Ben Skeggs [Wed, 20 Feb 2013 04:34:18 +0000 (14:34 +1000)]
drm/nv50-/kms: remove unnecessary wait-for-completion points
DP link training is now handled as part of the supervisor processing,
and can no longer race with it.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Tue, 19 Feb 2013 04:17:53 +0000 (23:17 -0500)]
drm/nv50-/disp: move DP link training to core and train from supervisor
We need to be able to do link training for PIOR-connected ANX9805 from
the third supervisor handler (due to script ordering in the bios, can't
have the "user" call train because some settings are overwritten from
the modesetting bios scripts).
This moves link training for SOR-connected DP encoders to the second
supervisor interrupt, *before* we call the modesetting scripts (yes,
different ordering from PIOR is necessary). This is useful since we
should now be able to remove some hacks to workaround races between
the supervisor and link training paths.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Mon, 18 Feb 2013 22:50:51 +0000 (17:50 -0500)]
drm/nv50-/disp: handle supervisor tasks from workqueue
i2c_algo_bit sleeps...
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sat, 16 Feb 2013 05:21:58 +0000 (15:21 +1000)]
drm/nouveau/i2c: create proper chipset-specific class implementations
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sat, 16 Feb 2013 03:49:21 +0000 (13:49 +1000)]
drm/nv50-/disp: 0x0000 is a valid udisp config value
Return an out-of-range value instead to signal a failure from
exec_clkcmp().
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sat, 16 Feb 2013 02:10:38 +0000 (12:10 +1000)]
drm/nv50/devinit: reverse the logic for running encoder init scripts
A single U encoder table can match multiple DCB entries, whereas the
reverse is not true and can lead to us not matching a DCB entry at
all, and fail to initialise some encoders.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sat, 16 Feb 2013 02:01:59 +0000 (12:01 +1000)]
drm/nouveau/bios: store a type/mask hash in parsed dcb data
Matches format used by a couple of other vbios tables, useful
to have laying around already calculated.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sat, 16 Feb 2013 03:19:18 +0000 (13:19 +1000)]
drm/nouveau/i2c: extend type to 16-bits, add lookup-by-type function
For off-chip transmitters we won't necessarily have an i2c table entry
to lookup, but we can do it instead by encoding the type to include
the extdev type and looking that up instead.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Fri, 15 Feb 2013 01:59:41 +0000 (20:59 -0500)]
drm/nouveau/i2c: aux channels not necessarily on nvio
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Fri, 15 Feb 2013 00:54:42 +0000 (19:54 -0500)]
drm/nouveau/i2c: fix a bit of a thinko in nv_wri2cr helper functions
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Tue, 12 Feb 2013 00:16:31 +0000 (10:16 +1000)]
drm/nouveau/bios: parse external transmitter type if off-chip
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Mon, 11 Feb 2013 10:15:03 +0000 (20:15 +1000)]
drm/nouveau: store i2c port pointer directly in nouveau_encoder
This is about to become somewhat more complicated to determine in a
number of cases, so store the "common" case (DDC/AUX) directly inside
the encoder structure.
Pre-nv50 code not touched except to fill the pointer, don't care.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Mon, 11 Feb 2013 10:06:04 +0000 (20:06 +1000)]
drm/nouveau/i2c: handle i2c/aux mux outside of port lookup function
Not quite how I want it yet, but, I'll fix that at some point. For
right now, it's needed because find() won't necessarily be used right
before a transaction anymore.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sun, 17 Feb 2013 01:31:17 +0000 (11:31 +1000)]
drm/nv50/graph: avoid touching 400724, it doesn't exist
Harmless, but we now get MMIO fault reports, so silence it.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Stefan de Konink [Sat, 22 Dec 2012 17:04:37 +0000 (18:04 +0100)]
drm/nouveau: Fix DPMS 1 on G4 Snowball, from snow white to coal black.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=40275.
Signed-off-by: Stefan de Konink <stefan@konink.de>
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Dan Carpenter [Wed, 23 Jan 2013 08:27:56 +0000 (11:27 +0300)]
drm/nouveau/disp: sizeof() wrong pointer
"data" is a void pointer and "args" is "data" after we have casted it to
a struct. We care about the size of the struct here. Btw,
sizeof(*data) is 1.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 14 Feb 2013 03:43:21 +0000 (13:43 +1000)]
drm/nv84-/fence: prepare for emit/sync support of sysram sequences
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 14 Feb 2013 03:20:17 +0000 (13:20 +1000)]
drm/nouveau/fence: make internal hooks part of the context
A step towards being able to provide fences from other engines not
connected to PFIFO.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 14 Feb 2013 02:59:36 +0000 (12:59 +1000)]
drm/nv17/fence: split from nv10 code
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Wed, 13 Feb 2013 23:37:35 +0000 (09:37 +1000)]
drm/nv84-/fence: abstract class emit/sync functions to virt+sequence
Now can be used to operate on any buffer mapped into the GPU virtual
address and not just the main inter-channel sync buffer.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Wed, 13 Feb 2013 23:28:37 +0000 (09:28 +1000)]
drm/nv84/fence: access fences with full virtual address, not offset
Allows most of the code to be shared between nv84/nvc0 implementations,
and paves the way for doing emit/sync on non-VRAM buffers (multi-gpu,
dma-buf).
