When creating a fence for a tiled object, only fence the area that
makes up the actual tiles. The object may be larger than the tiled
area and if we allow those extra addresses to be fenced, they'll
get converted to addresses beyond where the object is mapped. This
opens up the possiblity of writes beyond the end of object.
To prevent this, we adjust the size of the fence to only encompass
the area that makes up the actual tiles. The extra space is considered
un-tiled and now behaves as if it was a linear object.
Testcase: igt/gem_tiled_fence_overflow
Reported-by: Dan Hettena <danh@ghs.com>
Signed-off-by: Bob Paauwe <bob.j.paauwe@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
commit 6dda730e55
Author: Jani Nikula <jani.nikula@intel.com>
Date: Tue Jun 24 18:27:40 2014 +0300
drm/i915: respect the VBT minimum backlight brightness
introduced a bug which resulted in inconsistent brightness levels on
different machines. If a suspended was entered with the screen off some
machines would resume with the screen at minimum brightness and others
at maximum brightness.
The following commands can be used to produce this behavior.
xset dpms force off
sleep 1
sudo systemctl suspend
(resume ...)
The root cause of this problem is a comparison which checks to see if
the backlight level is zero when the panel is enabled. If it is zero,
it is set to the maximum level. Unfortunately, not all machines have a
minimum level of zero. On those machines the level is left at the
minimum instead of begin set to the maximum.
Fix the bug by updating the comparison to check for the minimum
backlight level instead of zero. Also, expand the comparison for
the possible case when the level is less than the minimum.
Fixes: 6dda730e55 ("respect the VBT minimum backlight brightness")
Signed-off-by: Jeremiah Mahler <jmmahler@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Commit 82460d972 ("drm/i915: Rework ppgtt init to no require an aliasing
ppgtt") introduced a regression on Broadwell, triggering the following
IOMMU fault at startup:
vgaarb: device changed decodes: PCI:0000:00:02.0,olddecodes=io+mem,decodes=io+mem:owns=io+mem
dmar: DRHD: handling fault status reg 2
dmar: DMAR:[DMA Write] Request device [00:02.0] fault addr 880000
DMAR:[fault reason 23] Unknown
fbcon: inteldrmfb (fb0) is primary device
Further commentary from Daniel:
I sugggested this change to David after staring at the offending patch
for a while. I have no idea and theory whatsoever why this would upset
the gpu less than the other way round. But it seems to work. David
promised to chase hw people a bit more to get a more meaningful answer.
Wrt the comment that this deletes: I've done some digging and afaict
loading context before ppgtt enable was once required before our recent
restructuring of the context/ppgtt init code: Before that context sw
setup (i.e. allocating the default context) and hw setup was smashed
together. Also the setup of the default context was the bit that
actually allocated the aliasing ppgtt structures. Which is the reason
for the context before ppgtt depency.
Or was, since with all the untangling there's no no real depency any
more (functional, who knows what the hw is doing), so the comment is
just stale.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: stable@vger.kernel.org
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Today we expect that all the bank are enabled, and count the number of banks
used by the pinctrl based on it instead of using the last bank id enabled.
So switch to it, set the chained IRQ at runtime based on enabled banks
and wait only the number of enabled gpio controllers at probe time.
Cc: <stable@vger.kernel.org> # 3.18
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This is a driver for the pinmux and GPIO controller available in
Amlogic Meson SoCs. It currently supports only Meson8, however the
common code should be generic enough to work also for other SoCs after
having defined the proper set of functions and groups.
GPIO interrupts are not supported at the moment due to lack of
documentation.
Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Similar to omap_gpio_irq_type() let's make sure that the GPIO
is usable as an interrupt if the platform init code did not
call gpio_request(). Otherwise we can get invalid device access
after setup_irq():
WARNING: CPU: 0 PID: 1 at drivers/bus/omap_l3_noc.c:147 l3_interrupt_handler+0x214/0x340()
44000000.ocp:L3 Custom Error: MASTER MPU TARGET L4CFG (Idle): Data Access in Supervisor mode during Functional access
...
