By using drm_gem_flink/drm_gem_open on an object using the same fd, it
is possible for a client to create multiple handles pointing to the same
object (tied to the same contexts and VMA), as exemplified by
igt::gem_handle_to_libdrm_bo(). Since this duplication has been possible
since forever, we cannot assume that the handle:(fpriv, object) is
unique and so must handle the multiple users of a single VMA.
v2: Added commentary noise.
Testcase: igt/gem_close
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102355
Fixes: d1b48c1e71 ("drm/i915: Replace execbuf vma ht with an idr")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170822110517.22277-3-chris@chris-wilson.co.uk
Tested-by: Marta Lofstedt <marta.lofstedt@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
(cherry-picked from commit 3ffff01749)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Sometimes we know we are the only user of the bo, but since we take a
protective pin_pages early on, an attempt to change the vmap on the
object is denied because it is busy. i915_gem_object_pin_map() cannot
tell from our single pin_count if the operation is safe. Instead we must
pass that information down from the caller in the manner of
I915_MAP_OVERRIDE.
This issue has existed from the introduction of the mapping, but was
never noticed as the only place where this conflict might happen is for
cached kernel buffers (such as allocated by i915_gem_batch_pool_get()).
Until recently there was only a single user (the cmdparser) so no
conflicts ever occurred. However, we now use it to allocate batches for
different operations (using MAP_WC on !llc for writes) in addition to the
existing shadow batch (using MAP_WB for reads).
We could either keep both mappings cached, or use a different write
mechanism if we detect a MAP_WB already exists (i.e. clflush
afterwards), but as we haven't seen this issue in the wild (it requires
hitting the GPU reloc path in addition to the cmdparser) for simplicity
just allow the mappings to be recreated.
v2: Include the i915_MAP_OVERRIDE bit in the enum so the compiler knows
about all the valid values.
Fixes: 7dd4f6729f ("drm/i915: Async GPU relocation processing")
Testcase: igt/gem_lut_handle # byt, completely by accident
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170828104631.8606-1-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
(cherry picked from commit a575c67617)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
If we are using the cmdparser, we will have to copy the batch and so
stall for the relocations. Rather than prolong that stall by adding more
relocation requests, just use CPU relocations and do the stall upfront.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170826135620.25949-1-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Sometimes we know we are the only user of the bo, but since we take a
protective pin_pages early on, an attempt to change the vmap on the
object is denied because it is busy. i915_gem_object_pin_map() cannot
tell from our single pin_count if the operation is safe. Instead we must
pass that information down from the caller in the manner of
I915_MAP_OVERRIDE.
This issue has existed from the introduction of the mapping, but was
never noticed as the only place where this conflict might happen is for
cached kernel buffers (such as allocated by i915_gem_batch_pool_get()).
Until recently there was only a single user (the cmdparser) so no
conflicts ever occurred. However, we now use it to allocate batches for
different operations (using MAP_WC on !llc for writes) in addition to the
existing shadow batch (using MAP_WB for reads).
We could either keep both mappings cached, or use a different write
mechanism if we detect a MAP_WB already exists (i.e. clflush
afterwards), but as we haven't seen this issue in the wild (it requires
hitting the GPU reloc path in addition to the cmdparser) for simplicity
just allow the mappings to be recreated.
v2: Include the i915_MAP_OVERRIDE bit in the enum so the compiler knows
about all the valid values.
Fixes: 7dd4f6729f ("drm/i915: Async GPU relocation processing")
Testcase: igt/gem_lut_handle # byt, completely by accident
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170828104631.8606-1-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
By using drm_gem_flink/drm_gem_open on an object using the same fd, it
is possible for a client to create multiple handles pointing to the same
object (tied to the same contexts and VMA), as exemplified by
igt::gem_handle_to_libdrm_bo(). Since this duplication has been possible
since forever, we cannot assume that the handle:(fpriv, object) is
unique and so must handle the multiple users of a single VMA.
v2: Added commentary noise.
Testcase: igt/gem_close
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102355
Fixes: d1b48c1e71 ("drm/i915: Replace execbuf vma ht with an idr")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170822110517.22277-3-chris@chris-wilson.co.uk
Tested-by: Marta Lofstedt <marta.lofstedt@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
This was the competing idea long ago, but it was only with the rewrite
of the idr as an radixtree and using the radixtree directly ourselves,
along with the realisation that we can store the vma directly in the
radixtree and only need a list for the reverse mapping, that made the
patch performant enough to displace using a hashtable. Though the vma ht
is fast and doesn't require any extra allocation (as we can embed the node
inside the vma), it does require a thread for resizing and serialization
and will have the occasional slow lookup. That is hairy enough to
investigate alternatives and favour them if equivalent in peak performance.
One advantage of allocating an indirection entry is that we can support a
single shared bo between many clients, something that was done on a
first-come first-serve basis for shared GGTT vma previously. To offset
the extra allocations, we create yet another kmem_cache for them.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-5-chris@chris-wilson.co.uk
Since the introduction of being able to perform a lockless lookup of an
object (i915_gem_object_get_rcu() in fbbd37b36f ("drm/i915: Move object
release to a freelist + worker") we no longer need to split the
object/vma lookup into 3 phases and so combine them into a much simpler
single loop.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-4-chris@chris-wilson.co.uk
When userspace is doing most of the work, avoiding relocs (using
NO_RELOC) and opting out of implicit synchronisation (using ASYNC), we
still spend a lot of time processing the arrays in execbuf, even though
we now should have nothing to do most of the time. One issue that
becomes readily apparent in profiling anv is that iterating over the
large execobj[] is unfriendly to the loop prefetchers of the CPU and it
much prefers iterating over a pair of arrays rather than one big array.
v2: Clear vma[] on construction to handle errors during vma lookup
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-3-chris@chris-wilson.co.uk
Since we keep the context around across the slow lookup where we may
drop the struct_mutex, we should double check that the context is still
valid upon reacquisition.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-2-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
MI_STORE_DWORD_IMM just doesn't work on the video decode engine under
Sandybridge, so refrain from using it. Then switch the selftests over to
using the now common test prior to using MI_STORE_DWORD_IMM.
Fixes: 7dd4f6729f ("drm/i915: Async GPU relocation processing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: <drm-intel-fixes@lists.freedesktop.org> # v4.13-rc1+
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
This commit adds support for waiting on or signaling DRM syncobjs as
part of execbuf. It does so by hijacking the currently unused cliprects
pointer to instead point to an array of i915_gem_exec_fence structs
which containe a DRM syncobj and a flags parameter which specifies
whether to wait on it or to signal it. This implementation
theoretically allows for both flags to be set in which case it waits on
the dma_fence that was in the syncobj and then immediately replaces it
with the dma_fence from the current execbuf.
v2:
- Rebase on new syncobj API
v3:
- Pull everything out into helpers
- Do all allocation in gem_execbuffer2
- Pack the flags in the bottom 2 bits of the drm_syncobj*
v4:
- Prevent a potential race on syncobj->fence
Testcase: igt/gem_exec_fence/syncobj*
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Link: https://patchwork.freedesktop.org/patch/msgid/1499289202-25441-1-git-send-email-jason.ekstrand@intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20170815145733.4562-1-chris@chris-wilson.co.uk
Another month, another story in the cache coherency saga. This time, we
come to the realisation that i915_gem_object_is_coherent() has been
reporting whether we can read from the target without requiring a cache
invalidate; but we were using it in places for testing whether we could
write into the object without requiring a cache flush. So split the
tracking into two, one to decide before reads, one after writes.