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Wed, 13 Feb 2013 03:34:39 +0000 (13:34 +1000)]
drm/nouveau/gpio/nve0: interrupt regs moved on kepler apparently
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sun, 3 Feb 2013 02:56:16 +0000 (12:56 +1000)]
drm/nouveau/gpio: use event interfaces for interrupt signalling
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sun, 3 Feb 2013 21:08:20 +0000 (07:08 +1000)]
drm/nouveau/gpio: pass number of on-die gpio lines to base
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Sun, 3 Feb 2013 10:29:53 +0000 (20:29 +1000)]
drm/nouveau/drm: store full dcb gpio function data in connector
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 31 Jan 2013 04:57:33 +0000 (14:57 +1000)]
drm/nouveau/fence/nv84-: put processes to sleep while waiting on fences
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 31 Jan 2013 22:43:55 +0000 (17:43 -0500)]
drm/nouveau/fifo/nvc0: bash some magic reg to make uevent interrupt work
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 31 Jan 2013 12:41:07 +0000 (22:41 +1000)]
drm/nouveau/fifo/nv84: support user event trigger
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 31 Jan 2013 03:51:20 +0000 (13:51 +1000)]
drm/nouveau/fifo/nvc0-: use interrupt 31 as an event trigger
Generated if you try and use fifo method 0x20 on any subchannel, appears
that it can be safely masked off without stalling the whole GPU.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Wed, 30 Jan 2013 23:23:34 +0000 (09:23 +1000)]
drm/nouveau/disp: port vblank handling to event interface
This removes the nastiness with the interactions between display and
software engines when handling vblank semaphore release interrupts.
Now, all the semantics are handled in one place (sw) \o/.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Wed, 30 Jan 2013 23:04:48 +0000 (09:04 +1000)]
drm/nouveau/disp/nv04: implement a base display object class
Will be used for upcoming vblank event interfaces.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Wed, 30 Jan 2013 14:38:22 +0000 (00:38 +1000)]
drm/nouveau/core: basic event interface between core and drm
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Fri, 1 Feb 2013 00:49:33 +0000 (19:49 -0500)]
drm/nouveau/fifo/nvc0: improve interrupt handler somewhat
Logs extra info for interrupts that have a sub-status register, and
handles the "special" ack from INTR bit 31.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Mon, 11 Feb 2013 03:56:31 +0000 (13:56 +1000)]
drm/nv50/disp: fix missing sor modectrl sync flags
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Tue, 5 Feb 2013 19:44:19 +0000 (20:44 +0100)]
drm/nouveau/therm: reduce stack usage of nouveau_therm_ic_ctor
Before: 1496 bytes
After: 152 bytes
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 3 Feb 2013 18:28:14 +0000 (19:28 +0100)]
drm/nouveau/therm: use workqueue to shutdown the machine
orderly_poweroff cannot be called from atomic context.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Acked-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 3 Feb 2013 18:12:49 +0000 (19:12 +0100)]
drm/nv40/therm: reset temperature sensor on init
Current uninitialized sensor detection does not work for me on nv4b and
sensor returns crazy values (>190°C). It stabilises later, but it's too
late - therm code shutdowns the machine...
Let's just reset it on init.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Acked-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 3 Feb 2013 17:17:41 +0000 (18:17 +0100)]
drm/nouveau/therm: turn on fan only when threshold hit in positive direction
+ the same for shutdown threshold - seems impossible, but shutdown can fail.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Acked-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 3 Feb 2013 21:02:47 +0000 (22:02 +0100)]
drm/nouveau: report channel owner in ioctl error paths
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Wed, 30 Jan 2013 21:21:31 +0000 (22:21 +0100)]
drm/nouveau/therm: always initialize alarm_program_lock
Fixes "BUG: spinlock bad magic" on module load for nva3+ cards.
Introduced in commit "drm/nouveau/therm: implement support for temperature
alarms".
