This reverts commit 964a9b0f61 ("drm/i915/gem: Use chained reloc batches")
and commit 0e97fbb080 ("drm/i915/gem: Use a single chained reloc batches
for a single execbuf").
When adding ww locking to execbuf, it's hard enough to deal with a
single BO that is part of relocation execution. Chaining is hard to
get right, and with GPU relocation deprecated, it's best to drop this
altogether, instead of trying to fix something we will remove.
This is not a completely 1:1 revert, we reset rq_size to 0 in
reloc_cache_init, this was from e3d291301f ("drm/i915/gem: Implement legacy
MI_STORE_DATA_IMM"), because we don't want to break the selftests. (Daniel)
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200819140904.1708856-3-maarten.lankhorst@linux.intel.com
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
This reverts commit 9e0f9464e2 ("drm/i915/gem: Async GPU relocations only"),
and related commit 7ac2d2536d ("drm/i915/gem: Delete unused code").
Async GPU relocations are not the path forward, we want to remove
GPU accelerated relocation support eventually when userspace is fixed
to use VM_BIND, and this is the first step towards that. We will keep
async gpu relocations around for now, until userspace is fixed.
Relocation support will be disabled completely on platforms where there
was never any userspace that depends on it, as the hardware doesn't
require it from at least gen9+ onward. For older platforms, the plan
is to use cpu relocations only.
The igt side is fixed in igt commit 39e9aa1032a4e ("tests/i915: Remove
subtests that rely on async relocation behavior").
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200819140904.1708856-2-maarten.lankhorst@linux.intel.com
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
If dma_fence_chain_find_seqno() reports an error, it does so in its
preamble before it disposes of the input fence. On handling the
error, we need to drop the reference to the fence.
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2292
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 13149e8baf ("drm/i915: add syncobj timeline support")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200806161056.17593-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Sometimes we have to be very careful not to allocate underneath a mutex
(or spinlock) and yet still want to track activity. Enter
i915_active_acquire_for_context(). This raises the activity counter on
i915_active prior to use and ensures that the fence-tree contains a slot
for the context.
v2: Refactor active_lookup() so it can be called again before/after
locking to resolve contention. Since we protect the rbtree until we
idle, we can do a lockfree lookup, with the caveat that if another
thread performs a concurrent insertion, the rotations from the insert
may cause us to not find our target. A second pass holding the treelock
will find the target if it exists, or the place to perform our
insertion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Thomas Hellström <thomas.hellstrom@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200731085015.32368-3-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
I915_GEM_THROTTLE dates back to the time before contexts where there was
just a single engine, and therefore a single timeline and request list
globally. That request list was in execution/retirement order, and so
walking it to find a particular aged request made sense and could be
split per file.
That is no more. We now have many timelines with a file, as many as the
user wants to construct (essentially per-engine, per-context). Each of
those run independently and so make the single list futile. Remove the
disordered list, and iterate over all the timelines to find a request to
wait on in each to satisfy the criteria that the CPU is no more than 20ms
ahead of its oldest request.
It should go without saying that the I915_GEM_THROTTLE ioctl is no
longer used as the primary means of throttling, so it makes sense to push
the complication into the ioctl where it only impacts upon its few
irregular users, rather than the execbuf/retire where everybody has to
pay the cost. Fortunately, the few users do not create vast amount of
contexts, so the loops over contexts/engines should be concise.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200728152010.30701-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Introduces a new parameters to execbuf so that we can specify syncobj
handles as well as timeline points.
v2: Reuse i915_user_extension_fn
v3: Check that the chained extension is only present once (Chris)
v4: Check that dma_fence_chain_find_seqno returns a non NULL fence (Lionel)
v5: Use BIT_ULL (Chris)
v6: Fix issue with already signaled timeline points,
dma_fence_chain_find_seqno() setting fence to NULL (Chris)
v7: Report ENOENT with invalid syncobj handle (Lionel)
v8: Check for out of order timeline point insertion (Chris)
v9: After explanations on
https://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
v11: Rebase
v12: Allow multiple extension nodes of timeline syncobj (Chris)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Co-authored-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> (v11)
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804085954.350343-3-lionel.g.landwerlin@intel.com
Link: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2901
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
v4: Rebase
Move array fence parsing in i915_gem_do_execbuffer()
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804085954.350343-2-lionel.g.landwerlin@intel.com
Link: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2901
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Since the engines belong to the GT, move the runtime-updated list of
available engines to the intel_gt struct. The original mask has been
renamed to indicate it contains the maximum engine list that can be
found on a matching device.