See commit e27ab73d17 ("drm/i915: Mark CPU cache as dirty on every
transition for CPU writes") for the previous entry in this saga.
v2: Be verbose
v3: Remove unused function (i915_gem_object_is_coherent)
v4: Fix inverted coherency check prior to execbuf (from v2)
v5: Add comment for nasty code where we are optimising on gcc's behalf.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101109
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101555
Testcase: igt/kms_mmap_write_crc
Testcase: igt/kms_pwrite_crc
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Tested-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170811111116.10373-1-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
The goal here was to minimise doing any thing or any check inside the
kernel that was not strictly required. For a userspace that assumes
complete control over the cache domains, the kernel is usually using
outdated information and may trigger clflushes where none were
required.
However, swapping is a situation where userspace has no knowledge of the
domain transfer, and will leave the object in the CPU cache. The kernel
must flush this out to the backing storage prior to use with the GPU. As
we use an asynchronous task tracked by an implicit fence for this, we
also need to cancel the ASYNC flag on the object so that the object will
wait for the clflush to complete before being executed. This also absolves
userspace of the responsibility imposed by commit 77ae995789 ("drm/i915:
Enable userspace to opt-out of implicit fencing") that its needed to ensure
that the object was out of the CPU cache prior to use on the GPU.
Fixes: 77ae995789 ("drm/i915: Enable userspace to opt-out of implicit fencing")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101571
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20170721145037.25105-5-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
I was being overly paranoid in not updating the execobject.offset after
performing the fallback copy where we set reloc.presumed_offset to -1.
The thinking was to ensure that a subsequent NORELOC execbuf would be
forced to process the invalid relocations. However this is overkill so
long as we *only* update the execobject.offset following a successful
update of the relocation value witin the batch. If we have to repeat the
execbuf due to a later interruption, then we may skip the relocations on
the second pass (honouring NORELOC) since the execobject.offset match
the actual offsets (even though reloc.presumed_offset is garbage).
Subsequent calls to execbuf with NORELOC should themselves ensure that
the reloc.presumed_offset have been corrected in case of future
migration.
Reporting back the actual execobject.offset, even when
reloc.presumed_offset is garbage, ensures that reuse of those objects
use the latest information to avoid relocations.
Fixes: 2889caa923 ("drm/i915: Eliminate lots of iterations over the execobjects array")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101635
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170721145037.25105-4-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Fix the sizeof(ptr) vs. sizeof(*ptr) typo.
Fixes: 2889caa923 ("drm/i915: Eliminate lots of iterations over the execobjects array")
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170714151242.517-2-imre.deak@intel.com
When computing a hash for looking up relocation target handles in an
execbuf, we start with a large size for the hashtable and proceed to
halve it until the allocation succeeds. The final attempt is with an
order of 0 (i.e. a single element). This means that we then pass bits=0
to hash_32() which then computes "hash >> (32 - 0)" to lookup the single
element. Right shifting a value by the width of the operand is
undefined, so limit the smallest hash table we use to order 1.
v2: Keep the retry allocation flag for the final pass
Fixes: 4ff4b44cbb ("drm/i915: Store a direct lookup from object handle to vma")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170629150425.27508-1-chris@chris-wilson.co.uk
commit 2889caa923 ("drm/i915: Eliminate lots of iterations over the
execobjects array") jiggled around the error handling and replace a test
that we cleaned up properly after ourselves with an assertion. That
assertion failed because in the release function (moments after the
assertion) we were indeed forgetting to mark the vma as cleared. The
consequence was when testing an invalid relocation address, we would try
to release the vma twice (following the couple of attempts to verify the
address) and on the second release notice that the first release was
incomplete.
Testcase: igt/gem_reloc_overflow/invalid-address
Fixes: 2889caa923 ("drm/i915: Eliminate lots of iterations over the execobjects array")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170622104722.2583-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Whilst the contents of the context is still protected by the big
struct_mutex, this is not much of an improvement. It is just one tiny
step towards reducing our BKL.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620110547.15947-3-chris@chris-wilson.co.uk
During execbuf, a mandatory step is that we add this request (this
fence) to each object's reservation_object. Inside execbuf, we track the
vma, and to add the fence to the reservation_object then means having to
first chase the obj, incurring another cache miss. We can reduce the
number of cache misses by stashing a pointer to the reservation_object
in the vma itself.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170616140525.6394-1-chris@chris-wilson.co.uk
If the user requires patching of their batch or auxiliary buffers, we
currently make the alterations on the cpu. If they are active on the GPU
at the time, we wait under the struct_mutex for them to finish executing
before we rewrite the contents. This happens if shared relocation trees
are used between different contexts with separate address space (and the
buffers then have different addresses in each), the 3D state will need
to be adjusted between execution on each context. However, we don't need
to use the CPU to do the relocation patching, as we could queue commands
to the GPU to perform it and use fences to serialise the operation with
the current activity and future - so the operation on the GPU appears
just as atomic as performing it immediately. Performing the relocation
rewrites on the GPU is not free, in terms of pure throughput, the number
of relocations/s is about halved - but more importantly so is the time
under the struct_mutex.
v2: Break out the request/batch allocation for clearer error flow.
v3: A few asserts to ensure rq ordering is maintained
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Currently, the last object in the execlist is the always the batch.
However, when building the batch buffer we often know the batch object
first and if we can use the first slot in the execlist we can emit
relocation instructions relative to it immediately and avoid a separate
pass to adjust the relocations to point to the last execlist slot.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
This simply hides the EAGAIN caused by userptr when userspace causes
resource contention. However, it is quite beneficial with highly
contended userptr users as we avoid repeating the setup costs and
kernel-user context switches.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
When choosing a slot for an execbuffer, we ideally want to use the same
address as last time (so that we don't have to rebind it) and the same
address as expected by the user (so that we don't have to fixup any
relocations pointing to it). If we first try to bind the incoming
execbuffer->offset from the user, or the currently bound offset that
should hopefully achieve the goal of avoiding the rebind cost and the
relocation penalty. However, if the object is not currently bound there
we don't want to arbitrarily unbind an object in our chosen position and
so choose to rebind/relocate the incoming object instead. After we
report the new position back to the user, on the next pass the
relocations should have settled down.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtien@linux.intel.com>
If we take a reference to the object/vma when it is first used in an
execbuf, we can keep that reference until the object's file-local handle
is closed. Thereby saving a frequent ref/unref pair.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
The major scaling bottleneck in execbuffer is the processing of the
execobjects. Creating an auxiliary list is inefficient when compared to
using the execobject array we already have allocated.
Reservation is then split into phases. As we lookup up the VMA, we
try and bind it back into active location. Only if that fails, do we add
it to the unbound list for phase 2. In phase 2, we try and add all those
objects that could not fit into their previous location, with fallback
to retrying all objects and evicting the VM in case of severe
fragmentation. (This is the same as before, except that phase 1 is now
done inline with looking up the VMA to avoid an iteration over the
execobject array. In the ideal case, we eliminate the separate reservation
phase). During the reservation phase, we only evict from the VM between
passes (rather than currently as we try to fit every new VMA). In
testing with Unreal Engine's Atlantis demo which stresses the eviction
logic on gen7 class hardware, this speed up the framerate by a factor of
2.
