In order to avoid confusing the HW, we must never submit an empty ring
during lite-restore, that is we should always advance the RING_TAIL
before submitting to stay ahead of the RING_HEAD.
Normally this is prevented by keeping a couple of spare NOPs in the
request->wa_tail so that on resubmission we can advance the tail. This
relies on the request only being resubmitted once, which is the normal
condition as it is seen once for ELSP[1] and then later in ELSP[0]. On
preemption, the requests are unwound and the tail reset back to the
normal end point (as we know the request is incomplete and therefore its
RING_HEAD is even earlier).
However, if this w/a should fail we would try and resubmit the request
with the RING_TAIL already set to the location of this request's wa_tail
potentially causing a GPU hang. We can spot when we do try and
incorrectly resubmit without advancing the RING_TAIL and spare any
embarrassment by forcing the context restore.
In the case of preempt-to-busy, we leave the requests running on the HW
while we unwind. As the ring is still live, we cannot rewind our
rq->tail without forcing a reload so leave it set to rq->wa_tail and
only force a reload if we resubmit after a lite-restore. (Normally, the
forced reload will be a part of the preemption event.)
Fixes: 22b7a426bb ("drm/i915/execlists: Preempt-to-busy")
Closes: https://gitlab.freedesktop.org/drm/intel/issues/673
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: stable@kernel.vger.org
Link: https://patchwork.freedesktop.org/patch/msgid/20191209023215.3519970-1-chris@chris-wilson.co.uk
Add missing endpoint-type sanity checks to probe.
This specifically prevents a warning in USB core on URB submission when
fuzzing USB descriptors.
Signed-off-by: Johan Hovold <johan@kernel.org>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20191202085610.12719-4-johan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver failed to initialise its receive-buffer pointer, something
which could lead to an illegal free on late probe errors.
Fix this by making sure to clear all driver data at allocation.
Fixes: 2032e2c230 ("usb_gigaset: code cleanup")
Cc: stable <stable@vger.kernel.org> # 2.6.33
Cc: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: Johan Hovold <johan@kernel.org>
Link: https://lore.kernel.org/r/20191202085610.12719-3-johan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix a general protection fault when accessing the endpoint descriptors
which could be triggered by a malicious device due to missing sanity
checks on the number of endpoints.
Reported-by: syzbot+35b1c403a14f5c89eba7@syzkaller.appspotmail.com
Fixes: 07dc1f9f2f ("[PATCH] isdn4linux: Siemens Gigaset drivers - M105 USB DECT adapter")
Cc: stable <stable@vger.kernel.org> # 2.6.17
Cc: Hansjoerg Lipp <hjlipp@web.de>
Cc: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: Johan Hovold <johan@kernel.org>
Link: https://lore.kernel.org/r/20191202085610.12719-2-johan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This undoes the previous call to alloc_chrdev_region() on failure,
and is probably what was meant originally given the label name.
Signed-off-by: Marcelo Diop-Gonzalez <marcgonzalez@google.com>
Cc: stable <stable@vger.kernel.org>
Fixes: 187ac53e59 ("staging: vchiq_arm: rework probe and init functions")
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Link: https://lore.kernel.org/r/20191203153921.70540-1-marcgonzalez@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
`rename_file' was exported but not properly namespaced causing a
multiple definition error because `rename_file' is already defined in
fs/hostfs/hostfs_user.c:
ld: drivers/staging/exfat/exfat_core.o: in function `rename_file':
drivers/staging/exfat/exfat_core.c:2327: multiple definition of
`rename_file'; fs/hostfs/hostfs_user.o:fs/hostfs/hostfs_user.c:350:
first defined here
make: *** [Makefile:1077: vmlinux] Error 1
This error can be reproduced on ARCH=um by selecting:
CONFIG_EXFAT_FS=y
CONFIG_HOSTFS=y
Add a namespace prefix exfat_* to fix this error.
