem with the block variables is that they provoke shadowing,
which is actually exactly what has happened in my other tree which
has the class patch.
Signed-off-by: Pete Zaitcev <zaitcev@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Pete Zaitcev [Sun, 25 Feb 2007 03:27:33 +0000 (19:27 -0800)]
usbmon: Extended text API
This patch adds a new text API, codenamed '1u', which captures more URB
fields than old '1t' interface did. Also the '1u' text API is compatible
with the future "bus zero" extension.
Signed-off-by: Pete Zaitcev <zaitcev@redhat.com> Acked-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Oliver Neukum [Thu, 1 Mar 2007 13:31:02 +0000 (14:31 +0100)]
USB: fix skeleton driver
compilation of the skeleton driver is currently broken. It doesn't compile.
So while I am it:
- fix typo
- add comments to answer common questions
- actually allow autosuspend in the driver struct
- increase paralellism by restricting code under locks
Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
David Brownell [Sat, 24 Feb 2007 20:23:52 +0000 (12:23 -0800)]
pxa2xx_udc: cleanups, use platform_get_irq
Make the pxa2xx_udc driver fetch its IRQ from platform resources
rather than using compile-time constants, so that it works properly
on IXP4xx systems not just PXA21x/25x/26x.
Other updates:
- Do that using platform_get_irq()
- Switch to platform_driver_probe()
- Handle device_add() errors
- Remove "function" sysfs attribute and its potential errors
- Whitespace cleanups
Signed-off-by: Milan Svoboda <msvoboda@ra.rockwell.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/driver-2.6: (46 commits)
dev_dbg: check dev_dbg() arguments
drivers/base/attribute_container.c: use mutex instead of binary semaphore
mod_sysfs_setup() doesn't return errno when kobject_add_dir() failure occurs
s2ram: add arch irq disable/enable hooks
define platform wakeup hook, use in pci_enable_wake()
security: prevent permission checking of file removal via sysfs_remove_group()
device_schedule_callback() needs a module reference
s390: cio: Delay uevents for subchannels
sysfs: bin.c printk fix
Driver core: use mutex instead of semaphore in DMA pool handler
driver core: bus_add_driver should return an error if no bus
debugfs: Add debugfs_create_u64()
the overdue removal of the mount/umount uevents
kobject: Comment and warning fixes to kobject.c
Driver core: warn when userspace writes to the uevent file in a non-supported way
Driver core: make uevent-environment available in uevent-file
kobject core: remove rwsem from struct subsystem
qeth: Remove usage of subsys.rwsem
PHY: remove rwsem use from phy core
IEEE1394: remove rwsem use from ieee1394 core
...
[akpm@linux-foundation.org: fix a couple of things which broke] Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Akinobu Mita [Thu, 26 Apr 2007 07:12:09 +0000 (00:12 -0700)]
mod_sysfs_setup() doesn't return errno when kobject_add_dir() failure occurs
mod_sysfs_setup() doesn't return an errno when kobject_add_dir() for module
"holders" directory fails. So caller of mod_sysfs_setup() will keep going
and get oops.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
For powermac, we need to do some things between suspending devices and
device_power_off, for example setting the decrementer. This patch
allows architectures to define arch_s2ram_{en,dis}able_irqs in their
asm/suspend.h to have control over this step.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Acked-by: Pavel Machek <pavel@ucw.cz> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
David Brownell [Thu, 26 Apr 2007 07:12:06 +0000 (00:12 -0700)]
define platform wakeup hook, use in pci_enable_wake()
This defines a platform hook to enable/disable a device as a wakeup event
source. It's initially for use with ACPI, but more generally it could be used
whenever enable_irq_wake()/disable_irq_wake() don't suffice.
The hook is called -- if available -- inside pci_enable_wake(); and the
semantics of that call are enhanced so that support for PCI PME# is no longer
needed. It can now work for devices with "legacy PCI PM", when platform
support allows it. (That support would use some board-specific signal for for
the same purpose as PME#.)