The second loop amalgamation is between move_to_gpu and move_to_active.
As we always submit the request, even if incomplete, we can use the
current request to track active VMA as we perform the flushes and
synchronisation required.
The next big advancement is to avoid copying back to the user any
execobjects and relocations that are not changed.
v2: Add a Theory of Operation spiel.
v3: Fall back to slow relocations in preparation for flushing userptrs.
v4: Document struct members, factor out eb_validate_vma(), add a few
more comments to explain some magic and hide other magic behind macros.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
If we write a relocation into the buffer, we require our own implicit
synchronisation added after the start of the execbuf, outside of the
user's control. As we may end up clflushing, or doing the patch itself
on the GPU, asynchronously we need to look at the implicit serialisation
on obj->resv and hence need to disable EXEC_OBJECT_ASYNC for this
object.
If the user does trigger a stall for relocations, we make sure the stall
is complete enough so that the batch is not submitted before we complete
those relocations.
Fixes: 77ae995789 ("drm/i915: Enable userspace to opt-out of implicit fencing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
We can simplify our tracking of pending writes in an execbuf to the
single bit in the vma->exec_entry->flags, but that requires the
relocation function knowing the object's vma. Pass it along.
Note we have only been using a single bit to track flushing since
commit cc889e0f6c
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Wed Jun 13 20:45:19 2012 +0200
drm/i915: disable flushing_list/gpu_write_list
unconditionally flushed all render caches before the breadcrumb and
commit 6ac42f4148
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Sat Jul 21 12:25:01 2012 +0200
drm/i915: Replace the complex flushing logic with simple invalidate/flush all
did away with the explicit GPU domain tracking. This was then codified
into the ABI with NO_RELOC in
commit ed5982e6ce
Author: Daniel Vetter <daniel.vetter@ffwll.ch> # Oi! Patch stealer!
Date: Thu Jan 17 22:23:36 2013 +0100
drm/i915: Allow userspace to hint that the relocations were known
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
The advent of full-ppgtt lead to an extra indirection between the object
and its binding. That extra indirection has a noticeable impact on how
fast we can convert from the user handles to our internal vma for
execbuffer. In order to bypass the extra indirection, we use a
resizable hashtable to jump from the object to the per-ctx vma.
rhashtable was considered but we don't need the online resizing feature
and the extra complexity proved to undermine its usefulness. Instead, we
simply reallocate the hastable on demand in a background task and
serialize it before iterating.
In non-full-ppgtt modes, multiple files and multiple contexts can share
the same vma. This leads to having multiple possible handle->vma links,
so we only use the first to establish the fast path. The majority of
buffers are not shared and so we should still be able to realise
speedups with multiple clients.
v2: Prettier names, more magic.
v3: Many style tweaks, most notably hiding the misuse of execobj[].rsvd2
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
For ease of use (i.e. avoiding a few checks and function calls), store
the object's cache coherency next to the cache is dirty bit.
Specifically this patch aims to reduce the frequency of no-op calls to
i915_gem_object_clflush() to counter-act the increase of such calls for
GPU only objects in the previous patch.
v2: Replace cache_dirty & ~cache_coherent with cache_dirty &&
!cache_coherent as gcc generates much better code for the latter
(Tvrtko)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Tested-by: Dongwon Kim <dongwon.kim@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170616105455.16977-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Currently, we only mark the CPU cache as dirty if we skip a clflush.
This leads to some confusion where we have to ask if the object is in
the write domain or missed a clflush. If we always mark the cache as
dirty, this becomes a much simply question to answer.
The goal remains to do as few clflushes as required and to do them as
late as possible, in the hope of deferring the work to a kthread and not
block the caller (e.g. execbuf, flips).
v2: Always call clflush before GPU execution when the cache_dirty flag
is set. This may cause some extra work on llc systems that migrate dirty
buffers back and forth - but we do try to limit that by only setting
cache_dirty at the end of the gpu sequence.
v3: Always mark the cache as dirty upon a level change, as we need to
invalidate any stale cachelines due to external writes.
Reported-by: Dongwon Kim <dongwon.kim@intel.com>
Fixes: a6a7cc4b7d ("drm/i915: Always flush the dirty CPU cache when pinning the scanout")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Tested-by: Dongwon Kim <dongwon.kim@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170615123850.26843-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Currently the vma has one link member that is used for both holding its
place in the execbuf reservation list, and in any eviction list. This
dual property is quite tricky and error prone.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170615081435.17699-3-chris@chris-wilson.co.uk
This has the benefit of not requiring us to manipulate the
vma->exec_link list when tearing down the execbuffer, and is a
marginally cheaper test to detect the user error.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170615081435.17699-2-chris@chris-wilson.co.uk
More stuff for 4.13:
- skl+ wm fixes from Mahesh Kumar
- some refactor and tests for i915_sw_fence (Chris)
- tune execlist/scheduler code (Chris)
- g4x,g33 gpu reset improvements (Chris, Mika)
- guc code cleanup (Michal Wajdeczko, Michał Winiarski)
- dp aux backlight improvements (Puthikorn Voravootivat)
- buffer based guc/host communication (Michal Wajdeczko)
* tag 'drm-intel-next-2017-05-29' of git://anongit.freedesktop.org/git/drm-intel: (253 commits)
drm/i915: Update DRIVER_DATE to 20170529
drm/i915: Keep the forcewake timer alive for 1ms past the most recent use
drm/i915/guc: capture GuC logs if FW fails to load
drm/i915/guc: Introduce buffer based cmd transport
drm/i915/guc: Disable send function on fini
drm: Add definition for eDP backlight frequency
drm/i915: Drop AUX backlight enable check for backlight control
drm/i915: Consolidate #ifdef CONFIG_INTEL_IOMMU
drm/i915: Only GGTT vma may be pinned and prevent shrinking
drm/i915: Serialize GTT/Aperture accesses on BXT
drm/i915: Convert i915_gem_object_ops->flags values to use BIT()
drm/i915/selftests: Silence compiler warning in igt_ctx_exec
drm/i915/guc: Skip port assign on first iteration of GuC dequeue
drm/i915: Remove misleading comment in request_alloc
drm/i915/g33: Improve reset reliability
Revert "drm/i915: Restore lost "Initialized i915" welcome message"
drm/i915/huc: Update GLK HuC version
drm/i915: Check for allocation failure
drm/i915/guc: Remove action status and statistics from debugfs
drm/i915/g4x: Improve gpu reset reliability
...
Now that drm_[cm]alloc* helpers are simple one line wrappers around
kvmalloc_array and drm_free_large is just kvfree alias we can drop
them and replace by their native forms.
This shouldn't introduce any functional change.
