| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Add a forgotten include that I've by mistake omitted when resolving
merge conflict in ead0687fe30 ("HID: i2c-hid: support regulator power
on/off").
Fixes: ead0687fe30 ("HID: i2c-hid: support regulator power on/off")
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
The HID report buffers that are initially allocated on i2c_hid_probe()
might not be big enough to hold the HID reports from a specific device,
in which case they will be freed and new ones will be allocated in
i2c_hid_start(), at point which the device's report size is known. But
at this point ihid->irq is already running, and may call
i2c_hid_get_input() which passes ihid->inbuf to i2c_master_recv(). Since
this handler runs in a separate thread, ihid->inbuf may be freed at this
very moment, and i2c_master_recv() will write on memory which may be
already owned by a different part of the kernel, corrupting its data.
This problem has been observed on an Asus UX360UA laptop which has an
I2C touchpad, and results in a complete system freeze or an unusable
slowness with a lof of "BUG: unable to handle kernel paging request at
<address>" warnings. Enabling SLUB debugging shows a use-after-free
warning on memory allocated in i2c_hid_alloc_buffers() and freed in
i2c_hid_free_buffers():
=============================================================================
BUG kmalloc-64 (Not tainted): Poison overwritten
-----------------------------------------------------------------------------
Disabling lock debugging due to kernel taint
INFO: 0xffff880264083273-0xffff88026408329e. first byte 0x0 instead of 0x6b
INFO: Allocated in i2c_hid_alloc_buffers+0x25/0xa0 [i2c_hid] age=35793 cpu=2 pid=430
___slab_alloc+0x41e/0x460
__slab_alloc+0x20/0x40
__kmalloc+0x210/0x280
i2c_hid_alloc_buffers+0x25/0xa0 [i2c_hid]
i2c_hid_probe+0x12f/0x5e0 [i2c_hid]
i2c_device_probe+0x10a/0x1b0
driver_probe_device+0x220/0x4a0
__device_attach_driver+0x71/0xa0
bus_for_each_drv+0x67/0xb0
__device_attach+0xdc/0x170
device_initial_probe+0x13/0x20
bus_probe_device+0x92/0xa0
device_add+0x4aa/0x670
device_register+0x1a/0x20
i2c_new_device+0x18e/0x230
acpi_i2c_add_device+0x1a0/0x210
INFO: Freed in i2c_hid_free_buffers+0x16/0x60 [i2c_hid] age=7552 cpu=1 pid=1473
__slab_free+0x221/0x330
kfree+0x139/0x160
i2c_hid_free_buffers+0x16/0x60 [i2c_hid]
i2c_hid_start+0x2a9/0x2df [i2c_hid]
mt_probe+0x160/0x22e [hid_multitouch]
hid_device_probe+0xd7/0x150 [hid]
driver_probe_device+0x220/0x4a0
__driver_attach+0x84/0x90
bus_for_each_dev+0x6c/0xc0
driver_attach+0x1e/0x20
bus_add_driver+0x1c3/0x280
driver_register+0x60/0xe0
__hid_register_driver+0x53/0x90 [hid]
0xffffffffc004f01e
do_one_initcall+0xb3/0x1f0
do_init_module+0x5f/0x1d0
INFO: Slab 0xffffea0009902080 objects=20 used=20 fp=0x (null) flags=0x17fff8000004080
INFO: Object 0xffff880264083260 @offset=4704 fp=0x (null)
Bytes b4 ffff880264083250: 8d e6 fe ff 00 00 00 00 5a 5a 5a 5a 5a 5a 5a 5a ........ZZZZZZZZ
Object ffff880264083260: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkkkkkkkkkkkkk
Object ffff880264083270: 6b 6b 6b 00 00 00 00 00 00 00 00 00 00 00 00 00 kkk.............
Object ffff880264083280: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Object ffff880264083290: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Redzone ffff8802640832a0: bb bb bb bb bb bb bb bb ........