[akpm@linux-foundation.org: Make it compile with CONFIG_PM=n] Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Zhang Rui <rui.zhang@intel.com> Cc: Len Brown <lenb@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
James Morris [Thu, 26 Apr 2007 07:12:05 +0000 (00:12 -0700)]
security: prevent permission checking of file removal via sysfs_remove_group()
Prevent permission checking from being performed when the kernel wants to
unconditionally remove a sysfs group, by introducing an kernel-only variant
of lookup_one_len(), lookup_one_len_kern().
Additionally, as sysfs_remove_group() does not check the return value of
the lookup before using it, a BUG_ON has been added to pinpoint the cause
of any problems potentially caused by this (and as a form of annotation).
Signed-off-by: James Morris <jmorris@namei.org> Cc: Nagendra Singh Tomar <nagendra_tomar@adaptec.com> Cc: Tejun Heo <htejun@gmail.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: Eric Paris <eparis@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Thu, 26 Apr 2007 07:12:04 +0000 (00:12 -0700)]
device_schedule_callback() needs a module reference
This patch (as896b) fixes an oversight in the design of
device_schedule_callback(). It is necessary to acquire a reference to the
module owning the callback routine, to prevent the module from being
unloaded before the callback can run.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Cc: Satyam Sharma <satyam.sharma@gmail.com> Cc: Neil Brown <neilb@suse.de> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
We often have the situation that we register a subchannel and start device
recognition, only to find out that the device is not usable after all, which
triggers an unregister of the subchannel. This often happens on hundreds of
subchannels on a LPAR, leading to a storm of events which aren't of any use.
Therefore, use uevent_suppress to delay the KOBJ_ADD uevent for a subchannel
until we know that its ccw_device is to be registered.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: Eric Rannaud <eric.rannaud@gmail.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kay Sievers [Thu, 5 Apr 2007 23:40:38 +0000 (01:40 +0200)]
Driver core: warn when userspace writes to the uevent file in a non-supported way
In the future we will allow the uevent type to be written to the uevent
file to trigger the different types of uevents. But for now, as we only
support the ADD event, warn if userspace tries to write anything else to
this file.
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kay Sievers [Thu, 5 Apr 2007 23:40:38 +0000 (01:40 +0200)]
Driver core: make uevent-environment available in uevent-file
This allows sysfs to show the environment variables that are available
if the uevent happens. This lets userspace not have to cache all of
this information as the kernel already knows it.
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
It isn't used at all by the driver core anymore, and the few usages of
it within the kernel have now all been fixed as most of them were using
it incorrectly. So remove it.
Now the whole struct subsys can be removed from the system, but that's
for a later patch...
Cc: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
the current driver tree contains the removal of subsys.rwsem.
Unfortunately, this breaks qeth. However, it should be no problem to
fix the walking of the devices for /proc/qeth:
No need to take subsys.rwsem during walking the devices,
driver_find_devices() should already suffice.
The subsystem rwsem is not used by the driver core at all, so the use of
it in the phy code doesn't make any sense. They might possibly
want to use a local lock, but I am unsure about that.
The subsystem rwsem is not used by the driver core at all, so the use of
it in the ieee1394 code doesn't make any sense. They might possibly
want to use a local lock, but as most of these operations are already
protected by a local lock, it really doesn't look like it would be
needed.
Cc: Ben Collins <bcollins@debian.org> Cc: Stefan Richter <stefanr@s5r6.in-berlin.de> Cc: linux1394-devel <linux1394-devel@lists.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The subsystem rwsem is not used by the driver core at all, so the use of
it in the ide-proc code of it doesn't make any sense. Perhaps a local
lock might be needed, but I do not really think so.
Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: linux ide <linux-ide@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The rwsem is not used to protect anything, so the use of it by the PNP
subsystem isn't really useful, and it's doubtful if it really did anything or
not. So I've removed it.
Cc: Adam Belay <ambx1@neo.rr.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
USB: remove use of the bus rwsem, as it doesn't really protect anything.