Changes since v1
- fix typo in drivers/gpu//drm/etnaviv/etnaviv_gem.c - noticed by 0day
build robot
Suggested-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Michal Hocko <mhocko@suse.com>drm: drop drm_[cm]alloc* helpers
[danvet: Fixup vgem which grew another user very recently.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: Christian König <christian.koenig@amd.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170517122312.GK18247@dhcp22.suse.cz
Introduce a new execobject.flag (EXEC_OBJECT_CAPTURE) that userspace may
use to indicate that it wants the contents of this buffer preserved in
the error state (/sys/class/drm/cardN/error) following a GPU hang
involving this batch.
Use this at your discretion, the contents of the error state. although
compressed, are allocated with GFP_ATOMIC (i.e. limited) and kept for all
eternity (until the error state is destroyed).
Based on an earlier patch by Ben Widawsky <ben@bwidawsk.net>
Testcase: igt/gem_exec_capture
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <ben@bwidawsk.net>
Cc: Matt Turner <mattst88@gmail.com>
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170415093902.22581-1-chris@chris-wilson.co.uk
Old devices have quite severe restrictions for using fences, and unlike
more recent device (anything from Pineview onwards) we need to enforce
those restrictions even for unfenced tiled access from the render
pipeline.
Fixes: 944397f04f ("drm/i915: Store required fence size/alignment for GGTT vma")
Reported-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: <drm-intel-fixes@lists.freedesktop.org> # v4.11-rc1+
Link: http://patchwork.freedesktop.org/patch/msgid/20170325113243.16438-1-chris@chris-wilson.co.uk
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Tested-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Adding to the tail of the client request list as the only other user is
in the throttle ioctl that iterates forwards over the list. It only
needs protection against deletion of a request as it reads it, it simply
won't see a new request added to the end of the list, or it would be too
early and rejected. We can further reduce the number of spinlocks
required when throttling by removing stale requests from the client_list
as we throttle.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170302122525.19675-1-chris@chris-wilson.co.uk
This patch makes the I915_PARAM_HAS_EXEC_CONSTANTS getparam return 0
(indicating the optional feature is not supported), and makes execbuf
always return -EINVAL if the flags are used.
Apparently, no userspace ever shipped which used this optional feature:
I checked the git history of Mesa, xf86-video-intel, libva, and Beignet,
and there were zero commits showing a use of these flags. Kernel commit
72bfa19c8d apparently introduced the feature prematurely. According
to Chris, the intention was to use this in cairo-drm, but "the use was
broken for gen6", so I don't think it ever happened.
'relative_constants_mode' has always been tracked per-device, but this
has actually been wrong ever since hardware contexts were introduced, as
the INSTPM register is saved (and automatically restored) as part of the
render ring context. The software per-device value could therefore get
out of sync with the hardware per-context value. This meant that using
them is actually unsafe: a client which tried to use them could damage
the state of other clients, causing the GPU to interpret their BO
offsets as absolute pointers, leading to bogus memory reads.
These flags were also never ported to execlist mode, making them no-ops
on Gen9+ (which requires execlists), and Gen8 in the default mode.
On Gen8+, userspace can write these registers directly, achieving the
same effect. On Gen6-7.5, it likely makes sense to extend the command
parser to support them. I don't think anyone wants this on Gen4-5.
Based on a patch by Dave Gordon.
v3: Return -ENODEV for the getparam, as this is what we do for other
obsolete features. Suggested by Chris Wilson.
Cc: stable@vger.kernel.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92448
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170215093446.21291-1-kenneth@whitecape.org
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Flushing the cachelines for an object is slow, can be as much as 100ms
for a large framebuffer. We currently do this under the struct_mutex BKL
on execution or on pageflip. But now with the ability to add fences to
obj->resv for both flips and execbuf (and we naturally wait on the fence
before CPU access), we can move the clflush operation to a workqueue and
signal a fence for completion, thereby doing the work asynchronously and
not blocking the driver or its clients.
v2: Introduce i915_gem_clflush.h and use a new name, split out some
extras into separate patches.
Suggested-by: Akash Goel <akash.goel@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170222114049.28456-5-chris@chris-wilson.co.uk
The change_domain tracepoint has been inaccurate for a few years - it
doesn't fully capture the domains, especially with userspace bypassing
them. It is defunct, misleading and time to be removed.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170222114049.28456-1-chris@chris-wilson.co.uk
Rename it to i915_gem_request_queue and fix the logged info
equivalent to the i915_gem_request even class. Also moved it
a bit further apart from the i915_gem_request_add tracepoint
since they otherwise provide similar information too close in
time.
v2: Remove sw fence singalling. We will rely on the soon to
come GuC scheduling backend to enable that. (Chris Wilson)
v3: Log hex with 0x prefix for clarity.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
This removes the usage of intel_ring_emit in favour of
directly writing to the ring buffer.
intel_ring_emit was preventing the compiler for optimising
fetch and increment of the current ring buffer pointer and
therefore generating very verbose code for every write.
It had no useful purpose since all ringbuffer operations
are started and ended with intel_ring_begin and
intel_ring_advance respectively, with no bail out in the
middle possible, so it is fine to increment the tail in
intel_ring_begin and let the code manage the pointer
itself.
Useless instruction removal amounts to approximately
two and half kilobytes of saved text on my build.
Not sure if this has any measurable performance
implications but executing a ton of useless instructions
on fast paths cannot be good.
v2:
* Change return from intel_ring_begin to error pointer by
popular demand.
* Move tail increment to intel_ring_advance to enable some
error checking.
v3:
* Move tail advance back into intel_ring_begin.
* Rebase and tidy.
v4:
* Complete rebase after a few months since v3.
v5:
* Remove unecessary cast and fix !debug compile. (Chris Wilson)
v6:
* Make intel_ring_offset take request as well.
* Fix recording of request postfix plus a sprinkle of asserts.
(Chris Wilson)
v7:
* Use intel_ring_offset to get the postfix. (Chris Wilson)
* Convert GVT code as well.
v8:
* Rename *out++ to *cs++.
v9:
* Fix GVT out to cs conversion in GVT.
v10:
* Rebase for new intel_ring_begin in selftests.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Zhi Wang <zhi.a.wang@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170214113242.29241-1-tvrtko.ursulin@linux.intel.com
Chris Wilson needs the new drm_driver->release callback to make sure
the shiny new dma-buf testcases don't oops the driver on unload.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
We're using non-canonical addresses in drm_mm, and we're making sure that
userspace is using canonical addressing - both in case of softpin
(verifying incoming offset) and when relocating (converting to canonical
when updating offset returned to userspace).
Unfortunately when considering the need for relocations, we're comparing
offset from userspace (in canonical form) with drm_mm node (in
non-canonical form), and as a result, we end up always relocating if our
offsets are in the "problematic" range.
Let's always convert the offsets to avoid the performance impact of
relocations.
Fixes: a5f0edf63b ("drm/i915: Avoid writing relocs with addresses in non-canonical form")
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Reported-by: Michał Pyrzowski <michal.pyrzowski@intel.com>
Signed-off-by: Michał Winiarski <michal.winiarski@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170207195559.18798-1-michal.winiarski@intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Fences are creted/checked before the pm ref is taken, so if we jump to
pre_mutex_err we will uncorrectly call intel_runtime_pm_put.
v2: Massage unwind error paths
Fixes: fec0445caa (drm/i915: Support explicit fencing for execbuf)
Testcase: igt/gem_exec_params
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1486161930-11764-1-git-send-email-daniele.ceraolospurio@intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The drm_mm range manager claimed to support top-down insertion, but it
was neither searching for the top-most hole that could fit the
allocation request nor fitting the request to the hole correctly.