Padding ffff8802640833e0: 5a 5a 5a 5a 5a 5a 5a 5a ZZZZZZZZ
CPU: 1 PID: 1503 Comm: python3 Tainted: G B 4.4.21+ #10
Hardware name: ASUSTeK COMPUTER INC. UX360UA/UX360UA, BIOS UX360UA.200 05/05/2016
0000000000000086 00000000622d48a2 ffff88026061ba38 ffffffff813f6044
ffff880264082010 ffff880264083260 ffff88026061ba78 ffffffff811e8eab
0000000000000008 ffff880200000001 ffff88026408329f ffff88026a007700
Call Trace:
[<ffffffff813f6044>] dump_stack+0x63/0x8f
[<ffffffff811e8eab>] print_trailer+0x14b/0x1f0
[<ffffffff811e94c1>] check_bytes_and_report+0xc1/0x100
[<ffffffff811e96c4>] check_object+0x1c4/0x240
[<ffffffff81293fde>] ? ext4_htree_store_dirent+0x3e/0x120
[<ffffffff811e9b44>] alloc_debug_processing+0x104/0x180
[<ffffffff811eb7be>] ___slab_alloc+0x41e/0x460
[<ffffffff81293fde>] ? ext4_htree_store_dirent+0x3e/0x120
[<ffffffff8124590b>] ? __getblk_gfp+0x2b/0x60
[<ffffffff8129b969>] ? ext4_getblk+0xa9/0x190
[<ffffffff811eb820>] __slab_alloc+0x20/0x40
[<ffffffff811ed320>] __kmalloc+0x210/0x280
[<ffffffff81293fde>] ? ext4_htree_store_dirent+0x3e/0x120
[<ffffffff812c1602>] ? ext4fs_dirhash+0xc2/0x2a0
[<ffffffff81293fde>] ext4_htree_store_dirent+0x3e/0x120
[<ffffffff812a4f47>] htree_dirblock_to_tree+0x187/0x1b0
[<ffffffff812a5fd2>] ext4_htree_fill_tree+0xb2/0x2e0
[<ffffffff811ebb7a>] ? kmem_cache_alloc_trace+0x1fa/0x220
[<ffffffff81293e45>] ? ext4_readdir+0x775/0x8b0
[<ffffffff81293cb1>] ext4_readdir+0x5e1/0x8b0
[<ffffffff81221c82>] iterate_dir+0x92/0x120
[<ffffffff81222118>] SyS_getdents+0x98/0x110
[<ffffffff81221d10>] ? iterate_dir+0x120/0x120
[<ffffffff818157f2>] entry_SYSCALL_64_fastpath+0x16/0x71
FIX kmalloc-64: Restoring 0xffff880264083273-0xffff88026408329e=0x6b
FIX kmalloc-64: Marking all objects used
Signed-off-by: João Paulo Rechi Vita <jprvita@endlessm.com>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Function hid_post_reset() should return negative error codes on failures.
However, in its implementation, it incorrectly returns 1. This patch fixes the
bug, returning proper error codes on failures.
Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
post input_sync only when there are input events posted
Signed-off-by: Ping Cheng <ping.cheng@wacom.com>
Reviewed-By: Jason Gerecke <jason.gerecke@wacom.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Make sure everything reported from pad are registered
Signed-off-by: Ping Cheng <ping.cheng@wacom.com>
Reviewed-By: Jason Gerecke <jason.gerecke@wacom.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Sometime valid events may not be supported by the driver yet. Make
sure we don't process them when the code is not ready.
This fix prevents a kernel panic due to unsupported HID events.
Signed-off-by: Ping Cheng <ping.cheng@wacom.com>
Reviewed-By: Jason Gerecke <jason.gerecke@wacom.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
It is unnecessary to return a value since nothing is expecting a
value from it.
