The plan is to pass requests around as the basic submission tracking structure
rather than rings and contexts. This patch updates the move_to_gpu() code paths.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Updated a couple of trace points to use the now cached request pointer rather
than extracting it from the ring.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The alloc_request() function does not actually return the newly allocated
request. Instead, it must be pulled from ring->outstanding_lazy_request. This
patch fixes this so that code can create a request and start using it knowing
exactly which request it actually owns.

v2: Updated for new i915_gem_request_alloc() scheme.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Shrunk the parameter list of i915_gem_execbuffer_retire_commands() to a single
structure as everything it requires is available in the execbuff_params object.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The do_execbuf() function takes quite a few parameters. The actual set of
parameters is going to change with the conversion to passing requests around.
Further, it is due to grow massively with the arrival of the GPU scheduler.

This patch simplifies the prototype by passing a parameter structure instead.
Changing the parameter set in the future is then simply a matter of
adding/removing items to the structure.

Note that the structure does not contain absolutely everything that is passed
in. This is because the intention is to use this structure more extensively
later in this patch series and more especially in the GPU scheduler that is
coming soon. The latter requires hanging on to the structure as the final
hardware submission can be delayed until long after the execbuf IOCTL has
returned to user land. Thus it is unsafe to put anything in the structure that
is local to the IOCTL call itself - such as the 'args' parameter. All entries
must be copies of data or pointers to structures that are reference counted in
some way and guaranteed to exist for the duration of the batch buffer's life.

v2: Rebased to newer tree and updated for changes to the command parser.
Specifically, a code shuffle has required saving the batch start address in the
params structure.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

In execlist mode, the context object pointer is written in to the request
structure (and reference counted) at the point of request creation. In legacy
mode, this only happens inside i915_add_request().

This patch updates the legacy code path to match the execlist version. This
allows all the intermediate code between request creation and request submission
to get at the context object given only a request structure. Thus negating the
need to pass context pointers here, there and everywhere.

v2: Moved the context reference so it does not need to be undone if the
get_seqno() fails.

v3: Fixed execlist mode always hitting a warning about invalid last_contexts
(which don't exist in execlist mode).

v4: Updated for new i915_gem_request_alloc() scheme.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The i915_add_request() function is called to keep track of work that has been
written to the ring buffer. It adds epilogue commands to track progress (seqno
updates and such), moves the request structure onto the right list and other
such house keeping tasks. However, the work itself has already been written to
the ring and will get executed whether or not the add request call succeeds. So
no matter what goes wrong, there isn't a whole lot of point in failing the call.

At the moment, this is fine(ish). If the add request does bail early on and not
do the housekeeping, the request will still float around in the
ring->outstanding_lazy_request field and be picked up next time. It means
multiple pieces of work will be tagged as the same request and driver can't
actually wait for the first piece of work until something else has been
submitted. But it all sort of hangs together.

This patch series is all about removing the OLR and guaranteeing that each piece
of work gets its own personal request. That means that there is no more
'hoovering up of forgotten requests'. If the request does not get tracked then
it will be leaked. Thus the add request call _must_ not fail. The previous patch
should have already ensured that it _will_ not fail by removing the potential
for running out of ring space. This patch enforces the rule by actually removing
the early exit paths and the return code.

Note that if something does manage to fail and the epilogue commands don't get
written to the ring, the driver will still hang together. The request will be
added to the tracking lists. And as in the old case, any subsequent work will
generate a new seqno which will suffice for marking the old one as complete.

v2: Improved WARNings (Tomas Elf review request).

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

It is a bad idea for i915_add_request() to fail. The work will already have been
send to the ring and will be processed, but there will not be any tracking or
management of that work.

The only way the add request call can fail is if it can't write its epilogue
commands to the ring (cache flushing, seqno updates, interrupt signalling). The
reasons for that are mostly down to running out of ring buffer space and the
problems associated with trying to get some more. This patch prevents that
situation from happening in the first place.