In order to search the range efficiently, we create a secondary index
for the holes using either their size or their address. This index
allows us to find the smallest hole or the hole at the bottom or top of
the range efficiently, whilst keeping the hole stack to rapidly service
evictions.
v2: Search for holes both high and low. Rename flags to mode.
v3: Discover rb_entry_safe() and use it!
v4: Kerneldoc for enum drm_mm_insert_mode.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: David Airlie <airlied@linux.ie>
Cc: Russell King <rmk+kernel@armlinux.org.uk>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Sean Paul <seanpaul@chromium.org>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Stephen Warren <swarren@wwwdotorg.org>
Cc: Alexandre Courbot <gnurou@gmail.com>
Cc: Eric Anholt <eric@anholt.net>
Cc: Sinclair Yeh <syeh@vmware.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com> # vmwgfx
Reviewed-by: Lucas Stach <l.stach@pengutronix.de> #etnaviv
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/20170202210438.28702-1-chris@chris-wilson.co.uk
Now that the user can opt-out of implicit fencing, we need to give them
back control over the fencing. We employ sync_file to wrap our
drm_i915_gem_request and provide an fd that userspace can merge with
other sync_file fds and pass back to the kernel to wait upon before
future execution.
Testcase: igt/gem_exec_fence
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: Chad Versace <chadversary@chromium.org>
Link: http://patchwork.freedesktop.org/patch/msgid/20170127094008.27489-2-chris@chris-wilson.co.uk
Userspace is faced with a dilemma. The kernel requires implicit fencing
to manage resource usage (we always must wait for the GPU to finish
before releasing its PTE) and for third parties. However, userspace may
wish to avoid this serialisation if it is either using explicit fencing
between parties and wants more fine-grained access to buffers (e.g. it
may partition the buffer between uses and track fences on ranges rather
than the implicit fences tracking the whole object). It follows that
userspace needs a mechanism to avoid the kernel's serialisation on its
implicit fences before execbuf execution.
The next question is whether this is an object, execbuf or context flag.
Hybrid users (such as using explicit EGL_ANDROID_native_sync fencing on
shared winsys buffers, but implicit fencing on internal surfaces)
require a per-object level flag. Given that this flag need to be only
set once for the lifetime of the object, this reduces the convenience of
having an execbuf or context level flag (and avoids having multiple
pieces of uABI controlling the same feature).
Incorrect use of this flag will result in rendering corruption and GPU
hangs - but will not result in use-after-free or similar resource
tracking issues.
Serious caveat: write ordering is not strictly correct after setting
this flag on a render target on multiple engines. This affects all
subsequent GEM operations (execbuf, set-domain, pread) and shared
dma-buf operations. A fix is possible - but costly (both in terms of
further ABI changes and runtime overhead).
Testcase: igt/gem_exec_async
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: Chad Versace <chadversary@chromium.org>
Link: http://patchwork.freedesktop.org/patch/msgid/20170127094008.27489-1-chris@chris-wilson.co.uk
Whilst writing testcases to exercise the VMA API, some oddities came to
light, such as i915_gem_obj_lookup_or_create(). Joonas suggested
i915_vma_instance() as a neat replacement, so rename them, move them to
i915_vma.c and add some kerneldoc as a sugary bonus.
s/i915_gem_obj_to_vma/i915_vma_lookup/
s/i915_gem_obj_lookup_or_create_vma/i915_vma_instance/
Suggested-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170116152131.18089-2-chris@chris-wilson.co.uk
Start converting over from the byte count to its semantic macro, either
we want to allocate the size of a physical page in main memory or we
want the size of a virtual page in the GTT. 4096 could mean either, but
PAGE_SIZE and I915_GTT_PAGE_SIZE are explicit and should help improve
code comprehension and future changes. In the future, we may want to use
variable GTT page sizes and so have the challenge of knowing which
hardcoded values were used to represent a physical page vs the virtual
page.
v2: Look for a few more 4096s to convert, discover IS_ALIGNED().
v3: 4096ul paranoia, make fence alignment a distinct value of 4096, keep
bdw stolen w/a as 4096 until we know better.
v4: Add asserts that i915_vma_insert() start/end are aligned to GTT page
sizes.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170110144734.26052-1-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Soft-pinning depends upon being able to check for availabilty of an
interval and evict overlapping object from a drm_mm range manager very
quickly. Currently it uses a linear list, and so performance is dire and
not suitable as a general replacement. Worse, the current code will oops
if it tries to evict an active buffer.
It also helps if the routine reports the correct error codes as expected
by its callers and emits a tracepoint upon use.
For posterity since the wrong patch was pushed (i.e. that missed these
key points and had known bugs), this is the changelog that should have
been on commit 506a8e87d8 ("drm/i915: Add soft-pinning API for
execbuffer"):
Userspace can pass in an offset that it presumes the object is located
at. The kernel will then do its utmost to fit the object into that
location. The assumption is that userspace is handling its own object
locations (for example along with full-ppgtt) and that the kernel will
rarely have to make space for the user's requests.
This extends the DRM_IOCTL_I915_GEM_EXECBUFFER2 to do the following:
* if the user supplies a virtual address via the execobject->offset
*and* sets the EXEC_OBJECT_PINNED flag in execobject->flags, then
that object is placed at that offset in the address space selected
by the context specifier in execbuffer.
* the location must be aligned to the GTT page size, 4096 bytes
* as the object is placed exactly as specified, it may be used by this
execbuffer call without relocations pointing to it
It may fail to do so if:
* EINVAL is returned if the object does not have a 4096 byte aligned
address
* the object conflicts with another pinned object (either pinned by
hardware in that address space, e.g. scanouts in the aliasing ppgtt)
or within the same batch.
EBUSY is returned if the location is pinned by hardware
EINVAL is returned if the location is already in use by the batch
* EINVAL is returned if the object conflicts with its own alignment (as meets
the hardware requirements) or if the placement of the object does not fit
within the address space
All other execbuffer errors apply.
Presence of this execbuf extension may be queried by passing
I915_PARAM_HAS_EXEC_SOFTPIN to DRM_IOCTL_I915_GETPARAM and checking for
a reported value of 1 (or greater).
v2: Combine the hole/adjusted-hole ENOSPC checks
v3: More color, more splitting, more blurb.
