We've lost the +1 required for correct timeouts in
commit 5ed0bdf21a
Author: Thomas Gleixner <tglx@linutronix.de>
Date: Wed Jul 16 21:05:06 2014 +0000
drm: i915: Use nsec based interfaces
Use ktime_get_raw_ns() and get rid of the back and forth timespec
conversions.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: John Stultz <john.stultz@linaro.org>
So fix this up by reinstating our handrolled _timeout function. While
at it bother with handling MAX_JIFFIES.
v2: Convert to usecs (we don't care about the accuracy anyway) first
to avoid overflow issues Dave Gordon spotted.
v3: Drop the explicit MAX_JIFFY_OFFSET check, usecs_to_jiffies should
take care of that already. It might be a bit too enthusiastic about it
though.
v4: Chris has a much nicer color, so use his implementation.
This requires to export nsec_to_jiffies from time.c.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=82749
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: John Stultz <john.stultz@linaro.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Acked-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Silence some pch fifo underrun reports and panel locking backtraces,
both cc: stable.
* tag 'drm-intel-fixes-2014-12-04' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Unlock panel even when LVDS is disabled
drm/i915: More cautious with pch fifo underruns
Fixes for 3.20. I did stick the gen3/4 reset work from Ville in because we
have an awful lot of gen4 mesa hangs, and with this reset should also work
on vintage i965g/gm (we already have reset for g4x/gen4.5). So should help
to appease users suffering from these hangs. Otherwise all over.
This is the last 3.20 pull from me, from here on Jani will take over. By Ville Syrjälä (8) and others
* tag 'drm-intel-next-fixes-2014-12-04' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Reject modeset when the same digital port is used more than once
drm/i915: mask RPS IRQs properly when disabling RPS
drm/i915: Tune down spurious CRC interrupt warning
drm/i915: Fix context object leak for legacy contexts
drm/i915/skl: Update in Gen9 multi-engine forcewake range
drm/i915/eDP: When enabling panel VDD cancel pending disable worker
drm/i915: Handle runtime pm in the CRC setup code
drm/i915: Disable crtcs gracefully before GPU reset on gen3/4
drm/i915: Grab modeset locks for GPU rest on pre-ctg
drm/i915: Implement GPU reset for g33
drm/i915: Implement GPU reset for 915/945
drm/i915: Restore the display config after a GPU reset on gen4
drm/i915: Fix gen4 GPU reset
drm/i915: Stop gathering error states for CS error interrupts
drm/i915: Disallow pin ioctl completely for kms drivers
drm/i915: Only warn the first time we attempt to mmio whilst suspended
drm/i915/chv: Enable AVI, SPD and HDMI infoframes for CHV.
drm/i915: Don't clobber crtc->new_config when nothing changes
Just three more fixes for 3.19.
This is the last request until -rc1. I will have our QA team run a full
HSA stack test on 3.19-rc1 and if we find problems, I will send a fixes pull
request.
* 'amdkfd-next-3.19' of git://people.freedesktop.org/~gabbayo/linux:
amdkfd: Set *buffer_ptr to NULL in case of error
amdkfd: use atomic allocations within srcu callbacks
amdkfd: use sizeof(long) granularity for the pasid bitmask
Note that the read manpages explicitly states that the read position
is undefined on error. Since EFAULT is just a userspace bug we are
therefore fine with just dropping the event on the floor.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
[danvet: Add note that just dropping the event is ok.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In function acquire_packet_buffer() we may return -ENOMEM. In that case, we
should set the *buffer_ptr to NULL, so that calling functions which check the
*buffer_ptr value as a criteria for success, will know that
acquire_packet_buffer() failed.
Reviewed-by: Alexey Skidanov <alexey.skidanov@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
This patch is in preparation of DSI dual link panels. For dual link
panels, few packets needs to be sent to Port A or Port C or both. Based
on the portno from MIPI Sequence Block#53, these sequences needs to be
sent accordingly.
v2: Addressed review comments by Jani
- port variables named properly
Signed-off-by: Gaurav K Singh <gaurav.k.singh@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch is in preparation for the DSI dual link
port enable and disable related changes.
Signed-off-by: Gaurav K Singh <gaurav.k.singh@intel.com>
Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When playing around with debugfs and a HSW machine I noticed that we
were displaying some garbled value in i915_ddb_info. This debugfs file
is only meaningful for gen9+, so don't display anything on earlier
platforms.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Multiple GGTT VMAs per object will be introduced in the near future which will
make it impossible to guarantee normal GGTT view is at the head of the list.
Purpose of this patch is to break this assumption straight away so any
potential hidden assumptions in the code base can be bisected to this
simple patch.
For: VIZ-4544
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Suggested-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This seems to work well on CI boards after fixing the
last few bugs noticed by Chernovsky Oleg.
On boards with a high default fan speed this should
reduce fan noise. Manual fan control is not enabled
yet.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
After commit b2b49ccbdd (PM: Kconfig: Set PM_RUNTIME if PM_SLEEP is
selected) PM_RUNTIME is always set if PM is set, so #ifdef blocks
depending on CONFIG_PM_RUNTIME may now be changed to depend on
CONFIG_PM.
Replace CONFIG_PM_RUNTIME with CONFIG_PM in 4 files under
gpu/drm/exynos/.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Jingoo Han <jg1.han@samsung.com>
Nobody is interested at which index the chunk is. What's needed is
a pointer to the chunk. Remove unused chunk_id field as well.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Completely unnecessary since the ww_mutex used to reserve a buffer
can detect double reservations from the same thread anyway.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch adds an optional list_head parameter to ttm_eu_reserve_buffers.
If specified duplicates in the execbuf list are no longer reported as errors,
but moved to this list instead.
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Check the that ring we are using for copies is functional
rather than the GFX ring. On newer asics we use the DMA
ring for bo moves.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Since we are now preserving the cursor across modesets, the cursor could
be left over in console if e.g. X crashed.
v2: add comment about universal plane support
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Stop using the VM mutex for this
v2: fix typo in comment
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The BO_VA contains everything necessary.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Better match what it is actually doing.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It's only used for duplicate check and that
can be done on the original as well.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It's only used once after initializing and that
ptr can be calculated from the BO as well.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Not just the userspace relocs, otherwise we won't wait
for a swapped out page tables to be swapped in again.
Signed-off-by: Christian König <christian.koenig@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It's only needed in radeon_crtc_cursor_set2.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Setting a mode seems to clear the cursor registers, so we need to
re-program them to make sure the cursor is visible.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Hardware team updated the recommended translation values for DP/eDP 1.3.
This should help with some stability and HBR2 issues.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Satheeshakrishna M<satheeshakrishna.m@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Commit 18df89fef2 ("drm: Decouple EDID parsing from I2C adapter")
renamed the adapter parameter of the drm_do_probe_ddc_edid function
to data but didn't update the kerneldoc accordingly.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
srcu callbacks are running in atomic context, we can't allocate using
__GFP_WAIT.
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
All the bit operations (such as find_first_zero_bit()) read sizeof(long) bytes
at a time. If we allocated less than sizeof(long) bytes for the bitmask we
would be accessing invalid memory when working with the bitmask.
Change the allocator to allocate sizeof(long) multiples for the bitmask.
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
This codepath is mostly hit when rebinding after a backup buffer swapout. It's
amazing that this error hasn't been more obvious but probably the shaders are
not reread from guest memory that often..
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
The commit "vmwgfx: Rework fence event action" introduced a number of bugs
that are fixed with this commit:
a) A forgotten return stateemnt.
b) An if statement with identical branches.
Cc: <stable@vger.kernel.org>
Reported-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
Kernel side fence objects are used when unbinding resources and may thus be
created as part of a memory reclaim operation. This might trigger recursive
memory reclaims and result in the kernel running out of stack space.
So a simple way out is to avoid accounting of these fence objects.
In principle this is OK since while user-space can trigger the creation of
such objects, it can't really hold on to them. However, their lifetime is
quite long, so some form of accounting should perhaps be implemented in the
future.
Fixes kernel crashes when running, for example viewperf11 ensight-04 test 3
with low system memory settings.
Cc: <stable@vger.kernel.org>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
The function vmw_master_check() might return -ERESTARTSYS if there is a
signal pending, indicating that the IOCTL should be rerun, potentially from
user-space. At that point we shouldn't print out an error message since that
is not an error condition. In short, avoid bloating the kernel log when a
process refuses to die on SIGTERM.
Cc: <stable@vger.kernel.org>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
We need to do that every time we resume the rings, not just at load.
I've overlooked this in my untangling of the ring init code.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
A previous commit introduced engine init changes:
commit 372ee59699d9 ("drm/i915: Only init engines once")
This broke execlists as intel_lr_context_render_state_init was trying to emit
commands to the RCS for the default context before the ring->init_hw was called.
Made a new gen8_init_rcs_context function and assign in to render ring
init_context. Moved call to intel_logical_ring_workarounds_emit into
gen8_init_rcs_context to maintain previous functionality.
Moved call to render_state_init from lr_context_deferred_create into
gen8_init_rcs_context, and modified deferred_create to call ring->init_context
for non-default contexts.
Modified i915_gem_context_enable to call ring->init_context for the default
context.
So init_context will now always be called when the hw is ready - in
i915_gem_context_enable for the default context and in lr_context_deferred_create
for other contexts.
Signed-off-by: Thomas Daniel <thomas.daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The conversion table can be replaced with simple enough function.
text data bss dec hex filename
839688 10987 24 850699 cfb0b drivers/gpu/drm/i915/i915.ko
839224 10987 24 850235 cf93b drivers/gpu/drm/i915/i915.ko
Result is 494 saved bytes (.05525%).
v2: - no run on sentences from subject (Chris, Jani)
- be verbose about the savings (Chris, Daniel)
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> (v1)
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that sanity prevails and we have the clean split between software
init and starting the engines we can drop all the "have we allocate
this struct already?" nonsense.
Execlist code could benefit quite a bit more still, but that's for
another patch.
Reviewed-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The same logic can be implemented without it, and it even saves a line
of code.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We can do this.
And now there's finally the clean split between software setup and
hardware setup I kinda wanted since multi-ring support was merged
aeons ago. It only took almost 5 years.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With this all the ->init_hw hooks really only set up hw state needed
to start the ring, all the software state setup and memory/buffer
allocations happen beforehand.
v2: We need to call intel_init_pipe_control after the ring init since
otherwise engine->dev is NULL and it falls over. Currently that's
now after the hw ring is enabled but a) we'll be fine as long as no
one submits a batch b) this will change soon.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is (mostly, some exceptions that need fixing) the hw setup
function which starts the ring. And not the function which allocates
all the resources.
Make this clear by giving it a better name.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
There are numerous places in the code where the driver's idea of
how much space is left in a ring is updated using the driver's
latest notions of the positions of 'head' and 'tail' for the ring.
Among them are some that update one or both of these values before
(re)doing the calculation. In particular, there are four different
places in the code where 'last_retired_head' is copied to 'head'
and then set to -1; and two of these do not have a guard to check
that it has actually been updated since last time it was consumed,
leaving the possibility that the dummy -1 can be transferred from
'last_retired_head' to 'head', causing the space calculation to
produce 'impossible' results (previously seen on Android/VLV).
This code therefore consolidates all the calculation and updating of
these values, such that there is only one place where the ring space
is updated, and it ALWAYS uses (and consumes) 'last_retired_head' if
(and ONLY if) it has been updated since the last call.
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The used space in a ring is given by the cyclic distance from the
consumer (HEAD) to the producer (TAIL), i.e. ((tail-head) MOD size);
conversely, the available space in a ring is the cyclic distance
from the producer to the consumer, MINUS the amount reserved for a
"gap" that is supposed to guarantee that the producer never catches
up with or overruns the consumer. Note that some GEN h/w requires
that TAIL never approach to within one cacheline of HEAD, so the gap
is usually set to twice the cacheline size to ensure this.
While the existing code gives the correct answer for correct inputs,
if the producer HAS overrun into the reserved space, the result can
be a value larger than the maximum valid value (size-reserved). We
can improve this by reorganising the calculation, so that in the
event of overrun the result will be negative rather than over-large.
This means that the commonly-used test (available >= required)
will then reject further writes into the ring after an overrun,
giving some chance that we can recover from or at least diagnose
the original problem; whereas allowing more writes would likely both
confuse the h/w and destroy the evidence of what went wrong.
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Later on this can include multiple ports (e.g. (1 << PORT_A) | (1 <<
PORT_C)) to describe dual link DSI.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Gaurav K Singh <gaurav.k.singh@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
MIPI DSI works on ports A and C, which map to pipes A and B,
respectively. Things are going to get more complicated with the
introduction of dual link DSI support, so clean up the register defines
and code to match reality.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Gaurav K Singh <gaurav.k.singh@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Clear the video overlay state on GPU reset. Any pending overlay request
in the ring has been nuked, and the display itself gets reset. So we
pretty much lose all state here. Adjust the software state to match so
that the next "putimage" will restore things to working order.
v2: Ass a locking check into intel_overlay_release_old_vid() (Daniel)
Cc: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
[danvet: s/0/NULL/ to appease sparse, reported by 0-day tester.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the trace_irq code to use requests instead of seqnos. This includes
reference counting the request object to ensure it sticks around when required.
Note that getting access to the reference counting functions means moving the
inline i915_trace_irq_get() function from intel_ringbuffer.h to i915_drv.h.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
[danvet: Resolve conflict due to shuffled merge order.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Similar to the patch from John which removed obj->ring.
Cc: John Harrison <John.C.Harrison@Intel.com>
Cc: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The ring member of the object structure was always updated with the
last_read_seqno member. Thus with the conversion to last_read_req, obj->ring is
now a direct copy of obj->last_read_req->ring. This makes it somewhat redundant
and potentially misleading (especially as there was no comment to explain its
purpose).
This checkin removes the redundant field. Many uses were simply testing for
non-null to see if the object is active on the GPU. Some of these have been
converted to check 'obj->active' instead. Others (where the last_read_req is
about to be used anyway) have been changed to check obj->last_read_req. The rest
simply pull the ring out from the request structure and proceed as before.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Almost everywhere that caled i915_seqno_passed() was really asking 'has the
given seqno popped out of the hardware yet?'. Thus it had to query the current
hardware seqno and then do a signed delta comparison (which copes with wrapping
around zero but not with seqno values more than 2GB apart, although the latter
is unlikely!).
Now that the majority of seqno instances have been replaced with request
structures, it is possible to convert this test to be request based as well.