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 27 Jan 2013 16:13:52 +0000 (17:13 +0100)]
drm/nouveau: handle backlight_device_register failure
Found by smatch.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 27 Jan 2013 16:04:48 +0000 (17:04 +0100)]
drm/nouveau: use kmemdup for edid allocation/copying
Avoids potential null pointer dereference.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 27 Jan 2013 15:43:00 +0000 (16:43 +0100)]
drm/nouveau: use drm_property_create_range helper
Avoids potential null pointer dereference.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Cong Ding [Tue, 15 Jan 2013 17:19:49 +0000 (18:19 +0100)]
drm/nouveau: remove unnecessary null pointer check from nouveau_fence_new
the variable chan is dereferenced in line 190, so it is no reason to check
null again in line 198.
Signed-off-by: Cong Ding <dinggnu@gmail.com>
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 27 Jan 2013 15:11:18 +0000 (16:11 +0100)]
drm/nvc0/graph: remove redundant null checks
It's safe to call kfree(NULL).
Found by smatch.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 27 Jan 2013 14:51:09 +0000 (15:51 +0100)]
drm/nouveau/fan: fix selection of fan speed when fan->get returns an error
fan->get returns int, but we write it to unsigned variable, and then check
whether it's >= 0 (it always is)
Found by smatch:
drivers/gpu/drm/nouveau/core/subdev/therm/fan.c:61 nouveau_fan_update() warn: always true condition '(duty >= 0) => (0-u32max >= 0)'
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 27 Jan 2013 14:01:55 +0000 (15:01 +0100)]
drm/nouveau: quiet static-related sparse noise
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 30 Dec 2012 15:35:24 +0000 (16:35 +0100)]
drm/nouveau/bios: tiny debugging messages fixes
COPY_ZM_REG: destination and source addresses were swapped
RAM_RESTRICT_ZM_REG_GROUP: missing 0x prefix for register address
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sat, 29 Dec 2012 15:24:37 +0000 (16:24 +0100)]
drm/nouveau: mark nv_printk_ as printf-like function
...and fix all warnings
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Tue, 25 Dec 2012 16:50:43 +0000 (17:50 +0100)]
drm/nouveau: share fence structures between nv10+ and nv50 implementations
We already rely on them having the same fields and layout.
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Thu, 20 Dec 2012 00:32:09 +0000 (01:32 +0100)]
drm/nouveau/fan: handle the cases where we are outside of the linear zone
This fixes a bug where, when temperature is outside of the linear range, fan
pwm would be outside of the allowed range ([0, 100]) and could get negative in
some cases.
It seems like a regression that happened when we re-worked the fan management
logic before merging.
Tested-by: Ozan Çağlayan <ozancag@gmail.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 9 Dec 2012 22:00:34 +0000 (23:00 +0100)]
drm/nouveau: report channel owner in error messages
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 9 Dec 2012 14:45:21 +0000 (15:45 +0100)]
drm/nouveau: prepare for reporting channel owner
- record channel owner process name
- add some helpers for accessing this information
- let nouveau_enum hold additional value (will be needed in the next patch)
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 9 Dec 2012 14:45:20 +0000 (15:45 +0100)]
drm/nouveau: use pr_cont
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Marcin Slusarz [Sun, 9 Dec 2012 14:45:19 +0000 (15:45 +0100)]
drm/nouveau: split fifo interrupt handler
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Sat, 8 Dec 2012 13:00:38 +0000 (14:00 +0100)]
drm/nouveau/therm: force a minimum hysteresis on temperature alarm thresholds
This should avoid the situation where a user gets its kernel logs flooded when
temperature oscillates around a threshold with 0°C hysteresis.
This patch is just meant to fix broken vbios (as reported on a nv4e on
sysfs hwmon interface.
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Emil Velikov [Fri, 7 Dec 2012 17:26:44 +0000 (17:26 +0000)]
drm/nouveau: set legacy bios data before parsing the structure
Commit
767baf82 drm/nouveau: remove some more unnecessary legacy bios code
has introduced a regression my misplacing the code that sets the major/chip
versions, which are used whist parsing the bmp/bit structure in vbios
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Fri, 7 Dec 2012 01:37:28 +0000 (11:37 +1000)]
drm/nouveau/therm: don't try pwm/toggle control if GPIO_FAN is input
My GTX660 has the GPIO_FAN function, but it's configured in input-mode;
presumably to monitor the frequency set by an I2C fan controller?