Reported-by: Brendan Higgins <brendanhiggins@google.com>
Signed-off-by: Brendan Higgins <brendanhiggins@google.com>
Cc: stable <stable@vger.kernel.org>
Cc: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Tested-by: David Gow <davidgow@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Link: https://lore.kernel.org/r/20191204234522.42855-1-brendanhiggins@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
wlan-ng uses the function crc32_le,
but CRC32 wasn't a dependency of wlan-ng
Co-developed-by: Michael Kupfer <michael.kupfer@fau.de>
Signed-off-by: Michael Kupfer <michael.kupfer@fau.de>
Signed-off-by: Kay Friedrich <kay.friedrich@fau.de>
Link: https://lore.kernel.org/r/20191127112457.2301-1-kay.friedrich@fau.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It should depends on ETHERNET, otherwise building fails:
drivers/staging/hp/hp100.o: In function `hp100_pci_remove':
hp100.c:(.text+0x165): undefined reference to `unregister_netdev'
hp100.c:(.text+0x214): undefined reference to `free_netdev'
Fixes: 52340b82cf ("hp100: Move 100BaseVG AnyLAN driver to staging")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Link: https://lore.kernel.org/r/20191113021306.35464-1-yuehaibing@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current use of the mode flag SPI_CS_HIGH is fragile: it
overwrites anything already assigned by the SPI core.
Assign ^= SPI_CS_HIGH since we might be active high
already, and that is usually the case with GPIOs used
for chip select, even if they are in practice active low.
Add a comment clarifying why ^= SPI_CS_HIGH is the right
choice here.
Reported-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20191204233230.22309-1-linus.walleij@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently the discard code in clr_alloc_bitmap() is just dead code.
Move code around so that the discard operation is properly attempted
when enabled.
Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Link: https://lore.kernel.org/r/20191205152913.GJ3276@xps-13
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add the Performance Monitoring Unit (PMU) device tree node to the H3
.dtsi, which tells DT users which interrupts are triggered by PMU
overflow events on each core. The numbers come from the manual and have
been checked in U-Boot and with perf in Linux.
Tested with perf record and taskset on an OrangePi Zero.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Add the Performance Monitoring Unit (PMU) device tree node to the H5
.dtsi, which tells DT users which interrupts are triggered by PMU
overflow events on each core.
As with the A64, the interrupt numbers from the manual were wrong (off
by 4), the actual SPI IDs have been gathered in U-Boot, and were
verified with perf in Linux.
Tested with perf record and taskset on an OrangePi PC2.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Add the Performance Monitoring Unit (PMU) device tree node to the H6
.dtsi, which tells DT users which interrupts are triggered by PMU
overflow events on each core. The numbers come from the manual and have
been checked in U-Boot and with perf in Linux.
Tested with perf record and taskset on a Pine H64.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Beelink X2 box comes with a remote. Add a mapping for it.
Signed-off-by: Jernej Skrabec <jernej.skrabec@siol.net>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Tanix TX6 box comes with a remote. Add a mapping for it.
Suggested-by: Michael Lange <linuxstuff@milaw.biz>
Signed-off-by: Jernej Skrabec <jernej.skrabec@siol.net>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Add new entry for rc-tanix-tx5max in linux,rc-map-name
Signed-off-by: Jernej Skrabec <jernej.skrabec@siol.net>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
This patch adds the model B of the PineH64.
The model B is smaller than the pine64 model A and has no PCIE slot.
The only devicetree difference with the pineH64 model A, is the PHY
regulator and the HDMI connector node.
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
The current sun50i-h6-pine-h64 DT does not specify which model (A or B)
it supports.
When this file was created, only modelA was existing, but now both model
exists and with the time, this DT drifted to support the model B since it is
the most common one.
Furtheremore, some part of the model A does not work with it like ethernet and
HDMI connector (as confirmed by Jernej on IRC).
So it is time to settle the issue, and the easiest way was to state that
this DT is for model B.
Easiest since only a small name changes is required.
Doing the opposite (stating this file is for model A) will add changes (for
ethernet and HDMI) and so, will break too many setup.
But as asked by the maintainer this patch state this file is for model A.
In the process this patch adds the missing compoments to made it work on
model A.
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Adds bindings for the new Emlid Neutis N5H3 board.
Signed-off-by: Georgii Staroselskii <georgii.staroselskii@emlid.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Emlid Neutis N5H3 is a version of Emlid Neutis SoM with H3 instead of H5
inside.
6eeb4180d4 ("ARM: dts: sunxi: h3-h5: Add Bananapi M2+ v1.2 device")
was used as reference.