When a request is created, it marks sufficient space as reserved for the
epilogue commands. Thus guaranteeing that by the time the epilogue is written,
there will be plenty of space for it. Note that a ring_begin() call is required
to actually reserve the space (and do any potential waiting). However, that is
not currently done at request creation time. This is because the ring_begin()
code can allocate a request. Hence calling begin() from the request allocation
code would lead to infinite recursion! Later patches in this series remove the
need for begin() to do the allocate. At that point, it becomes safe for the
allocate to call begin() and really reserve the space.

Until then, there is a potential for insufficient space to be available at the
point of calling i915_add_request(). However, that would only be in the case
where the request was created and immediately submitted without ever calling
ring_begin() and adding any work to that request. Which should never happen. And
even if it does, and if that request happens to fall down the tiny window of
opportunity for failing due to being out of ring space then does it really
matter because the request wasn't doing anything in the first place?

v2: Updated the 'reserved space too small' warning to include the offending
sizes. Added a 'cancel' operation to clean up when a request is abandoned. Added
re-initialisation of tracking state after a buffer wrap to keep the sanity
checks accurate.

v3: Incremented the reserved size to accommodate Ironlake (after finally
managing to run on an ILK system). Also fixed missing wrap code in LRC mode.

v4: Added extra comment and removed duplicate WARN (feedback from Tomas).

For: VIZ-5115
CC: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

In Indirect context w/a batch buffer,
+WaFlushCoherentL3CacheLinesAtContextSwitch:bdw

v2: Add LRI commands to set/reset bit that invalidates coherent lines,
update WA to include programming restrictions and exclude CHV as
it is not required (Ville)

v3: Avoid unnecessary read when it can be done by reading register once (Chris).

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: NRafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: NArun Siluvery <arun.siluvery@linux.intel.com>
Acked-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

In Indirect and Per context w/a batch buffer,
+WaDisableCtxRestoreArbitration

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: NRafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: NArun Siluvery <arun.siluvery@linux.intel.com>
Acked-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Some of the WA applied using WA batch buffers perform writes to scratch page.
In the current flow WA are initialized before scratch obj is allocated.
This patch reorders intel_init_pipe_control() to have a valid scratch obj
before we initialize WA.

v2: Check for valid scratch page before initializing WA as some of them
perform writes to it.

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: NMichel Thierry <michel.thierry@intel.com>
Signed-off-by: NArun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Some of the WA are to be applied during context save but before restore and
some at the end of context save/restore but before executing the instructions
in the ring, WA batch buffers are created for this purpose and these WA cannot
be applied using normal means. Each context has two registers to load the
offsets of these batch buffers. If they are non-zero, HW understands that it
need to execute these batches.

v1: In this version two separate ring_buffer objects were used to load WA
instructions for indirect and per context batch buffers and they were part
of every context.

v2: Chris suggested to include additional page in context and use it to load
these WA instead of creating separate objects. This will simplify lot of things
as we need not explicity pin/unpin them. Thomas Daniel further pointed that GuC
is planning to use a similar setup to share data between GuC and driver and
WA batch buffers can probably share that page. However after discussions with
Dave who is implementing GuC changes, he suggested to use an independent page
for the reasons - GuC area might grow and these WA are initialized only once and
are not changed afterwards so we can share them share across all contexts.

The page is updated with WA during render ring init. This has an advantage of
not adding more special cases to default_context.

We don't know upfront the number of WA we will applying using these batch buffers.
For this reason the size was fixed earlier but it is not a good idea. To fix this,
the functions that load instructions are modified to report the no of commands
inserted and the size is now calculated after the batch is updated. A macro is
introduced to add commands to these batch buffers which also checks for overflow
and returns error.
We have a full page dedicated for these WA so that should be sufficient for
good number of WA, anything more means we have major issues.
The list for Gen8 is small, same for Gen9 also, maybe few more gets added
going forward but not close to filling entire page. Chris suggested a two-pass
approach but we agreed to go with single page setup as it is a one-off routine
and simpler code wins.

One additional option is offset field which is helpful if we would like to
have multiple batches at different offsets within the page and select them
based on some criteria. This is not a requirement at this point but could
help in future (Dave).

Chris provided some helpful macros and suggestions which further simplified
the code, they will also help in reducing code duplication when WA for
other Gen are added. Add detailed comments explaining restrictions.
Use do {} while(0) for wa_ctx_emit() macro.

