If intel context create failed, the perf_request_latency() will return 0
rather than error, because we doesn't initialize the return value.

Fixes: 25c26f18 ("drm/i915/selftests: Measure dispatch latency")
Reported-by: NHulk Robot <hulkci@huawei.com>
Signed-off-by: NZhang Xiaoxu <zhangxiaoxu5@huawei.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20201116143540.3648870-1-zhangxiaoxu5@huawei.com
(cherry picked from commit 19384452)
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

If intel context create failed, the perf_series_engines() will return 0
rather than error, because we doesn't initialize the return value.

Fixes: cbfd3a0c ("drm/i915/selftests: Add request throughput measurement to perf")
Reported-by: NHulk Robot <hulkci@huawei.com>
Signed-off-by: NZhang Xiaoxu <zhangxiaoxu5@huawei.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20201116144112.3673011-1-zhangxiaoxu5@huawei.com
(cherry picked from commit 01d70884)
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

We are incorrectly limiting the max allocation size as per the mm
max_order, which is effectively the largest power-of-two that we can fit
in the region size. However, it's normal to setup the region or
allocator with a non-power-of-two size(for example 3G), which we should
already handle correctly, except it seems for the early too-big-check.

v2: make sure we also exercise the I915_BO_ALLOC_CONTIGUOUS path, which
is quite different, since for that we are actually limited by the
largest power-of-two that we can fit within the region size. (Chris)

Fixes: b908be54 ("drm/i915: support creating LMEM objects")
Signed-off-by: NMatthew Auld <matthew.auld@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: CQ Tang <cq.tang@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20201021103606.241395-1-matthew.auld@intel.com
(cherry picked from commit 83ebef47)
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

Since we store a pointer to the fake iommu device that is allocated on
the stack, as soon as we leave the function it goes out of scope and any
future dereference is undefined behaviour. Just in case we may need to
look at the fake iommu device after initialiation, move the allocation
from the stack into the data.

Fixes: 01b9d4e2 ("iommu/vt-d: Use dev_iommu_priv_get/set()")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200916105022.28316-2-chris@chris-wilson.co.uk
(cherry picked from commit 9f9f4101)
Signed-off-by: NJani Nikula <jani.nikula@intel.com>

To avoid having to create all the device and driver scaffolding we
just manually create and destroy a devres_group.

v2: Rebased

v3: use devres_open/release_group so we can use devm without real
hacks in the driver core or having to create an entire fake bus for
testing drivers. Might want to extract this into helpers eventually,
maybe as a mock_drm_dev_alloc or test_drm_dev_alloc.

v4:
- Fix IS_ERR handling (Matt)
- Delete surplus put_device() in mock_device_release (intel-gfx-ci)

v5:
- do not switch to device_add - it breaks runtime pm in the tests and
  with the devres_group_add/release no longer needed for automatic
  cleanup (CI). Update commit message to match.
- print correct error in pr_err (Matt)

v6: Remove now unused err variable (CI).

v7: More warning fixes ...
Acked-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> (v3)
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.william.auld@gmail.com> (v4)
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200919134032.2488403-1-daniel.vetter@ffwll.ch

Just some prep work before we rework the lifetime handling, which
requires replacing all the drm_dev_put in selftests by something else.

v2: Don't go with a static inline, upsets the header tests and
separation.
Acked-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200918132505.2316382-2-daniel.vetter@ffwll.ch

Since we store a pointer to the fake iommu device that is allocated on
the stack, as soon as we leave the function it goes out of scope and any
future dereference is undefined behaviour. Just in case we may need to
look at the fake iommu device after initialiation, move the allocation
from the stack into the data.

Fixes: 01b9d4e2 ("iommu/vt-d: Use dev_iommu_priv_get/set()")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200916105022.28316-2-chris@chris-wilson.co.uk

Make sure vma_lock is not used as inner lock when kernel context is used,
and add ww handling where appropriate.

Ensure that execbuf selftests keep passing by using ww handling.

Changes since v2:
- Fix i915_gem_context finally.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NThomas Hellström <thomas.hellstrom@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200819140904.1708856-22-maarten.lankhorst@linux.intel.comSigned-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

i915_gem_ww_ctx is used to lock all gem bo's for pinning and memory
eviction. We don't use it yet, but lets start adding the definition
first.

To use it, we have to pass a non-NULL ww to gem_object_lock, and don't
unlock directly. It is done in i915_gem_ww_ctx_fini.