Signed-off-by: Georgii Staroselskii <georgii.staroselskii@emlid.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
A new variant of Emlid Neutis has been inroduced. This one uses H3
instead of H5. The boards are essentially the same. This commit moves
non-SoC-specific parts out so that the common parts could be reused with
ease.
Signed-off-by: Georgii Staroselskii <georgii.staroselskii@emlid.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Function rtw_pwr_unassociated_idle returns boolean values.
Remove the return variable ret and the exit label to simplify the
function a little bit. Return true or false directly instead.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20191123151635.155138-3-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Replace tabs with spaces in declarations to cleanup whitespace.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20191123151635.155138-2-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch addresses CamelCase warnings in qlge.h under struct
mpi_coredump_global_header and mpi_coredump_segment_header. As
well ass addresses CamelCase warnings in qlge_dbg.c when the
structs are used.
Signed-off-by: Scott Schafer <schaferjscott@gmail.com>
Link: https://lore.kernel.org/r/20191209101908.23878-1-schaferjscott@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Create a static atomic counter for numerating cameras.
Use the Media Subsystem Kernel Internal API to create distinct
device-names, so that the camera-number (given by the counter)
matches the camera-name.
Co-developed-by: Kay Friedrich <kay.friedrich@fau.de>
Signed-off-by: Kay Friedrich <kay.friedrich@fau.de>
Signed-off-by: Michael Kupfer <michael.kupfer@fau.de>
Link: https://lore.kernel.org/r/20191206085432.19962-1-michael.kupfer@fau.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
match_string() returns the array index of a matching string.
Use it instead of the open-coded implementation.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Link: https://lore.kernel.org/r/20191107143223.44696-1-yuehaibing@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes:
"WARNING: block comment use * on subsequent lines"
"WARNING: block comments should align the * on each line"
"WARNING: block comments use a trailing */ on a separate line"
by checkpatch.pl.
Signed-off-by: Dorothea Ehrl <dorothea.ehrl@fau.de>
Co-developed-by: Vanessa Hack <vanessa.hack@fau.de>
Signed-off-by: Vanessa Hack <vanessa.hack@fau.de>
Link: https://lore.kernel.org/r/20191127123052.16424-5-dorothea.ehrl@fau.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes "WARNING: braces {} are not necessary for single
statement blocks" and "WARNING: braces {} are not necessary for any arm
of this statement" by checkpatch.pl.
Signed-off-by: Dorothea Ehrl <dorothea.ehrl@fau.de>
Co-developed-by: Vanessa Hack <vanessa.hack@fau.de>
Signed-off-by: Vanessa Hack <vanessa.hack@fau.de>
Link: https://lore.kernel.org/r/20191127123052.16424-4-dorothea.ehrl@fau.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes "CHECK: braces {} should be used on all arms of this
statement" by checkpatch.pl.
Signed-off-by: Dorothea Ehrl <dorothea.ehrl@fau.de>
Co-developed-by: Vanessa Hack <vanessa.hack@fau.de>
Signed-off-by: Vanessa Hack <vanessa.hack@fau.de>
Link: https://lore.kernel.org/r/20191127123052.16424-3-dorothea.ehrl@fau.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver forgets to call pci_release_regions() in probe failure
and remove.
Add the missed calls to fix it.
Signed-off-by: Chuhong Yuan <hslester96@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20191206075515.18581-1-hslester96@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ath.git patches for v5.6. Major changes:
ath11k
* a new driver for Qualcomm Wi-Fi 6 (IEEE 802.11ax) devices
ath10k
* significant improvements on receive throughput and firmware download
with SDIO bus
* report signal strength for each chain also on SDIO
* set max mtu to 1500 on SDIO devices
If an ocxl device is unbound through sysfs at the same time its AFU is
being opened by a user process, the open code may dereference freed
stuctures, which can lead to kernel oops messages. You'd have to hit a
tiny time window, but it's possible. It's fairly easy to test by
making the time window bigger artificially.
Fix it with a combination of 2 changes:
- when an AFU device is found in the IDR by looking for the device
minor number, we should hold a reference on the device until after
the context is allocated. A reference on the AFU structure is kept
when the context is allocated, so we can release the reference on
the device after the context allocation.