(Many thanks to Chris, Dave and Thomas for their reviews and inputs)

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: NRafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: NArun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

commit 53292cdb ("drm/i915: Workaround
to avoid lite restore with HEAD==TAIL") added a check for req0 != null
which is unnecessary.

The only way req0 could be null is if the list was empty, and this is
already addressed at the beginning of execlists_context_unqueue().
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMichel Thierry <michel.thierry@intel.com>
Reviewed-by: NJani Nikula <jani.nikula@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This trims a little overhead from the common case of not needing to
synchronize between rings.

v2: execlists is special and likes to duplicate code.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Currently, we only track the last request globally across all engines.
This prevents us from issuing concurrent read requests on e.g. the RCS
and BCS engines (or more likely the render and media engines). Without
semaphores, we incur costly stalls as we synchronise between rings -
greatly impacting the current performance of Broadwell versus Haswell in
certain workloads (like video decode). With the introduction of
reference counted requests, it is much easier to track the last request
per ring, as well as the last global write request so that we can
optimise inter-engine read read requests (as well as better optimise
certain CPU waits).

v2: Fix inverted readonly condition for nonblocking waits.
v3: Handle non-continguous engine array after waits
v4: Rebase, tidy, rewrite ring list debugging
v5: Use obj->active as a bitfield, it looks cool
v6: Micro-optimise, mostly involving moving code around
v7: Fix retire-requests-upto for execlists (and multiple rq->ringbuf)
v8: Rebase
v9: Refactor i915_gem_object_sync() to allow the compiler to better
optimise it.

Benchmark: igt/gem_read_read_speed
hsw:gt3e (with semaphores):
Before: Time to read-read 1024k:		275.794µs
After:  Time to read-read 1024k:		123.260µs

hsw:gt3e (w/o semaphores):
Before: Time to read-read 1024k:		230.433µs
After:  Time to read-read 1024k:		124.593µs

bdw-u (w/o semaphores):             Before          After
Time to read-read 1x1:            26.274µs       10.350µs
Time to read-read 128x128:        40.097µs       21.366µs
Time to read-read 256x256:        77.087µs       42.608µs
Time to read-read 512x512:       281.999µs      181.155µs
Time to read-read 1024x1024:    1196.141µs     1118.223µs
Time to read-read 2048x2048:    5639.072µs     5225.837µs
Time to read-read 4096x4096:   22401.662µs    21137.067µs
Time to read-read 8192x8192:   89617.735µs    85637.681µs

Testcase: igt/gem_concurrent_blit (read-read and friends)
Cc: Lionel Landwerlin <lionel.g.landwerlin@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> [v8]
[danvet: s/\<rq\>/req/g]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

If an batch ends while the IRQs are not turned on the notification can
go missing and the GPU can hang. So generate a warning in this case.
Signed-off-by: NPeter Antoine <peter.antoine@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We switched from calling i915_gem_alloc_context_obj() to calling
i915_gem_alloc_object() so the error handling needs to be updated to
check for NULL instead of IS_ERR().

Fixes: 149c86e7 ('drm/i915: Allocate context objects from stolen')
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

WaIdleLiteRestore is an execlists-only workaround, and requires the driver
to ensure that any context always has HEAD!=TAIL when attempting lite
restore.

Add two extra MI_NOOP instructions at the end of each request, but keep
the requests tail pointing before the MI_NOOPs. We may not need to
executed them, and this is why request->tail is sampled before adding
these extra instructions.

If we submit a context to the ELSP which has previously been submitted,
move the tail pointer past the MI_NOOPs. This ensures HEAD!=TAIL.

v2: Move overallocation to gen8_emit_request, and added note about
sampling request->tail in commit message (Chris).

v3: Remove redundant request->tail assignment in __i915_add_request, in
lrc mode this is already set in execlists_context_queue.
Do not add wa implementation details inside gem (Chris).

v4: Apply the wa whenever the req has been resubmitted and update
comment (Chris).