Changes since v1:
- Change ww_ctx and obj order in locking functions (Jonas Lahtinen)
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NThomas Hellström <thomas.hellstrom@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200819140904.1708856-6-maarten.lankhorst@linux.intel.comSigned-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

When igt_random_offset() is a given a range of [0, PAGE_SIZE], it is
allowed to return 0. However, attempting to use a size of 0 for the
igt_lmem_write_cpu() byte poking, leads to call igt_random_offset() with
a range of [offset, offset + 0] and ask it to find a length of 4 within
it. This triggers the bug on that the requested length should fit within
the range!
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200806145728.16495-1-chris@chris-wilson.co.ukSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

The GEM object is grossly overweight for the practicality of tracking
large numbers of individual pages, yet it is currently our only
abstraction for tracking DMA allocations. Since those allocations need
to be reserved upfront before an operation, and that we need to break
away from simple system memory, we need to ditch using plain struct page
wrappers.

In the process, we drop the WC mapping as we ended up clflushing
everything anyway due to various issues across a wider range of
platforms. Though in a future step, we need to drop the kmap_atomic
approach which suggests we need to pre-map all the pages and keep them
mapped.

v2: Verify our large scratch page is suitably DMA aligned; and manually
clear the scratch since we are allocating plain struct pages full of
prior content.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200729164219.5737-2-chris@chris-wilson.co.ukSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

We need to make the DMA allocations used for page directories to be
performed up front so that we can include those allocations in our
memory reservation pass. The downside is that we have to assume the
worst case, even before we know the final layout, and always allocate
enough page directories for this object, even when there will be overlap.
This unfortunately can be quite expensive, especially as we have to
clear/reset the page directories and DMA pages, but it should only be
required during early phases of a workload when new objects are being
discovered, or after memory/eviction pressure when we need to rebind.
Once we reach steady state, the objects should not be moved and we no
longer need to preallocating the pages tables.

It should be noted that the lifetime for the page directories DMA is
more or less decoupled from individual fences as they will be shared
across objects across timelines.

v2: Only allocate enough PD space for the PTE we may use, we do not need
to allocate PD that will be left as scratch.
v3: Store the shift unto the first PD level to encapsulate the different
PTE counts for gen6/gen8.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200729164219.5737-1-chris@chris-wilson.co.ukSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

igt_mm_config() calls ilog2() on the (pseudo)random 21-bit number
s>>12.  Once in 2 million seeds, this is zero and ilog2 summons
the nasal demons.

There was an attempt to handle this case with a max(), but that's
too late; ms could already be something bizarre.

Given that the low 12 bits of s and ms are always zero, it's a lot
simpler just to divide them by 4096, then everything fits into 32
bits, and we can easily generate a random number 1 <= s <= 0x1fffff.

Fixes: 14d1b9a6 ("drm/i915: buddy allocator")
Signed-off-by: NGeorge Spelvin <lkml@sdf.org>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: intel-gfx@lists.freedesktop.org
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200325192429.GA8865@SDF.ORGSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
(cherry picked from commit 21118e8e56479ef33460fbd63a5ad0535843b666)
Signed-off-by: NJani Nikula <jani.nikula@intel.com>

Since we use the module parameters stored inside the drm_i915_device
itself, we need to ensure the mock i915_device also sets up the right
defaults.

Fixes: 8a25c4be ("drm/i915/params: switch to device specific parameters")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200728150600.4509-1-chris@chris-wilson.co.ukSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>
(cherry picked from commit 98ef067453709444a264939940f7b3a5dfdfa09e)
Signed-off-by: NJani Nikula <jani.nikula@intel.com>

igt_mm_config() calls ilog2() on the (pseudo)random 21-bit number
s>>12.  Once in 2 million seeds, this is zero and ilog2 summons
the nasal demons.

There was an attempt to handle this case with a max(), but that's
too late; ms could already be something bizarre.

Given that the low 12 bits of s and ms are always zero, it's a lot
simpler just to divide them by 4096, then everything fits into 32
bits, and we can easily generate a random number 1 <= s <= 0x1fffff.

Fixes: 14d1b9a6 ("drm/i915: buddy allocator")
Signed-off-by: NGeorge Spelvin <lkml@sdf.org>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: intel-gfx@lists.freedesktop.org
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200325192429.GA8865@SDF.ORGSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

In function i915_active_acquire_preallocate_barrier(), not all
paths have the return value set correctly, and in case of memory
allocation failure, a negative error code should be returned.

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NTianjia Zhang <tianjia.zhang@linux.alibaba.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200802115655.25568-1-chris@chris-wilson.co.ukSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

Since we use the module parameters stored inside the drm_i915_device
itself, we need to ensure the mock i915_device also sets up the right
defaults.