[<c05f21e4>] (__irq_svc) from [<c05f1974>] (_raw_spin_unlock_irqrestore+0x34/0x44)
[<c05f1974>] (_raw_spin_unlock_irqrestore) from [<c00914a8>] (__setup_irq+0x244/0x530)
[<c00914a8>] (__setup_irq) from [<c00917d4>] (setup_irq+0x40/0x8c)
[<c00917d4>] (setup_irq) from [<c0039c8c>] (omap_system_dma_probe+0x1d4/0x2b4)
[<c0039c8c>] (omap_system_dma_probe) from [<c03b2200>] (platform_drv_probe+0x44/0xa4)
...
We can fix this the same way omap_gpio_irq_type() is handling it.
Note that the long term solution is to change the gpio-omap driver
to handle the banks as separate driver instances. This will allow
us to rely on just runtime PM for tracking the bank specific state.
Reported-by: Russell King <rmk+kernel@arm.linux.org.uk>
Cc: Javier Martinez Canillas <javier@dowhile0.org>
Cc: Kevin Hilman <khilman@kernel.org>
Cc: Santosh Shilimkar <ssantosh@kernel.org>
Tested-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Do not wait for channel command buffers in IPA commands.
The potential wait could be done while holding a spin lock and causes
in recent kernels such a bug if kernel lock debugging is enabled:
kernel: BUG: sleeping function called from invalid context at drivers/s390/net/qeth_core_main.c:
794
kernel: in_atomic(): 1, irqs_disabled(): 0, pid: 2031, name: NetworkManager
kernel: 2 locks held by NetworkManager/2031:
kernel: #0: (rtnl_mutex){+.+.+.}, at: [<00000000006e0d7a>] rtnetlink_rcv+0x32/0x50
kernel: #1: (_xmit_ETHER){+.....}, at: [<00000000006cfe90>] dev_set_rx_mode+0x30/0x50
kernel: CPU: 0 PID: 2031 Comm: NetworkManager Not tainted 3.18.0-rc5-next-20141124 #1
kernel: 00000000275fb1f0 00000000275fb280 0000000000000002 0000000000000000
00000000275fb320 00000000275fb298 00000000275fb298 00000000007e326a
0000000000000000 000000000099ce2c 00000000009b4988 000000000000000b
00000000275fb2e0 00000000275fb280 0000000000000000 0000000000000000
0000000000000000 00000000001129c8 00000000275fb280 00000000275fb2e0
kernel: Call Trace:
kernel: ([<00000000001128b0>] show_trace+0xf8/0x158)
kernel: [<000000000011297a>] show_stack+0x6a/0xe8
kernel: [<00000000007e995a>] dump_stack+0x82/0xb0
kernel: [<000000000017d668>] ___might_sleep+0x170/0x228
kernel: [<000003ff80026f0e>] qeth_wait_for_buffer+0x36/0xd0 [qeth]
kernel: [<000003ff80026fe2>] qeth_get_ipacmd_buffer+0x3a/0xc0 [qeth]
kernel: [<000003ff80105078>] qeth_l3_send_setdelmc+0x58/0xf8 [qeth_l3]
kernel: [<000003ff8010b1fe>] qeth_l3_set_ip_addr_list+0x2c6/0x848 [qeth_l3]
kernel: [<000003ff8010bbb4>] qeth_l3_set_multicast_list+0x434/0xc48 [qeth_l3]
kernel: [<00000000006cfe9a>] dev_set_rx_mode+0x3a/0x50
kernel: [<00000000006cff90>] __dev_open+0xe0/0x140
kernel: [<00000000006d02a0>] __dev_change_flags+0xa0/0x178
kernel: [<00000000006d03a8>] dev_change_flags+0x30/0x70
kernel: [<00000000006e14ee>] do_setlink+0x346/0x9a0
...