Signed-off-by: Ping Cheng <ping.cheng@wacom.com>
Reviewed-By: Jason Gerecke <jason.gerecke@wacom.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
We have the %pU printf extension for doing exactly this. Saves some
.text, and is likely also a little faster.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Since the gpio-as-irq support has been added, the driver now depends
on proper support being available in gpiolib.
Fixes: 13de9cca514e ("HID: cp2112: add IRQ chip handling")
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Update the hid-asus module to add multitouch support for the Asus i2c touchpad.
This patch aims to resolve the issue raised here:
https://bugzilla.kernel.org/show_bug.cgi?id=120181
The issue is in relation to an Asus touchpad device which currently does not
have multitouch support.
The device currently falls through to the hid-generic driver which
treats the device as a mouse.
This patch aims to add the multitouch support.
[jkosina@suse.cz: move most of the 'patch comment' into actual changelog]
[jkosina@suse.cz: drop hunk that changes ->name of the driver]
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Brendan McGrath <redmcg@redmandi.dyndns.org>
Signed-off-by: Victor Vlasenko <victor.vlasenko@sysgears.com>
Signed-off-by: Frederik Wenigwieser <frederik.wenigwieser@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Although unlikely but it is possible that when a connect or disconnect
request is issued to the firmware, before the response comes, user
terminates the client session. In this case when the response is arrived
there is no matching client instance in the list of currently active
clients. In this case, don't issue call to wake up a waiting client.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Add support for USB based DS4 dongle device, which allows connecting
a DS4 through Bluetooth, but hides Bluetooth from the host system.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
The DS4 side of hid-sony used the hid-core layer to assign buttons
and axes based on the HID report descriptors. The default mapping
was strange e.g. right stick using ABS_Z/ABS_RZ or the physical
'south button' being reported as BTN_EAST etcetera.
This patch makes the DS4 side ofi the hid-sony driver comply to
the Linux game controller spec as suggested in a discussion with
Dmitry on the linux-input list.
Currently the main user of the DS4 is the SDL2 library, which has
a mapping table using vendor/device/version as a key. In order to
not break SDL2 we discussed adjusting the version number, so it
can have both mappings. This was discust on linux-input and we
discussed privately with SDL2 developers.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
The dualshock 4 supports both analog sticks of which one uses
ABS_X/_Y and a touchpad. In a recent discussion with Dmitry about
some input-mt changes we proposed for disabling pointer emulation from
input_mt_sync_frame, Dmitry mentioned ABS_X/_Y should report the
same data as ABS_MT_POSITION_X/_Y. The current driver is mixing axes
for different subdevices. It was suggested to make the touchpad
its own sub-device.
This patch turns the touchpad into its own device. In addition
this patch also moves the button underneath the touchpad into
the new device. It felt like this button should be part of the
device. No known user space application (not even SDL2) seems to
be using it.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
A previous patch moved most input initialization from sony_probe to
sony_input_configured to avoid some race conditions. The driver has some
special logic to prevent the device to get registered twice in case the
user connects it both over Bluetooth and USB. When this condition
happens sony_input_configured returns a failure, but sony_probe continues
as hid_hw_start doesn't fail. As was discussed on linux-input, it is
acceptable for this function to fail.
This patch adds a check for the HID_CLAIMED_INPUT flag within sony_probe
to determine whether initialization succeeded correctly. The flag is
not set by the HID layer when sony_input_configured fails.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
The GPIO part doesn't provide interrupts when GPIO are toggled.
So use a polling mechanism if someone requests a GPIO as an IRQ.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Instead of forcing the level trigger of the IRQ, we can count
on ACPI or OF to set it up for us.
The first release of the HID over I2C specification mentioned
that the level trigger needed to be active low. In the latest
version of the specification, there is no such explicit mention,
so it's better to not assume one.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
We already have in place a quirk for Windows 8 devices, but it looks
like the Surface Cover are not conforming to it.
Given that we are only interested in 3 feature reports (the ones that
the Windows driver retrieves), we should be safe to unconditionally apply
the quirk to everybody.