Fixes: 8a25c4be ("drm/i915/params: switch to device specific parameters")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200728150600.4509-1-chris@chris-wilson.co.ukSigned-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

The error code needs to be set on this path.  It currently returns
success.

Fixes: ed2690a9 ("drm/i915/selftest: Check that GPR are restored across noa_wait")
Reported-by: NLionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: NLionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200714143652.GA337376@mwanda

There is an error condition where err is not being set and an uninitialized
garbage value in err is being returned.  Fix this by assigning err to an
appropriate error return value before taking the error exit path.

Addresses-Coverity: ("Uninitialized scalar value")
Fixes: ed2690a9 ("drm/i915/selftest: Check that GPR are restored across noa_wait")
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200713142551.423649-1-colin.king@canonical.com

Perf implements a GPU delay (noa_wait) by looping until the CS timestamp
has passed a certain point. This use MI_MATH and the general purpose
registers of the user's context, and since it is clobbering the user
state it must carefully save and restore the user's data around the
noa_wait. We can verify this by loading some values in the GPR that we
know will be clobbered by the noa_wait, and then inspecting the GPR after
the noa_wait completes and confirming that they have been restored.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: NLionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200709224504.11345-2-chris@chris-wilson.co.uk

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: NDaniele 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: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200708003952.21831-5-daniele.ceraolospurio@intel.com

Reuse the ppgtt_bind_vma() for aliasing_ppgtt_bind_vma() so we can
reduce some code near-duplication. The catch is that we need to then
pass along the i915_address_space and not rely on vma->vm, as they
differ with the aliasing-ppgtt.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NAndi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200703102519.26539-1-chris@chris-wilson.co.uk

Remove the use of dev->archdata.iommu and use the private per-device
pointer provided by IOMMU core code instead.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>
Reviewed-by: NJerry Snitselaar <jsnitsel@redhat.com>
Reviewed-by: NLu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20200625130836.1916-3-joro@8bytes.org

Return the monotonic timestamp (ktime_get()) at the time of sampling the
busy-time. This is used in preference to taking ktime_get() separately
before or after the read seqlock as there can be some large variance in
reported timestamps. For selftests trying to ascertain that we are
reporting accurate to within a few microseconds, even a small delay
leads to the test failing.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200617130916.15261-2-chris@chris-wilson.co.uk

A couple of very simple tests to ensure that the basic properties of
per-engine busyness accounting [0% and 100% busy] are faithful.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200617130916.15261-1-chris@chris-wilson.co.uk

In commit 5ba32c7b ("drm/i915/execlists: Always force a context
reload when rewinding RING_TAIL"), we placed the check for rewinding a
context on actually submitting the next request in that context. This
was so that we only had to check once, and could do so with precision
avoiding as many forced restores as possible. For example, to ensure
that we can resubmit the same request a couple of times, we include a
small wa_tail such that on the next submission, the ring->tail will
appear to move forwards when resubmitting the same request. This is very
common as it will happen for every lite-restore to fill the second port
after a context switch.

However, intel_ring_direction() is limited in precision to movements of
upto half the ring size. The consequence being that if we tried to
unwind many requests, we could exceed half the ring and flip the sense
of the direction, so missing a force restore. As no request can be
greater than half the ring (i.e. 2048 bytes in the smallest case), we
can check for rollback incrementally. As we check against the tail that
would be submitted, we do not lose any sensitivity and allow lite
restores for the simple case. We still need to double check upon
submitting the context, to allow for multiple preemptions and
resubmissions.

Fixes: 5ba32c7b ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: NBruce Chang <yu.bruce.chang@intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200609151723.12971-1-chris@chris-wilson.co.uk
(cherry picked from commit e36ba817)
Signed-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

Reduce the smoke depth by trimming the number of contexts, repetitions
and wait times. This is in preparation for a less greedy scheduler that
tries to be fair across contexts, resulting in a great many more context
switches. A thousand context switches may be 50-100ms, causing us to
timeout as the HW is not fast enough to complete the deep smoketests.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NAndi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200607222108.14401-5-chris@chris-wilson.co.uk

In commit 5ba32c7b ("drm/i915/execlists: Always force a context
reload when rewinding RING_TAIL"), we placed the check for rewinding a
context on actually submitting the next request in that context. This
was so that we only had to check once, and could do so with precision
avoiding as many forced restores as possible. For example, to ensure
that we can resubmit the same request a couple of times, we include a
small wa_tail such that on the next submission, the ring->tail will
appear to move forwards when resubmitting the same request. This is very
common as it will happen for every lite-restore to fill the second port
after a context switch.