Fixes: 506a8e87d8 ("drm/i915: Add soft-pinning API for execbuffer")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161205142941.21965-2-chris@chris-wilson.co.uk
We need to distinguish between full i915_vma structs and simple
drm_mm_nodes when considering eviction (i.e. we must be careful not to
treat a mere drm_mm_node as a much larger i915_vma causing memory
corruption, if we are lucky). To do this, color these not-a-vma with -1
(I915_COLOR_UNEVICTABLE).
v2...v200: New name for -1.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161205142941.21965-1-chris@chris-wilson.co.uk
As i915.enable_cmd_parser is an unsafe option, make it read-only at
runtime. Now that it is constant, we can use the value determined during
initialisation as to whether we need the cmdparser at execbuffer time.
v2: Remove the inline for its single user, it is clear enough (and
shorter) without!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161124125851.6615-1-chris@chris-wilson.co.uk
I tried to avoid having to track the write for every VMA by only
tracking writes to the ggtt. However, for the purposes of frontbuffer
tracking this is insufficient as we need to invalidate around writes not
just to the the ggtt but all aliased ppgtt views of the framebuffer. By
moving the critical section to the object and only doing so for
framebuffer writes we can reduce the tracking even further by only
watching framebuffers and not vma.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161116190704.5293-1-chris@chris-wilson.co.uk
Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
A small selection of macros which can only accept dev_priv from
now on and a resulting trickle of fixups.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: David Weinehall <david.weinehall@linux.intel.com>
A small selection of macros which can only accept dev_priv from
now on and a resulting trickle of fixups.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: David Weinehall <david.weinehall@linux.intel.com>
On LLC, or even snooped, machines rendering via the GPU ends up in the CPU
cache. This cacheline dirt also needs to be flushed to main memory when
moving to an incoherent domain, such as the display's scanout engine.
Mostly, this happens because either the object is marked as dirty from
its first use or is avoided by setting the object into the display
domain from the start.
v2: Treat WT as not requiring a clflush prior to use on the display
engine as well.
Fixes: 0f71979ab7 ("drm/i915: Performed deferred clflush inside set-cache-level")
References: https://bugs.freedesktop.org/show_bug.cgi?id=95414
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v4.0+
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161107165204.7008-1-chris@chris-wilson.co.uk
Move has_64bit_reloc into dev_priv->info. This will make it visible
in the feature listing debug output.
v2:
- Keep the struct member to keep GCC fragile but happy (Chris)
v3:
- More detailed commit message (Chris)
- Include forgotten CHV and BXT (Chris)
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/1478162386-5018-1-git-send-email-joonas.lahtinen@linux.intel.com
In preparation to support many distinct timelines, we need to expand the
activity tracking on the GEM object to handle more than just a request
per engine. We already use the struct reservation_object on the dma-buf
to handle many fence contexts, so integrating that into the GEM object
itself is the preferred solution. (For example, we can now share the same
reservation_object between every consumer/producer using this buffer and
skip the manual import/export via dma-buf.)
v2: Reimplement busy-ioctl (by walking the reservation object), postpone
the ABI change for another day. Similarly use the reservation object to
find the last_write request (if active and from i915) for choosing
display CS flips.
Caveats:
* busy-ioctl: busy-ioctl only reports on the native fences, it will not
warn of stalls (in set-domain-ioctl, pread/pwrite etc) if the object is
being rendered to by external fences. It also will not report the same
busy state as wait-ioctl (or polling on the dma-buf) in the same
circumstances. On the plus side, it does retain reporting of which
*i915* engines are engaged with this object.
* non-blocking atomic modesets take a step backwards as the wait for
render completion blocks the ioctl. This is fixed in a subsequent
patch to use a fence instead for awaiting on the rendering, see
"drm/i915: Restore nonblocking awaits for modesetting"
* dynamic array manipulation for shared-fences in reservation is slower
than the previous lockless static assignment (e.g. gem_exec_lut_handle
runtime on ivb goes from 42s to 66s), mainly due to atomic operations
(maintaining the fence refcounts).
* loss of object-level retirement callbacks, emulated by VMA retirement
tracking.
* minor loss of object-level last activity information from debugfs,
could be replaced with per-vma information if desired
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-21-chris@chris-wilson.co.uk
The plan is to make obtaining the backing storage for the object avoid
struct_mutex (i.e. use its own locking). The first step is to update the
API so that normal users only call pin/unpin whilst working on the
backing storage.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-12-chris@chris-wilson.co.uk
We only need the active reference to keep the object alive after the
handle has been deleted (so as to prevent a synchronous gem_close). Why
then pay the price of a kref on every execbuf when we can insert that
final active ref just in time for the handle deletion?
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-6-chris@chris-wilson.co.uk
We will need to wait on DMA completion (as signaled via struct fence)
before executing our i915_gem_request. Therefore we want to expose a
method for adding the await on the fence itself to the request.
v2: Add a comment detailing a failure to handle a signal-on-any
fence-array.
v3: Pretend that magic numbers don't exist.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-1-chris@chris-wilson.co.uk
We are not allowed to touch the GTT entries underneath an atomic section,
as they take a rpm wakelock (which is illegal from atomic context) and
in the near future acquiring the DMA address for a page within an object
may sleep for an allocation. This makes the current shortcircuit in
relocation_iomap() for performing a second relocation on an adjacent page
illegal, and we need to release the atomic iomapping, lookup the DMA,
insert it into the GTT before reentering the atomic iomap section.
As it happens, this is precisely what we do on if we are using an
iomapping over the full object and not just a single page and by
removing the shortcut, we do the right thing.
Fixes: 9c870d0367 ("drm/i915: Use RPM as the barrier for controlling...")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028142756.3850-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
- first slice of the gvt device model (Zhenyu et al)
- compression support for gpu error states (Chris)
- sunset clause on gpu errors resulting in dmesg noise telling users
how to report them
- .rodata diet from Tvrtko
- switch over lots of macros to only take dev_priv (Tvrtko)
- underrun suppression for dp link training (Ville)
- lspcon (hmdi 2.0 on skl/bxt) support from Shashank Sharma, polish
from Jani
- gen9 wm fixes from Paulo&Lyude
- updated ddi programming for kbl (Rodrigo)
- respect alternate aux/ddc pins (from vbt) for all ddi ports (Ville)
* tag 'drm-intel-next-2016-10-24' of git://anongit.freedesktop.org/drm-intel: (227 commits)
drm/i915: Update DRIVER_DATE to 20161024
drm/i915: Stop setting SNB min-freq-table 0 on powersave setup
drm/i915/dp: add lane_count check in intel_dp_check_link_status
drm/i915: Fix whitespace issues
drm/i915: Clean up DDI DDC/AUX CH sanitation
drm/i915: Respect alternate_ddc_pin for all DDI ports
drm/i915: Respect alternate_aux_channel for all DDI ports
drm/i915/gen9: Remove WaEnableYV12BugFixInHalfSliceChicken7
drm/i915: KBL - Recommended buffer translation programming for DisplayPort
drm/i915: Move down skl/kbl ddi iboost and n_edp_entires fixup
drm/i915: Add a sunset clause to GPU hang logging
drm/i915: Stop reporting error details in dmesg as well as the error-state
drm/i915/gvt: do not ignore return value of create_scratch_page
drm/i915/gvt: fix spare warnings on odd constant _Bool cast
drm/i915/gvt: mark symbols static where possible
drm/i915/gvt: fix sparse warnings on different address spaces
drm/i915/gvt: properly access enabled intel_engine_cs
drm/i915/gvt: Remove defunct vmap_batch()
drm/i915/gvt: Use common mapping routines for shadow_bb object
drm/i915/gvt: Use common mapping routines for indirect_ctx object
...
When handling execbuf relocations, we play a delicate dance with
pagefault. We first try to access the user pages underneath our
struct_mutex. However, if those pages were inside a GEM object, we may
trigger a pagefault and deadlock as i915_gem_fault() tries to
recursively acquire struct_mutex. Instead, we choose to disable
pagefaulting around the copy_from_user whilst inside the struct_mutex
and handle the EFAULT by falling back to a copy outside the
struct_mutex.