In case there is an issue with a controller, we can always mark it as such
in the transport driver, and hid-multitouch won't try to retrieve the
feature report.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
There is no reasons to filter out keyboard and consumer control collections
in hid-multitouch.
With the previous hid-input fix, there is now a full support of the Type
Cover and we can remove all specific bits from hid-core and hid-microsoft.
hid-multitouch will automatically set HID_QUIRK_NO_INIT_REPORTS so we can
also remove it from the list of ushbid quirks.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
The purpose of HID_QUIRK_MULTI_INPUT is to have an input device per
report id. This is useful when the HID device presents several HID
collections of different device types.
The current implementation of hid-input creates one input node per id per
type (input or output). This is problematic for the LEDs of a keyboard as
they are often set through an output report. The current code creates
one input node with all the keyboard keys, and one other with only the
LEDs.
To solve this, we use a two-passes way:
- first, we initialize all input nodes and associate one per report id
- then, we register all the input nodes
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
According to https://msdn.microsoft.com/en-us/library/windows/hardware/mt604195(v=vs.85).aspx
external buttons have some weird usage mapping:
- Button 2 Indicates Button State for external button for primary
(default left) clicking.
- Button 3 Indicates Button State for external button for secondary
(default right) clicking.
So in the current state, the buttons are mapped to right and middle.
Move the usage by one to correctly map the external buttons.
Tested-by: Chris Chiu <chiu@endlessm.com>
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
One more device requiring a quirk :/
Reported-by: Christian-Nils Boda <christian-nils.boda@gadz.org>
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
One more device requiring a quirk :/
[jkosina@suse.cz: update comment based on Bastien's remark]
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Tested-by: Bastien Nocera <hadess@hadess.net>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Otherwise each individual rotator char would be printed in a new line:
(...)
[ 0.642350] -
[ 0.644374] |
[ 0.646367] -
(...)
Signed-off-by: Nicolas Schichan <nicolas.schichan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
For large values of "mult" and long uptimes, the intermediate
result of "cycles * mult" can overflow 64 bits. For example,
the tile platform calls clocksource_cyc2ns with a 1.2 GHz clock;
we have mult = 853, and after 208.5 days, we overflow 64 bits.
Since clocksource_cyc2ns() is intended to be used for relative
cycle counts, not absolute cycle counts, performance is more
importance than accepting a wider range of cycle values. So,
just use mult_frac() directly in tile's sched_clock().
Commit 4cecf6d401a0 ("sched, x86: Avoid unnecessary overflow
in sched_clock") by Salman Qazi results in essentially the same
generated code for x86 as this change does for tile. In fact,
a follow-on change by Salman introduced mult_frac() and switched
to using it, so the C code was largely identical at that point too.
Peter Zijlstra then added mul_u64_u32_shr() and switched x86
to use it. This is, in principle, better; by optimizing the
64x64->64 multiplies to be 32x32->64 multiplies we can potentially
save some time. However, the compiler piplines the 64x64->64
multiplies pretty well, and the conditional branch in the generic
mul_u64_u32_shr() causes some bubbles in execution, with the
result that it's pretty much a wash. If tilegx provided its own
implementation of mul_u64_u32_shr() without the conditional branch,
we could potentially save 3 cycles, but that seems like small gain
for a fair amount of additional build scaffolding; no other platform
currently provides a mul_u64_u32_shr() override, and tile doesn't
currently have an <asm/div64.h> header to put the override in.
Additionally, gcc currently has an optimization bug that prevents
it from recognizing the opportunity to use a 32x32->64 multiply,
and so the result would be no better than the existing mult_frac()
until such time as the compiler is fixed.
For now, just using mult_frac() seems like the right answer.
Cc: stable@kernel.org [v3.4+]
Signed-off-by: Chris Metcalf <cmetcalf@mellanox.com>
|
|
When user tried to read some fields like hysteresis from IIO sysfs on some
systems, it fails. The reason is that this field is a byte field and caller
of sensor_hub_get_feature() passes a buffer of 4 bytes. Here the function
sensor_hub_get_feature() copies the single byte from the report to the
caller buffer and returns "1" as the number of bytes copied. So caller
can use the return value.