There is now a 'i915_gem_request_completed()' function which takes a request and
returns true or false as appropriate. Note that this currently just wraps up the
original _passed() test but a later patch in the series will reduce this to
simply returning a cached internal value, i.e.:
_completed(req) { return req->completed; }'
This checkin converts almost all _seqno_passed() calls. The only one left is in
the semaphore code which still requires seqnos not request structures.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
[danvet: Drop hunk touching the trace_irq code since I've dropped the
patch which converts that, and resolve resulting conflict.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It makes a lot more sense (and makes future seqno -> request conversion patches
simpler) to fill in the 'ring' field of the request structure at the point of
creation rather than submission. Given that the request structure is assigned by
ring specific code and thus is locked to a ring from the start, there really is
no reason to defer this assignment.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
More seqno value to request structure conversions. Note, this change temporarily
moves the 'get_seqno()' call inside ring_idle() but this will disappear again in
a later patch when i915_seqno_passed() itself is converted.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
All the code above is now using requests not seqnos so it is possible to convert
the trace functions across. Note that rather than get into problematic reference
counting issues, the trace code only saves the seqno and ring values from the
request structure not the structure pointer itself.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Converted the flip_queued_seqno value to be a request structure as part of the
on going seqno to request changes. This includes reference counting the request
being saved away to ensure it can not be retired and freed while the flip code
is still waiting on it.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
[danvet: Again get rid of the _irq request unref by simply moving that
into the unpin worker. Doesn't matter when we hang onto the request
for a bit longer, and in the unpin worker we already grab the
dev->struct_mutex anyway.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
There is no longer any need to retrieve a seqno value from an i915_add_request()
call. The calling code already knows which request structure is being processed
(it can only be ring->OLR). And as the request itself is now used in preference
to the basic seqno value, the latter is now redundant in this situation.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that all code above is using request structures instead of seqno values, it
is possible to convert __wait_seqno() itself. Internally, it is still calling
i915_seqno_passed(), this will be updated later in the series. This step is just
changing the parameter list and function name.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Converted the mmio_flip 'seqno' value to be a request structure as part of the
on going seqno to request changes. This includes reference counting the request
being saved away to ensure it can not be retired and freed while the flip code
is still waiting on it.
v2: Used the IRQ friendly request dereference call in the notify handler as that
code is called asynchronously without holding any useful mutex locks.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
[danvet: Drop the _irq variant and use the normal reques unref,
wrapped in dev->struct_mutex per the discussion on the m-l.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With refcounting it looks like you can just drop that refcount, but
that's not really the case. So make sure no one forgets.
Motivated by the unlocked call in the mmio flip code.
Cc: John Harrison <John.C.Harrison@Intel.com>
Cc: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Updated i915_wait_seqno() to take a request structure instead of a seqno value
and renamed it accordingly. Internally, it just pulls the seqno out of the
request and calls on to __wait_seqno() as before. However, all the code further
up the stack is now simplified as it can just pass the request object straight
through without having to peek inside.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
[danvet: Squash in hunk from an earlier patch which was rebased
wrongly.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Converted 'last_flip_req' to be an actual request rather than a seqno value as
part of the on going seqno to request changes. This includes reference counting
the request being saved away to ensure it can not be retired and freed while the
overlay code is still waiting on it.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the _check_olr() function to actually take a request object and compare
it to the OLR rather than extracting seqnos and comparing those.
Note that there is one use case where the request object being processed is no
longer available at that point in the call stack. Hence a temporary copy of the
original function is still present (but called _check_ols() instead). This will
be removed in a subsequent patch.
Also, downgraded a BUG_ON to a WARN_ON as apparently the former is frowned upon
for shipping code.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The OLS value is now obsolete. Exactly the same value is guarateed to be always
available as PLR->seqno. Thus it is safe to remove the OLS completely. And also
to rename the PLR to OLR to keep the 'outstanding lazy ...' naming convention
valid.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Added reference counting of the request structure around __wait_seqno() calls.
This is a precursor to updating the wait code itself to take the request rather
than a seqno. At that point, it would be a Bad Idea for a request object to be
retired and freed while the wait code is still using it.
v3:
Note that even though the mutex lock is held during a call to i915_wait_seqno(),
it is still necessary to explicitly bump the reference count. It appears that
the shrinker can asynchronously retire items even though the mutex is locked.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
[danvet: Remove wrongly squashed hunk which breaks the build.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Convert the throttle code to use the request structure rather than extracting a
ring/seqno pair from it and using those. This is in preparation for
__wait_seqno() becoming __wait_request().
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The object structure contains the last read, write and fenced seqno values for
use in syncrhonisation operations. These have now been replaced with their
request structure counterparts.
Note that to ensure that objects do not end up with dangling pointers, the
assignments of last_*_req include reference count updates. Thus a request cannot
be freed if an object is still hanging on to it for any reason.
v2: Corrected 'last_rendering_' to 'last_read_' in a number of comments that did
not get updated when 'last_rendering_seqno' became 'last_read|write_seqno'
several millenia ago.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Added helper functions for retrieving the ring and seqno entries from a request
structure. This allows the internal workings of the request structure to be
hidden from code that is using these. It also allows for useful
workarounds/debug code to be added as or when necessary.
Note that it is intended that the majority (if not all) uses of the seqno
accessor will disappear eventually as code is updated to use the request
structure itself rather than working with seqno values.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The plan is to use request structures everywhere that seqno values were
previously used. This means saving pointers to structures in places that used to
be simple integers. In turn, that means that the target structure now needs much
more stringent lifetime tracking. That is, it must not be freed while some other
random object still holds a pointer to it.
To achieve this tracking, a reference count needs to be added. Whenever a
pointer to the structure is saved away, the count must be incremented and the
free must only occur when all references have been released.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The aim is to replace seqno values with request structures. A step along the way
is to switch to using the PLR in preference to the OLS. That requires the PLR to
only be valid when and only when the OLS is also valid. I.e., the two must be
kept in lock step. Then, code which was using the OLS can be safely switched
over to using the PLR instead.
For: VIZ-4377
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <Thomas.Daniel@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
That's the version actually taking the dev_priv->power_domains lock.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Before suspending, we wait upon the outstanding GPU requests and flush
our pending idle handlers. This should downclock the GPU to its lowest
power state. Add a WARN to check that the delayed tasks were run and did
their job properly.
Suggested-by: Akash Goel <akash.goel@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The logical ring code was updating the software ring 'head' value
by reading the hardware 'HEAD' register. In LRC mode, this is not
valid as the hardware is not necessarily executing the same context
that is being processed by the software. Thus reading the h/w HEAD
could put an unrelated (undefined, effectively random) value into
the s/w 'head' -- A Bad Thing for the free space calculations.
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The request queue is per-engine, and may therefore contain requests
from several different contexts/ringbuffers. In determining which
request to wait for, this function should only consider requests
from the ringbuffer that it's checking for space, and ignore any
that it finds that belong to other contexts.
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch is the last in series of VLV/CHV PSR,
that finally enable PSR by adding it to HAS_PSR
and calling the proper enable and disable
functions on the right places.
Although it is still disabled by default.
v2: Rebase over intel_psr and merge Durgadoss's fixes.
v3: Fix typo.
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Add debugfs support for Valleyview and Cherryview considering that
we have PSR per pipe and we don't have any kind of
performance counter as we have on other platforms that support PSR.
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch introduces exit/activate functions for PSR
on VLV+. Since on VLV+ HW cannot track frame updates and force PSR
exit let's use fully SW tracking available.
v2: Rebase over intel_psr.c;
Remove Single Frame update transitioning from state 3 to 5 directly;
Fake a software invalidation for sprites and cursor so we don't miss
any screen update;
v3: As pointed out by Durgadoss msecs_to_jiffies used on wait_for only uses int,
so let's use 1 instead. Althought the 1/4 of this is needed for the
transition let's use 1 for simplicity;
Also fix comments as suggested by Durgadoss
Cc: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The biggest difference from HSW/BDW PSR here is that VLV enable_source
function enables PSR but let it in Inactive state. So it might be called
on early stage along with setup and enable_sink ones.
v2: Rebase over intel_psr.c;
Remove docs from static functions;
Merge vlv_psr_active_on_pipe;
Timeout for psr transition is 250us;
Remove SRC_TRASMITTER_STATE;
v3: Rebase after is_psr_enabled function got removed;
Get SRC_TRANSMITTER_STATE back to be on the safe side since
default for panels is to require link training on exit when
main link off;
As pointed out by Durgadoss msecs_to_jiffies used on wait_for only uses int,
so let's use 1 instead. Althought the 1/4 of this is needed for the
transition let's use 1 for simplicity;
Cc: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Baytrail (Valleyview) and Braswell (Cherryview) uses a complete different
implementation of PSR that we currently have supported for
Haswell and Broadwell. So let's start by adding registers definitions.
I usually don't like commit that adds just registers without using,
but after I put all in one commit I realized that no one would want
to take the AR to review it so I decided to split in order to make
reviewer's life easier. Only last commit in this series will actually
enable the PSR on intel enable panel path.
But as it happens currently with HSW/BDW the plan is to let it
disabled by default (protected by kernel parameter)
while we are able to fully validate it.
v2: Remove a unused bit definition that isn't used on vlv and
reserved on chv as pointed out by Durgadoss.
Cc: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This function was in use to check if PSR feature got enabled.
However on HSW and BDW we currently force psr exit by disabling
EDP_PSR_ENABLE bit at EDP_PSR_CTL(dev). So this function was actually
returning the active/inactive state that is different from the enable/disable
meaning and had the risk of false negative.
But anyway this check with DRRS was dangerous, since DRRS could try to get enabled
before PSR gets there. So let's just remove it for now.
A proper synchronization mechanism must be implemented later probably
using pipe config.
Cc: Daniel Vetter <daniel@ffwll.ch>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Single frame update is a feature available on BDW for PSR that allows
Source to send Sink only one frame and get it updated. Usually useful
when page flipping. However with our frontbuffer tracking where we force
psr exit on flips we don't need this feature.
Also after it got added here many workaround was added to documentation
to mask some bits when using single frame update. So the safest thing
is to just stop using it.
v2: Rebase after removing skip aux one and fixing typo on commit message.
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
OEMs can specify if full_link might be always enabled, i.e. only_standby
over VBT.
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Let's use VBT + 1 now we parse it.
v2: fix subject
v3: rebase over intel_psr and without counting on previous fix
Cc: Arthur Runyan <arthur.j.runyan@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
PSR (aka SRD) block is defined at VBT and currently being used.
Mainly/At-least to configure the amount of idle_frames require to get
back to PSR Entry.
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Durgadoss R <durgadoss.r@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Because the plane registers are different in Skylake we need to adapt
the MMIO code as well.
v2: Don't introduce yet another vfunc when the direction is do
consolidate the plane updates to use the same code path (Daniel)
v3:
- Use enum pipe instead of int (Ville)
- Also update PLANE_STRIDE when the tiling has changed (Ville)
- Put intel_mark_page_flip_active() in the shared code (Damien)
v4:
- Remove unused variable
v5:
- Fix whitespace Vs tabs (Ville)
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
v2: Put the DPLL0 state readout in skylake_get_ddi_pll(), closer to
where the PLL assignement read out is done rather than the frequency
readout function. (Daniel)
v3: Remove stray new line (Damien)
Add Paulo's r-b tag for v1
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> (v1)
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
On pre-HSW we have two encoders per digital port: one HDMI, one DP.
However they are the same physical port in hardware and we can't enable
both at the same time. Reject the modeset if the user attempts this.
So far we've been saved by the fact that we never see both HDMI and DP
connectors as connected. But if the user decides to force a mode anyway,
all kinds of funny stuff might happen.
Unfortunately we don't seem to have any way to inform userspace that
such configurations are invalid except by returning an error from
setcrtc. possible_clones only covers real cloning situations, and
looking at the connector names doesn't work either since we don't
always register both connectors for the same port. I suppose the
only way to fix that would be to expose only a single encoder per
digital port like we do on HSW+ but that would be a fairly large
undertaking for little gain.
kms_setmode hits this since it forces modes on non-connected VGA and
HDMI connectors. Previosuly it just resulted in weirdness such as
failed link training. With this patch it will now get an error back
from the kernel and will die with an assert since it thinks that the
configuration should be fine.
v2: Deal with INTEL_OUTPUT_UNKNOWN (Paulo)
Cc: Paulo Zanoni <przanoni@gmail.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Atm, igt/gem_reset_stats can trigger the recently added WARN on
left-over PM_IIR bits in gen6_enable_rps_interrupts(). There are two
reasons for this:
1. we call intel_enable_gt_powersave() without a preceeding
intel_disable_gt_powersave()
2. gen6_disable_rps_interrupts() doesn't mask interrupts in PM_IMR
1. means RPS interrupts will remain enabled and can be serviced during
the HW initialization after a GPU reset. 2. means even if we called
gen6_disable_rps_interrupts() any new RPS interrupt during RPS
initialization would still propagate to PM_IIR too early (though
wouldn't be serviced).
This patch solves the 2. issue by also masking interrupts in PM_IMR, the
following patch fixes 1. getting rid of the WARN. This also makes
intel_enable_gt_powersave() and intel_disable_gt_powersave() more
symmetric.
Since gen6_disable_rps_interrupts() is called during driver loading with
i915 interrupts disabled add a new version of gen6_disable_pm_irq() that
doesn't WARN for this.
Also while at it, get the irq_lock around the whole PM_IMR/IER/IIR
programming sequence and make sure that any queued PM_IIR bit is also
cleared.
The WARN was caught by PRTS after I sent my previous RPS sanitizing
patchset and I could easily reproduce it on HSW. To actually fix it we
also need the next patch.
Reported-by: He, Shuang <shuang.he@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We don't really synchronously turn them off from debugfs. We try to
avoid hitting them too badly by waiting one vblank, but apparently the
irq handler can still race through that gap.
Since this isn't really all that important for testcases, only for
debugging CRC issues let's tune it down to a debug message.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=82602
Cc: Damien Lespiau <damien.lespiau@intel.com>
Acked-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Dynamic context pinning for LRCs introduced a leak in legacy mode.
Reinstate context unreference in i915_gem_free_request for legacy contexts.
Leak reported by i-g-t/drv_module_reload fixed by this patch.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=86507
Signed-off-by: Thomas Daniel <thomas.daniel@intel.com>
Reviewed-by: John Harrison<John.C.Harrison@Intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updates in forcewake range for Render/Media/Common
power wells for Gen9.