We however presumed that disabling pagefaults would be expensive. It is
just an operation on the local current task. Cheap enough that we can
restrict the disable/enable to the critical section around the copy, and
so avoid having to handle the atomic sections within the relocation
handling itself.
v2: Just illustrate the broken error handling rather than argue why it
is safer to ignore it, for now.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161018120251.25043-4-chris@chris-wilson.co.uk
We never used any invalid ptes, those were put in place for
a possibility of doing gpu faults. However our batchbuffers are not
restricted in length, so everything needs to be pointing to something
and thus out-of-bounds is pointing to scratch.
Remove the valid flag as it is always true.
v2: Expand commit msg, patch reorder (Mika)
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Michał Winiarski <michal.winiarski@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/1476360162-24062-1-git-send-email-michal.winiarski@intel.com
With the possibility of addition of many more number of rings in future,
the drm_i915_private structure could bloat as an array, of type
intel_engine_cs, is embedded inside it.
struct intel_engine_cs engine[I915_NUM_ENGINES];
Though this is still fine as generally there is only a single instance of
drm_i915_private structure used, but not all of the possible rings would be
enabled or active on most of the platforms. Some memory can be saved by
allocating intel_engine_cs structure only for the enabled/active engines.
Currently the engine/ring ID is kept static and dev_priv->engine[] is simply
indexed using the enums defined in intel_engine_id.
To save memory and continue using the static engine/ring IDs, 'engine' is
defined as an array of pointers.
struct intel_engine_cs *engine[I915_NUM_ENGINES];
dev_priv->engine[engine_ID] will be NULL for disabled engine instances.
There is a text size reduction of 928 bytes, from 1028200 to 1027272, for
i915.o file (but for i915.ko file text size remain same as 1193131 bytes).
v2:
- Remove the engine iterator field added in drm_i915_private structure,
instead pass a local iterator variable to the for_each_engine**
macros. (Chris)
- Do away with intel_engine_initialized() and instead directly use the
NULL pointer check on engine pointer. (Chris)
v3:
- Remove for_each_engine_id() macro, as the updated macro for_each_engine()
can be used in place of it. (Chris)
- Protect the access to Render engine Fault register with a NULL check, as
engine specific init is done later in Driver load sequence.
v4:
- Use !!dev_priv->engine[VCS] style for the engine check in getparam. (Chris)
- Kill the superfluous init_engine_lists().
v5:
- Cleanup the intel_engines_init() & intel_engines_setup(), with respect to
allocation of intel_engine_cs structure. (Chris)
v6:
- Rebase.
v7:
- Optimize the for_each_engine_masked() macro. (Chris)
- Change the type of 'iter' local variable to enum intel_engine_id. (Chris)
- Rebase.
v8: Rebase.
v9: Rebase.
v10:
- For index calculation use engine ID instead of pointer based arithmetic in
intel_engine_sync_index() as engine pointers are not contiguous now (Chris)
- For appropriateness, rename local enum variable 'iter' to 'id'. (Joonas)
- Use for_each_engine macro for cleanup in intel_engines_init() and remove
check for NULL engine pointer in cleanup() routines. (Joonas)
v11: Rebase.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Akash Goel <akash.goel@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1476378888-7372-1-git-send-email-akash.goel@intel.com
Pull drm updates from Dave Airlie:
"Core:
- Fence destaging work
- DRIVER_LEGACY to split off legacy drm drivers
- drm_mm refactoring
- Splitting drm_crtc.c into chunks and documenting better
- Display info fixes
- rbtree support for prime buffer lookup
- Simple VGA DAC driver
Panel:
- Add Nexus 7 panel
- More simple panels
i915:
- Refactoring GEM naming
- Refactored vma/active tracking
- Lockless request lookups
- Better stolen memory support
- FBC fixes
- SKL watermark fixes
- VGPU improvements
- dma-buf fencing support
- Better DP dongle support
amdgpu:
- Powerplay for Iceland asics
- Improved GPU reset support
- UVD/VEC powergating support for CZ/ST
- Preinitialised VRAM buffer support
- Virtual display support
- Initial SI support
- GTT rework
- PCI shutdown callback support
- HPD IRQ storm fixes
amdkfd:
- bugfixes
tilcdc:
- Atomic modesetting support
mediatek:
- AAL + GAMMA engine support
- Hook up gamma LUT
- Temporal dithering support
imx:
- Pixel clock from devicetree
- drm bridge support for LVDS bridges
- active plane reconfiguration
- VDIC deinterlacer support
- Frame synchronisation unit support
- Color space conversion support
analogix:
- PSR support
- Better panel on/off support
rockchip:
- rk3399 vop/crtc support
- PSR support
vc4:
- Interlaced vblank timing
- 3D rendering CPU overhead reduction
- HDMI output fixes
tda998x:
- HDMI audio ASoC support
sunxi:
- Allwinner A33 support
- better TCON support
msm:
- DT binding cleanups
- Explicit fence-fd support
sti:
- remove sti415/416 support
etnaviv:
- MMUv2 refactoring
- GC3000 support
exynos:
- Refactoring HDMI DCC/PHY
- G2D pm regression fix
- Page fault issues with wait for vblank
There is no nouveau work in this tree, as Ben didn't get a pull
request in, and he was fighting moving to atomic and adding mst
support, so maybe best it waits for a cycle"
* tag 'drm-for-v4.9' of git://people.freedesktop.org/~airlied/linux: (1412 commits)
drm/crtc: constify drm_crtc_index parameter
drm/i915: Fix conflict resolution from backmerge of v4.8-rc8 to drm-next
drm/i915/guc: Unwind GuC workqueue reservation if request construction fails
drm/i915: Reset the breadcrumbs IRQ more carefully
drm/i915: Force relocations via cpu if we run out of idle aperture
drm/i915: Distinguish last emitted request from last submitted request
drm/i915: Allow DP to work w/o EDID
drm/i915: Move long hpd handling into the hotplug work
drm/i915/execlists: Reinitialise context image after GPU hang
drm/i915: Use correct index for backtracking HUNG semaphores
drm/i915: Unalias obj->phys_handle and obj->userptr
drm/i915: Just clear the mmiodebug before a register access
drm/i915/gen9: only add the planes actually affected by ddb changes
drm/i915: Allow PCH DPLL sharing regardless of DPLL_SDVO_HIGH_SPEED
drm/i915/bxt: Fix HDMI DPLL configuration
drm/i915/gen9: fix the watermark res_blocks value
drm/i915/gen9: fix plane_blocks_per_line on watermarks calculations
drm/i915/gen9: minimum scanlines for Y tile is not always 4
drm/i915/gen9: fix the WaWmMemoryReadLatency implementation
drm/i915/kbl: KBL also needs to run the SAGV code
...
If we run out of enough aperture space to fit the entire object, we
fallback to trying to insert a single page. However, if that also fails,
we currently fail to userspace with an unexpected ENOSPC. (ENOSPC means
to userspace that their batch could not be fitted within the GTT.) Prior
to commit e8cb909ac3 ("drm/i915: Fallback to single page GTT
mmappings for relocations") the approach is to fallback to using the
slow CPU relocation path in case of iomapping failure, and that is the
behaviour we need to restore.