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Fri, 7 Dec 2012 01:01:25 +0000 (11:01 +1000)]
drm/nouveau/bios: rename DCB_GPIO_PWM_FAN to DCB_GPIO_FAN
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Fri, 7 Dec 2012 00:31:09 +0000 (10:31 +1000)]
drm/nouveau/bios: add support for parsing xpio table data
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 6 Dec 2012 05:44:23 +0000 (15:44 +1000)]
drm/nouveau: remove some more unnecessary legacy bios code
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 6 Dec 2012 05:32:56 +0000 (15:32 +1000)]
drm/nouveau: remove legacy vbios type detection
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 6 Dec 2012 05:17:12 +0000 (15:17 +1000)]
drm/nouveau/hwmon: create hwmon attributes under hwmon device in sysfs
From browsing my /sys, a few other things seem to do this, and it looks
cleaner this way too :)
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 6 Dec 2012 05:13:06 +0000 (15:13 +1000)]
drm/nouveau/hwmon: s/fan0/fan1/
Fan speed info now shown by sensors.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Thu, 6 Dec 2012 00:28:34 +0000 (10:28 +1000)]
drm/nouveau/therm: better transitions and debug logging
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Martin Peres [Tue, 20 Nov 2012 01:29:22 +0000 (02:29 +0100)]
drm/nouveau/doc: document the sysfs thermal management interface
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Tue, 20 Nov 2012 00:14:13 +0000 (01:14 +0100)]
drm/nouveau/hwmon: add missing alarm thresholds
Expose all the hysteresis parameters + shutdown (emergency) +
fan_boost (fixed pwm trip point).
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Sun, 4 Nov 2012 23:18:49 +0000 (00:18 +0100)]
drm/nouveau/therm: implement support for temperature alarms
For now, we only boost the fan speed to the maximum and auto-mode
when hitting the FAN_BOOST threshold and halt the computer when it
reaches the shutdown temperature. The downclock and critical thresholds
do nothing.
On nv43:50 and nva3+, temperature is polled because of the limited hardware.
I'll improve the nva3+ situation by implementing alarm management in PDAEMON
whenever I can but polling once every second shouldn't be such a problem.
v2 (Ben Skeggs):
- rebased
v3: fixed false-detections and threshold reprogrammation handling on nv50:nvc0
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Martin Peres [Fri, 7 Dec 2012 01:26:02 +0000 (02:26 +0100)]
drm/nv41/bus: report useful data on mmio fault
Based on Ben Skeggs's nvc0 patch. Tested on my nv4b, 84 and 92.
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Ben Skeggs [Tue, 4 Dec 2012 05:25:28 +0000 (15:25 +1000)]
drm/nvc0/bus: report useful data on mmio fault
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Sun, 4 Nov 2012 00:01:53 +0000 (01:01 +0100)]
drm/nouveau/pbus: add a PBUS subdev that hands IRQs to the right subdevs
We are going to use PTHERM's IRQs for thermal monitoring but we need to route
them first.
On nv31-50, PBUS's IRQ line is shared with GPIOs IRQs.
It seems like nv10-31 GPIO interruptions aren't well handled. I kept the
original behaviour but it is wrong and may lead to an IRQ storm.
Since we enable all PBUS IRQs, we need a way to avoid being stormed if we
don't handle them. The solution I used was to mask the IRQs that have not been
handled. This will also print one message in the logs to let us know.
v2: drop the shared intr handler because of was racy
v3: style fixes
v4: drop a useless construct in the chipset-dependent INTR
v5: add BUS to the disable mask
v6 (Ben Skeggs):
- general tidy to match the rest of the driver's style
- nva3->nvc0, nva3 can be serviced just fine with nv50.c, rnndb even notes
that the THERM_ALARM bit got left in the hw until fermi anyway.. so, it's
not going to conflict
- removed the peephole and user stuff, for the moment.. will handle them
later if we find a good reason to actually care..
- limited INTR_EN to just what we can handle for now, mostly to prevent
spam of unknown status bits (seen on at least nv4x)
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Martin Peres [Wed, 5 Dec 2012 10:28:09 +0000 (20:28 +1000)]
drm/nouveau/therm: implement automatic fan management
v2: improved design but drops safety monitoring (to be in a later patch)
v3: fix locking and mode management
v4: gently fallback to the no-control mode when temperature cannot be got
and use kernel-provided min/max macros
v5 (Ben Skeggs):
- rebased on my previous patches
v6: fix hysterisis management in trip-based auto fan management
This commit also forbids access to fan management to nvc0+ chipsets as
fan management is already taken care of my PDAEMON's default fw.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Martin Peres [Wed, 5 Dec 2012 09:46:35 +0000 (19:46 +1000)]
drm/nouveau/fan: obey fan bump/slow periods as defined by vbios
v2 (Ben Skeggs):
- split from larger patch
- fixed to not require alarm resched patch