- with the fix above, there's still another even tinier window,
between the time the AFU device is found in the IDR and the
reference on the device is taken. We can fix this one by removing
the IDR entry earlier, when the device setup is removed, instead
of waiting for the 'release' device callback. With proper locking
around the IDR.
Fixes: 75ca758adb ("ocxl: Create a clear delineation between ocxl backend & frontend")
Cc: stable@vger.kernel.org # v5.2+
Signed-off-by: Frederic Barrat <fbarrat@linux.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190624144148.32022-1-fbarrat@linux.ibm.com
To parse the mac header make use of *struct ieee80211_hdr* instead of
extracting individual fields separately using pointer operation.
Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Link: https://lore.kernel.org/r/20191122205153.30723-3-adham.abozaeid@microchip.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed the unused compile type featurization. It's not recommended to
have compile type feature. Currently removing these defines as they are
not used. If any of these parameters are needed later should be added
using run time feature.
Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Link: https://lore.kernel.org/r/20191122205153.30723-2-adham.abozaeid@microchip.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
All multipliers and dividers are small.
Storing them in u8 instead of unsigned int reduces kernel size for a
generic kernel by ca. 0.5 KiB.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Link: https://lore.kernel.org/r/20191206133254.23800-1-geert+renesas@glider.be
Move the definition of acpi_get_wakeup_address() into sleep.c to break
linux/acpi.h's dependency (by way of asm/acpi.h) on asm/realmode.h.
Everyone and their mother includes linux/acpi.h, i.e. modifying
realmode.h results in a full kernel rebuild, which makes the already
inscrutable real mode boot code even more difficult to understand and is
positively rage inducing when trying to make changes to x86's boot flow.
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-13-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Convert acpi_wakeup_address from a raw variable into a function so that
x86 can wrap its dereference of the real mode boot header in a function
instead of broadcasting it to the world via a #define. This sets the
stage for a future patch to move x86's definition of the new function,
acpi_get_wakeup_address(), out of asm/acpi.h and thus break acpi.h's
dependency on asm/realmode.h.
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-12-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
None of the declarations in x86's acpi/sleep.h are in any way dependent
on the real mode boot code. Remove sleep.h's include of asm/realmode.h
to limit the dependencies on realmode.h to code that actually interacts
with the boot code.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-11-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Through a labyrinthian sequence of includes, usage of virt_to_phys() is
dependent on the include of asm/io.h in x86's asm/realmode.h, which is
included in x86's asm/acpi.h and thus by linux/acpi.h. Explicitly
include linux/io.h to break the dependency on realmode.h so that a
future patch can remove the realmode.h include from acpi.h without
breaking the build.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Acked-by: Mark Brown <broonie@kernel.org>
Link: https://lkml.kernel.org/r/20191126165417.22423-10-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Through a labyrinthian sequence of includes, usage of virt_to_phys() is
dependent on the include of asm/io.h in x86's asm/realmode.h, which is
included in x86's asm/acpi.h and thus by linux/acpi.h. Explicitly
include linux/io.h to break the dependency on realmode.h so that a
future patch can remove the realmode.h include from acpi.h without
breaking the build.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-9-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Through a labyrinthian sequence of includes, usage of page_to_phys(),
virt_to_phys() and out*() is dependent on the include of asm/io.h in
x86's asm/realmode.h, which is included in x86's asm/acpi.h and thus by
linux/acpi.h. Explicitly include linux/io.h to break the dependency on
realmode.h so that a future patch can remove the realmode.h include from
acpi.h without breaking the build.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-8-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Through a labyrinthian sequence of includes, usage of page_to_phys() is
dependent on the include of asm/io.h in x86's asm/realmode.h, which is
included in x86's asm/acpi.h and thus by linux/acpi.h. Explicitly
include linux/io.h to break the dependency on realmode.h so that a
future patch can remove the realmode.h include from acpi.h without
breaking the build.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-7-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Through a labyrinthian sequence of includes, usage of virt_to_phys() is
dependent on the include of asm/io.h in asm/realmode.h via asm/acpi.h.
Explicitly include asm/io.h to break the dependency on realmode.h so
that a future patch can remove the realmode.h include from acpi.h
without breaking the build.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Link: https://lkml.kernel.org/r/20191126165417.22423-6-sean.j.christopherson@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