Fixes: e8cb909ac3 ("drm/i915: Fallback to single page GTT mmappings...")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98101
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161007065327.24515-2-chris@chris-wilson.co.uk
(cherry picked from commit d7f7633557)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
If we run out of enough aperture space to fit the entire object, we
fallback to trying to insert a single page. However, if that also fails,
we currently fail to userspace with an unexpected ENOSPC. (ENOSPC means
to userspace that their batch could not be fitted within the GTT.) Prior
to commit e8cb909ac3 ("drm/i915: Fallback to single page GTT
mmappings for relocations") the approach is to fallback to using the
slow CPU relocation path in case of iomapping failure, and that is the
behaviour we need to restore.
Fixes: e8cb909ac3 ("drm/i915: Fallback to single page GTT mmappings...")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98101
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161007065327.24515-2-chris@chris-wilson.co.uk
* the only remaining callers of "short" fault-ins are just as happy with generic
variants (both in lib/iov_iter.c); switch them to multipage variants, kill the
"short" ones
* rename the multipage variants to now available plain ones.
* get rid of compat macro defining iov_iter_fault_in_multipage_readable by
expanding it in its only user.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Now that we can wait upon fences before emitting the request, it becomes
trivial to wait upon any implicit fence provided by the dma-buf
reservation object.
To protect against failure, we force any asynchronous waits on a foreign
fence to timeout after 10s - so that a stall in another driver does not
permanently cripple ourselves. Still unpleasant though!
Testcase: igt/prime_vgem/fence-wait
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: John Harrison <john.c.harrison@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160909131201.16673-21-chris@chris-wilson.co.uk
We are about to specialize object synchronisation to enable nonblocking
execbuf submission. First we make a copy of the current object
synchronisation for execbuffer. The general i915_gem_object_sync() will
be removed following the removal of CS flips in the near future.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: John Harrison <john.c.harrison@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160909131201.16673-16-chris@chris-wilson.co.uk
Use atomic type and operands for dev_priv->mm.bsd_engine_dispatch_index
to avoid one struct_mutex locking scenario.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Zhao Yakui <yakui.zhao@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/1472731101-21982-1-git-send-email-joonas.lahtinen@linux.intel.com
With full-ppgtt, we want the user to have full control over their memory
layout, with a separate instance per context. Forcing them to use a
shared memory layout for !RCS not only duplicates the amount of work we
have to do, but also defeats the memory segregation on offer.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20160822080350.4964-4-chris@chris-wilson.co.uk
Reviewed-by: John Harrison <john.c.harrison@intel.com>
Reviewed-by: Thomas Daniel <thomas.daniel@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If userspace is asynchronously streaming into the batch or other
execobjects, we may not flush those writes along with a change in cache
domain (as there is no change). Therefore those writes may end up in
internal chipset buffers and not visible to the GPU upon execution. We
must issue a flush command or otherwise we encounter incoherency in the
batchbuffers and the GPU executing invalid commands (i.e. hanging) quite
regularly.
v2: Throw a paranoid wmb() into the general flush so that we remain
consistent with before.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90841
Fixes: 1816f92363 ("drm/i915: Support creation of unbound wc user...")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Akash Goel <akash.goel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Tested-by: Matti Hämäläinen <ccr@tnsp.org>
Cc: stable@vger.kernel.org
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-1-chris@chris-wilson.co.uk
(cherry picked from commit 600f436801)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
As io_mapping.h now always allocates the struct, we can avoid that
allocation and extra pointer dance by embedding the struct inside
drm_i915_private
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160819155428.1670-5-chris@chris-wilson.co.uk
The single largest factor in the overhead of parsing the commands is the
setup of the virtual mapping to provide a continuous block for the batch
buffer. If we keep those vmappings around (against the better judgement
of mm/vmalloc.c, which we offset by handwaving and looking suggestively
at the shrinker) we can dramatically improve the performance of the
parser for small batches (such as media workloads). Furthermore, we can
use the prepare shmem read/write functions to determine how best we
need to clflush the range (rather than every page of the object).
The impact of caching both src/dst vmaps is +80% on ivb and +140% on byt
for the throughput on small batches. (Caching just the dst vmap and
iterating over the src, doing a page by page copy is roughly 5% slower
on both platforms. That may be an acceptable trade-off to eliminate one
cached vmapping, and we may be able to reduce the per-page copying overhead
further.) For *this* simple test case, the cmdparser is now within a
factor of 2 of ideal performance.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-33-chris@chris-wilson.co.uk
In order to handle tiled partial GTT mmappings, we need to associate the
fence with an individual vma.
v2: A couple of silly drops replaced spotted by Joonas
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-21-chris@chris-wilson.co.uk
Our current practice is to only name the actual list (here
dev_priv->fence_list) using "list", and elements upon that list are
referred to as "link". Further, the lru nature is of the list and not of
the node and including in the name does not disambiguate the link from
anything else.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-20-chris@chris-wilson.co.uk
By moving map-and-fenceable tracking from the object to the VMA, we gain
fine-grained tracking and the ability to track individual fences on the VMA
(subsequent patch).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-16-chris@chris-wilson.co.uk
If we cannot pin the entire object into the mappable region of the GTT,
try to pin a single page instead. This is much more likely to succeed,
and prevents us falling back to the clflush slow path.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-14-chris@chris-wilson.co.uk
With the introduction of the reloc page cache, we are just one step away
from refactoring the relocation write functions into one. Not only does
it tidy the code (slightly), but it greatly simplifies the control logic
much to gcc's satisfaction.
v2: Add selftests to document the relationship between the clflush
flags, the KMAP bit and packing into the page-aligned pointer.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-13-chris@chris-wilson.co.uk
When doing relocations, we have to obtain a mapping to the page
containing the target address. This is either a kmap or iomap depending
on GPU and its cache coherency. Neighbouring relocation entries are
typically within the same page and so we can cache our kmapping between
them and avoid those pesky TLB flushes.
Note that there is some sleight-of-hand in how the slow relocate works
as the reloc_entry_cache implies pagefaults disabled (as we are inside a
kmap_atomic section). However, the slow relocate code is meant to be the
fallback from the atomic fast path failing. Fortunately it works as we
already have performed the copy_from_user for the relocation array (no
more pagefaults there) and the kmap_atomic cache is enabled after we
have waited upon an active buffer (so no more sleeping in atomic).
Magic!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-7-chris@chris-wilson.co.uk
If userspace is asynchronously streaming into the batch or other
execobjects, we may not flush those writes along with a change in cache
domain (as there is no change). Therefore those writes may end up in
internal chipset buffers and not visible to the GPU upon execution. We
must issue a flush command or otherwise we encounter incoherency in the
batchbuffers and the GPU executing invalid commands (i.e. hanging) quite
regularly.
v2: Throw a paranoid wmb() into the general flush so that we remain
consistent with before.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90841
Fixes: 1816f92363 ("drm/i915: Support creation of unbound wc user...")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Akash Goel <akash.goel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Tested-by: Matti Hämäläinen <ccr@tnsp.org>
Cc: stable@vger.kernel.org
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160818161718.27187-1-chris@chris-wilson.co.uk
Chris Wilson