The driver core stopped using the rwsem a long time ago, yet the USB
core still grabbed the lock, thinking it protected something. This
patch removes that useless use.
Cc: Alan Stern <stern@rowland.harvard.edu> Cc: Oliver Neukum <oneukum@suse.de> Cc: David Brownell <david-b@pacbell.net> Cc: linux-usb-devel <linux-usb-devel@lists.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Driver core: add suspend() and resume() to struct device_type
Driver core: add suspend() and resume() to struct device_type
In cases when there are devices of different types in the same class
we can't use class's implementation of suspend and resume methods and
we need to add them to struct device_type instead.
Also fix error handling in resume code (we should not try to call
class's resume method iof bus's resume method for the device failed.
Cornelia Huck [Thu, 29 Mar 2007 09:12:14 +0000 (11:12 +0200)]
Driver core: switch firmware_class to uevent_suppress.
Use uevent_suppress instead of returning an error code in
firmware_uevent(). Get rid of the now unneeded FW_STATUS_READY
and FW_STATUS_READY_NOHOTPLUG.
Cornelia Huck [Thu, 29 Mar 2007 09:12:11 +0000 (11:12 +0200)]
Driver core: suppress uevents via filter
Suppress uevents for devices if uevent_suppress is set via
dev_uevent_filter(). This makes the driver core suppress all device
uevents, not just the add event in device_add().
Oliver Neukum [Mon, 2 Apr 2007 12:47:59 +0000 (14:47 +0200)]
kref: fix CPU ordering with respect to krefs
some atomic operations are only atomic, not ordered. Thus a CPU is allowed
to reorder memory references to an object to before the reference is
obtained. This fixes it.
Signed-off-by: Oliver Neukum <oneukum@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Driver core: remove unneeded completion from driver release path
The completion in the driver release path is due to ancient history in
the _very_ early 2.5 days when we were not tracking the module reference
count of attributes. It is not needed at all and can be removed.
Note, we now have an empty release function for the driver structure.
This is due to the fact that drivers are statically allocated in the
system at this point in time, something which I want to change in the
future. But remember, drivers are really code, which is reference
counted by the module, unlike devices, which are data and _must_ be
reference counted properly in order to work correctly.
Cc: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Cornelia Huck [Tue, 6 Feb 2007 00:15:26 +0000 (16:15 -0800)]
driver core: don't fail attaching the device if it cannot be bound
Don't fail bus_attach_device() if the device cannot be bound.
If dev->driver has been specified, reset it to NULL if device_bind_driver()
failed and add the device as an unbound device. As a result,
bus_attach_device() now cannot fail, and we can remove some checking from
device_add().
Also remove an unneeded check in bus_rescan_devices_helper().
Cornelia Huck [Tue, 6 Feb 2007 00:15:25 +0000 (16:15 -0800)]
driver core: per-subsystem multithreaded probing
Make multithreaded probing work per subsystem instead of per driver.
It doesn't make much sense to probe the same device for multiple drivers in
parallel (after all, only one driver can bind to the device). Instead, create
a probing thread for each device that probes the drivers one after another.
Also make the decision to use multi-threaded probe per bus instead of per
device and adapt the pci code.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kay Sievers [Mon, 12 Mar 2007 20:08:57 +0000 (21:08 +0100)]
Driver core: add name to device_type
If "name" of a device_type is specified, the uevent will
contain the device_type name in the DEVTYPE variable.
This helps userspace to distingiush between different types
of devices, belonging to the same subsystem.
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Dmitry Torokhov [Sat, 10 Mar 2007 06:37:34 +0000 (01:37 -0500)]
driver core: Use attribute groups in struct device_type
Driver core: use attribute groups in struct device_type
Attribute groups are more flexible than attribute lists
(an attribute list can be represented by anonymous group)
so switch struct device_type to use them.
Also rework attribute creation for devices so that they all
cleaned up properly in case of errors.
The flag "drivers_autoprobe" controls the behavior of the bus to bind
devices by default, or just initialize the device and leave it alone.