In preparation for other info being moved to the gt in follow up patches
(sseu), introduce an intel_gt_info structure to group all gt-related
runtime info.
v2: s/max_engine_mask/platform_engine_mask (tvrtko), fix selftest
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Andi Shyti <andi.shyti@intel.com>
Cc: Venkata Sandeep Dhanalakota <venkata.s.dhanalakota@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> #v1
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200708003952.21831-5-daniele.ceraolospurio@intel.com
Rather than reuse the common ctx->mutex for locking the execbuffer LUT,
split it into its own lock to avoid being taken [as part of ctx->mutex]
at inappropriate times. In particular to avoid the inversion from taking
the timeline->mutex for the whole execbuf submission in the next patch.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200703004306.11117-1-chris@chris-wilson.co.uk
The obj->lut_list is traversed when the object is closed as the file
table is destroyed during process termination. As this occurs before we
kill any outstanding context if, due to some bug or another, the closure
is blocked, then we fail to shootdown any inflight operations
potentially leaving the GPU spinning forever. As we only need to guard
the list against concurrent closures and insertions, the hold is short
and merits being treated as a simple spinlock.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200701084439.17025-1-chris@chris-wilson.co.uk
Catch up with upstream, in particular to get c1e8d7c6a7 ("mmap locking
API: convert mmap_sem comments").
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
core:
- fix race in connectors sending hotplug
i915:
- Avoid use after free in cmdparser
- Avoid NULL dereference when probing all display encoders
- Fixup to module parameter type
sun4i:
- clock divider fix
ast:
- 24/32 bpp mode setting fix
-----BEGIN PGP SIGNATURE-----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=IBzW
-----END PGP SIGNATURE-----
Merge tag 'drm-next-2020-06-11-1' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Dave Airlie:
"One sun4i fix and a connector hotplug race The ast fix is for a
regression in 5.6, and one of the i915 ones fixes an oops reported by
dhowells.
core:
- fix race in connectors sending hotplug
i915:
- Avoid use after free in cmdparser
- Avoid NULL dereference when probing all display encoders
- Fixup to module parameter type
sun4i:
- clock divider fix
ast:
- 24/32 bpp mode setting fix"
* tag 'drm-next-2020-06-11-1' of git://anongit.freedesktop.org/drm/drm:
drm/ast: fix missing break in switch statement for format->cpp[0] case 4
drm/sun4i: hdmi ddc clk: Fix size of m divider
drm/i915/display: Only query DP state of a DDI encoder
drm/i915/params: fix i915.reset module param type
drm/i915/gem: Mark the buffer pool as active for the cmdparser
drm/connector: notify userspace on hotplug after register complete
If the execbuf is interrupted after building the cmdparser pipeline, and
before we commit to submitting the request to HW, we would attempt to
clean up the cmdparser early. While we held active references to the vma
being parsed and constructed, we did not hold an active reference for
the buffer pool itself. The result was that an interrupted execbuf could
still have run the cmdparser pipeline, but since the buffer pool was
idle, its target vma could have been recycled.
Note this problem only occurs if the cmdparser is running async due to
pipelined waits on busy fences, and the execbuf is interrupted.
Fixes: 686c7c35ab ("drm/i915/gem: Asynchronous cmdparser")
Fixes: 16e8745967 ("drm/i915/gt: Move the batch buffer pool from the engine to the gt")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200604103751.18816-1-chris@chris-wilson.co.uk
(cherry picked from commit 57a78ca4ec)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reduce the 3 relocation paths down to the single path that accommodates
all. The primary motivation for this is to guard the relocations with a
natural fence (derived from the i915_request used to write the
relocation from the GPU).