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Martin Peres [Wed, 5 Dec 2012 08:42:00 +0000 (18:42 +1000)]
drm/nouveau/bios: parse fan bump/slow periods, and trip points
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Martin Peres [Sun, 2 Sep 2012 23:37:02 +0000 (01:37 +0200)]
drm/nouveau/fan: add toggle fan support
v2: change percent from int to atomic_t
v3: random fixes
v4 (Ben Skeggs):
- adapted for split-out fan-control "protocol" structure
- removed need for timer resched
- support for forcing 'toggle' control on PWM boards
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Ben Skeggs [Wed, 5 Dec 2012 06:21:59 +0000 (16:21 +1000)]
drm/nouveau/therm: cleanly separate pwm control logic from therm
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Ben Skeggs [Wed, 5 Dec 2012 04:56:37 +0000 (14:56 +1000)]
drm/nouveau/therm: add interfaces to allow forcing off pwm fan control
Mostly to allow for the possibility of testing 'toggle' fan control easily.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Ben Skeggs [Tue, 4 Dec 2012 02:10:19 +0000 (12:10 +1000)]
drm/nvd0/therm: add support for hardware fan tachometer
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Tue, 4 Dec 2012 02:18:59 +0000 (12:18 +1000)]
drm/nva3/therm: add support for hardware fan tachometer
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs [Tue, 4 Dec 2012 03:25:26 +0000 (13:25 +1000)]
drm/nouveau/therm: collect fan tach info in common fan constructor
This info will be used by two more implementations in upcoming commits.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Ben Skeggs [Tue, 4 Dec 2012 01:35:40 +0000 (11:35 +1000)]
drm/nouveau/therm: fix various style issues, make more consistent
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Ben Skeggs [Mon, 3 Dec 2012 23:50:33 +0000 (09:50 +1000)]
nvd0/therm: implement more appropriate pwm fan control functions
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Linus Torvalds [Fri, 1 Feb 2013 01:08:14 +0000 (12:08 +1100)]
Linux 3.8-rc6
Linus Torvalds [Fri, 1 Feb 2013 01:04:22 +0000 (12:04 +1100)]
Merge tag 'dm-3.8-fixes-2' of git://git./linux/kernel/git/agk/linux-dm
Pull more device-mapper fixes from Alasdair G Kergon:
"A fix for stacked dm thin devices and a fix for the new dm WRITE SAME
support."
* tag 'dm-3.8-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm:
dm: fix write same requests counting
dm thin: fix queue limits stacking
Linus Torvalds [Thu, 31 Jan 2013 21:44:59 +0000 (08:44 +1100)]
Merge branch 'for-3.8/upstream-fixes' of git://git./linux/kernel/git/jikos/hid
PullHID fixes from Jiri Kosina:
- fix i2c-hid and hidraw interaction, by Benjamin Tissoires
- a quirk to make a particular device (Formosa IR receiver) work
properly, by Nicholas Santos
* 'for-3.8/upstream-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid:
HID: i2c-hid: fix i2c_hid_output_raw_report
HID: usbhid: quirk for Formosa IR receiver
HID: remove x bit from sensor doc
Linus Torvalds [Thu, 31 Jan 2013 21:43:52 +0000 (08:43 +1100)]
Merge tag 'nfs-for-3.8-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull NFS client bugfixes from Trond Myklebust:
- Error reporting in nfs_xdev_mount incorrectly maps all errors to
ENOMEM
- Fix an NFSv4 refcounting issue
- Fix a mount failure when the server reboots during NFSv4 trunking
discovery
- NFSv4.1 mounts may need to run the lease recovery thread.
- Don't silently fail setattr() requests on mountpoints
- Fix a SUNRPC socket/transport livelock and priority queue issue
- We must handle NFS4ERR_DELAY when resetting the NFSv4.1 session.
* tag 'nfs-for-3.8-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFSv4.1: Handle NFS4ERR_DELAY when resetting the NFSv4.1 session
SUNRPC: When changing the queue priority, ensure that we change the owner
NFS: Don't silently fail setattr() requests on mountpoints
NFSv4.1: Ensure that nfs41_walk_client_list() does start lease recovery
NFSv4: Fix NFSv4 trunking discovery
NFSv4: Fix NFSv4 reference counting for trunked sessions
NFS: Fix error reporting in nfs_xdev_mount
Linus Torvalds [Thu, 31 Jan 2013 21:43:04 +0000 (08:43 +1100)]
Merge branch 'upstream' of git://git.linux-mips.org/ralf/upstream-linus
Pull MIPS updates from Ralf Baechle:
"A number of fixes all across the MIPS tree. No area is particularly
standing out and things have cooled down quite nicely for a release."
* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
MIPS: Function tracer: Fix broken function tracing
mips: Move __virt_addr_valid() to a place for MIPS 64
MIPS: Netlogic: Fix UP compilation on XLR
MIPS: AR71xx: Fix AR71XX_PCI_MEM_SIZE
MIPS: AR724x: Fix AR724X_PCI_MEM_SIZE
MIPS: Lantiq: Fix cp0_perfcount_irq mapping
MIPS: DSP: Fix DSP mask for registers.
MIPS: Fix build failure by adding definition of pfn_pmd().
MIPS: Octeon: Fix warning.
MIPS: delay.c: Check BITS_PER_LONG instead of __SIZEOF_LONG__
MIPS: PNX833x: Fix comment.