The device driver has plenty of command buffers available
per channel for channel command communication.
In the extremely rare case when there is no command buffer
available, return a NULL pointer and issue a warning
in the kernel log. The caller handles the case when
a NULL pointer is encountered and returns an error.
In the case the wait for command buffer is possible
(because no lock is held as in the OSN case), still wait
until a channel command buffer is available.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Reviewed-by: Eugene Crosser <Eugene.Crosser@ru.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the functions that are registering and unregistering MAC
addresses in the qeth-handled hardware, remove callback functions
that are unnesessary, as only the return code is analyzed.
Translate hardware response codes to semi-standard 'errno'-like
codes for readability.
Add kernel-doc description to the internal API function
qeth_send_control_data().
Signed-off-by: Eugene Crosser <Eugene.Crosser@ru.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Reviewed-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
spin_event_timeout() is PPC dependent, use an arch independent
equivalent instead.
Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 5a7baa7885 ("bonding: Advertize vxlan offload features when
supported"), Or Gerlitz added support conditional vxlan offload.
In this patch I also add support for all kind of tunnels,
but we allow a bonding device to not require segmentation,
as it is always better to make this segmentation at the very last stage,
if a particular slave device requires it.
Tested:
Setup a GRE tunnel,
on a physical NIC not having tx-gre-segmentation.
Results on bnx2x are even better, as we no longer have to segment
in software.
ethtool -K bond0 tx-gre-segmentation off
super_netperf 50 --google-pacing-rate 30000000 -H 10.7.8.152 -l 15
7538.32
ethtool -K bond0 tx-gre-segmentation on
super_netperf 50 --google-pacing-rate 30000000 -H 10.7.8.152 -l 15
10200.5
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When commiting a plane update where the framebuffer doesn't change, we
can skip the prepare_fb/cleanup_fb steps. This also allows us to avoid
an unnecessary vblank wait at the end of the operation when we're just
moving a plane and not changing its image (e.g., for a cursor).
At the moment, i915 is the only upstream driver using the transitional
plane helpers, and thus the only driver affected by this change.
Note that this replicates a corresponding change in the atomic helpers
implemented in
commit ab58e3384b
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Mon Nov 24 20:42:42 2014 +0100
drm/atomic-helper: Skip vblank waits for unchanged fbs
Reported-by: Jeremiah Mahler <jmmahler@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88540
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by: Tested-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The driver supports negative high and critical limits, it can return
negative hysteresis values, so there is no good reason to not let the
user write negative hysteresis values.
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
data->temp[index] has type s16. Because of C's promotion rules,
(data->temp[index] << 7) >> 7 is exactly the same as
data->temp[index]. The intention was to use bit 8 as a sign bit, so do
that using the existing API.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The comment above (data << 2) >> 2 explains what the intention is: To
use bit 13 of the 14-bit value data as the sign bit. However, this
doesn't work due to C's promotion rules. data has type s16, but data
<< 2 has type int. To get sign extension, that expression would have
to be cast back to an s16 before being shifted (at which point C's
promotion rules would then kick in again and promote the left operand
to int). As it stands, both expressions are no-ops for any value of
data.
Avoid these subtleties by using the existing API for
this. sign_extend32 works equally well for 8 and 16 bit types.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Fixed the following warnings (reported by cppcheck):
[drivers/hwmon/abx500.c:224]: (warning) %ld in format string (no. 1)
requires 'long' but the argument type is 'unsigned long'.
[drivers/hwmon/abx500.c:233]: (warning) %ld in format string (no. 1)
requires 'long' but the argument type is 'unsigned long'.
[drivers/hwmon/abx500.c:242]: (warning) %ld in format string (no. 1)
requires 'long' but the argument type is 'unsigned long'.