The tradeoff in using async gpu relocations is that it increases latency
over using direct CPU relocations, for the cases where the target is
idle and accessible by the CPU. The benefit is greatly reduced lock
contention and improved concurrency by pipelining.
Note that forcing the async gpu relocations does reveal a few issues
they have. Firstly, is that they are visible as writes to gem_busy,
causing to mark some buffers are being to written to by the GPU even
though userspace only reads. Secondly is that, in combination with the
cmdparser, they can cause priority inversions. This should be the case
where the work is being put into a common workqueue losing our priority
information and so being executed in FIFO from the worker, denying us
the opportunity to reorder the requests afterwards.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200604211457.19696-1-chris@chris-wilson.co.uk
If the execbuf is interrupted after building the cmdparser pipeline, and
before we commit to submitting the request to HW, we would attempt to
clean up the cmdparser early. While we held active references to the vma
being parsed and constructed, we did not hold an active reference for
the buffer pool itself. The result was that an interrupted execbuf could
still have run the cmdparser pipeline, but since the buffer pool was
idle, its target vma could have been recycled.
Note this problem only occurs if the cmdparser is running async due to
pipelined waits on busy fences, and the execbuf is interrupted.
Fixes: 686c7c35ab ("drm/i915/gem: Asynchronous cmdparser")
Fixes: 16e8745967 ("drm/i915/gt: Move the batch buffer pool from the engine to the gt")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200604103751.18816-1-chris@chris-wilson.co.uk
We infrequently use the direct i915 backpointer from the i915_request,
so do we really need to waste the space in the struct for it? 8 bytes
from the most frequently allocated struct vs an 3 bytes and pointer
chasing in using rq->engine->i915?
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200602220953.21178-1-chris@chris-wilson.co.uk
struct drm_device specific drm_WARN* macros include device information
in the backtrace, so we know what device the warnings originate from.
Prefer drm_WARN* over WARN* at places where struct drm_device pointer
can be extracted.
Signed-off-by: Pankaj Bharadiya <pankaj.laxminarayan.bharadiya@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200504181600.18503-6-pankaj.laxminarayan.bharadiya@intel.com
Now that we have fast timeslicing on semaphores, we no longer need to
prioritise none-semaphore work as we will yield any work blocked on a
semaphore to the next in the queue. Previously with no timeslicing,
blocking on the semaphore caused extremely bad scheduling with multiple
clients utilising multiple rings. Now, there is no impact and we can
remove the complication.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200513173504.28322-1-chris@chris-wilson.co.uk
Since there can only be one of in_fence/exec_fence, just use the single
in_fence local.
v2: Consolidate lookup
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200513180937.28992-1-chris@chris-wilson.co.uk
Upon waiting a request (when asked), we gave that request a small
priority boost, not enough for it to cause preemption, but enough for it
to be scheduled next before all equals. We also used that bit to give
new clients a small priority boost, similar to FQ_CODEL, such that we
favoured short interactive tasks ahead of long running streams.
However, this is causing lots of complications with timeslicing where we
both want to honour the boost and yet ignore it. Those complications
cause unexpected user behaviour (tasks not being timesliced and run
concurrently as epxected), and the easiest way to resolve that is to
remove the boost. Hopefully, we can find a compromise again if we need
to, but in theory timeslicing itself and future more advanced schedulers
should give us the interactivity boost we seek.
Testcase: igt/gem_exec_schedule/lateslice
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200507152338.7452-3-chris@chris-wilson.co.uk
The older arches did not convert MI_STORE_DATA_IMM to using the GTT, but
left them writing to a physical address. The notes suggest that the
primary reason would be so that the writes were cache coherent, as the
CPU cache uses physical tagging. As such we did not implement the
legacy variant of MI_STORE_DATA_IMM and so left all the relocations
synchronous -- but with a small function to convert from the vma address
into the physical address, we can implement asynchronous relocs on these
older arches, fixing up a few tests that require them.