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NThomas Daniel <thomas.daniel@intel.com>
Signed-off-by: NMichel Thierry <michel.thierry@intel.com>
Signed-off-by: NJani Nikula <jani.nikula@intel.com>

On GEN9+ per specification a NULL PIPE_CONTROL needs to be emitted
before any PIPE_CONTROL command with the VS_INVALIDATE flag set.
Signed-off-by: NImre Deak <imre.deak@intel.com>
Reviewed-by: NNick Hoath <nicholas.hoath@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

After commit d7b9ca2f
("drm/i915: Remove request->uniq")

dev_priv is no longer needed.

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NMichel Thierry <michel.thierry@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

As we never expose context objects directly to userspace, we can forgo
allocating a first-class GEM object for them and prefer to use the
limited resource of reserved/stolen memory for them. Note this means
that their initial contents are undefined.

However, a downside of using stolen objects for execlists is that we
cannot access the physical address directly (thanks MCH!) which prevents
their use.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We already assign a unique identifier to every request: seqno. That
someone felt like adding a second one without even mentioning why and
tweaking ABI smells very fishy.

Fixes regression from
commit b3a38998
Author: Nick Hoath <nicholas.hoath@intel.com>
Date:   Thu Feb 19 16:30:47 2015 +0000

    drm/i915: Fix a use after free, and unbalanced refcounting

v2: Rebase
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Nick Hoath <nicholas.hoath@intel.com>
Cc: Thomas Daniel <thomas.daniel@intel.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Jani Nikula <jani.nikula@intel.com>
[danvet: Fixup because different merge order.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This eliminates six needless spin lock/unlock pairs when writing out
ELSP.

v2: Respin with my preferred colour.
v3: Mostly back to the original colour

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> [v1]
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Already tagged this one and 0-day builder is failing me.
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>

After the removal of DRI1, all access to the rings are through requests
and so we can always be sure that there is a request to wait upon to
free up available space. The fallback code only existed so that we could
quiesce the GPU following unmediated access by DRI1.

v2: Rebase
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

When we submit a request to the GPU, we first take the rpm wakelock, and
only release it once the GPU has been idle for a small period of time
after all requests have been complete. This means that we are sure no
new interrupt can arrive whilst we do not hold the rpm wakelock and so
can drop the individual get/put around every single request inside
execlists.

Note: to close one potential issue we should mark the GPU as busy
earlier in __i915_add_request.

To elaborate: The issue is that we emit the irq signalling sequence
before we grab the rpm reference, which means we could miss the
resulting interrupt (since that's not set up when suspended). The only
bad side effect is a missed interrupt, gt mmio writes automatically
wake up the hw itself. But otoh we have an umbrella rpm reference for
the entirety of execbuf, as long as that's there we're covered.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
[danvet: Explain a bit more about the add_request issue, which after
some irc chatting with Chris turns out to not be an issue really.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We can use the simpler spinlock form to disable interrupts as we are
always outside of an irq/softirq handler.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This finishes off the dynamic page tables allocations, in the legacy 3
level style that already exists. Most everything has already been setup
to this point, the patch finishes off the enabling by setting the
appropriate function pointers.

In LRC mode, contexts need to know the PDPs when they are populated. With
dynamic page table allocations, these PDPs may not exist yet. Check if
PDPs have been allocated and use the scratch page if they do not exist yet.

Before submission, update the PDPs in the logic ring context as PDPs
have been allocated.

v2: Update aliasing/true ppgtt allocate/teardown/clear functions for
gen 6 & 7.

v3: Rebase.

v4: Remove BUG() from ppgtt_unbind_vma, but keep checking that either
teardown_va_range or clear_range functions exist (Daniel).

v5: Similar to gen6, in init, gen8_ppgtt_clear_range call is only needed
for aliasing ppgtt. Zombie tracking was originally added for teardown
function and is no longer required.

v6: Update err_out case in gen8_alloc_va_range (missed from lastest
rebase).

v7: Rebase after s/page_tables/page_table/.

v8: Updated scratch_pt check after scratch flag was removed in previous
patch.

v9: Note that lrc mode needs to be updated to support init state without
any PDP.

v10: Unmap correct page_table in gen8_alloc_va_range's error case,  clean-up
gen8_aliasing_ppgtt_init (remove duplicated map), and initialize PTs
during page table allocation.

v11: Squashed LRC enabling commit, otherwise LRC mode would be left broken
until it was updated to handle the init case without any PDP.

v12: Do not overallocate new_pts bitmap, make alloc_gen8_temp_bitmaps
static and don't abuse of inline functions. (Mika)

Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v2+)
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

When we do dynamic page table allocations for gen8, we'll need to have
more control over how and when we map page tables, similar to gen6.
In particular, DMA mappings for page directories/tables occur at allocation
time.