Signed-off-by: Asaf Vertz <asaf.vertz@tandemg.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Subtracting an unsigned long from a signed value causes an overflow with large
values. Use clamp_val() to reduce the number range prior to subtracting it
from the temperature limit.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Jean Delvare <jdelvare@suse.de>
Mixed use of long and int caused an integer overflow when writing large limits.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Jean Delvare <jdelvare@suse.de>
Despite the name, sign_extend32 works just fine for 16 bit variables,
so it is safe to use.
Cc: Martin Kepplinger <martink@posteo.de>
Reviewed-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for "ina231" as compatible string, and update
Documentation and Kconfig accordingly.
Tested with the Exynos5422-based odroid-xu3 board which has on-board
INA231 sensors.
Signed-off-by: Kevin Hilman <khilman@linaro.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Use DIV_ROUND_CLOSEST() when dealing with the calibration values to make the
calculations less error prone.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
This attribute allows to configure the update interval of ina226. Although
the bus and shunt voltage conversion times remain hardcoded to 1.1 ms, we can
now modify said interval by changing the averaging rate.
While we're at it - add an additional variable to ina2xx_data, which holds
the current configuration settings - this way we'll be able to restore the
configuration in case of an unexpected chip reset.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The shunt resistance can only be set via platform_data or device tree. This
isn't suitable for devices in which the shunt resistance can change/isn't
known at boot-time.
Add a sysfs attribute that allows to read and set the shunt resistance.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Shunt resistance values greater than the chip's calibration factor make no
sense since the actual value written to the register equals:
<calibration factor> / <shunt>
Bail-out from ina2xx_probe() if the configured value is greater than the
calibration factor.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Chips from the ina family don't like to be uninitialized. In case the power
is cut-off and restored again the calibration register will be reset
to 0 and both the power and current registers will remain at 0.
Check the calibration register in ina2xx_update_device() and reinitialize
the chip if needed.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The regmap_config struct may be const because it is not modified by the
driver and regmap_init() accepts pointer to const.
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The regmap_config struct may be const because it is not modified by the
driver and regmap_init() accepts pointer to const.
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
it's not only badly racy, it's actually dead code - the call in
ll_invalidate_aliases() is unreachable. For an alias of our inode
to be root dentry, the inode would have to be its ->d_inode, aka
the root inode, and we never call ll_invalidate_aliases() for that.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Pull x86 fixes from Thomas Gleixner:
"Hopefully the last round of fixes for 3.19
- regression fix for the LDT changes
- regression fix for XEN interrupt handling caused by the APIC
changes
- regression fixes for the PAT changes
- last minute fixes for new the MPX support
- regression fix for 32bit UP
- fix for a long standing relocation issue on 64bit tagged for stable
- functional fix for the Hyper-V clocksource tagged for stable
- downgrade of a pr_err which tends to confuse users
Looks a bit on the large side, but almost half of it are valuable
comments"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/tsc: Change Fast TSC calibration failed from error to info
x86/apic: Re-enable PCI_MSI support for non-SMP X86_32
x86, mm: Change cachemode exports to non-gpl
x86, tls: Interpret an all-zero struct user_desc as "no segment"
x86, tls, ldt: Stop checking lm in LDT_empty
x86, mpx: Strictly enforce empty prctl() args
x86, mpx: Fix potential performance issue on unmaps
x86, mpx: Explicitly disable 32-bit MPX support on 64-bit kernels
x86, hyperv: Mark the Hyper-V clocksource as being continuous
x86: Don't rely on VMWare emulating PAT MSR correctly