But this is done by multiple callers, so if we just change the
sensor_hub_get_feature so that caller buffer is initialized with 0s
then we don't to change all functions.
Signed-off-by: Song Hongyan <hongyan.song@intel.com>
Acked-by: Jonathan Cameron <jic23@kernel.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Adding support for not JP versions of the Microsoft Surface 4 Type Cover Pro
[jkosina@suse.cz: The identical patch has been sent by Jeff Farthing, so I am
including his signoff as well]
Signed-off-by: Jeff Farthing <jeff@jfarthing.com>
Signed-off-by: Daniel Keller <daniel.keller@gcd.de>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Kernel v4.9 strictly enforces DMA capable buffers, so we need to remove
buffers allocated on the stack.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Kernel v4.9 strictly enforces DMA capable buffers, so we need to remove
buffers allocated on the stack.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Kernel v4.9 strictly enforces DMA capable buffers, so we need to remove
buffers allocated on the stack.
[jkosina@suse.cz: fix up second usage of hid_hw_raw_request(), spotted by
0day build bot]
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
Kernel v4.9 strictly enforces DMA capable buffers, so we need to remove
buffers allocated on the stack.
Use a spinlock to prevent concurrent accesses to the buffer.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
|
|
A correct bugfix introduced a harmless warning that shows up with gcc-7:
fs/nfs/callback.c: In function 'nfs_callback_up':
fs/nfs/callback.c:214:14: error: array subscript is outside array bounds [-Werror=array-bounds]
What happens here is that the 'minorversion == 0' check tells the
compiler that we assume minorversion can be something other than 0,
but when CONFIG_NFS_V4_1 is disabled that would be invalid and
result in an out-of-bounds access.
The added check for IS_ENABLED(CONFIG_NFS_V4_1) tells gcc that this
really can't happen, which makes the code slightly smaller and also
avoids the warning.
The bugfix that introduced the warning is marked for stable backports,
we want this one backported to the same releases.
Fixes: 98b0f80c2396 ("NFSv4.x: Fix a refcount leak in nfs_callback_up_net")
Cc: stable@vger.kernel.org # v3.7+
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Group validation expects all events to be of the same PMU; however
is_uncore_pmu() is too wide, it matches _all_ uncore events, even
across PMUs.
This triggers failure when we group different events from different
uncore PMUs, like:
perf stat -vv -e '{uncore_cbox_0/config=0x0334/,uncore_qpi_0/event=1/}' -a sleep 1
Fix is_uncore_pmu() by only matching events to the box at hand.
Note that generic code; ran after this step; will disallow this
mixture of PMU events.
Reported-by: Jiri Olsa <jolsa@redhat.com>
Tested-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vince@deater.net>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/20161118125354.GQ3117@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Vince Weaver reported that perf_fuzzer + KASAN detects that PEBS event
unwinds sometimes do 'weird' things. In particular, we seemed to be
ending up unwinding from random places on the NMI stack.
While it was somewhat expected that the event record BP,SP would not
match the interrupt BP,SP in that the interrupt is strictly later than
the record event, it was overlooked that it could be on an already
overwritten stack.
Therefore, don't copy the recorded BP,SP over the interrupted BP,SP
when we need stack unwinds.
Note that its still possible the unwind doesn't full match the actual
event, as its entirely possible to have done an (I)RET between record
and interrupt, but on average it should still point in the general
direction of where the event came from. Also, it's the best we can do,
considering.
The particular scenario that triggered the bogus NMI stack unwind was
a PEBS event with very short period, upon enabling the event at the
tail of the PMI handler (FREEZE_ON_PMI is not used), it instantly
triggers a record (while still on the NMI stack) which in turn
triggers the next PMI. This then causes back-to-back NMIs and we'll
try and unwind the stack-frame from the last NMI, which obviously is
now overwritten by our own.