This patch adds the functionality and calls it at init, which should
have no functional change.

The PDPEs are still a special case for now. We'll need a function for
that in the future as well.

v2: Handle renamed unmap_and_free_page functions.
v3: Updated after teardown_va logic was removed.
v4: Rebase after s/page_tables/page_table/.
v5: No longer allocate all PDPs in GEN8+ systems with less than 4GB of
memory, and update populate_lr_context to handle this new case (proper
tracking will be added later in the patch series).
v6: Assign lrc page directory pointer addresses using a macro. (Mika)

Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v2+)
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

I woke up one morning and found 50k objects sitting in the batch pool
and every search seemed to iterate the entire list... Painting the
screen in oils would provide a more fluid display.

One issue with the current design is that we only check for retirements
on the current ring when preparing to submit a new batch. This means
that we can have thousands of "active" batches on another ring that we
have to walk over. The simplest way to avoid that is to split the pools
per ring and then our LRU execution ordering will also ensure that the
inactive buffers remain at the front.

v2: execlists still requires duplicate code.
v3: execlists requires more duplicate code
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

According to Spec this is a reserved bit for Gen9+ and should not be set.

Change-Id: I0215fb7057b94139b7a2f90ecc7a0201c0c93ad4
Signed-off-by: NArun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The legacy and LRC code paths have an almost identical procedure for waiting for
space in the ring buffer. They both search for a request in the free list that
will advance the tail to a point where sufficient space is available. They then
wait for that request, retire it and recalculate the free space value.

Unfortunately, a bug in the LRC side meant that the resulting free space might
not be as large as expected and indeed, might not be sufficient. This is because
it was testing against the value of request->tail not request->postfix. Whereas,
when a request is retired, ringbuf->tail is updated to req->postfix not
req->tail.

Another significant difference between the two is that the LRC one did not trust
the wait for request to work! It redid the is there enough space available test
and would fail the call if insufficient. Whereas, the legacy version just said
'return 0' - it assumed the preceeding code works. This difference meant that
the LRC version still worked even with the bug - it just fell back to the
polling wait path.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NThomas Daniel <thomas.daniel@intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The request allocation code is largely duplicated between legacy mode and
execlist mode. The actual difference between the two versions of the code is
pretty minimal.

This patch moves the common code out into a separate function. This is then
called by the execution specific version prior to setting up the one different
value.

For: VIZ-5190
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The only usage of intel_logical_ring_begin() is within intel_lrc.c so it can be
made static. To avoid a forward declaration at the top of the file, it and bunch
of other functions have been shuffled upwards.

For: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Reviewed-by: NTomas Elf <tomas.elf@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

In execlist mode, the ringbuf is a function of the ring and context whereas in
legacy mode, it is derived from the ring alone. Thus the calculation required to
determine the ringbuf pointer from the ring (and context) also needs to test
execlist mode or not. This is messy.

Further, the request structure holds a pointer to both the ring and the context
for which it was created. Thus, given a request, it is possible to derive the
ringbuf in either legacy or execlist mode. Hence it is necessary to pass just
the request in to all the low level functions rather than some combination of
request, ring, context and ringbuf. However, rather than recalculating it each
time, it is much simpler to just cache the ringbuf pointer in the request
structure itself.

Caching the pointer means the calculation is done once at request creation time
and all further code and simply read it directly from the request structure.

OTC-Jira: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
[danvet: Drop contentless comment in lrc alloc request entirely. And
spelling fix in the commit message.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

There is a trace point in the legacy execbuffer execution path that is missing
from the execlist path. Trace points are extremely useful for debugging and are
used by various automated validation tests. Hence, this patch adds the missing
trace point back in.