Signed-off-by: Akash Goel <akash.goel@intel.com>
Signed-off-by: Zhe Wang <zhe1.wang@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Before testing if the panel VDD is enabled on eDP cancel any pending
disable worker. This makes sure the worker will be triggered with a
delay from the last time edp_panel_vdd_schedule_off() is called, not
the first time. This avoids unnecessary overhead.
https://bugs.freedesktop.org/show_bug.cgi?id=86201
v2: use cancel_delayed_work() instead of cancel_delayed_work_sync()
as the pps_mutexes will provide the required serialization with
edp_panel_vdd_work() while the sync variant may deadlock. Suggested
by Ville Syrjälä <ville.syrjala@linux.intel.com>.
Made commit message a bit clearer.
Signed-off-by: Egbert Eich <eich@suse.de>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The crc code doesn't handle anything really that could drop the
register state (by design so that we have less complexity). Which
means userspace may only start crc capture once the pipe is fully set
up.
With an i-g-t patch this will be the case, but there's still the
problem that this results in obscure unclaimed register write
failures. Which is a pain to debug.
So instead make sure we don't have the basic unclaimed register write
failure by grabbing runtime pm references. And reject completely
invalid requests with -EIO. This is still racy of course, but for a
test library we don't really care - if userspace shuts down the pipe
right afterwards the entire setup will be lost anyway.
v2: Put instead of get, spotted by Damien. Also explain the runtime pm
dance.
v3: There's really no need for rpm get/put since power_is_enabled only
checks software state (Damien).
References: https://bugs.freedesktop.org/show_bug.cgi?id=86092
Cc: Damien Lespiau <damien.lespiau@intel.com> (v2)
Tested-by: lu hua <huax.lu@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The GPU reset also resets the display on gen3/4. The g33 docs say we
should disable all planes before flipping the reset switch. Just
disable all the crtcs instead. That seems a nicer thing to do anyway.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
On gen4 and earlier the GPU reset also resets the display, so we should
protect against concurrent modeset operations. Grab all the modeset locks
around the entire GPU reset dance, remebering first ti dislogde any
pending page flip to make sure we don't deadlock. Any pageflip coming
in between these two steps should fail anyway due to reset_in_progress,
so this should be safe.
This fixes a lot of failed asserts in the modeset code when there's a
modeset racing with the reset. Naturally the asserts aren't happy when
the expected state has disappeared.
v2: Drop UMS checks, complete pending flips after the reset (Daniel)
Cc: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
g33 seems to sit somewhere between the 915/945/965 style and the
g4x style. The bits look like g4x, but we still need to do a full
reset including display.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
915/945 have the same reset registers as 965, so share the code.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
On pre-ctg GPU reset also resets the display hardware. Force a mode
restore after the GPU reset, and also re-init clock gating.
v2: Use intel_modeset_init_hw() instead of intel_init_clock_gating()
in case more relevant stuff gets added there at some point
Restore interrupts after the reset as well
Tested-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
On pre-ctg the reset bit directly controls the reset signal. We must
assert it for >=20usec and then deassert it. Bit 1 is a RO status bit
which should also go down when the reset is no longer asserted.
Tested-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
There's quite a few bug reports with error states where the error
reasons makes just about no sense at all. Like dying on tlbs for a
display plane that's not even there. Also users don't really report a
lot of bad side effects generally, just the error states.
Furthermore we don't even enable these interrupts any more on gen5+
(though the handling code is still there). So this mostly concerns old
platforms.
Given all that lets make our lives a bit easier and stop capturing
error states, in the hopes that we can just ignore them. In case
that's not true and the gpu indeed dies the hangcheck should
eventually kick in. And I've left some debug log in to make this case
noticeble. Referenced bug is just an example.
v2: Fix missing \n Jani spotted.
References: https://bugs.freedesktop.org/show_bug.cgi?id=82095
References: https://bugs.freedesktop.org/show_bug.cgi?id=85944
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The problem here is that SNA pins batchbuffers to etch out a bit more
performance. Iirc it started out as a w/a for i830M (which we've
implemented in the kernel since a long time already). The problem is
that the pin ioctl wasn't added in
commit d23db88c3a
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri May 23 08:48:08 2014 +0200
drm/i915: Prevent negative relocation deltas from wrapping
Fix this by simply disallowing pinning from userspace so that the
kernel is in full control of batch placement again. Especially since
distros are moving towards running X as non-root, so most users won't
even be able to see any benefits.
UMS support is dead now, but we need this minimal patch for
backporting. Follow-up patch will remove the pin ioctl code
completely.
Note to backporters: You must have both
commit b45305fce5
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Mon Dec 17 16:21:27 2012 +0100
drm/i915: Implement workaround for broken CS tlb on i830/845
which laned in 3.8 and
commit c4d69da167
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon Sep 8 14:25:41 2014 +0100
drm/i915: Evict CS TLBs between batches
which is also marked cc: stable. Otherwise this could introduce a
regression by disabling the userspace w/a without the kernel w/a being
fully functional on i830/45.
References: https://bugs.freedesktop.org/show_bug.cgi?id=76554#c116
Cc: stable@vger.kernel.org # requires c4d69da167 and v3.8
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
In all likelihood we will do a few hundred errnoneous register
operations if we do a single invalid register access whilst the device
is suspended. As each instance causes a WARN, this floods the system
logs and can make the system unresponsive.
The warning was first introduced in
commit b2ec142cb0
Author: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Fri Feb 21 13:52:25 2014 -0300
drm/i915: call assert_device_not_suspended at gen6_force_wake_work
and despite the claims the WARN is still encountered in the wild today.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Imre Deak <imre.deak@intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
CHV infoframes were not being enabled.
Signed-off-by: Clint Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When doing a nop modeset we currently leave crtc->new_config point at
the already freed temporary pipe_config. That will anger the sanity
checks in intel_modeset_update_state() when the nop modeset gets
followed by a GPU reset on gen3/4 where the display block gets fully
reinitialized during the reset.
So leave crtc->new_config alone until we know a modeset is actually
required.
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The DRM connector's encoder pointer is managed internally by the DRM
core and set to NULL when the DRM connector is disconnected from the
CRTC it was attached to. This results in a NULL pointer dereference in
the HDMI connector functions when trying to call the associated slave
encoder's operations.
Fix this by retrieving the slave encoder pointer from the R-Car
connector structure instead of the DRM connector structure.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
drm-intel-next-2014-11-21:
- infoframe tracking (for fastboot) from Jesse
- start of the dri1/ums support removal
- vlv forcewake timeout fixes (Imre)
- bunch of patches to polish the rps code (Imre) and improve it on bdw (Tom
O'Rourke)
- on-demand pinning for execlist contexts
- vlv/chv backlight improvements (Ville)
- gen8+ render ctx w/a work from various people
- skl edp programming (Satheeshakrishna et al.)
- psr docbook (Rodrigo)
- piles of little fixes and improvements all over, as usual
* tag 'drm-intel-next-2014-11-21-fixed' of git://anongit.freedesktop.org/drm-intel: (117 commits)
drm/i915: Don't pin LRC in GGTT when dumping in debugfs
drm/i915: Update DRIVER_DATE to 20141121
drm/i915/g4x: fix g4x infoframe readout
drm/i915: Only call mod_timer() if not already pending
drm/i915: Don't rely upon encoder->type for infoframe hw state readout
drm/i915: remove the IRQs enabled WARN from intel_disable_gt_powersave
drm/i915: Use ggtt error obj capture helper for gen8 semaphores
drm/i915: vlv: increase timeout when setting idle GPU freq
drm/i915: vlv: fix cdclk setting during modeset while suspended
drm/i915: Dump hdmi pipe_config state
drm/i915: Gen9 shadowed registers
drm/i915/skl: Gen9 multi-engine forcewake
drm/i915: Read power well status before other registers for drpc info
drm/i915: Pin tiled objects for L-shaped configs
drm/i915: Update ring freq for full gpu freq range
drm/i915: change initial rps frequency for gen8
drm/i915: Keep min freq above floor on HSW/BDW
drm/i915: Use efficient frequency for HSW/BDW
drm/i915: Can i915_gem_init_ioctl
drm/i915: Sanitize ->lastclose
...
Not just the userspace relocs, otherwise we won't wait
for a swapped out page tables to be swapped in again.
v2: rebased on Alex current drm-fixes-3.18
Signed-off-by: Christian König <christian.koenig@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Original code sent always 0 as the index number of the node. This patch fixes
this bug by sending a variable which is incremented per node.
Signed-off-by: Ben Goz <ben.goz@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
This patch fixes a device QCM bug, where the number of queues were not
counted correctly for the operation of update queue. The count was incorrect
as there was no regard to the previous state of the queue.
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
This patch starts to add support for the VI APU in the KQ (kernel queue)
module.
Because most (more than 90%) of the KQ code is shared among AMD's APUs, we
chose a design that performs most/all the code in the shared KQ file
(kfd_kernel_queue.c). If there is H/W specific code to be executed,
than it is written in an asic-specific extension function for that H/W.
That asic-specific extension function is called from the shared function at the
appropriate time. This requires that for every asic-specific extension function
that is implemented in a specific ASIC, there will be an equivalent
implementation in ALL ASICs, even if those implementations are just stubs.
That way we achieve:
- Maintainability: by having one copy of most of the code, we only need to
fix bugs at one locations
- Readability: very clear what is the shared code and what is done per ASIC
- Extensibility: very easy to add new H/W specific files/functions
Signed-off-by: Ben Goz <ben.goz@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
This patch does some re-org on the kernel_queue structure. It takes out
all the function pointers from the structure and puts them in a new structure,
called kernel_queue_ops. Then, it puts an instance of that structure
inside kernel_queue.
This re-org is done to prepare the KQ module to support more than one AMD APU
(Kaveri).
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
This patch starts to add support for the VI APU in the DQM module.
Because most (more than 90%) of the DQM code is shared among AMD's APUs, we
chose a design that performs most/all the code in the shared DQM file
(kfd_device_queue_manager.c). If there is H/W specific code to be executed,
than it is written in an asic-specific extension function for that H/W.
That asic-specific extension function is called from the shared function at the
appropriate time. This requires that for every asic-specific extension function
that is implemented in a specific ASIC, there will be an equivalent
implementation in ALL ASICs, even if those implementations are just stubs.
That way we achieve:
- Maintainability: by having one copy of most of the code, we only need to
fix bugs at one locations
- Readability: very clear what is the shared code and what is done per ASIC
- Extensibility: very easy to add new H/W specific files/functions
Signed-off-by: Ben Goz <ben.goz@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
This patch does some re-org on the device_queue_manager structure. It takes out
all the function pointers from the structure and puts them in a new structure,
called device_queue_manager_ops. Then, it puts an instance of that structure
inside device_queue_manager.
This re-org is done to prepare the DQM module to support more than one AMD APU
(Kaveri).
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Instead of creating a BUG if trying to free a NULL GART sub-allocation object,
just return 0 (success).
This is done to mirror behavior of kfree.
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
LRC object does not need to be mapped into the GGTT when dumping. A side-effect
of this patch is that a compiler warning goes away (not checking return value
of i915_gem_obj_ggtt_pin).
v2: Broke out individual context dumping into a new function as the indentation
was getting a bit crazy. Added notification of contexts with no gem object for
debugging purposes. Removed unnecessary pin_pages and unpin_pages, replaced
with explicit get_pages for the context object as there may be no backing store
allocated at this time (Comment for get_pages says "Ensure that the associated
pages are gathered from the backing storage and pinned into our object").
Improved error checking - get_pages and get_page are checked for failure.
Signed-off-by: Thomas Daniel <thomas.daniel@intel.com>
[danvet: Align paramter continuation lines properly. Also add some
braces to the nested loops again for readability.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Otherwise we'll have backtraces in assert_panel_unlocked because the
BIOS locks the register. In the reporter's case this regression was
introduced in
commit c31407a367
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu Oct 18 21:07:01 2012 +0100
drm/i915: Add no-lvds quirk for Supermicro X7SPA-H
Reported-by: Alexey Orishko <alexey.orishko@gmail.com>
Cc: Alexey Orishko <alexey.orishko@gmail.com>
Cc: stable@vger.kernel.org
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Francois Tigeot <ftigeot@wolfpond.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Tested-by: Alexey Orishko <alexey.orishko@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This patch adds the basic structure of a DRM Driver for Rockchip Socs.
Signed-off-by: Mark Yao <mark.yao@rock-chips.com>
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
Acked-by: Daniel Vetter <daniel@ffwll.ch>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Apparently PCH fifo underruns are tricky, we have plenty reports that
we see the occasional underrun (especially at boot-up).
So for a change let's see what happens when we don't re-enable pch
fifo underrun reporting when the pipe is disabled. This means that the
kernel can't catch pch fifo underruns when they happen (except when
all pipes are on on the pch). But we'll still catch underruns when
disabling the pipe again. So not a terrible reduction in test
coverage.
Since the DRM_ERROR is new and hence a regression plan B would be to
revert it back to a debug output. Which would be a lot worse than this
hack for underrun test coverage in the wild. See the referenced
discussions for more.
References: http://mid.gmane.org/CA+gsUGRfGe3t4NcjdeA=qXysrhLY3r4CEu7z4bjTwxi1uOfy+g@mail.gmail.com
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85898
References: https://bugs.freedesktop.org/show_bug.cgi?id=85898
References: https://bugs.freedesktop.org/show_bug.cgi?id=86233
References: https://bugs.freedesktop.org/show_bug.cgi?id=86478
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Tested-by: lu hua <huax.lu@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
- Tegra K1 voltage support, and coherency improvements
- GM204 support (modesetting, still waiting on NVIDIA for signed fw to
proceed further), and a lot of bios/i2c/devinit adjustments needed to
support it
- GT21x memory reclocking work
- Various other bits and pieces, most of which are prep-work for a
couple of bigger projects I didn't get finished in time
* 'linux-3.19' of git://anongit.freedesktop.org/git/nouveau/linux-2.6: (73 commits)
drm/nv50/kms: drop requirement that framebuffer bos be contig up-front
drm/nv50/kms: directly use cursor image from userspace buffer
drm/nouveau/kms: when pinning display-related buffers, force contig vram
drm/nouveau: teach nouveau_bo_pin() how to force a contig vram allocation
drm/nouveau/volt: add support for GK20A
drm/nouveau/platform: add GPU speedo information to nouveau platform
drm/nouveau/volt: allow non-bios voltage scaling
drm/gf100-/gr: return non-fatal error code when fw not present
drm/nouveau/devinit: bump priv ring timeouts before executing scripts
drm/nouveau/bios: translate ramcfg strap through M0203
drm/nouveau/fb: make use of M0203 routines for ram type determination
drm/nouveau/bios: add parsing of BIT M(v2) +0x03 table
drm/nouveau/core: allow vbios parsing without knowing chipset type
drm/nouveau/lib: add null backend
drm/nouveau/device: store revision
drm/nouveau/core: add some forgotten subdevs to disable mask
drm/gk20a/clk: fix max VCO value
drm/nouveau: we need pin_refcnt for nouveau_bo_placement_set()
drm/nv50-/kms: add some evo tracing ability for debugging
drm/nv50/kms: use sclass() instead of trial-and-error
...