MIPS: Add struct p_format to union mips_instruction.
MIPS: Export <asm/break.h>.
MIPS: BCM47xx: Enable SSB prerequisite SSB_DRIVER_PCICORE.
MIPS: BCM47xx: Select GPIOLIB for BCMA on bcm47xx platform
MIPS: vpe.c: Fix null pointer dereference in print arguments.
Benjamin Tissoires [Thu, 31 Jan 2013 16:50:02 +0000 (17:50 +0100)]
HID: i2c-hid: fix i2c_hid_output_raw_report
i2c_hid_output_raw_report is used by hidraw to forward set_report requests.
The current implementation of i2c_hid_set_report needs to take the
report_id as an argument. The report_id is stored in the first byte
of the buffer in argument of i2c_hid_output_raw_report.
Not removing the report_id from the given buffer adds this byte 2 times
in the command, leading to a non working command.
Reported-by: Andrew Duggan <aduggan@synaptics.com>
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Al Cooper [Wed, 16 Jan 2013 22:43:28 +0000 (22:43 +0000)]
MIPS: Function tracer: Fix broken function tracing
Function tracing is currently broken for all 32 bit MIPS platforms.
When tracing is enabled, the kernel immediately hangs on boot.
This is a result of commit
b732d439cb43336cd6d7e804ecb2c81193ef63b0
that changes the kernel/trace/Kconfig file so that is no longer
forces FRAME_POINTER when FUNCTION_TRACING is enabled.
MIPS frame pointers are generally considered to be useless because
they cannot be used to unwind the stack. Unfortunately the MIPS
function tracing code has bugs that are masked by the use of frame
pointers. This commit fixes the bugs so that MIPS frame pointers
don't need to be enabled.
The bugs are a result of the odd calling sequence used to call the trace
routine. This calling sequence is inserted into every traceable function
when the tracing CONFIG option is enabled. This sequence is generated
for 32bit MIPS platforms by the compiler via the "-pg" flag.
Part of the sequence is "addiu sp,sp,-8" in the delay slot after every
call to the trace routine "_mcount" (some legacy thing where 2 arguments
used to be pushed on the stack). The _mcount routine is expected to
adjust the sp by +8 before returning. So when not disabled, the original
jalr and addiu will be there, so _mcount has to adjust sp.
The problem is that when tracing is disabled for a function, the
"jalr _mcount" instruction is replaced with a nop, but the
"addiu sp,sp,-8" is still executed and the stack pointer is left
trashed. When frame pointers are enabled the problem is masked
because any access to the stack is done through the frame
pointer and the stack pointer is restored from the frame pointer when
the function returns.
This patch writes two nops starting at the address of the "jalr _mcount"
instruction whenever tracing is disabled. This means that the
"addiu sp,sp.-8" will be converted to a nop along with the "jalr". When
disabled, there will be two nops.
This is SMP safe because the first time this happens is during
ftrace_init() which is before any other processor has been started.
Subsequent calls to enable/disable tracing when other CPUs ARE running
will still be safe because the enable will only change the first nop
to a "jalr" and the disable, while writing 2 nops, will only be changing
the "jalr". This patch also stops using stop_machine() to call the
tracer enable/disable routines and calls them directly because the
routines are SMP safe.
When the kernel first boots we have to be able to handle the gcc
generated jalr, addui sequence until ftrace_init gets a chance to run
and change the sequence. At this point mcount just adjusts the stack
and returns. When ftrace_init runs, we convert the jalr/addui to nops.
Then whenever tracing is enabled we convert the first nop to a "jalr
mcount+8". The mcount+8 entry point skips the stack adjust.
[ralf@linux-mips.org: Folded in Steven Rostedt's build fix.]
Signed-off-by: Al Cooper <alcooperx@gmail.com>
Cc: rostedt@goodmis.org
Cc: ddaney.cavm@gmail.com
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/4806/
Patchwork: https://patchwork.linux-mips.org/patch/4841/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Alasdair G Kergon [Thu, 31 Jan 2013 14:23:36 +0000 (14:23 +0000)]
dm: fix write same requests counting
When processing write same requests, fix dm to send the configured
number of WRITE SAME requests to the target rather than the number of
discards, which is not always the same.
Device-mapper WRITE SAME support was introduced by commit
23508a96cd2e857d57044a2ed7d305f2d9daf441 ("dm: add WRITE SAME support").
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Acked-by: Mike Snitzer <snitzer@redhat.com>
Steven Rostedt [Fri, 25 Jan 2013 17:13:15 +0000 (17:13 +0000)]
mips: Move __virt_addr_valid() to a place for MIPS 64
Commit
d3ce88431892 "MIPS: Fix modpost error in modules attepting to use
virt_addr_valid()" moved __virt_addr_valid() from a macro in a header
file to a function in ioremap.c. But ioremap.c is only compiled for MIPS
32, and not for MIPS 64.