In order to be able to test the legacy paths, refactor the gpu
relocations so that we can hook them up to a selftest.
v2: Use an array of offsets not enum labels for the selftest
v3: Refactor the common igt_hexdump()
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/757
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200504140629.28240-1-chris@chris-wilson.co.uk
It is required that a chained batch be in the same address domain as its
parent, and also that must be specified in the command for earlier gen
as it is not inferred from the chaining until gen6.
Fixes: 964a9b0f61 ("drm/i915/gem: Use chained reloc batches")
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/20200504125149.4396-1-chris@chris-wilson.co.uk
If at first we don't succeed, try try again.
Not all engines may support the MI ops we need to perform asynchronous
relocation patching, and so we end up falling back to a synchronous
operation that has a liability of blocking. However, Tvrtko pointed out
we don't need to use the same engine to perform the relocations as we
are planning to execute the execbuf on, and so if we switch over to a
working engine, we can perform the relocation asynchronously. The user
execbuf will be queued after the relocations by virtue of fencing.
This patch creates a new context per execbuf requiring asynchronous
relocations on an unusable engines. This is perhaps a bit excessive and
can be ameliorated by a small context cache, but for the moment we only
need it for working around a little used engine on Sandybridge, and only
if relocations are actually required to an active batch buffer.
Now we just need to teach the relocation code to handle physical
addressing for gen2/3, and we should then have universal support!
Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Testcase: igt/gem_exec_reloc/basic-spin # snb
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200501192945.22215-3-chris@chris-wilson.co.uk
As we can now keep chaining together a relocation batch to process any
number of relocations, we can keep building that relocation batch for
all of the target vma. This avoiding emitting a new request into the
ring for each target, consuming precious ring space and a potential
stall.
v2: Propagate the failure from submitting the relocation batch.
Testcase: igt/gem_exec_reloc/basic-wide-active
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200501192945.22215-2-chris@chris-wilson.co.uk
The ring is a precious resource: we anticipate to only use a few hundred
bytes for a request, and only try to reserve that before we start. If we
go beyond our guess in building the request, then instead of waiting at
the start of execbuf before we hold any locks or other resources, we
may trigger a wait inside a critical region. One example is in using gpu
relocations, where currently we emit a new MI_BB_START from the ring
every time we overflow a page of relocation entries. However, instead of
insert the command into the precious ring, we can chain the next page of
relocation entries as MI_BB_START from the end of the previous.
v2: Delay the emit_bb_start until after all the chained vma
synchronisation is complete. Since the buffer pool batches are idle, this
_should_ be a no-op, but one day we may some fancy async GPU bindings
for new vma!
v3: Use pool/batch consitently, once we start thinking in terms of the
batch vma, use batch->obj.
v4: Explain the magic number 4.
Tvrtko spotted that we lose propagation of the error for failing to
submit the relocation request; that's easier to fix up in the next
patch.
Testcase: igt/gem_exec_reloc/basic-many-active
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/20200501192945.22215-1-chris@chris-wilson.co.uk
When i915_gem_execbuffer2_ioctl() is using user_access_begin(),
that's only to perform unsafe_put_user() so use
user_write_access_begin() in order to only open write access.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/ebf1250b6d4f351469fb339e5399d8b92aa8a1c1.1585898438.git.christophe.leroy@c-s.fr
Since the introduction of 'soft-rc6', we aim to park the device quickly
and that results in frequent idling of the whole device. Currently upon
idling we free the batch buffer pool, and so this renders the cache
ineffective for many workloads. If we want to have an effective cache of
recently allocated buffers available for reuse, we need to decouple that
cache from the engine powermanagement and make it timer based. As there
is no reason then to keep it within the engine (where it once made
retirement order easier to track), we can move it up the hierarchy to the
owner of the memory allocations.
v2: Hook up to debugfs/drop_caches to clear the cache on demand.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200430111819.10262-2-chris@chris-wilson.co.uk
The history of i915_vma_close() is confusing, as is its use. As the
lifetime of the i915_vma is currently bounded by the object it is
attached to, we needed a means of identify when a vma was no longer in
use by userspace (via the user's fd). This is further complicated by
that only ppgtt vma should be closed at the user's behest, as the ggtt
were always shared.