Preparation for transition to planes, which use framebuffers for the
cursor image. We've always done copies from the userspace buffer up
until now for legacy reasons, there's no good reason to do so on the
chipsets this code covers.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
We have the ability to move buffers around in the kernel if necessary,
and should probably use it rather than failing if userspace passes us
a non-contig buffer for a plane.
The NOUVEAU_GEM_TILE_NONCONTIG flag from userspace will become a mere
initial placement hint once all the relevant paths have been updated.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
The voltage value are calculated by the hardware characterized
result.
Signed-off-by: Vince Hsu <vinceh@nvidia.com>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Martin Peres <martin.peres@free.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
For GK20A we need the GPU speedo value to calculate voltage levels.
Signed-off-by: Vince Hsu <vinceh@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Move the vbios parsing out of init() and call it conditionally if the
platform has a vbios. Non-vbios platforms can use the ctor() to init the
data structures.
Signed-off-by: Vince Hsu <vinceh@nvidia.com>
Acked-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Martin Peres <martin.peres@free.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
A machine has been spotted where the ramcfg strap is "8", and the ramcfg
xlat table goes 0-7,0-7, resulting in us selecting config 0 for memory
items. On this particular system, config "8" is available and supposed
to be used. It appears that starting from GT21x (where Mv2 appears),
we're supposed to use the value in this table instead.
One concern here is that not all the places we currently use ramcfg xlat
are supposed to be treated the same now. The strap xlat table wasn't
removed from the vbios either, presumably for some kind of good reason.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
We only support one kind of matching here (ramcfg strap), but it appears
alternate methods are possible. I wrote a tool to scan our vbios repo
for other types, but did not see any used. Hopefully this means there
aren't any in the wild that will now break.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
For some reason max_vco was set to a lower value that it can support,
which prevented some clock states to be applied. Fix this by setting it
to the same value as downstream.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
On architectures for which access to GPU memory is non-coherent,
caches need to be flushed and invalidated explicitly when BO control
changes between CPU and GPU.
This patch adds buffer synchronization functions which invokes the
correct API (PCI or DMA) to ensure synchronization is effective.
Based on the TTM DMA cache helper patches by Lucas Stach.
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Specify TTM_PL_FLAG_UNCACHED when allocating GPFIFOs and fences to
allow them to be safely accessed by the kernel without being synced
on non-coherent architectures.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Allow nouveau_bo_new() to recognize the TTM_PL_FLAG_UNCACHED flag, which
means that we want the allocated BO to be perfectly coherent between the
CPU and GPU. This is useful on non-coherent architectures for which we
do not want to manually sync some rarely-accessed buffers: typically,
fences and pushbuffers.
A TTM BO allocated with the TTM_PL_FLAG_UNCACHED on a non-coherent
architecture will be populated using the DMA API, and accesses to it
performed using the coherent mapping performed by dma_alloc_coherent().
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Add a function allowing us to know whether a device is CPU-coherent,
i.e. accesses performed by the CPU on GPU-mapped buffers will
be immediately visible on the GPU side and vice-versa.
For now, a device is considered to be coherent if it uses the PCI bus on
a non-ARM architecture.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Pinned BOs are supposed to remain in their current location until
unpinned. Display a warning for the supposedly-erroneous case where we
are trying to move such objects.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Probably missing something here, doesn't make a lot of sense to write
or+link data into a register whose offset is calculated by the same
or+link info..
This is the all I've witnessed the binary driver and vbios doing so
far, so it'll do.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
The binary driver has been doing this since GF119, and we've somehow
gotten away with it. But, TMDS that hasn't been initialised already
by the x86 vbios code is distorted without it on GM204.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Logging at trace level, rather than as en error, as it seems conceivable
that failure could be normal under certain circumstances (new bios,
older sink that doesn't support a particular DPCD address)
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Starting from GM204, certain registers are no longer accessible by the host
(or unsigned PMU firmware).
This commit implements devinit on PMU, using a signed microcode image, and
devinit data, from the VBIOS.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
The field at +0x2 is technically processor specific, though I don't know
that it's ever mattered in practice (yet).
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
We're about to need to be able to fetch additional chunks of data beyond
the primary bios image, which makes fetching a lot more complicated.
This splits out the verious shadowing routines to be nothing more than
very dumb "fetch this much data from this offset" routines, and leaves
the logic of what and how much to fetch in common code.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Just a couple of fixes for the fallout from the fence rework.
* 'linux-3.18' of git://anongit.freedesktop.org/git/nouveau/linux-2.6:
drm/nouveau/gf116: remove copy1 engine
drm/nouveau: prevent stale fence->channel pointers, and protect with rcu
drm/nouveau/fifo/g84-: ack non-stall interrupt before handling it
Indications are that no GF116's actually have a copy engine there, but
actually have the decompression engine. This engine can be made to do
copies, but that should be done separately.
Unclear why this didn't turn up on all GF116's, but perhaps the
non-mobile ones came with enough VRAM to not trigger ttm migrations in
test scenarios.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=85465
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=59168
Cc: stable@vger.kernel.org
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Closes a very unlikely race that can occur if another NonStallInterrupt
method passes between checking fences and acking the previous interrupt.
With this change, the interrupt will re-fire under such conditions.
Tested-by: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
add gpio bitbanging i2c adapter on LPC device of atom e6xx
gpu chipset to access lvds EDID
tested on SECO QuadMo747-E6xx-EXTREME Qseven platform
Reviewed-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Signed-off-by: Jan Safrata <jan.nikitenko@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
So here's a pile of atomic fixes and improvements from various people.
There's still more patches in-flight, so I think I'll keep collecting them
in a separate branch.
* tag 'topic/core-stuff-2014-11-28' of git://anongit.freedesktop.org/drm-intel:
drm/atomic: clear plane's CRTC and FB when shutting down
drm: Handle atomic state properly in kms getfoo ioctl
drm: use mode_object_find helpers
drm: fix indentation
drm/msm: switch to atomic-helpers iterator macros
drm/atomic: add plane iterator macros
drm/atomic: track bitmask of planes attached to crtc
drm: Free atomic state during cleanup
drm: Make drm_atomic.h standalone includible
drm: Make drm_atomic_helper.h standalone includible
drm/plane: Add missing kerneldoc
drm/plane: Pass old state to ->atomic_update()
drm/atomic_helper: Cope with plane->crtc == NULL in disable helper
drm/atomic: Drop per-plane locking TODO
drm/atomic-helper: Skip vblank waits for unchanged fbs
drm: Document that drm_dev_alloc doesn't need a parent
The function will also be used by a later patch, so factor it out.
V2: make raw_edid const, define/declare before first use
V3: fix erroneuos removal of csum variable
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
There is no need to dump the whole EDID block in case it contains no
information. Just print a single line stating the block is empty instead
of 8 lines containing only zeroes.
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm_edid_is_zero will be used by drm_edid_block valid, move it up.
raw_edid argument can be const.
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Pull drm fixes from Dave Airlie:
"Two i915 regressions and one dual-gpu laptop radeon fix"
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
drm/radeon: report disconnected for LVDS/eDP with PX if ddc fails
drm/i915: Cancel vdd off work before suspend
drm/i915: Ignore SURFLIVE and flip counter when the GPU gets reset
Two regression fixes from Ville.
* tag 'drm-intel-fixes-2014-11-27' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Cancel vdd off work before suspend
drm/i915: Ignore SURFLIVE and flip counter when the GPU gets reset
Otherwise we'd still end up w/ the plane attached to the CRTC, and
seemingly active, but without an FB. Which ends up going *boom*
in the drivers.
Slightly modified version of Daniel's irc suggestion.
Note that the big problem isn't drivers going *boom* here (since we
already have the situation of planes being left enabled when the crtc
goes down). The real issue is that the core assumes the primary plane
always goes down when calling ->set_config with a NULL mode. Ignoring
that assumption leads to the legacy state pointers plane->fb/crtc
getting out of sync with atomic, and that then leads to the subsequent
*boom* all over the place.
CC: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Rob Clark <robdclark@gmail.com>
[danvet: Drop my opinion of what's going sidewides here into the
commit message as a note.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
So the problem with async commit (especially async modeset commit) is
that the legacy pointers only get updated after the point of no
return, in the async part of the modeset sequence. At least as
implemented by the current helper functions. This is done in the
set_routing_links function in drm_atomic_helper.c.
Which also means that access isn't protected by locks but only
coordinated by synchronizing with async workers. No problem thus far,
until we lock at the getconnector/encoder ioctls.
So fix this up by adding special cases for atomic drivers: For those
we need to look at state objects. Unfortunately digging out the
correct encoder->crtc link is a bit of work, so wrap this up in a
helper function.
Moving the assignments of connector->encoder and encoder->crtc earlier
isn't a good idea because the point of the atomic helpers is that we
stage the state updates. That way the disable functions can still
inspect the links and rely upon them.
v2: Extract full encoder->crtc lookup into helper (Rob).
v3: Extract drm_connector_get_encoder too since - we need to always
return state->best_encoder when there is a state otherwise we might
return stale data if there's a pending async disable (and chase
unlocked pointers, too). Same issue with encoder_get_crtc but there
it's a bit more tricky to handle.
Cc: Rob Clark <robdclark@gmail.com>
Cc: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Lightly-Tested-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Chasing plane->state->crtc of planes that are *not* part of the same
atomic update is racy, making it incredibly awkward (or impossible) to
do something simple like iterate over all planes and figure out which
ones are attached to a crtc.
Solve this by adding a bitmask of currently attached planes in the
crtc-state.
Note that the transitional helpers do not maintain the plane_mask. But
they only support the legacy ioctls, which have sufficient brute-force
locking around plane updates that they can continue to loop over all
planes to see what is attached to a crtc the old way.
Signed-off-by: Rob Clark <robdclark@gmail.com>
[danvet:
- Drop comments about locking in set_crtc_for_plane since they're a
bit misleading - we already should hold lock for the current crtc.
- Also WARN_ON if get_state on the old crtc fails since that should
have been done already.
- Squash in fixup to check get_plane_state return value, reported by
Dan Carpenter and acked by Rob Clark.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If ddc fails, presumably the i2c mux (and hopefully the signal
mux) are switched to the other GPU so don't fetch the edid from
the vbios so that the connector reports disconnected.
bug:
https://bugzilla.opensuse.org/show_bug.cgi?id=904417
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
The branch is based on a merge of drm-next and Simon's tags/renesas-dt-du-for-
v3.19 available at
git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git, the latter
having been pulled in the ARM SoC tree for v3.19.
Compared to v1, I've rebased my branch on a later drm-next, added Julia's
error return code fix, and documented the "drm: Decouple EDID parsing from I2C
adapter" patch properly.
v1:
Here's a pull request that adds HDMI support to the R-Car DU driver, including
a new slave encoder driver for the adv7511.
* 'drm/du/adv7511' of git://linuxtv.org/pinchartl/fbdev:
drm: Add adv7511 encoder driver
video: Add ADV751[13] DT bindings documentation
drm: Decouple EDID parsing from I2C adapter
drm: rcar-du: Add HDMI encoder and connector support
drm: rcar-du: Replace drm_encoder with drm_slave_encoder
drm: rcar-du: Replace direct DRM encoder access with cast macro
drm: rcar-du: Pass the encoder DT node to rcar_du_encoder_init()
drm: rcar-du: Remove platform data support
drm: rcar-du: fix error return code
ARM: shmobile: koelsch: Enable DU device in DT
ARM: shmobile: koelsch-reference: Remove DU platform device
ARM: shmobile: lager: Enable DU device in DT
ARM: shmobile: lager-reference: Remove DU platform device
ARM: shmobile: marzen: Enable DU device in DT
ARM: shmobile: dts: Add common file for AA104XD12 panel
ARM: shmobile: r8a7791: Add DU node to device tree
ARM: shmobile: r8a7790: Add DU node to device tree
ARM: shmobile: r8a7779: Add DU node to device tree
ARM: shmobile: Remove FSF address from copyright headers
- Fixes for sparse warnings
- Memory leak fixes
- Fix for deadlock between amdkfd and iommu
* 'amdkfd-next-3.19' of git://people.freedesktop.org/~gabbayo/linux:
amdkfd: delete some dead code
amdkfd: Fix memory leak of mqds on dqm fini
amdkfd: fix an error handling bug in pqm_create_queue()
amdkfd: fix some error handling in ioctl
amdkfd: Remove DRM_AMDGPU dependency from Kconfig
amdkfd: explicitely include io.h in kfd_doorbell.c
amdkfd: Clear ctx cb before suspend
amdkfd: Instead of using get function, use container_of
amdkfd: use schedule() in sync_with_hw
amdkfd: Fix memory leak on process deregistration
amdkfd: add __iomem attribute to doorbell_ptr
amdkfd: fence_wait_timeout() can be static
amdkfd: is_occupied() can be static
amdkfd: Fix sparse warnings in kfd_flat_memory.c
amdkfd: pqm_get_kernel_queue() can be static
amdkfd: test_kq() can be static
amdkfd: Fix sparse warnings in kfd_topology.c
amdkfd: Fix sparse warnings in kfd_chardev.c
This patch adds a driver for the Analog Devices adv7511. The adv7511 is
a standalone HDMI transmitter chip. It features a HDMI output interface
on one end and video and audio input interfaces on the other.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
The drm_get_edid() function performs direct I2C accesses to read EDID
blocks, assuming that the monitor DDC interface is directly connected to
the I2C bus. It can't thus be used with HDMI encoders that control the
DDC bus and expose EDID blocks through a different interface.
Refactor drm_do_get_edid() to take a block read callback function
instead of an I2C adapter, and export it for direct use by drivers.
As in the general case the DDC bus is accessible by the kernel at the
I2C level, drivers must make all reasonable efforts to expose it as an
I2C adapter and use drm_get_edid() instead of abusing this function.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
SoCs that integrate the DU have no internal HDMI encoder, support
external encoders only.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
DRM slave encoders require their associated struct drm_encoder instance
to be embedded in a struct drm_slave_encoder. This makes processing
encoders regardless of their types needlessly and painfully complex in
drivers that use a mix of slave encoders and custom encoders. Such a
driver will need to either create drm_slave_encoder instances that fake
their embedded encoder instance, or to turn all drm_encoder instances
into drm_slave_encoder instances.