When compiling for my yeeloong2, which supposedly supports hibernation,
which compiles kernel/power/snapshot.c which calls virt_addr_valid(), I
got this error:
LD init/built-in.o
kernel/built-in.o: In function `memory_bm_free':
snapshot.c:(.text+0x4c9c4): undefined reference to `__virt_addr_valid'
snapshot.c:(.text+0x4ca58): undefined reference to `__virt_addr_valid'
kernel/built-in.o: In function `snapshot_write_next':
(.text+0x4e44c): undefined reference to `__virt_addr_valid'
kernel/built-in.o: In function `snapshot_write_next':
(.text+0x4e890): undefined reference to `__virt_addr_valid'
make[1]: *** [vmlinux] Error 1
make: *** [sub-make] Error 2
I suspect that __virt_addr_valid() is fine for mips 64. I moved it to
mmap.c such that it gets compiled for mips 64 and 32.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/4842/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Mike Snitzer [Thu, 31 Jan 2013 14:11:14 +0000 (14:11 +0000)]
dm thin: fix queue limits stacking
thin_io_hints() is blindly copying the queue limits from the thin-pool
which can lead to incorrect limits being set. The fix here simply
deletes the thin_io_hints() hook which leaves the existing stacking
infrastructure to set the limits correctly.
When a thin-pool uses an MD device for the data device a thin device
from the thin-pool must respect MD's constraints about disallowing a bio
from spanning multiple chunks. Otherwise we can see problems. If the raid0
chunksize is 1152K and thin-pool chunksize is 256K I see the following
md/raid0 error (with extra debug tracing added to thin_endio) when
mkfs.xfs is executed against the thin device:
md/raid0:md99: make_request bug: can't convert block across chunks or bigger than 1152k 6688 127
device-mapper: thin: bio sector=2080 err=-5 bi_size=130560 bi_rw=17 bi_vcnt=32 bi_idx=0
This extra DM debugging shows that the failing bio is spanning across
the first and second logical 1152K chunk (sector 2080 + 255 takes the
bio beyond the first chunk's boundary of sector 2304). So the bio
splitting that DM is doing clearly isn't respecting the MD limits.
max_hw_sectors_kb is 127 for both the thin-pool and thin device
(queue_max_hw_sectors returns 255 so we'll excuse sysfs's lack of
precision). So this explains why bi_size is 130560.
But the thin device's max_hw_sectors_kb should be 4 (PAGE_SIZE) given
that it doesn't have a .merge function (for bio_add_page to consult
indirectly via dm_merge_bvec) yet the thin-pool does sit above an MD
device that has a compulsory merge_bvec_fn. This scenario is exactly
why DM must resort to sending single PAGE_SIZE bios to the underlying
layer. Some additional context for this is available in the header for
commit
8cbeb67a ("dm: avoid unsupported spanning of md stripe boundaries").
Long story short, the reason a thin device doesn't properly get
configured to have a max_hw_sectors_kb of 4 (PAGE_SIZE) is that
thin_io_hints() is blindly copying the queue limits from the thin-pool
device directly to the thin device's queue limits.
Fix this by eliminating thin_io_hints. Doing so is safe because the
block layer's queue limits stacking already enables the upper level thin
device to inherit the thin-pool device's discard and minimum_io_size and
optimal_io_size limits that get set in pool_io_hints. But avoiding the
queue limits copy allows the thin and thin-pool limits to be different
where it is important, namely max_hw_sectors_kb.
Reported-by: Daniel Browning <db@kavod.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Linus Torvalds [Thu, 31 Jan 2013 06:10:36 +0000 (17:10 +1100)]
Merge branch 'x86-efi-for-linus' of git://git./linux/kernel/git/tip/tip
Pull x86 EFI fixes from Peter Anvin:
"This is a collection of fixes for the EFI support. The controversial
bit here is a set of patches which bumps the boot protocol version as
part of fixing some serious problems with the EFI handover protocol,
used when booting under EFI using a bootloader as opposed to directly
from EFI. These changes should also make it a lot saner to support
cross-mode 32/64-bit EFI booting in the future. Getting these changes
into 3.8 means we avoid presenting an inconsistent ABI to bootloaders.
Other changes are display detection and fixing efivarfs."
