While reviewing various users of kernel memory allocation functions I came
across drivers/input/serio/hil_mlc.c::hil_mlc_register() and noticed that:
- it calls kzalloc() but fails to check for a NULL return before use.
- it makes several allocations and if one fails it doesn't free the
previous ones.
- It doesn't return -ENOMEM in the failed memory allocation case (it just
crashes).
This patch corrects all of the above and also reworks the only caller of
this function that I could find
(drivers/input/serio/hp_sdc_mlc.c::hp_sdc_mlc_out()) so that it now checks
the return value of hil_mlc_register() and properly propagates it on
failure and I also restructured the code to remove some labels and goto's
to make it, IMHO nicer to read.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Tested-by: Helge Deller <deller@gmx.de>
Acked-by: Helge Deller <deller@gmx.de>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
Input: fix typo in keycode validation supporting large scancodes
Input: aiptek - tighten up permissions on sysfs attributes
Input: sysrq - pass along lone Alt + SysRq
Instead of creating an exclusive thread to handle gameport events (which
happen rarely), let's switch to common workqueue. With the arrival
of concurrency-managed workqueue infrastructure we are not concerned
that our callers or callees also using workqueue (no deadlocks anymore)
and it should reduce total number of threads in the system.
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Instead of creating an exclusive thread to handle serio events (which
happen rarely), let's switch to using common workqueue. With the arrival
of concurrency-managed workqueue infrastructure we are not concerned
that our callers or callees also using workqueue (no deadlocks anymore)
and it should reduce total number of threads in the system.
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
The kcalloc call for the object table is using sizeof(struct qt602240_data)
when it should be using sizeof(struct qt6602240_object), resulting in a larger
allocation than is required.
Signed-off-by: Chris Leech <christopher.leech@linux.intel.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
As pointed out by Oliver Neukum:
xpad->irq_in is currently submitted before xpad->bulk_out is allocated.
That however is a race, because the callback for irq_in can call
xpad360w_process_packet(), which will in turn submit the bulk URB.
This patch moves initialization for xpad->bulk_out earlier, so we can
ensure xpad->bulk_out is initialized before submitting urb.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
The big kernel lock has been removed from all these files at some point,
leaving only the #include.
Remove this too as a cleanup.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Sysfs attributes affecting device behavior should not be, by default,
world-writeable. If distributions want to allow console users access
these attributes they need to employ udev and friends to adjust
permissions as needed.
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
This patch adds support for the MacBookAir3,1 and MacBookAir3,2
models.
[rydberg@euromail.se: touchpad range calibration]
Cc: stable@kernel.org
Signed-off-by: Edgar (gimli) Hucek <gimli@dark-green.com>
Signed-off-by: Henrik Rydberg <rydberg@euromail.se>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Recent testing of this codepath showed that it wasn't working,
perhaps due to changes within the input layer. This fixes it.
Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Disable the recalibration guard where new recalibrations are triggered
if we detect a packet too soon after calibrating - we found that this
results in erroneous recalibrations, and if the recalibration failed
then the rest of our badness-detection code will request another.
Add a module option disabling all of the recalibration code, in case
an OLPC deployment thinks all of the workarounds we have are doing
more damage than good and wants to experiment with them all disabled.
Based on work by Paul Fox.
Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
In addition to forcing recalibrations upon detection of cursor jumps (and
performing them quicker than before), detect and discard errant 'jump'
packets caused by a firmware bug, which are then repeated with each one
being approximately half the delta of the one previously (as if it is
averaging out)
Based on original work by Paul Fox.
Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
The old implementation of spew detection simply tracked the overall
position delta of the cursor over every 100 packets. We found that
this causes occasional false positives in spew detection, and also
that the conditions of the spewy packets are perhaps more fixed than
we once thought.
Rework the spew detection to look for packets of specific small
delta, and only recalibrating if the overall movement delta stays
within expected bounds.
Also discard duplicate packets in the advanced mode, which appear
to be very common. If we don't, the spew detection kicks in far
too early. If we get a large spew of duplicates, request a
recalibration straight up.
Based on earlier work by Paul Fox.
Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Add a "hgpk_mode" sysfs attribute that allows selection between 3 options:
Mouse (the existing option), GlideSensor and PenTablet.
GlideSensor is an enhanced protocol for the regular touchpad mode that
additionally reports pressure and uses absolute coordinates. We suspect
that it may be more reliable than mouse mode in some environments.
PenTablet mode puts the touchpad into resistive mode, you must then use
a stylus as an input. We suspect this is the most reliable way to drive
the touchpad.
The GlideSensor and PenTablet devices expose themselves with the
intention of being combined with the synaptics X11 input driver.
Based on earlier work by Paul Fox.
Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
ixpad can never be NULL here; if it is NULL we would not have been bound to
the interface and then why would we be called?
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Properly free the resources in error path by the reverse order of resource
allocation.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
In xpad_led_disconnect(), what we really want is to kfree(xpad_led).
In xpad_disconnect(), add a missing kfree(xpad->bdata) to fix the memory leak.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
In current implementation, xpad_probe return 0 when
usb_alloc_urb failed for xpad->bulk_out and kzalloc failed for xpad->bdata.
This patch removes the initialization for error variable,
assign the error code at the place the error happens instead.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Sometimes input handlers (as opposed to input devices) have a need to
inject (or re-inject) events back into input core. For example sysrq
filter may want to inject previously suppressed Alt-SysRq so that user
can take a screen print. In this case we do not want to pass such events
back to the same same handler that injected them to avoid loops.
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
It is not allowed to call input_free_device() after calling
input_unregister_device() because input devices are refcounted and
unregister will free the device if we were holding he last referenc.
The preferred style in input/ is to make input_register_device() the
last function in the probe which can fail. That way we don't need to
call input_unregister_device().
Also do not need to call input_set_drvdata() as nothing in the driver
uses the data.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