Between the two evils, one must choose the lesser. Use drm_slave_encoder
everywhere.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Add a new macro to downcast an rcar_du_encoder pointer to a drm_encoder
pointer and use it. This prepares for the replacement of the
rcar_drm_encoder encoder field with a drm_slave_encoder.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
The encoder DT node will be needed to register an external HDMI encoder.
Pass it to the rcar_du_encoder_init() function to prepare for HDMI
support.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
All platforms now instantiate the DU through DT, platform data support
isn't needed anymore.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Propagate the error code on failure.
A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)
// <smpl>
@@
identifier ret; expression e1,e2;
@@
(
if (\(ret < 0\|ret != 0\))
{ ... return ret; }
|
ret = 0
)
... when != ret = e1
when != &ret
*if(...)
{
... when != ret = e2
when forall
return ret;
}
// </smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Currently we just make sure vdd is off before suspending, but we don't
cancel the vdd off work. The work wil not touch vdd if
want_panel_vdd==false so in theory this is fine.
In the past that was perfectly fine since the vdd off work didn't do
anything when want_panel_vdd==false, so even if the work would have been
run during system resume before i915 has resumed, nothing would happen.
However since pps_lock() will now grab the power domain references before
it can check want_panel_vdd, we may end up toggling the power wells on/off
already before the driver has resumed. That is not really acceptable, so
cancel the vdd off work when suspending the encoder.
The problem appeared when pps_lock() was introduced in:
commit 773538e860
Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
Date: Thu Sep 4 14:54:56 2014 +0300
drm/i915: Reset power sequencer pipe tracking when disp2d is off
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
The drm_prime_pages_to_sg() function never returns NULL pointers, only
error pointers and valid pointers.
Signed-off-by: Haixia Shi <hshi@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Fixes dmabuf export failure with -E_NOMEM when the page is not mapped.
Signed-off-by: Haixia Shi <hshi@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The imx-drm driver was put into staging mostly for the following reasons,
all of which have been addressed or superseded:
- convert the irq driver to use linear irq domains
- work out the device tree bindings, this lead to the common of_graph
bindings being used
- factor out common helper functions, this mostly resulted in the
component framework and drm of_graph helpers.
Before adding new fixes, and certainly before adding new features,
move it into its proper place below drivers/gpu/drm.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Add Exynos4415 SoC support, some fixups and cleanups.
Summary:
- Resolve kernel lockup issue incurred by probe request in probe context.
. For this, it moves all register codes of sub drivers into init function
and adds component binding support for vidi driver.
- Add Exynos4415 SoC support.
- Make each manager and display object to be embedded
in each driver context.
- Fix and clean up FIMD and MIPI-DSI drivers.
- Clean up unnecesary or wrong descriptions.
- And trivial cleanups.
* 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos: (58 commits)
drm/exynos: avoid leak if exynos_dpi_probe() fails
drm/exynos: Fix exynos_dpi_remove() parameter
drm/exynos: vidi: add component support
drm/exynos: fix exynos_drm_component_del
drm/exynos/ipp: fix error return code
drm/exynos: clean up machine compatible string check
drm/exynos: move Exynos platform drivers registration to init
Revert "drm/exynos: fix null pointer dereference issue"
drm/exynos/dpi: stop using display->ctx pointer
drm/exynos/dpi: embed display into private context
drm/exynos/dp: stop using display->ctx pointer
drm/exynos/dp: embed display into private context
drm/exynos/vidi: stop using display->ctx pointer
drm/exynos/vidi: embed display into private context
drm/exynos/hdmi: stop using display->ctx pointer
drm/exynos/hdmi: embed display into private context
drm/exynos/fimd: stop using manager->ctx pointer
drm/exynos/fimd: embed manager into private context
drm/exynos/vidi: stop using manager->ctx pointer
drm/exynos/vidi: embed manager into private context
...
This is dead code. We don't need to unbind here, we can just return
directly.
Reviewed-by: Oded Gabbay <oded.gabbay@amd.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
The mqds array members are not freed when dqm is uninitialized.
Reviewed-by: Ben Goz <Ben.Goz@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
The current state of CRTCs, planes and connectors currently leaks during
DRM driver ->unload() unless drivers explicitly clean it up. Since there
is nothing driver-specific about it, that cleanup can be done within the
DRM core.
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In most situations it will be useful to have the old state passed to the
->atomic_update() callback. For example if a plane is being disabled the
new state's .crtc field will be NULL, but some drivers may rely on this
field to program the CRTCs registers.
v2: rename variable to old_plane_state and remove redundant comment as
suggested by Daniel Vetter, remove an Exynos hunk that doesn't apply to
drm-next and add a hunk for pending MSM mdp5 changes
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The drm core can call the plane disable hook multiple times, which
means it can get called when plane->crtc is already NULL. That in turn
means we can't get at the implicit acquire ctx we use in the atomic
helpers for legacy entries points.
We could try to pass drm_modeset_legacy_acquire_ctx a drm_device
pointer so that it can cope with a NULL crtc. But that still doesn't
work since the cursor ioctls (remapped with the universal cursor plane
support code) only grabs the crtc locks. So the global acquire context
isn't set eitehr.
The real solution here would be to bite the bullet and wire up
explicit acquire context parameters to all relevant functions. We need
to do that anyway (to be able to get rid of some small allocations
which we can't cope with failing). But that's a lot of work and better
done once atomic has settled a bit.
So meanwhile just catch this case in the helper and bail out.
Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
[danvet: Completely rewrite commit message and comment but keep
Jasper's logic and author credits since his patch is the only
short-term solution that works.]
Tested-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
I've forgotten to remove that in my per-plane locking patch.
Reported-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Especially with legacy cursor ioctls existing userspace assumes that
you can pile up lots of updates in one go. The super-proper way to
support this would be a special commit mode which overwrites the last
update. But getting there will be quite a bit of work.
Meanwhile do what pretty much all the drivers have done for the plane
update functions: Simply skip the vblank wait for the buffer cleanup
if the buffer is the same. Since the universal cursor plane code will
not recreate framebuffers needlessly this allows us to not slow down
legacy pageflip events while someone moves the cursor around.
v2: Drop the async plane update hunk from a previous attempt at this
issue.
v3: Fix up kerneldoc.
v4: Don't oops so badly. Reported by Jasper.
Cc: Rob Clark <robdclark@gmail.com>
Cc: "Jasper St. Pierre" <jstpierre@mecheye.net>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net>
Tested-by: Jasper St. Pierre <jstpierre@mecheye.net>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The call to kernel_queue_uninit(NULL) will trigger a BUG(), and also the
error code is incorrect.
Fixes: 45102048f7 ('amdkfd: Add process queue manager module')
Reviewed-by: Oded Gabbay <oded.gabbay@amd.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
There is a typo here so the errors from kfd_bind_process_to_device()
are not detected.
Reviewed-by: Oded Gabbay <oded.gabbay@amd.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
The component must be deleted if the probe fails.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_dpi_remove() should receive a exynos_drm_display but when
DRM_EXYNOS_DPI was disabled it was receiving a struct device resulting in
ia compiler warning.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This patch adds component support for vidi driver.
vidi driver is a kms driver so it doesn't need to be registered
to exynos_drm_subdrv_list. For this, it changes for the component
framework to be used for vidi driver.
This patch fixes below error also,
# echo 1 > /sys/devices/platform/exynos-drm-vidi/connection
[ 55.618529] ------------[ cut here ]------------
[ 55.621960] WARNING: CPU: 0 PID: 1397 at drivers/gpu/drm/drm_irq.c:1203 exynos_drm_crtc_dpms+0x88/0x17c()
[ 55.631268] Modules linked in:
[ 55.634278] CPU: 0 PID: 1397 Comm: sh Not tainted 3.18.0-rc2-146253-g31449d7 #1154
[ 55.641885] [<c0014400>] (unwind_backtrace) from [<c0011570>] (show_stack+0x10/0x14)
[ 55.649597] [<c0011570>] (show_stack) from [<c04764f4>] (dump_stack+0x84/0xc4)
[ 55.656802] [<c04764f4>] (dump_stack) from [<c00218b8>] (warn_slowpath_common+0x6c/0x88)
[ 55.664866] [<c00218b8>] (warn_slowpath_common) from [<c0021970>] (warn_slowpath_null+0x1c/0x24)
[ 55.673632] [<c0021970>] (warn_slowpath_null) from [<c027a780>] (exynos_drm_crtc_dpms+0x88/0x17c)
[ 55.682482] [<c027a780>] (exynos_drm_crtc_dpms) from [<c027a910>] (exynos_drm_crtc_commit+0x14/0x44)
[ 55.691622] [<c027a910>] (exynos_drm_crtc_commit) from [<c025521c>] (drm_crtc_helper_set_mode+0x3d0/0x51c)
[ 55.701233] [<c025521c>] (drm_crtc_helper_set_mode) from [<c0255d68>] (drm_crtc_helper_set_config+0x87c/0x9dc)
[ 55.711230] [<c0255d68>] (drm_crtc_helper_set_config) from [<c026afa8>] (drm_mode_set_config_internal+0x58/0xd4)
[ 55.721380] [<c026afa8>] (drm_mode_set_config_internal) from [<c025c208>] (restore_fbdev_mode+0xcc/0xec)
[ 55.730834] [<c025c208>] (restore_fbdev_mode) from [<c025c244>] (drm_fb_helper_restore_fbdev_mode_unlocked+0x1c/0x30)
[ 55.741424] [<c025c244>] (drm_fb_helper_restore_fbdev_mode_unlocked) from [<c025e0a8>] (drm_fb_helper_set_par+0x1c/0x60)
[ 55.752271] [<c025e0a8>] (drm_fb_helper_set_par) from [<c025e174>] (drm_fb_helper_hotplug_event+0x88/0xc4)
[ 55.761906] [<c025e174>] (drm_fb_helper_hotplug_event) from [<c02571c4>] (drm_helper_hpd_irq_event+0xc8/0x134)
[ 55.771898] [<c02571c4>] (drm_helper_hpd_irq_event) from [<c028e27c>] (vidi_store_connection+0x90/0xc8)
[ 55.781268] [<c028e27c>] (vidi_store_connection) from [<c0125f80>] (kernfs_fop_write+0xc0/0x180)
[ 55.790045] [<c0125f80>] (kernfs_fop_write) from [<c00cdf60>] (vfs_write+0xa0/0x1ac)
[ 55.797757] [<c00cdf60>] (vfs_write) from [<c00ce468>] (SyS_write+0x44/0x9c)
[ 55.804790] [<c00ce468>] (SyS_write) from [<c000e6a0>] (ret_fast_syscall+0x0/0x30)
[ 55.812328] ---[ end trace 3c0fe4386702d4dd ]---
This issue occurs when modeset to vidi is tried in case that drm_vblank_init
is called prior to crtc creation of vidi driver. In this case, crtc number
of vidi is invalid so any requests with the crtc number will fail.
This patch guarantees drm_vblank_init to be called after all kms drivers
are ready by using component framework.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This patch resolves the issue that component object isn't removed
correctly.
A given component object couldn't be placed to head of drm_component_list
so all component objects added to the drm_component_list should be checked
to remove the given component object.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Propagate the returned error code on failure.
A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)
// <smpl>
@@
identifier ret; expression e1,e2;
@@
(
if (\(ret < 0\|ret != 0\))
{ ... return ret; }
|
ret = 0
)
... when != ret = e1
when != &ret
*if(...)
{
... when != ret = e2
when forall
return ret;
}
// </smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Registering the Exynos DRM subdevices platform drivers in the probe
function is causing an infinite loop. Fix this by moving it to the
exynos_drm_init() function to register the drivers on module init.
Registering drivers in the probe functions causes a deadlock in the parent
device lock. See Grant Likely explanation on the topic:
"I think the problem is that exynos_drm_init() is registering a normal
(non-OF) platform device, so the parent will be /sys/devices/platform.
It immediately gets bound against exynos_drm_platform_driver which
calls the exynos drm_platform_probe() hook. The driver core obtains
device_lock() on the device *and on the device parent*.
Inside the probe hook, additional platform_drivers get registered.
Each time one does, it tries to bind against every platform device in
the system, which includes the ones created by OF. When it attempts to
bind, it obtains device_lock() on the device *and on the device
parent*.
Before the change to move of-generated platform devices into
/sys/devices/platform, the devices had different parents. Now both
devices have /sys/devices/platform as the parent, so yes they are
going to deadlock.
The real problem is registering drivers from within a probe hook. That
is completely wrong for the above deadlock reason. __driver_attach()
will deadlock. Those registrations must be pulled out of .probe().
Registering devices in .probe() is okay because __device_attach()
doesn't try to obtain device_lock() on the parent."
INFO: task swapper/0:1 blocked for more than 120 seconds.
Not tainted 3.18.0-rc3-next-20141105 #794
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
swapper/0 D c052534c 0 1 0 0x00000000
[<c052534c>] (__schedule) from [<c0525b34>] (schedule_preempt_disabled+0x14/0x20)
[<c0525b34>] (schedule_preempt_disabled) from [<c0526d44>] (mutex_lock_nested+0x1c4/0x464
[<c0526d44>] (mutex_lock_nested) from [<c02be908>] (__driver_attach+0x48/0x98)
[<c02be908>] (__driver_attach) from [<c02bcc00>] (bus_for_each_dev+0x54/0x88)
[<c02bcc00>] (bus_for_each_dev) from [<c02bdce0>] (bus_add_driver+0xe4/0x200)
[<c02bdce0>] (bus_add_driver) from [<c02bef94>] (driver_register+0x78/0xf4)
[<c02bef94>] (driver_register) from [<c029e99c>] (exynos_drm_platform_probe+0x34/0x234)
[<c029e99c>] (exynos_drm_platform_probe) from [<c02bfcf0>] (platform_drv_probe+0x48/0xa4)
[<c02bfcf0>] (platform_drv_probe) from [<c02be680>] (driver_probe_device+0x13c/0x37c)
[<c02be680>] (driver_probe_device) from [<c02be954>] (__driver_attach+0x94/0x98)
[<c02be954>] (__driver_attach) from [<c02bcc00>] (bus_for_each_dev+0x54/0x88)
[<c02bcc00>] (bus_for_each_dev) from [<c02bdce0>] (bus_add_driver+0xe4/0x200)
[<c02bdce0>] (bus_add_driver) from [<c02bef94>] (driver_register+0x78/0xf4)
[<c02bef94>] (driver_register) from [<c029e938>] (exynos_drm_init+0x70/0xa0)
[<c029e938>] (exynos_drm_init) from [<c00089b0>] (do_one_initcall+0xac/0x1f0)
[<c00089b0>] (do_one_initcall) from [<c074bd90>] (kernel_init_freeable+0x10c/0x1d8)
[<c074bd90>] (kernel_init_freeable) from [<c051eabc>] (kernel_init+0x8/0xec)
[<c051eabc>] (kernel_init) from [<c000f268>] (ret_from_fork+0x14/0x2c)
3 locks held by swapper/0/1:
#0: (&dev->mutex){......}, at: [<c02be908>] __driver_attach+0x48/0x98
#1: (&dev->mutex){......}, at: [<c02be918>] __driver_attach+0x58/0x98
#2: (&dev->mutex){......}, at: [<c02be908>] __driver_attach+0x48/0x98
Changelog v2:
- call platform_driver_register after all kms and non kms drivers are
registered
- rebased it to exynos-drm-next
Signed-off-by: Javier Martinez Canillas <javier.martinez@collabora.co.uk>
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This reverts commit cea24824ab432f8acabb254d6805e9aa756de6af.