* 'x86-efi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86, efi: remove attribute check from setup_efi_pci
x86, build: Dynamically find entry points in compressed startup code
x86, efi: Fix PCI ROM handing in EFI boot stub, in 32-bit mode
x86, efi: Fix 32-bit EFI handover protocol entry point
x86, efi: Fix display detection in EFI boot stub
x86, boot: Define the 2.12 bzImage boot protocol
x86/boot: Fix minor fd leakage in tools/relocs.c
x86, efi: Set runtime_version to the EFI spec revision
x86, efi: fix 32-bit warnings in setup_efi_pci()
efivarfs: Delete dentry from dcache in efivarfs_file_write()
efivarfs: Never return ENOENT from firmware
efi, x86: Pass a proper identity mapping in efi_call_phys_prelog
efivarfs: Drop link count of the right inode
Linus Torvalds [Thu, 31 Jan 2013 06:08:43 +0000 (17:08 +1100)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip
Pull x86 fixes from Peter Anvin:
"This is a collection of miscellaneous fixes, the most important one is
the fix for the Samsung laptop bricking issue (auto-blacklisting the
samsung-laptop driver); the efi_enabled() changes you see below are
prerequisites for that fix.
The other issues fixed are booting on OLPC XO-1.5, an UV fix, NMI
debugging, and requiring CAP_SYS_RAWIO for MSR references, just as
with I/O port references."
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
samsung-laptop: Disable on EFI hardware
efi: Make 'efi_enabled' a function to query EFI facilities
smp: Fix SMP function call empty cpu mask race
x86/msr: Add capabilities check
x86/dma-debug: Bump PREALLOC_DMA_DEBUG_ENTRIES
x86/olpc: Fix olpc-xo1-sci.c build errors
arch/x86/platform/uv: Fix incorrect tlb flush all issue
x86-64: Fix unwind annotations in recent NMI changes
x86-32: Start out cr0 clean, disable paging before modifying cr3/4
Linus Torvalds [Thu, 31 Jan 2013 06:04:46 +0000 (17:04 +1100)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull console lockdep checking revert from Dave Airlie.
The lockdep splat this showed was interesting, but it's very very old,
and we won't be fixing it until 3.9. In the meantime, undo the lockdep
annotation so that we don't generate the (known) console lockdep issue,
and then possibly hide any potential other (unknown) lockdep problems
that got disabled by the first one that triggered.
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
Revert "console: implement lockdep support for console_lock"
Dave Airlie [Thu, 31 Jan 2013 03:27:03 +0000 (14:27 +1100)]
Revert "console: implement lockdep support for console_lock"
This reverts commit
daee779718a319ff9f83e1ba3339334ac650bb22.
I'll requeue this after the console locking fixes, so lockdep
is useful again for people until fbcon is fixed.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Trond Myklebust [Wed, 30 Jan 2013 18:04:10 +0000 (13:04 -0500)]
NFSv4.1: Handle NFS4ERR_DELAY when resetting the NFSv4.1 session
NFS4ERR_DELAY is a legal reply when we call DESTROY_SESSION. It
usually means that the server is busy handling an unfinished RPC
request. Just sleep for a second and then retry.
We also need to be able to handle the NFS4ERR_BACK_CHAN_BUSY return
value. If the NFS server has outstanding callbacks, we just want to
similarly sleep & retry.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
Trond Myklebust [Sun, 27 Jan 2013 19:20:49 +0000 (14:20 -0500)]
SUNRPC: When changing the queue priority, ensure that we change the owner
This fixes a livelock in the xprt->sending queue where we end up never
making progress on lower priority tasks because sleep_on_priority()
keeps adding new tasks with the same owner to the head of the queue,
and priority bumps mean that we keep resetting the queue->owner to
whatever task is at the head of the queue.
Regression introduced by commit
c05eecf636101dd4347b2d8fa457626bf0088e0a
(SUNRPC: Don't allow low priority tasks to pre-empt higher priority ones).
Reported-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
H. Peter Anvin [Wed, 30 Jan 2013 22:43:05 +0000 (14:43 -0800)]
Merge tag 'efi-for-3.8' into x86/efi
Various urgent EFI fixes and some warning cleanups for v3.8
* EFI boot stub fix for Macbook Pro's from Maarten Lankhorst
* Fix an oops in efivarfs from Lingzhu Xiang
* 32-bit warning cleanups from Jan Beulich
* Patch to Boot on >512GB RAM systems from Nathan Zimmer
* Set efi.runtime_version correctly
* efivarfs updates
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Trond Myklebust [Tue, 22 Jan 2013 05:17:06 +0000 (00:17 -0500)]
NFS: Don't silently fail setattr() requests on mountpoints
Ensure that any setattr and getattr requests for junctions and/or
mountpoints are sent to the server. Ever since commit
0ec26fd0698 (vfs: automount should ignore LOOKUP_FOLLOW), we have
silently dropped any setattr requests to a server-side mountpoint.
For referrals, we have silently dropped both getattr and setattr
requests.
This patch restores the original behaviour for setattr on mountpoints,
and tries to do the same for referrals, provided that we have a
filehandle...
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org