Analyzed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Reported-by: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@gmail.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: davej@codemonkey.org.uk <davej@codemonkey.org.uk>
Cc: dvyukov@google.com <dvyukov@google.com>
Cc: stable@vger.kernel.org
Fixes: ca037701a025 ("perf, x86: Add PEBS infrastructure")
Link: http://lkml.kernel.org/r/20161117171731.GV3157@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
The following commit:
75925e1ad7f5 ("perf/x86: Optimize stack walk user accesses")
... switched from copy_from_user_nmi() to __copy_from_user_nmi() with a manual
access_ok() check.
Unfortunately, copy_from_user_nmi() does an explicit check against TASK_SIZE,
whereas the access_ok() uses whatever the current address limit of the task is.
We are getting NMIs when __probe_kernel_read() has switched to KERNEL_DS, and
then see vmalloc faults when we access what looks like pointers into vmalloc
space:
[] WARNING: CPU: 3 PID: 3685731 at arch/x86/mm/fault.c:435 vmalloc_fault+0x289/0x290
[] CPU: 3 PID: 3685731 Comm: sh Tainted: G W 4.6.0-5_fbk1_223_gdbf0f40 #1
[] Call Trace:
[] <NMI> [<ffffffff814717d1>] dump_stack+0x4d/0x6c
[] [<ffffffff81076e43>] __warn+0xd3/0xf0
[] [<ffffffff81076f2d>] warn_slowpath_null+0x1d/0x20
[] [<ffffffff8104a899>] vmalloc_fault+0x289/0x290
[] [<ffffffff8104b5a0>] __do_page_fault+0x330/0x490
[] [<ffffffff8104b70c>] do_page_fault+0xc/0x10
[] [<ffffffff81794e82>] page_fault+0x22/0x30
[] [<ffffffff81006280>] ? perf_callchain_user+0x100/0x2a0
[] [<ffffffff8115124f>] get_perf_callchain+0x17f/0x190
[] [<ffffffff811512c7>] perf_callchain+0x67/0x80
[] [<ffffffff8114e750>] perf_prepare_sample+0x2a0/0x370
[] [<ffffffff8114e840>] perf_event_output+0x20/0x60
[] [<ffffffff8114aee7>] ? perf_event_update_userpage+0xc7/0x130
[] [<ffffffff8114ea01>] __perf_event_overflow+0x181/0x1d0
[] [<ffffffff8114f484>] perf_event_overflow+0x14/0x20
[] [<ffffffff8100a6e3>] intel_pmu_handle_irq+0x1d3/0x490
[] [<ffffffff8147daf7>] ? copy_user_enhanced_fast_string+0x7/0x10
[] [<ffffffff81197191>] ? vunmap_page_range+0x1a1/0x2f0
[] [<ffffffff811972f1>] ? unmap_kernel_range_noflush+0x11/0x20
[] [<ffffffff814f2056>] ? ghes_copy_tofrom_phys+0x116/0x1f0
[] [<ffffffff81040d1d>] ? x2apic_send_IPI_self+0x1d/0x20
[] [<ffffffff8100411d>] perf_event_nmi_handler+0x2d/0x50
[] [<ffffffff8101ea31>] nmi_handle+0x61/0x110
[] [<ffffffff8101ef94>] default_do_nmi+0x44/0x110
[] [<ffffffff8101f13b>] do_nmi+0xdb/0x150
[] [<ffffffff81795187>] end_repeat_nmi+0x1a/0x1e
[] [<ffffffff8147daf7>] ? copy_user_enhanced_fast_string+0x7/0x10
[] [<ffffffff8147daf7>] ? copy_user_enhanced_fast_string+0x7/0x10
[] [<ffffffff8147daf7>] ? copy_user_enhanced_fast_string+0x7/0x10
[] <<EOE>> <IRQ> [<ffffffff8115d05e>] ? __probe_kernel_read+0x3e/0xa0
Fix this by moving the valid_user_frame() check to before the uaccess
that loads the return address and the pointer to the next frame.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: linux-kernel@vger.kernel.org
Fixes: 75925e1ad7f5 ("perf/x86: Optimize stack walk user accesses")
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Exactly because for_each_thread() in autogroup_move_group() can't see it
and update its ->sched_task_group before _put() and possibly free().