Moving subdriver probe to exynos_drm_platform_probe() was making
exynos_drm_device_subdrv_probe() fail because the platform data wasn't set
yet. It only gets set in exynos_drm_load.
We need to find a smarter way to fix this issue.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to display->ctx pointer by container_of
construct. The field is removed as well as dpi was the last user of it.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_display is used by internal Exynos DRM framework for
representing encoder:connector pair. As it should be mapped 1:1 to dpi
private context it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple dpi devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to display->ctx pointer by container_of
construct. It will allow to remove ctx field in the future.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_display is used by internal Exynos DRM framework for
representing encoder:connector pair. As it should be mapped 1:1 to dp
private context it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple dp devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to display->ctx pointer by container_of
construct. It will allow to remove ctx field in the future.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_display is used by internal Exynos DRM framework for
representing encoder:connector pair. As it should be mapped 1:1 to vidi
private context it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple vidi devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to display->ctx pointer by container_of
construct. It will allow to remove ctx field in the future.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_display is used by internal Exynos DRM framework for
representing encoder:connector pair. As it should be mapped 1:1 to hdmi
private context it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple hdmi devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to manager->ctx pointer by container_of
construct. As fimd was the last user of ctx the patch removes
this field as well.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_manager is used by internal Exynos DRM framework for
representing crtc. As it should be mapped 1:1 to fimd private context
it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple FIMD devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to manager->ctx pointer by container_of
construct. It will allow to remove ctx field in the future.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_manager is used by internal Exynos DRM framework for
representing crtc. As it should be mapped 1:1 to vidi private context
it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple mixer devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to manager->ctx pointer by container_of
construct. It will allow to remove ctx field in the future.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_drm_manager is used by internal Exynos DRM framework for
representing crtc. As it should be mapped 1:1 to fimd private context
it seems more reasonable to embed it directly in that context.
As a result further code simplification will be possible.
Moreover it will be possible to handle multiple mixer devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The exynos_dsi_te_irq_handler() works only dsi(DPMS) is on.
So it is enough to enable and disable TE GPIO IRQ in
exynos_dsi_enable(disable)_irq() like DSI IRQ.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>a
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The drm_helper_hpd_irq_event() does dpms control and
the panel is initialized and displayed on by it.
So the exynos_dsi_te_irq_handler() should be registered
beforehand.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The drm_handle_vblank should be called whenever be vsync, te interrupt
means vsync on i80 interface.
Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
There is a case like set config which requires triggering
but vblank is not enabled yet.
So triggering unset routine is required to exit from
triggering mode.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
For the I80 interface, the video interrupt pending register(VIDINTCON1)
should be handled in fimd_irq_handler() and the video interrupt control
register(VIDINTCON0) should be handled in fimd_enable_vblank() and
fimd_disable_vblank() like RGB interface.
So this patch moves each set / unset routines into proper positions.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This function is valid only the SoC has SHADOWCON register
and it should be used together with fimd_enable_video_output()
to match the ENWIN_F bit in WINCON# and C#_EN_F bit in SHADOWCON.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This bit is used for video output and logic signal control.
So it is better for readability.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The C#_EN_F in SHADOWCON register is updated per frame.
So it should be protected by fimd_shadow_protect_win().
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
It's better to be checking whether triggerring in fimd_trigger function.
Also it will return if in triggerring on fimd_te_handler, then it can't
execute remain codes.
Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The exynos fimd provides video type selection bits from system register
but exynos3 series don't has it, so needs has_vtsel flag and we can
distinguish whether set video type selection bits.
Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
During system resume from suspend to RAM the Exynos DRM driver forced
CRTC mode thus turning display on (DPMS_ON). This lead to runtime resuming
of DSI which failed because whole LCD power domain was off and it was
not allowed to turn on because of system resume in progress.
Forcing mode should not be needed and removing it solves this particular
problem.
This necessary fix for following scenario reproduced on Exynos DRM:
1. Power domain is off before suspending the system.
2. System is suspended to RAM.
3. Resuming starts. The Exynos DRM driver resume callback is called.
4. The Exynos DRM driver calls drm_helper_resume_force_mode() which turns
on the screen by calling exynos_dsi_dpms with DRM_MODE_DPMS_ON.
5. The Exynos DSI driver calls pm_runtime_get. The driver runtime
resumes and this should turn LCD power domain on.
6. Unfortunately the domain cannot be turned on because system resume is
in progress and genpd->prepared_count is positive.
Steps to reproduce:
1. Add runtime PM to Exynos DSI driver.
2. Build Exynos DRM/FB without FRAMEBUFFER_CONSOLE.
3. Enable the connector and screen (e.g. with modeset-vsync).
4. echo 3 > /sys/devices/platform/exynos-drm/graphics/fb0/blank
5. echo mem > /sys/power/state
6. Resume.
[ 77.712469] PM: early resume of devices complete after 3.854 msecs
[ 77.712739] exynos-dsi 11c80000.dsi: pm_genpd_resume()
[ 77.712758] exynos4-fimc 11800000.fimc: pm_genpd_resume()
[ 77.712774] exynos4-fimc 11810000.fimc: pm_genpd_resume()
[ 77.712787] exynos-drm-fimc 11820000.fimc: pm_genpd_resume()
[ 77.712802] exynos-drm-fimc 11830000.fimc: pm_genpd_resume()
[ 77.712815] s5p-mipi-csis 11880000.csis: pm_genpd_resume()
[ 77.712829] s5p-mipi-csis 11890000.csis: pm_genpd_resume()
[ 77.712843] exynos-fimc-lite 12390000.fimc-lite: pm_genpd_resume()
[ 77.712856] exynos-fimc-lite 123a0000.fimc-lite: pm_genpd_resume()
[ 77.713788] exynos4-fb 11c00000.fimd: pm_genpd_resume()
[ 77.713912] wake disabled for irq 184
[ 77.713923] wake disabled for irq 185
[ 77.714082] wake disabled for irq 173
[ 77.715676] wake disabled for irq 176
[ 77.718540] exynos4-fb 11c00000.fimd: pm_genpd_runtime_resume()
[ 77.718567] exynos4-fb 11c00000.fimd: state restore latency exceeded, new value 1708 ns
[ 77.718636] exynos-dsi 11c80000.dsi: pm_genpd_runtime_resume()
[ 77.892366] exynos-dsi 11c80000.dsi: PLL failed to stabilize
[ 77.892377] exynos-dsi 11c80000.dsi: failed to configure DSI PLL
[ 78.192168] exynos-dsi 11c80000.dsi: timeout waiting for reset
[ 78.211578] exynos-dsi 11c80000.dsi: waiting for bus lanes timed out
[ 78.307173] exynos-dsi 11c80000.dsi: xfer timed out: d1 00 (null)
[ 78.307190] panel_s6e8aa0 11c80000.dsi.0: error -110 reading dcs seq(0xd1)
[ 78.307199] panel_s6e8aa0 11c80000.dsi.0: read id failed
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The command mode panel should draw image earlier than the display
on command execution to prevent showing garbage GRAM screen data.
So should set dsi->state as DSIM_STATE_ENABLED between calling
exynos_dsi_set_display_enable() and drm_panel_enable() to transmit
image data before executing display on command.
And moves the display on command execution routine from prepare()
to enable() in drm_panel_funcs also.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
For providing VBLANK information, drm_handle_vblank() should
be called properly, but it is blocked by wait_vsync_event
condition which is set by manager_ops->wait_for_vblank().
So moves it out from wait_vsync_event routine.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
The exynos_drm_crtc_dpms() waits until pended page flip
queue is empty, calls the drm_vblank_off() then calls
manager->ops->dpms() when mode is DRM_MODE_DPMS_OFF.
The fimd_dpms() is one of manager->ops->dpms()s and
finally calls fimd_window_suspend().
But there is no active window and vblank is already off
when it is called.
So addtional waiting vblank is not necessary any more.
Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
Acked-by: Inki Dae <inki.dae@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Now that we have moved to generic phy based bindings,
we don't need to have any code related to older dptx-phy.
Nobody is using this dptx-phy anymore, so removing the
same.
Signed-off-by: Vivek Gautam <gautam.vivek@samsung.com>
Acked-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
They are not implemented anywhere, so wipe them out.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Description of the @create_connector callback was missing,
and the @manager was no longer needed.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Save a few bytes by compiling them all in the same byte.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This is a leftover, all code using this macro have been removed/
changed already.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Re-declare struct is not a good practice, let's use the original
drm and exynos declarations.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Re-declare struct is not a good practice, let's use the original
drm declarations.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The struct is defined in the same file, declare it here is just
unnecessary
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The struct is defined in the same file, declare it here is just
unnecessary.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
It is not even used in this header anymore.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces accesses to display->ctx pointer by container_of
construct. It will allow to remove ctx field in the future.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch removes redundant encoder field from private DSI context.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces multiple evaluation of device address
with local variable.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
exynos_dsi_display is used by internal Exynos DRM framework for
representing pair encoder->connecter. As it should be mapped 1:1 to dsi
private context it seems more reasonable to embed it directly
in that context. As a result further code simplification will be possible.
Moreover it will be possible to handle multiple DSI devices in the system.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The patch replaces separate calls to driver (de)registration by
loops over the array of drivers. As a result it significantly
decreases number of ifdefs. Additionally it moves device registration
related ifdefs to header file.
Changelog v2:
- Rebased.
- Consider non kms driver in respect to infinite loop issue.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This patch fixes null pointer dereference issue incurred
when ipp driver is enabled and Exynos drm driver is closed.
Non kms driver should register its own sub driver to setup necessary
resources, which is done by load(). So null pointer dereference
occurs when ipp driver is enabled and Exynos drm driver is closed
because ipp core device is registered after component_master_add_with_match
call.
This patch makes exynos_drm_device_subdrv_probe() to be called after all non
kms drivers are registered.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This patch fixes a null pointer dereference issue incurred by
calling g2d_remove when exynos_drm_platform_probe is failed.
cmdlist_pool of g2d is allocated when g2d sub driver is probed.
So if exynos_drm_platform_probe is failed, the g2d sub driver is
not probed and the cmdlist_pool is still NULL.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This patch resovles the infinite loop issue incurred
when Exyno drm driver is enabled but all kms drivers
are disabled on Exynos board by returning -EPROBE_DEFER
only in case that there is kms device registered.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
This patch resolves temporarily infinite loop issue incurred
when Exynos drm driver is enabled and multi-platform kernel
is used by registering Exynos drm device object only in case
of Exynos SoC. So this patch will be replaced with more generic
way later.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
During a GPU reset we need to get pending page flip cleared out
since the ring contents are gone and flip will never complete
on its own. This used to work until the mmio vs. CS flip race
detection came about. That piece of code is looking for a
specific surface address in the SURFLIVE register, but as
a flip to that address may never happen the check may never
pass. So we should just skip the SURFLIVE and flip counter
checks when the GPU gets reset.
intel_display_handle_reset() tries to effectively complete
the flip anyway by calling .update_primary_plane(). But that
may not satisfy the conditions of the mmio vs. CS race
detection since there's no guarantee that a modeset didn't
sneak in between the GPU reset and intel_display_handle_reset().
Such a modeset will not wait for pending flips due to the ongoing GPU
reset, and then the primary plane updates performed by
intel_display_handle_reset() will already use the new surface
address, and thus the surface address the flip is waiting for
might never appear in SURFLIVE. The result is that the flip
will never complete and attempts to perform further page flips
will fail with -EBUSY.
During the GPU reset intel_crtc_has_pending_flip() will return
false regardless, so the deadlock with a modeset vs. the error
work acquiring crtc->mutex was avoided. And the reset_counter
check in intel_crtc_has_pending_flip() actually made this bug
even less severe since it allowed normal modesets to go through
even though there's a pending flip.
This is a regression introduced by me here:
commit 75f7f3ec60
Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
Date: Tue Apr 15 21:41:34 2014 +0300
drm/i915: Fix mmio vs. CS flip race on ILK+
Testcase: igt/kms_flip/flip-vs-panning-vs-hang
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Some radeon ASICs don't support all 64 address bits of MSIs despite
advertising support for 64-bit MSIs in their configuration space.
This breaks on systems such as IBM POWER7/8, where 64-bit MSIs can
be assigned with some of the high address bits set.
This makes use of the newly introduced "no_64bit_msi" flag in structure
pci_dev to allow the MSI allocation code to fallback to 32-bit MSIs
on those adapters.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
CC: <stable@vger.kernel.org>
---
Adding Alex's review tag. Patch to the driver is identical to the
reviewed one, I dropped the arch/powerpc hunk rewrote the subject
and cset comment.
In disable_outputs() we need to shut down the outgoing encoder, not the
incoming one (we have already swapped-state at this point). Without
this, we end up telling the driver to crtc->dpms(OFF) without first
encoder->dpms(OFF), and that makes some hw quite unhappy.
v2: missing WARN_ON() hunk and comment
Reviewed-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Rob Clark <robdclark@gmail.com>
The intention is that drivers can set crtc_state->mode_changed in their
atomic_check() fxns if they encounter a scenario that requires full
modeset.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch removes the dependency of amdkfd upon DRM_AMDGPU symbol in amdkfd's
Kconfig file.
This is done because amdgpu driver is not yet upstreamed and therefore,
DRM_AMDGPU symbol is not present in any Kconfig file.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
On mdp4 there is a single global LAYERMIXER_IN_CFG register. The
previous logic to share that between multiple crtcs didn't actually
handle plane-disable very well. Easier just to look at all of the
crtcs each time.