However, intel_ring_direction() is limited in precision to movements of
upto half the ring size. The consequence being that if we tried to
unwind many requests, we could exceed half the ring and flip the sense
of the direction, so missing a force restore. As no request can be
greater than half the ring (i.e. 2048 bytes in the smallest case), we
can check for rollback incrementally. As we check against the tail that
would be submitted, we do not lose any sensitivity and allow lite
restores for the simple case. We still need to double check upon
submitting the context, to allow for multiple preemptions and
resubmissions.

Fixes: 5ba32c7b ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: NBruce Chang <yu.bruce.chang@intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200609151723.12971-1-chris@chris-wilson.co.uk

This has no code changes, but the typo is clearly getting copy/pasted,
so better to avoid this now and fix the typo. IS_ENABLED() takes full
names, and must have the "CONFIG_" prefix.
Reported-by: NJoe Perches <joe@perches.com>
Link: https://lore.kernel.org/lkml/b08611018fdb6d88757c6008a5c02fa0e07b32fb.camel@perches.comSigned-off-by: NKees Cook <keescook@chromium.org>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/202006050718.9D4FCFC2E@keescook

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: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NAkeem G Abodunrin <akeem.g.abodunrin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200602220953.21178-1-chris@chris-wilson.co.uk

Name the object classes and their offspring for easier lockdep
debugging.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200529183204.16850-2-chris@chris-wilson.co.uk

A useful metric of the system's health is how fast we can tell the GPU
to do various actions, so measure our latency.

v2: Refactor all the instruction building into emitters.
v3: Mark the error handling if not perfect, at least consistent.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200519130802.4067-1-chris@chris-wilson.co.uk

Pull the code to do the CS timestamp ns<->ticks conversion into
helpers and use them all over.

The check in i915_perf_noa_delay_set() seems a bit dubious,
so we switch it to do what I assume it wanted to do all along
(ie. make sure the resulting delay in CS timestamp ticks
doesn't exceed 32bits)?

Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200302143943.32676-5-ville.syrjala@linux.intel.comReviewed-by: NChris Wilson <chris@chris-wilson.co.uk>

kHz isn't accurate enough for storing the CS timestamp
frequency on some of the platforms. Store the value
in Hz instead.

Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200302143943.32676-2-ville.syrjala@linux.intel.comReviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NLionel Landwerlin <lionel.g.landwerlin@intel.com>

While this does not appear to fix any issues, the backend itself knows
when it wants to emit a breadcrumb, so let it make the final call.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200513074809.18194-16-chris@chris-wilson.co.uk

Be consistent, and even when we know we had used a WC, flush the mapped
object after writing into it. The flush understands the mapping type and
will only clflush if !I915_MAP_WC, but will always insert a wmb [sfence]
so that we can be sure that all writes are visible.

v2: Add the unconditional wmb so we are know that we always flush the
writes to memory/HW at that point.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200511141304.599-1-chris@chris-wilson.co.uk

Be consistent and ensure that we always emit the asynchronous waits
prior to issuing instructions that use the address. This ensures that if
we do emit GPU commands to do the await, they are before our use!
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200510102431.21959-1-chris@chris-wilson.co.uk

The presumption is that by using a circular counter that is twice as
large as the maximum ELSP submission, we would never reuse the same CCID
for two inflight contexts.

However, if we continually preempt an active context such that it always
remains inflight, it can be resubmitted with an arbitrary number of
paired contexts. As each of its paired contexts will use a new CCID,
eventually it will wrap and submit two ELSP with the same CCID.

Rather than use a simple circular counter, switch over to a small bitmap
of inflight ids so we can avoid reusing one that is still potentially
active.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796
Fixes: 2935ed53 ("drm/i915: Remove logical HW ID")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-2-chris@chris-wilson.co.uk
(cherry picked from commit 5c4a53e3)
(cherry picked from commit 134711240307d5586ae8e828d2699db70a8b74f2)
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

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/757Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200504140629.28240-1-chris@chris-wilson.co.uk

In the near future, we will utilize the busy-stats on each engine to
approximate the C0 cycles of each, and use that as an input to a manual
RPS mechanism. That entails having busy-stats always enabled and so we
can remove the enable/disable routines and simplify the pmu setup. As a
consequence of always having the stats enabled, we can also show the
current active time via sysfs/engine/xcs/active_time_ns.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200429205446.3259-1-chris@chris-wilson.co.uk