Now that we attach the vma to a lut on the user's context, the open
count does indicate how many unique and open context/vm are referencing
this vma from the user. As such, we can and should just use the
open_count to track when the vma is still in use by userspace.
It's a poor man's replacement for reference counting.
Closes: https://gitlab.freedesktop.org/drm/intel/issues/1193
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200422190558.30509-1-chris@chris-wilson.co.uk
__i915_gem_object_flush_map() takes a byte range, so feed it the written
bytes and do not mistake the u32 index as bytes!
Fixes: a679f58d05 ("drm/i915: Flush pages on acquisition")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Cc: <stable@vger.kernel.org> # v5.2+
Reviewed-by: Matthew Auld <matthew.william.auld@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200406114821.10949-1-chris@chris-wilson.co.uk
(cherry picked from commit 30c88a47f1)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
If the user passes in a readonly reloc[], by the time we notice we have
already committed to modifying the execobjects, or have indeed done so
already. Reporting the failure just compounds the issue as we have no
second pass to fall back to anymore.
"Be damned if you do, and damned if you don't."
Testcase: igt/gem_exec_reloc/readonly
Fixes: 7dc8f11437 ("drm/i915/gem: Drop relocation slowpath")
References: fddcd00a49 ("drm/i915: Force the slow path after a user-write error")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200331162150.3635-1-chris@chris-wilson.co.uk
(cherry picked from commit 97a37c919f)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
__i915_gem_object_flush_map() takes a byte range, so feed it the written
bytes and do not mistake the u32 index as bytes!
Fixes: a679f58d05 ("drm/i915: Flush pages on acquisition")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Cc: <stable@vger.kernel.org> # v5.2+
Reviewed-by: Matthew Auld <matthew.william.auld@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200406114821.10949-1-chris@chris-wilson.co.uk
If the current node/entry location is occupied, and the object is not
pinned, try assigning it some free space. We cannot wait here, so if in
doubt, we unreserve and try to grab all at once.
v2: Use the final pin_flags so that we won't have to move the object if
we find the wrong free space.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200401194135.5442-1-chris@chris-wilson.co.uk
If the user passes in a readonly reloc[], by the time we notice we have
already committed to modifying the execobjects, or have indeed done so
already. Reporting the failure just compounds the issue as we have no
second pass to fall back to anymore.
"Be damned if you do, and damned if you don't."
Testcase: igt/gem_exec_reloc/readonly
Fixes: 7dc8f11437 ("drm/i915/gem: Drop relocation slowpath")
References: fddcd00a49 ("drm/i915: Force the slow path after a user-write error")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200331162150.3635-1-chris@chris-wilson.co.uk
Use a separate array allocation for the execbuf vma, so that we can
track their lifetime independently from the copy of the user arguments.
With luck, this has a secondary benefit of splitting the malloc size to
within reason and avoid vmalloc. The downside is that we might require
two separate vmallocs -- but much less likely.
In the process, this prevents a memory leak on the ww_mutex error
unwind.
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1390
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200330133710.14385-1-chris@chris-wilson.co.uk
A recent commit in clang added -Wtautological-compare to -Wall, which is
enabled for i915 after -Wtautological-compare is disabled for the rest
of the kernel so we see the following warning on x86_64:
../drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c:1433:22: warning:
result of comparison of constant 576460752303423487 with expression of
type 'unsigned int' is always false
[-Wtautological-constant-out-of-range-compare]
if (unlikely(remain > N_RELOC(ULONG_MAX)))
~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~
../include/linux/compiler.h:78:42: note: expanded from macro 'unlikely'
# define unlikely(x) __builtin_expect(!!(x), 0)
^
1 warning generated.