Signed-off-by: Rob Clark <robdclark@gmail.com>
This patch fixes a compilation error when using certain configuration by
including the file io.h in kfd_doorbell.c
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
When disabling the interface (INTF), the change doesn't latch until next
vblank, so we need to wait for vblank.
Also, to be pedantic, in the crtc, set all the mixer stages to unused.
It shouldn't really matter, since at this point we have already disabled
the INTF and waited for necessary vblank.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Convert mdp5 over to atomic helpers. Extend/wrap drm_plane_state to
track plane zpos and to keep track of the needed when applying the
atomic update. In mdp5's plane->atomic_check() we also need to check
for updates which require SMP reallocation, in order to trigger full
modeset.
Signed-off-by: Rob Clark <robdclark@gmail.com>
For example, use 'struct mdp5_smp *' everywhere instead of 'void *', but
only declare it as 'struct mdp5_smp;' in common headers, so the struct
body is still private. The accomplishes the desired modularity while
still letting the compiler provide some type checking for us.
Signed-off-by: Rob Clark <robdclark@gmail.com>
MDP5 currently support one single CRTC with its private pipe.
This change allows the configuration of multiple CRTCs with
the possibility to attach several public planes to these CRTCs.
Signed-off-by: Stephane Viau <sviau@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
The hardware configuration modification from a version to another
is quite consequent. Introducing a configuration module
(mdp5_cfg) may make things more clear and easier to access when a
new hardware version comes up.
Signed-off-by: Stephane Viau <sviau@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
The Shared Memory Pool (SMP) has its own limitation, features and
state. Some examples are:
- the number of Memory Macro Block (MMB) and their size
- the number of lines that can be fetched
- the state of MMB currently allocated
- the computation of number of blocks required per plane
- client IDs ...
In order to avoid private data to be overwritten by other modules,
let's make these private to the SMP module.
Some of these depend on the hardware configuration, let's add them
to the mdp5_config struct.
In some hw configurations, some MMBs are statically tied to RGB
pipes and cannot be re-allocated dynamically. This change
introduces the concept of MMB static usage and makes sure that
dynamic MMB requests are dimensioned accordingly.
A note on passing a pipe pointer, instead of client IDs:
Client IDs are SMP-related information. Passing PIPE information
to SMP lets SMP module to find out which SMP client(s) are used.
This allows the SMP module to access the PIPE pointer, which can
be used for FIFO watermark configuration.
By the way, even though REG_MDP5_PIPE_REQPRIO_FIFO_WM_* registers
are part of the PIPE registers, their functionality is to reflect
the behavior of the SMP block. These registers access is now
restricted to the SMP module.
Signed-off-by: Stephane Viau <sviau@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
A left-over from prior to component framework. The original intent was
to deal with hdmi getting unloaded before the master component, but that
isn't really going to work anyways. These days with the component
framework taking care to unload the master component first, we don't
have to worry about this.
Signed-off-by: Rob Clark <robdclark@gmail.com>
The core clock rate depends on the hw configuration. Once we have
read the hardware revision, we can set the core clock to its
maximum value.
Before then, the clock is set at a rate supported by all MDP5
revisions.
Signed-off-by: Stephane Viau <sviau@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
For mdp5, the irqs of hdmi/eDP/dsi0/dsi1 blocks get routed through the
mdp block. In order to decouple hdmi/eDP/etc, register an irq domain
in mdp5. When hdmi/dsi/etc are used with mdp4, they can directly setup
their irqs in their DT nodes as normal. When used with mdp5, instead
set the mdp device as the interrupt-parent, as in:
mdp: qcom,mdss_mdp@fd900000 {
compatible = "qcom,mdss_mdp";
interrupt-controller;
#interrupt-cells = <1>;
...
};
hdmi: qcom,hdmi_tx@fd922100 {
compatible = "qcom,hdmi-tx-8074";
interrupt-parent = <&mdp>;
interrupts = <8 0>; /* MDP5_HW_INTR_STATUS.INTR_HDMI */
...
};
There is a slight awkwardness, in that we cannot disable child irqs
at the mdp level, they can only be cleared in the child block. So
you must not use threaded irq handlers in the child. I'm not sure
if there is a better way to deal with that.
Signed-off-by: Rob Clark <robdclark@gmail.com>
fix one regression and one endian issue.
* 'drm-fixes-3.18' of git://people.freedesktop.org/~agd5f/linux:
drm/radeon: fix endian swapping in vbios fetch for tdp table
drm/radeon: disable native backlight control on pre-r6xx asics (v2)
- More CI dpm fixes
- Initial DPM fan control for SI/CI (disabled by default)
- GPUVM multi-ring efficiency improvements
- Some cursor fixes
* 'drm-next-3.19' of git://people.freedesktop.org/~agd5f/linux: (22 commits)
drm/radeon: update the VM after setting BO address v4
drm/radeon: sync PT updates as shared v2
drm/radeon: sync PD updates as shared
drm/radeon: fence BO_VAs manually
drm/radeon: use one VMID for each ring
drm/radeon: track VM update fences separately
drm/radeon: fence PT updates manually v2
drm/radeon: split semaphore and sync object handling v2
drm/radeon: remove unnecessary VM syncs
drm/radeon: stop re-reserving the BO in radeon_vm_bo_set_addr
drm/radeon: rework vm_flush parameters
drm/radeon/ci: disable needless sclk changes
drm/radeon/ci: force pcie level before sclk and mclk
drm/radeon/ci: use different smc command for pcie dpm
drm/radeon/ci: apply disp voltage changes before clk changes
drm/radeon: fix PCC debugging message for CI DPM
drm/radeon/dpm: add thermal dpm support for CI
drm/radeon/dpm: add smc fan control for CI (v2)
drm/radeon/dpm: add smc fan control for SI (v2)
drm/radeon: work around a hw bug in MGCG on CIK
...
The vfree() function performes also input parameter validation. Thus the test
around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Reviewed-by: Thierry Reding <thierry.reding@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The vunmap() function performes also input parameter validation. Thus the test
around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Reviewed-by: Thierry Reding <thierry.reding@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The drm_fbdev_cma_hotplug_event() function tests whether its argument is NULL
and then returns immediately. Thus the test around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Reviewed-by: Thierry Reding <thierry.reding@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The release_firmware() function tests whether its argument is NULL and then
returns immediately. Thus the test around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Reviewed-by: Thierry Reding <thierry.reding@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is an oversight from
commit f52b69f1ec
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Wed Nov 19 18:38:08 2014 +0100
drm/atomic: Don't overrun the connector array when hotplugging
Cc: Dave Airlie <airlied@redhat.com>
Cc: Rob Clark <robdclark@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
It happens on occasion that developers of generic user-space applications
abuse the dumb buffer API to get hold of drm buffers that they can both
mmap() and use for GPU acceleration, using the assumptions that dumb buffers
and buffers available for GPU are
a) The same type and can be aribtrarily type-casted.
b) fully coherent.
This patch makes the most widely used drivers warn nicely when that happens,
the next step will be to fail.
v2: Move drmP.h changes to drm_gem.h. Fix Radeon dumb mmap breakage.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Need to check the port too.
Reported-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The final arrangement of updating timer->expires and calling mod_timer()
used in
commit 672e7b7c18
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed Nov 19 09:47:19 2014 +0000
drm/i915: Don't continually defer the hangcheck
turns out to be very unsafe. Try again.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
encoder->type can change underneath us and doesn't need to reflect
actual hw state (since we don't construct it from hw state like
e.g. encoder->crtc crtc->config).
And this can indeed happen:
1) Boot with plugged-in hdmi screen. Since we only set ->type in the
probe functions this means we won't detect any infoframes since
type is still unkown.
2) First probe sets type to HDMI.
3) If the first modeset now does _not_ happen on the HDMI pipe with
infoframes encoder->get_config suddenly sees infoframes and the
state checker gets angry.
Fix this by only relying on actual hw state when figuring out whether
the ddi port is in hdmi mode and sends infoframes.
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: Paulo Zanoni <przanoni@gmail.com>
Reported-by: Paulo Zanoni <przanoni@gmail.com>
Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This way the necessary VM update is kicked off immediately
if all BOs involved are in GPU accessible memory.
v2: fix vm lock
v3: immediately update unmaps as well
v4: use drm_free_large instead of kfree
Tested-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Only invalidating PTEs needs to be executed synchronized to using the PT.
v2: fix sync to uses
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We never invalidate PD entries and making them valid can
run with other users in parallel.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This allows us to finally remove the VM fence and
so allow concurrent use of it from different engines.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use multiple VMIDs for each VM, one for each ring. That allows
us to execute flushes separately on each ring, still not ideal
cause in a lot of cases rings can share IDs.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Note for each fence if it's a VM page table update or not. This allows
us to determine the last VM update in a sync object and so to figure
out if we need to flush the TLB or not.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This allows us to add the real execution fence as shared.
v2: fix typo
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Previously we just allocated space for four hardware semaphores
in each software semaphore object. Make software semaphore objects
represent only one hardware semaphore address again by splitting
the sync code into it's own object.
v2: fix typo in comment
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The PD/PTs reservation object now contains everything needed.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
That's useless when all callers drop the reservation
immediately after calling the function.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use ring structure instead of index and provide vm_id and pd_addr separately.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The current code always reprogrammed the sclk levels,
but we don't currently handle disp sclk requirements
so just skip it.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable smc fan control for CI boards. Should
reduce the fan noise on systems with a higher
default fan profile.
v2: disable by default, add additional fan setup, rpm control
bug:
https://bugs.freedesktop.org/show_bug.cgi?id=73338
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Enable smc fan control for SI boards. Should
reduce the fan noise on systems with a higher
default fan profile.
v2: disable by default, add rpm controls
bug:
https://bugs.freedesktop.org/show_bug.cgi?id=73338
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Just use the acpi interface. That's what windows uses on this
generation and it's the only thing that seems to work reliably
on these generation parts.
You can still force the native backlight interface by setting
radeon.backlight=1
Bug:
https://bugzilla.kernel.org/show_bug.cgi?id=88501
v2: merge into above if/else block
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Always need to set bit 0 of RLC_CGTT_MGCG_OVERRIDE
to avoid unreliable doorbell updates in some cases.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
The cursor_set2 hook provides the cursor hotspot position within the
cursor image. When the hotspot position changes, we can adjust the cursor
position such that the hotspot doesn't move on the screen. This prevents
the cursor from appearing to intermittently jump around on the screen
when the position of the hotspot within the cursor image changes.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This function can be called now with i915 interrupts enabled, so the
corresponding WARN is incorrect, remove it. I think this was spotted by
Paulo during his review, but since I already removed the same WARN
from intel_suspend_gt_powersave() I missed then his point.
Spotted-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Spotted while reading and trying to understand how our error capture
code deals with full ppgtt.
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
I saw punit timeouts in vlv_set_rps_idle() while running various
subtests of pm_rpm. Increasing the timeout to 100ms got rid of the
issue.
Testcase: igt/pm_rpm
Reference: https://bugs.freedesktop.org/show_bug.cgi?id=82939
Signed-off-by: Imre Deak <imre.deak@intel.com>
Tested-by: Guo Jinxian <jinxianx.guo@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Currently after doing DPMS-OFF on all outputs CDCLK won't be set to its
minimum value as it should. A subsequent modeset to turn off all outputs
will thus run with all power domains disabled, and notice that it needs
to change CDCLK to its minimum value. Since the power domains are
disabled this will emit a register-access-while-suspended WARN and fail
to set the minimum freq.
The proper solution for this is to set the minimum frequency during
DPMS-OFF. That needs a bigger rework that would take into account the
user DPMS setting too during the calculation of the new modesetting
configuration. Until that's done this stop-gap solution gets the PIPE-A
power domain during setting the CDCLK; this domain covers the HW blocks
needed for this.
Idea to use PIPE-A domain from Ville.
Testcase: igt/pm_rpm
Reference: https://bugs.freedesktop.org/show_bug.cgi?id=82939
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Makes it easier to debug infoframe mismatches.
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
For MMIO registers which are shadowed, force wake is not needed to
write to these registers.
v2: Rebase on top of nightly (Damien)
v3: Rebase on top of "Gen9 multiple-engine forcewake" changes
v4: (Mika, Bob, done by Damien)
- Reorder the shadowed registers by popularity
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Zhe Wang <zhe1.wang@intel.com>
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Enable multi-engine forcewake for Gen9.
v2: (Damien)
- Rebase on top of nightly
- Move the register range definitions to intel_uncore.c
- Whitespace fixes
v3: (Addressing Mika's comment, done by Damien)
- Use REG_RANGE() (introduced after the patch was written)
- Add a SKL_NEEDS_FORCE_WAKE() macro that gets rid of a useless
comparison to FORCEWAKE (reg 0xa18c is not used on SKL)
v4: (Damien)
- Use newly introduced ASSIGN_READ/WRITE_MMIO_VFUNCS() macros
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Zhe Wang <zhe1.wang@intel.com>
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
There's no such thing as "list_struct".
Signed-off-by: Andrey Utkin <andrey.krieger.utkin@gmail.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Trying to read the status of the power wells right after taking forcewake
for the other register reads makes little sense. Most of the time the
power wells will still be up due to the recent forcewake. Instead do the
power well status read first, and only then read the register needing
forcewake. This way the reported power well status can actually reflect
what's going on in the system.
Cc: Deepak S <deepak.s@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Deepak S <deepak.s@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Let's just throw in the towel on this one and take the cheap way out.
Based on a patch from Chris Wilson, but checking for a different bit.
Chris' patch checked for even bank layout, this one here for a magic
bit. Given the evidence we've gathered (not much) both work I think,
but checking for the magic bit might be more accurate.
Anyway, works on my gm45 here.
For paranoi restrict to gen4 (and mobile), since we've only ever seen
this on gm45 and i965gm.
Also add some debugfs output so that we can skip the tiled swapping
tests properly in these cases.
v2: Clean up the quirk'ed pin count in free_object to avoid upsetting
the WARN_ON. Spotted by Chris.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=28813
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45092
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In __gen6_update_ring_freq, use the full range of
possible gpu frequencies from max_freq to min_freq.
The actual gpu frequency could be outside the range
from max_freq_softlimit to min_freq_softlimit due
to power/thermal constraints.
Signed-off-by: Tom O'Rourke <Tom.O'Rourke@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In gen8_enable_rps, change the initial rps setting
to the min_freq_softlimit (same as gen6_enable_rps).
Signed-off-by: Tom O'Rourke <Tom.O'Rourke@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Set the min_freq_softlimit to max(RPe, 450MHz).