So the exiting task needs another sched_move_task() before exit_notify()
and we need to re-introduce the PF_EXITING (or similar) check removed by
the previous change for another reason.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hartsjc@redhat.com
Cc: vbendel@redhat.com
Cc: vlovejoy@redhat.com
Link: http://lkml.kernel.org/r/20161114184612.GA15968@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
The PF_EXITING check in task_wants_autogroup() is no longer needed. Remove
it, but see the next patch.
However the comment is correct in that autogroup_move_group() must always
change task_group() for every thread so the sysctl_ check is very wrong;
we can race with cgroups and even sys_setsid() is not safe because a task
running with task_group() == ag->tg must participate in refcounting:
int main(void)
{
int sctl = open("/proc/sys/kernel/sched_autogroup_enabled", O_WRONLY);
assert(sctl > 0);
if (fork()) {
wait(NULL); // destroy the child's ag/tg
pause();
}
assert(pwrite(sctl, "1\n", 2, 0) == 2);
assert(setsid() > 0);
if (fork())
pause();
kill(getppid(), SIGKILL);
sleep(1);
// The child has gone, the grandchild runs with kref == 1
assert(pwrite(sctl, "0\n", 2, 0) == 2);
assert(setsid() > 0);
// runs with the freed ag/tg
for (;;)
sleep(1);
return 0;
}
crashes the kernel. It doesn't really need sleep(1), it doesn't matter if
autogroup_move_group() actually frees the task_group or this happens later.
Reported-by: Vern Lovejoy <vlovejoy@redhat.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hartsjc@redhat.com
Cc: vbendel@redhat.com
Link: http://lkml.kernel.org/r/20161114184609.GA15965@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
The aliasing check in map_and_copy is no longer necessary because
the IPsec ESP code no longer provides an IV that points into the
actual request data. As this check is now triggering BUG checks
due to the vmalloced stack code, I'm removing it.
Reported-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Recently an init call was added to hash_recvmsg so as to reset
the hash state in case a sendmsg call was never made.
Unfortunately this ended up clobbering the result if the previous
sendmsg was done with a MSG_MORE flag. This patch fixes it by
excluding that case when we make the init call.
Fixes: a8348bca2944 ("algif_hash - Fix NULL hash crash with shash")
Reported-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
We need to zero out the private data area when application switches
connection to different algorithm (TCP_CONGESTION setsockopt).
When congestion ops get assigned at connect time everything is already
zeroed because sk_alloc uses GFP_ZERO flag. But in the setsockopt case
this contains whatever previous cc placed there.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The tc could return NET_XMIT_CN as one congestion notification, but
it does not mean the packe is lost. Other modules like ipvlan,
macvlan, and others treat NET_XMIT_CN as success too.
So l2tp_eth_dev_xmit should add the NET_XMIT_CN check.
Signed-off-by: Gao Feng <gfree.wind@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
While walking the list of lock_states, keep a reference on each
nfs4_lock_state to be checked, otherwise the lock state could be removed
while the check performs TEST_STATEID and possible FREE_STATEID.
Signed-off-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
There's not much point, except compile test, enabling the stmmac
platform drivers unless the STM32 SoC is enabled. It's not
useful without it.
Signed-off-by: Peter Robinson <pbrobinson@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Commit 3105f234e0aba43e44e277c20f9b32ee8add43d4 replaced module
cpu id table with a cpu feature check, which is logically correct.
But we need the module device table to allow module auto loading.