It is not wrong in the case where ULONG_MAX > UINT_MAX but it does not
account for the case where this file is built for 32-bit x86, where
ULONG_MAX == UINT_MAX and this check is still relevant.
Cast remain to unsigned long, which keeps the generated code the same
(verified with clang-11 on x86_64 and GCC 9.2.0 on x86 and x86_64) and
the warning is silenced so we can catch more potential issues in the
future.
Closes: https://github.com/ClangBuiltLinux/linux/issues/778
Suggested-by: Michel Dänzer <michel@daenzer.net>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200214054706.33870-1-natechancellor@gmail.com
I need to keep the GEM context around a bit longer so adding an explicit
flag for syncing execbuf with closed/abandonded contexts.
v2:
* Use already available context flags. (Chris)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@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/20200319170707.8262-1-chris@chris-wilson.co.uk
(cherry picked from commit 207e4a71fb)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
If the caller allows and we do not have to wait for any signals,
immediately execute the work within the caller's process. By doing so we
avoid the overhead of scheduling a new task, and the latency in
executing it, at the cost of pulling that work back into the immediate
context. (Sometimes we still prefer to offload the task to another cpu,
especially if we plan on executing many such tasks in parallel for this
client.)
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/20200325120227.8044-2-chris@chris-wilson.co.uk
Drop the pretense of kicking the tasklet (used only for the defunct guc
submission backend, it should just take ownership of the submit!) and so
remove the bh-kicking from around submission.
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/20200323092841.22240-5-chris@chris-wilson.co.uk
As we store the handle lookup inside a radix tree, we do not need the
gem_context->mutex except until we need to insert our lookup into the
common radix tree. This takes a small bit of rearranging to ensure that
the lut we insert into the tree is ready prior to actually inserting it
(as soon as it is exposed via the radixtree, it is visible to any other
submission).
v2: For brownie points, remove the goto spaghetti.
v3: Tighten up the closed-handle checks.
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/20200323092841.22240-8-chris@chris-wilson.co.uk
I need to keep the GEM context around a bit longer so adding an explicit
flag for syncing execbuf with closed/abandonded contexts.
v2:
* Use already available context flags. (Chris)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@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/20200319170707.8262-1-chris@chris-wilson.co.uk
Since the relocations are no longer performed under a global
struct_mutex, or any other lock, that is also held by pagefault handlers,
we can relax and allow our fast path to take a fault. As we no longer
need to abort the fast path for lock avoidance, we no longer need the
slow path handling at all.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200311160310.26711-1-chris@chris-wilson.co.uk
We only need to serialise the multiple pinning during the eb_reserve
phase. Ideally this would be using the vm->mutex as an outer lock, or
using a composite global mutex (ww_mutex), but at the moment we are
using struct_mutex for the group.
Closes: https://gitlab.freedesktop.org/drm/intel/issues/1381
Fixes: 003d8b9143 ("drm/i915/gem: Only call eb_lookup_vma once during execbuf ioctl")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200306071614.2846708-3-chris@chris-wilson.co.uk
The alignment is u64, and yet is_power_of_2() assumes unsigned long,
which might give different results between 32b and 64b kernel.
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
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/20200305203534.210466-1-matthew.auld@intel.com
Cc: stable@vger.kernel.org
Trying to use i915_request_skip() prior to i915_request_add() causes us
to try and fill the ring upto request->postfix, which has not yet been
set, and so may cause us to memset() past the end of the ring.
Instead of skipping the request immediately, just flag the error on the
request (only accepting the first fatal error we see) and then clear the
request upon submission.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200304121849.2448028-1-chris@chris-wilson.co.uk
As we no longer stash anything inside i915_vma under the exclusive
protection of struct_mutex, we do not need to revoke the i915_vma
stashes before dropping struct_mutex to handle pagefaults. Knowing that
we must drop the struct_mutex while keeping the eb->vma around, means
that we are required to hold onto to the object reference until we have
marked the vma as active.
Fixes: 155ab8836c ("drm/i915: Move object close under its own lock")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200303204345.1859734-3-chris@chris-wilson.co.uk
Maarten Lankhorst