x86, irq: Properly tag virtualization entry in /proc/interrupts
x86, boot: Skip relocs when load address unchanged
x86/xen: Override ACPI IRQ management callback __acpi_unregister_gsi
ACPI: pci: Do not clear pci_dev->irq in acpi_pci_irq_disable()
x86/xen: Treat SCI interrupt as normal GSI interrupt
Pull irq fixes from Thomas Gleixner:
"From the irqchip departement you get:
- regression fix for omap-intc
- regression fix for atmel-aic-common
- functional correctness fix for hip04
- type mismatch fix for gic-v3-its
- proper error pointer check for mtd-sysirq
Mostly one and two liners except for the omap regression fix which is
slightly larger than desired"
* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip: atmel-aic-common: Prevent clobbering of priority when changing IRQ type
irqchip: omap-intc: Fix legacy DMA regression
irqchip: gic-v3-its: Fix use of max with decimal constant
irqchip: hip04: Initialize hip04_cpu_map to 0xffff
irqchip: mtk-sysirq: Use IS_ERR() instead of NULL pointer check
Pull timer fixes from Thomas Gleixner:
"A set of small fixes:
- regression fix for exynos_mct clocksource
- trivial build fix for kona clocksource
- functional one liner fix for the sh_tmu clocksource
- two validation fixes to prevent (root only) data corruption in the
kernel via settimeofday and adjtimex. Tagged for stable"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
time: adjtimex: Validate the ADJ_FREQUENCY values
time: settimeofday: Validate the values of tv from user
clocksource: sh_tmu: Set cpu_possible_mask to fix SMP broadcast
clocksource: kona: fix __iomem annotation
clocksource: exynos_mct: Fix bitmask regression for exynos4_mct_write
Pull ARM SoC fixes from Olof Johansson:
"A week's worth of fixes for various ARM platforms. Diff wise, the
largest fix is for OMAP to deal with how GIC now registers interrupts
(irq_domain_add_legacy() -> irq_domain_add_linear() changes).
Besides this, a few more renesas platforms needed the GIC instatiation
done for legacy boards. There's also a fix that disables coherency of
mvebu due to issues, and a few other smaller fixes"
* tag 'armsoc-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
arm64: dts: add baud rate to Juno stdout-path
ARM: dts: imx25: Fix PWM "per" clocks
bus: mvebu-mbus: fix support of MBus window 13
Merge tag 'mvebu-fixes-3.19-3' of git://git.infradead.org/linux-mvebu into fixes
ARM: mvebu: completely disable hardware I/O coherency
ARM: OMAP: Work around hardcoded interrupts
ARM: shmobile: r8a7779: Instantiate GIC from C board code in legacy builds
ARM: shmobile: r8a7778: Instantiate GIC from C board code in legacy builds
arm: boot: dts: dra7: enable dwc3 suspend PHY quirk
Pull device mapper fixes from Mike Snitzer:
"Two stable fixes for dm-cache and one 3.19 DM core fix:
- fix potential for dm-cache metadata corruption via stale metadata
buffers being used when switching an inactive cache table to
active; this could occur due to each table having it's own bufio
client rather than sharing the client between tables.
- fix dm-cache target to properly account for discard IO while
suspending otherwise IO quiescing could complete prematurely.
- fix DM core's handling of multiple internal suspends by maintaining
an 'internal_suspend_count' and only resuming the device when this
count drops to zero"
* tag 'dm-3.19-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm: fix handling of multiple internal suspends
dm cache: fix problematic dual use of a single migration count variable
dm cache: share cache-metadata object across inactive and active DM tables
Pull two block layer fixes from Jens Axboe:
"Two small patches that should make it into 3.19:
- a fixup from me for NVMe, making the cq_vector a signed variable.
Otherwise our -1 comparison fails, and commit 2b25d98179 doesn't
do what it was supposed to.
- a fixup for the hotplug handling for blk-mq from Ming Lei, using
the proper kobject referencing to ensure we release resources at
the right time"
* 'for-linus' of git://git.kernel.dk/linux-block:
blk-mq: fix hctx/ctx kobject use-after-free
NVMe: cq_vector should be signed
irq_dispose_mapping is not called on all error paths from caam_jr_init.
This takes care of several clean-up issues by performing resource
clean-up and allocation at the same level.
Signed-off-by: Cristian Stoica <cristian.stoica@freescale.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