Cc: stable@vger.kernel.org # 4.8
Fixes:3105f234 thermal/powerclamp: correct cpu support check
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
|
|
The token table passed into match_token() must be null-terminated, which
it currently is not in the perf's address filter string parser, as caught
by Vince's perf_fuzzer and KASAN.
It doesn't blow up otherwise because of the alignment padding of the table
to the next element in the .rodata, which is luck.
Fixing by adding a null-terminator to the token table.
Reported-by: Vince Weaver <vincent.weaver@maine.edu>
Tested-by: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: dvyukov@google.com
Cc: stable@vger.kernel.org # v4.7+
Fixes: 375637bc524 ("perf/core: Introduce address range filtering")
Link: http://lkml.kernel.org/r/877f81f264.fsf@ashishki-desk.ger.corp.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Rename the watchdog platform library file to explicitly show that is used only
on Intel Merrifield platforms.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20161118172723.179761-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Since the bootloader may load the compressed x86 kernel at any address,
it should always be built as PIE, not just when CONFIG_RELOCATABLE=y.
Otherwise, linker in binutils 2.27 will optimize GOT load into the
absolute address when building the compressed x86 kernel as a non-PIE
executable.
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
[ Small wording changes. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Watchdog device in Intel Tangier relies on SCU to be present. It uses the SCU
IPC channel to send commands and receive responses. If watchdog driver is
initialized quite before SCU and a command has been sent the result is always
an error like the following:
intel_mid_wdt: Error stopping watchdog: 0xffffffed
Register watchdog device whne SCU is ready to avoid described issue.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20161118165224.175514-1-andriy.shevchenko@linux.intel.com
[ Small cleanups. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Robert O'Callahan reported that after an execve PTRACE_GETREGSET
NT_X86_XSTATE continues to return the pre-exec register values
until the exec'ed task modifies FPU state.
The test code is at:
https://bugzilla.redhat.com/attachment.cgi?id=1164286.
What is happening is fpu__clear() does not properly clear fpstate.
Fix it by doing just that.
Reported-by: Robert O'Callahan <robert@ocallahan.org>
Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
Cc: <stable@vger.kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@suse.de>
Cc: David Hansen <dave.hansen@linux.intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi V. Shankar <ravi.v.shankar@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1479402695-6553-1-git-send-email-yu-cheng.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Linux will have all kinds of sporadic problems on systems that don't
have the CPUID instruction unless CONFIG_M486=y. In particular,
sync_core() will explode.
I believe that these kernels had a better chance of working before
commit 05fb3c199bb0 ("x86/boot: Initialize FPU and X86_FEATURE_ALWAYS
even if we don't have CPUID"). That commit inadvertently fixed a
serious bug: we used to fail to detect the FPU if CPUID wasn't
present. Because we also used to forget to set X86_FEATURE_ALWAYS, we
end up with no cpu feature bits set at all. This meant that
alternative patching didn't do anything and, if paravirt was disabled,
we could plausibly finish the entire boot process without calling
sync_core().
Rather than trying to work around these issues, just have the kernel
fail loudly if it's running on a CPUID-less 486, doesn't have CPUID,
and doesn't have CONFIG_M486 set.
Reported-by: Matthew Whitehead <tedheadster@gmail.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/70eac6639f23df8be5fe03fa1984aedd5d40077a.1479598603.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
Jiri Kosina
João Paulo Rechi Vita
Pan Bian
Ping Cheng
Rasmus Villemoes
Brendan McGrath
Srinivas Pandruvada
Roderick Colenbrander
Benjamin Tissoires
Nicolas Schichan
Chris Metcalf
Song Hongyan
Daniel Keller
Arnd Bergmann
Peter Zijlstra
Johannes Weiner
Oleg Nesterov
Herbert Xu
Florian Westphal
Gao Feng
Benjamin Coddington
Peter Robinson
Jacob Pan
Alexander Shishkin
Andy Shevchenko
H.J. Lu
Yu-cheng Yu
Andy Lutomirski