Setting a floor can ensure a minimum experience
level. The 450MHz value came from a power and
performance study of various types of workloads
(3D, Media, GPGPU, idle, etc).
v2: rebased
Signed-off-by: Tom O'Rourke <Tom.O'Rourke@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Added gen6_init_rps_frequencies() to initialize
the rps frequency values. This function replaces
parse_rp_state_cap(). In addition to reading RPn,
RP0, and RP1 from RP_STATE_CAP register, the new
function reads efficient frequency (aka RPe) from
pcode for Haswell and Broadwell and sets the turbo
softlimits. The turbo minimum frequency softlimit
is set to RPe for Haswell and Broadwell and to RPn
otherwise.
For RPe, the efficiency is based on the frequency/power
ratio (MHz/W); this is considering GT power and not
package power. The efficent frequency is the highest
frequency for which the frequency/power ratio is within
some threshold of the highest frequency/power ratio.
A fixed decrease in frequency results in smaller
decrease in power at frequencies less than RPe than
at frequencies above RPe.
v2: Following suggestions from Chris Wilson and
Daniel Vetter to extend and rename parse_rp_state_cap
and to open-code a poorly named function.
Signed-off-by: Tom O'Rourke <Tom.O'Rourke@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Remove unused variables.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Found one more!
With this we can clear up the ggtt init code a bit, yay!
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
With this all the ums nonsense around gem setup/teardown has
disappeared, yay!
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Again just complicates gem init functions and makes a general mess out
of everything.
Good riddance!
v2: In my enthusiasm to start removing dri1/ums crud I went overboard a
bit and killed parts of hangcheck. Resurrect it.
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
This patch allows framebuffers for cirrus to be created with
32bpp pixel formats provided that they do not violate certain
restrictions of the cirrus hardware.
v2: Use pci resource length for vram size.
Signed-off-by: Zach Reizner <zachr@google.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Only importing an FD to a handle is currently supported on UDL,
but the exporting functionality is equally useful.
Signed-off-by: Haixia Shi <hshi@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
By default set udl_gem_object as cacheable, but set WC flag when attaching
dmabuf. In udl_gem_mmap() update cache attributes based on the flags, similar
to exynos_drm_gem_mmap().
Signed-off-by: Haixia Shi <hshi@chromium.org>
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>
Reviewed-by: Olof Johansson <olofj@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Before this patch, cirrus_device_init could have failed while
cirrus_mm_init succeeded and the driver would have reported overall
success on load. This patch causes cirrus_device_init to return on
the first error encountered.
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Oversight from my kerneldoc cleanup when doing the original atomic
helper series - I've only applied this clarification to the modeset
related helpers, and not the plane update code. Remedy this asap.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
I guess for hysterical raisins this was meant to be the way to read
blob properties. But that's done with the two-stage approach which
uses separate blob kms object and the special-purpose get_blob ioctl.
Shipping userspace seems to have never relied on this, and the kernel
also never put any blob thing onto that property. And nowadays it
would blow up, e.g. in drm_property_destroy. Also it makes no sense to
return values in an ioctl that only returns metadata about everything.
So let's ditch all the internal code for the blob list, rename the
list to be unambiguous and sprinkle comments all over the place to
explain this peculiar piece of api.
v2: Squash in fixup from Rob to remove now unused variables.
Cc: Rob Clark <robdclark@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
- Make it clear that it's a negative errno (more in line with
everything else).
- Clean up the confusion around get_properties vs. getproperty ioctls:
One reads per-obj property values, the other reads property
metadata.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Yet another fallout from not considering DP MST hotplug. With the
previous patches we have stable indices, but it might still happen
that a connector gets added between when we allocate the array and
when we actually add a connector. Especially when we back off due to
ww mutex contention or similar issues.
So store the sizes of the arrays in struct drm_atomic_state and double
check them. We don't really care about races except that we want to
use a consistent value, so ACCESS_ONCE is all we need. And if we
indeed notice that we'd overrun the array then just give up and
restart the entire ioctl.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Otherwise the connector might have been unplugged and destroyed while
we didn't look. Yet another fallout from DP MST hotplugging that I
didn't consider.
To make sure we get this right add an appropriate WARN_ON to
drm_atomic_state_clear (obviously only when we actually have a state
to clear up). And reorder all the state_clear and backoff calls to
make it work out properly.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
I've totally forgotten that with DP MST connectors can now be
hotplugged. And failed to adapt Rob's drm_atomic_state code (which
predates connector hotplugging) to the new realities.
The first step is to make sure that the connector indices used to
access the arrays of pointers are stable. The connection mutex gives
us enough guarantees for that, which means we won't unecessarily block
on concurrent modesets or background probing.
So add a locking WARN_ON and shuffle the code slightly to make sure we
always hold the right lock.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Let's make things a bit easier to debug when things go bad (potentially
under console_lock).
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Andrew Morton wrote:
> On Wed, 12 Nov 2014 13:08:55 +0900 Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp> wrote:
>
> > Andrew Morton wrote:
> > > Poor ttm guys - this is a bit of a trap we set for them.
> >
> > Commit a91576d791 ("drm/ttm: Pass GFP flags in order to avoid deadlock.")
> > changed to use sc->gfp_mask rather than GFP_KERNEL.
> >
> > - pages_to_free = kmalloc(npages_to_free * sizeof(struct page *),
> > - GFP_KERNEL);
> > + pages_to_free = kmalloc(npages_to_free * sizeof(struct page *), gfp);
> >
> > But this bug is caused by sc->gfp_mask containing some flags which are not
> > in GFP_KERNEL, right? Then, I think
> >
> > - pages_to_free = kmalloc(npages_to_free * sizeof(struct page *), gfp);
> > + pages_to_free = kmalloc(npages_to_free * sizeof(struct page *), gfp & GFP_KERNEL);
> >
> > would hide this bug.
> >
> > But I think we should use GFP_ATOMIC (or drop __GFP_WAIT flag)
>
> Well no - ttm_page_pool_free() should stop calling kmalloc altogether.
> Just do
>
> struct page *pages_to_free[16];
>
> and rework the code to free 16 pages at a time. Easy.
Well, ttm code wants to process 512 pages at a time for performance.
Memory footprint increased by 512 * sizeof(struct page *) buffer is
only 4096 bytes. What about using static buffer like below?
----------
>From d3cb5393c9c8099d6b37e769f78c31af1541fe8c Mon Sep 17 00:00:00 2001
From: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Date: Thu, 13 Nov 2014 22:21:54 +0900
Subject: [PATCH] drm/ttm: Avoid memory allocation from shrinker functions.
Commit a91576d791 ("drm/ttm: Pass GFP flags in order to avoid
deadlock.") caused BUG_ON() due to sc->gfp_mask containing flags
which are not in GFP_KERNEL.
https://bugzilla.kernel.org/show_bug.cgi?id=87891
Changing from sc->gfp_mask to (sc->gfp_mask & GFP_KERNEL) would
avoid the BUG_ON(), but avoiding memory allocation from shrinker
function is better and reliable fix.
Shrinker function is already serialized by global lock, and
clean up function is called after shrinker function is unregistered.
Thus, we can use static buffer when called from shrinker function
and clean up function.
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: stable <stable@kernel.org> [2.6.35+]
Signed-off-by: Dave Airlie <airlied@redhat.com>
Commit f9b9faf6d94dd29eab8c128905c7d091f955481d "drm: flip-work: change
drm_flip_work_init prototype" changed the drm_flip_work_init prototype
to a void function, which makes 'ret' an unused variable.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Implement crtc page_flip callback for bochsdrm. The qemu stdvga has no
vblank signaling, so we have to fake it. We do so by instantly calling
drm_send_vblank_event. Tested with kmscon.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Remove the mapping offset from the bo backing the fbdev framebuffer.
Wire up fbdev mmap function to map the backing bo using ttm_fbdev_mmap.
With that patch in place mmap(/dev/fb0) works as expected.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Recently (qemu 2.2+) the qemu stdvga got a register to switch the vga
framebuffer endianness. This patch adds code to explicitly set the
endianness of the framebuffer. In most cases this has no effect as
the default is guest architecture endianness. It is needed though in
case a architecture supports both big and little endian, i.e. for
ppc64le.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
We've killed ums support by now, it's time to reap the benefits. This
one here is getting in the way of doing some ring init cleanup.
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
KMS always intializes, this was only a valid check when userspace
was still in control of the kernel driver.
v2: Comment that we outright reject all dri1/ums params.
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Whether we'll reject them or no-op doesn't really matter ...
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With the deprecation of UMS, and by association DRI1, we have a tough
choice when updating the ring access routines. We either rewrite the
DRI1 routines blindly without testing (so likely to be broken) or take
the liberty of declaring them no longer supported and remove them
entirely. This takes the latter approach.
v2: Also remove the DRI1 sarea updates
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Fix rebase conflicts.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Same as with the context, pinning to GGTT regardless is harmful (it
badly fragments the GGTT and can even exhaust it).
Unfortunately, this case is also more complex than the previous one
because we need to map and access the ringbuffer in several places
along the execbuffer path (and we cannot make do by leaving the
default ringbuffer pinned, as before). Also, the context object
itself contains a pointer to the ringbuffer address that we have to
keep updated if we are going to allow the ringbuffer to move around.
v2: Same as with the context pinning, we cannot really do it during
an interrupt. Also, pin the default ringbuffers objects regardless
(makes error capture a lot easier).
v3: Rebased. Take a pin reference of the ringbuffer for each item
in the execlist request queue because the hardware may still be using
the ringbuffer after the MI_USER_INTERRUPT to notify the seqno update
is executed. The ringbuffer must remain pinned until the context save
is complete. No longer pin and unpin ringbuffer in
populate_lr_context() - this transient address is meaningless and the
pinning can cause a sleep while atomic.
v4: Moved ringbuffer pin and unpin into the lr_context_pin functions.
Downgraded pinning check BUG_ONs to WARN_ONs.
v5: Reinstated WARN_ONs for unexpected execlist states. Removed unused
variable.
Issue: VIZ-4277
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Thomas Daniel <thomas.daniel@intel.com>
Reviewed-by: Akash Goel <akash.goels@gmail.com>
Reviewed-by: Deepak S<deepak.s@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Up until now, we have pinned every logical ring context backing object
during creation, and left it pinned until destruction. This made my life
easier, but it's a harmful thing to do, because we cause fragmentation
of the GGTT (and, eventually, we would run out of space).
This patch makes the pinning on-demand: the backing objects of the two
contexts that are written to the ELSP are pinned right before submission
and unpinned once the hardware is done with them. The only context that
is still pinned regardless is the global default one, so that the HWS can
still be accessed in the same way (ring->status_page).
v2: In the early version of this patch, we were pinning the context as
we put it into the ELSP: on the one hand, this is very efficient because
only a maximum two contexts are pinned at any given time, but on the other
hand, we cannot really pin in interrupt time :(
v3: Use a mutex rather than atomic_t to protect pin count to avoid races.
Do not unpin default context in free_request.
v4: Break out pin and unpin into functions. Fix style problems reported
by checkpatch
v5: Remove unpin_lock as all pinning and unpinning is done with the struct
mutex already locked. Add WARN_ONs to make sure this is the case in future.
Issue: VIZ-4277
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Thomas Daniel <thomas.daniel@intel.com>
Reviewed-by: Akash Goel <akash.goels@gmail.com>
Reviewed-by: Deepak S<deepak.s@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
No longer create a work item to clean each execlist queue item.
Instead, move retired execlist requests to a queue and clean up the
items during retire_requests.
v2: Fix legacy ring path broken during overzealous cleanup
v3: Update idle detection to take execlists queue into account
v4: Grab execlist lock when checking queue state
v5: Fix leaking requests by freeing in execlists_retire_requests.
Issue: VIZ-4274
Signed-off-by: Thomas Daniel <thomas.daniel@intel.com>
Reviewed-by: Deepak S <deepak.s@linux.intel.com>
Reviewed-by: Akash Goel <akash.goels@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
So with all the code movement and extraction in intel_pm.c in -next
git is hopelessly confused with
commit 2208d655a9
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Fri Nov 14 09:25:29 2014 +0100
drm/i915: drop WaSetupGtModeTdRowDispatch:snb
from -fixes. Worse even small changes in -next move around the
conflict context so rerere is equally useless. Let's just backmerge
and be done with it.
Conflicts:
drivers/gpu/drm/i915/i915_drv.c
drivers/gpu/drm/i915/intel_pm.c
Except for git getting lost no tricky conflicts really.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
After the previous patch RPS disabling doesn't depend any more on the
first level interrupts being disabled, so we can move it everywhere
earlier. Doing so let's us think about the uninitialization steps
afterwards independently of any asynchronous RPS events that can happen
atm. It also makes the system/runtime suspend time RPS disabling more
uniform. Finally this gets rid of the WARN in
intel_suspend_gt_powersave(), which we can hit if a final RPS work runs
after we disabled the first level interrupts.
Testcase: igt/pm_rpm
Reference: https://bugs.freedesktop.org/show_bug.cgi?id=82939
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When disabling the RPS interrupts there is a tricky dependency between
the thread disabling the interrupts, the RPS interrupt handler and the
corresponding RPS work. The RPS work can reenable the interrupts, so
there is no straightforward order in the disabling thread to (1) make
sure that any RPS work is flushed and to (2) disable all RPS
interrupts. Currently this is solved by masking the interrupts using two
separate mask registers (first level display IMR and PM IMR) and doing
the disabling when all first level interrupts are disabled.
This works, but the requirement to run with all first level interrupts
disabled is unnecessary making the suspend / unload time ordering of RPS
disabling wrt. other unitialization steps difficult and error prone.
Removing this restriction allows us to disable RPS early during suspend
/ unload and forget about it for the rest of the sequence. By adding a
more explicit method for avoiding the above race, it also becomes easier
to prove its correctness. Finally currently we can hit the WARN in
snb_update_pm_irq(), when a final RPS work runs with the first level
interrupts already disabled. This won't lead to any problem (due to the
separate interrupt masks), but with the change in this and the next
patch we can get rid of the WARN, while leaving it in place for other
scenarios.
To address the above points, add a new RPS interrupts_enabled flag and
use this during RPS disabling to avoid requeuing the RPS work and
reenabling of the RPS interrupts. Since the interrupt disabling happens
now in intel_suspend_gt_powersave(), we will disable RPS interrupts
explicitly during suspend (and not just through the first level mask),
but there is no problem doing so, it's also more consistent and allows
us to unify more of the RPS disabling during suspend and unload time in
the next patch.
v2/v3:
- rebase on patch "drm/i915: move rps irq disable one level up" in the
patchset
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