The command "drivers_probe" accepts a bus_id and the bus tries to bind a
driver to this device.
Systems who want to control the driver binding with udev, switch off the
bus initiated probing:
echo 0 > /sys/subsystem/usb/drivers_autoprobe
echo 0 > /sys/subsystem/pcmcia/drivers_autoprobe
...
and initiate the probing with udev rules like:
ACTION=="add", SUBSYSTEM=="usb", ATTR{subsystem/drivers_probe}="$kernel"
ACTION=="add", SUBSYSTEM=="pcmcia", ATTR{subsystem/drivers_probe}="$kernel"
...
Custom driver binding can happen in earlier rules by something like:
ACTION=="add", SUBSYSTEM=="usb", \
ATTRS{idVendor}=="1234", ATTRS{idProduct}=="5678" \
ATTR{subsystem/drivers/<custom-driver>/bind}="$kernel"
This is intended to solve the modprobe.conf mess with "install-rules", custom
bind/unbind-scripts and all the weird things people invented over the years.
It should also provide the functionality "libusual" was supposed to do.
With udev, one can just write a udev rule to drive all USB-disks at the
third port of USB-hub by the "ub" driver, and everything else by
usb-storage. One can also instruct udev to bind different wireless
drivers to identical cards - just selected by the pcmcia slot-number, and
whatever ...
To use the mentioned rules, it needs udev version 106, to be able to
write ATTR{}="$kernel" to sysfs files.
Jean Delvare [Fri, 9 Mar 2007 15:33:10 +0000 (16:33 +0100)]
dev_printk and new-style class devices
As the new-style class devices (as opposed to old-style struct
class_device) are becoming more widely used, I noticed that the
dev_printk-based functions are not working properly with these.
New-style class devices have no driver nor bus, almost by definition,
and as a result dev_driver_string(), which is used as the first
parameter of dev_printk, resolves to an empty string. This causes
entries like the following to show in my logs:
i2c-2: adapter [SMBus stub driver] registered
Notice the unaesthetical leading whitespace. In order to fix this
problem, I suggest that we extend dev_driver_string to deal with
new-style class devices:
Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kay Sievers [Wed, 14 Mar 2007 02:25:56 +0000 (03:25 +0100)]
driver core: fix namespace issue with devices assigned to classes
- uses a kset in "struct class" to keep track of all directories
belonging to this class
- merges with the /sys/devices/virtual logic.
- removes the namespace-dir if the last member of that class
leaves the directory.
There may be locking or refcounting fixes left, I stopped when it seemed
to work with network and sound modules. :)
From: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/selinux-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/selinux-2.6:
selinux: preserve boolean values across policy reloads
selinux: change numbering of boolean directory inodes in selinuxfs
selinux: remove unused enumeration constant from selinuxfs
selinux: explicitly number all selinuxfs inodes
selinux: export initial SID contexts via selinuxfs
selinux: remove userland security class and permission definitions
SELinux: move security_skb_extlbl_sid() out of the security server
MAINTAINERS: update selinux entry
SELinux: rename selinux_netlabel.h to netlabel.h
SELinux: extract the NetLabel SELinux support from the security server
NetLabel: convert a BUG_ON in the CIPSO code to a runtime check
NetLabel: cleanup and document CIPSO constants
show_state() (SysRq-T) developed the buggy habbit of not showing
TASK_RUNNING tasks. This was due to the mistaken belief that state_filter
== -1 would be a pass-through filter - while in reality it did not let
TASK_RUNNING == 0 p->state values through.
Fix this by restoring the original '!state_filter means all tasks'
special-case i had in the original version. Test-built and test-booted on
i686, SysRq-T now works as intended.
Merge branch 'for-linus' of git://git.infradead.org/ubi-2.6
* 'for-linus' of git://git.infradead.org/ubi-2.6:
UBI: remove unused variable
UBI: add me to MAINTAINERS
JFFS2: add UBI support
UBI: Unsorted Block Images
projects / linux-2.6 / log