OTC-Jira: VIZ-5115
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

There is a flags word that is passed through the execbuffer code path all the
way from initial decoding of the user parameters down to the very final dispatch
buffer call. It is simply called 'flags'. Unfortuantely, there are many other
flags words floating around in the same blocks of code. Even more once the GPU
scheduler arrives.

This patch makes it more obvious exactly which flags word is which by renaming
'flags' to 'dispatch_flags'. Note that the bit definitions for this flags word
already have an 'I915_DISPATCH_' prefix on them and so are not quite so
ambiguous.

OTC-Jira: VIZ-1587
Signed-off-by: NJohn Harrison <John.C.Harrison@Intel.com>
[danvet: Resolve conflict with Chris' rework of the bb parsing.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

As we move toward dynamic page table allocation, it becomes much easier
to manage our data structures if break do things less coarsely by
breaking up all of our actions into individual tasks.  This makes the
code easier to write, read, and verify.

Aside from the dissection of the allocation functions, the patch
statically allocates the page table structures without a page directory.
This remains the same for all platforms,

The patch itself should not have much functional difference. The primary
noticeable difference is the fact that page tables are no longer
allocated, but rather statically declared as part of the page directory.
This has non-zero overhead, but things gain additional complexity as a
result.

This patch exists for a few reasons:
1. Splitting out the functions allows easily combining GEN6 and GEN8
code. Page tables have no difference based on GEN8. As we'll see in a
future patch when we add the DMA mappings to the allocations, it
requires only one small change to make work, and error handling should
just fall into place.

2. Unless we always want to allocate all page tables under a given PDE,
we'll have to eventually break this up into an array of pointers (or
pointer to pointer).

3. Having the discrete functions is easier to review, and understand.
All allocations and frees now take place in just a couple of locations.
Reviewing, and catching leaks should be easy.

4. Less important: the GFP flags are confined to one location, which
makes playing around with such things trivial.

v2: Updated commit message to explain why this patch exists

v3: For lrc, s/pdp.page_directory[i].daddr/pdp.page_directory[i]->daddr/

v4: Renamed free_pt/pd_single functions to unmap_and_free_pt/pd (Daniel)

v5: Added additional safety checks in gen8 clear/free/unmap.

v6: Use WARN_ON and return -EINVAL in alloc_pt_range (Mika).

v7: Make err_out loop symmetrical to the way we allocate in
alloc_pt_range. Also s/page_tables/page_table and correct commit
message (Mika)

Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v3+)
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Move the remaining members over to the new page table structures.

This can be squashed with the previous commit if desire. The reasoning
is the same as that patch. I simply felt it is easier to review if split.

v2: In lrc: s/ppgtt->pd_dma_addr[i]/ppgtt->pdp.page_directory[i].daddr/
v3: Rebase.
v4: Rebased after s/page_tables/page_table/.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v2+)
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

When converting from implicitly tracked execlist queue items to ref counted
requests, not all frees of requests were replaced with unrefs, and extraneous
refs/unrefs of contexts were added.
Correct the unbalanced refcount & replace the frees.
Remove a noisy warning when hitting the request creation path.

drm_i915_gem_request and intel_context are both kref reference counted
structures. Upon allocation, drm_i915_gem_request's ref count should be
bumped using kref_init. When a context is assigned to the request,
the context's reference count should be bumped using i915_gem_context_reference.
i915_gem_request_reference will reduce the context reference count when
the request is freed.

Problem introduced in
commit 6d3d8274
Author:     Nick Hoath <nicholas.hoath@intel.com>
AuthorDate: Thu Jan 15 13:10:39 2015 +0000

     drm/i915: Subsume intel_ctx_submit_request in to drm_i915_gem_request

v2: Added comments explaining how the ctx pointer and the request object should
be ref-counted. Removed noisy warning.

v3: Cleaned up the language used in the commit & the header
description (Thanks David Gordon)

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88652Signed-off-by: NNick Hoath <nicholas.hoath@intel.com>
Reviewed-by: NThomas Daniel <thomas.daniel@intel.com>
Reviewed-by: NDaniel Vetter <daniel@ffwll.ch>
Signed-off-by: NJani Nikula <jani.nikula@intel.com>