Since we wake the GT up before executing a request, and go to sleep as
soon as it is retired, the GT wake time not only represents how long the
device is powered up, but also provides a summary, albeit an overestimate,
of the device runtime (i.e. the rc0 time to compare against rc6 time).
v2: s/busy/awake/
v3: software-gt-awake-time and I915_PMU_SOFTWARE_GT_AWAKE_TIME
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Reported-by: kernel test robot <oliver.sang@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201215154456.13954-1-chris@chris-wilson.co.uk
Adding any kinds of "last" abi markers is usually a mistake which I
repeated when implementing the PMU because it felt convenient at the time.
This patch marks I915_PMU_LAST as deprecated and stops the internal
implementation using it for sizing the event status bitmask and array.
New way of sizing the fields is a bit less elegant, but it omits reserving
slots for tracking events we are not interested in, and as such saves some
runtime space. Adding sampling events is likely to be a special event and
the new plumbing needed will be easily detected in testing. Existing
asserts against the bitfield and array sizes are keeping the code safe.
First event which gets the new treatment in this new scheme are the
interrupts - which neither needs any tracking in i915 pmu nor needs
waking up the GPU to read it.
v2:
* Streamline helper names. (Chris)
v3:
* Comment which events need tracking. (Chris)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20201201131757.206367-1-tvrtko.ursulin@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>
This new parameter let's the application choose how often the OA
buffer should be checked on the CPU side for data availability. Longer
polling period tend to reduce CPU overhead if the application does not
care about somewhat real time data collection.
v2: Allow disabling polling completely with 0 value (Lionel)
v3: Version the new parameter (Joonas)
v4: Rebase (Umesh)
v5: Make poll delay value of 0 invalid (Umesh)
v6:
- Describe poll_oa_period (Ashutosh)
- Fix comment for new poll parameter (Lionel)
- Drop open_flags in read_properties_unlocked (Lionel)
- Rename uapi parameter (Ashutosh)
v7: Reword the comment in uapi (Ashutosh)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200324185457.14635-4-umesh.nerlige.ramappa@intel.com
On Gen11 powergating half the execution units is a functional
requirement when using the VME samplers. Not fullfilling this
requirement can lead to hangs.
This unfortunately plays fairly poorly with the NOA requirements. NOA
requires a stable power configuration to maintain its configuration.
As a result using OA (and NOA feeding into it) so far has required us
to use a power configuration that can work for all contexts. The only
power configuration fullfilling this is powergating half the execution
units.
This makes performance analysis for 3D workloads somewhat pointless.
Failing to find a solution that would work for everybody, this change
introduces a new i915-perf stream open parameter that punts the
decision off to userspace. If this parameter is omitted, the existing
Gen11 behavior remains (half EU array powergating).
This change takes the initiative to move all perf related sseu
configuration into i915_perf.c
v2: Make parameter priviliged if different from default
v3: Fix context modifying its sseu config while i915-perf is enabled
v4: Always consider global sseu a privileged operation (Tvrtko)
Override req_sseu point in intel_sseu_make_rpcs() (Tvrtko)
Remove unrelated changes (Tvrtko)
v5: Some typos (Tvrtko)
Process sseu param in read_properties_unlocked() (Tvrtko)
v6: Actually commit the bits from v5...
Fixup some checkpath warnings
v7: Only compare engine uabi field (Chris)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200317132222.2638719-3-lionel.g.landwerlin@intel.com
This is really just an alias of mmap_gtt. The 'mmap offset' nomenclature
comes from the value returned by this ioctl which is the offset into the
device fd which userpace uses with mmap(2).
mmap_gtt was our initial mmap_offset implementation, this extends
our CPU mmap support to allow additional fault handlers that depends on
the object's backing pages.
Note that we multiplex mmap_gtt and mmap_offset through the same ioctl,
and use the zero extending behaviour of drm to differentiate between
them, when we inspect the flags.
To support multiple mmap types on an object we need to support multiple
mmap_offsets for an object (each offset in the global device address
space corresponding to a unique instance of the object for a file + mmap
type). As we drop the simplified drm core idea of a single mmap_offset,
we need to provide replacement hooks for the dumb mmap interface as
well.
Link: https://gitlab.freedesktop.org/mesa/mesa/merge_requests/1675
Testcase: igt/gem_mmap_offset
Signed-off-by: Abdiel Janulgue <abdiel.janulgue@linux.intel.com>
Signed-off-by: Matthew Auld <matthew.auld@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/20191204120032.3682839-1-chris@chris-wilson.co.uk
Our existing behaviour is to allow contexts and their GPU requests to
persist past the point of closure until the requests are complete. This
allows clients to operate in a 'fire-and-forget' manner where they can
setup a rendering pipeline and hand it over to the display server and
immediately exit. As the rendering pipeline is kept alive until
completion, the display server (or other consumer) can use the results
in the future and present them to the user.
The compute model is a little different. They have little to no buffer
sharing between processes as their kernels tend to operate on a
continuous stream, feeding the results back to the client application.
These kernels operate for an indeterminate length of time, with many
clients wishing that the kernel was always running for as long as they
keep feeding in the data, i.e. acting like a DSP.
Not all clients want this persistent "desktop" behaviour and would prefer
that the contexts are cleaned up immediately upon closure. This ensures
that when clients are run without hangchecking (e.g. for compute kernels
of indeterminate runtime), any GPU hang or other unexpected workloads
are terminated with the process and does not continue to hog resources.
The default behaviour for new contexts is the legacy persistence mode,
as some desktop applications are dependent upon the existing behaviour.
New clients will have to opt in to immediate cleanup on context
closure. If the hangchecking modparam is disabled, so is persistent
context support -- all contexts will be terminated on closure.
We expect this behaviour change to be welcomed by compute users, who
have often been caught between a rock and a hard place. They disable
hangchecking to avoid their kernels being "unfairly" declared hung, but
have also experienced true hangs that the system was then unable to
clean up. Naturally, this leads to bug reports.
Testcase: igt/gem_ctx_persistence
Link: https://github.com/intel/compute-runtime/pull/228
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Jon Bloomfield <jon.bloomfield@intel.com>
Reviewed-by: Jon Bloomfield <jon.bloomfield@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20191029202338.8841-1-chris@chris-wilson.co.uk
We would like to make use of perf in Vulkan. The Vulkan API is much
lower level than OpenGL, with applications directly exposed to the
concept of command buffers (pretty much equivalent to our batch
buffers). In Vulkan, queries are always limited in scope to a command
buffer. In OpenGL, the lack of command buffer concept meant that
queries' duration could span multiple command buffers.
With that restriction gone in Vulkan, we would like to simplify
measuring performance just by measuring the deltas between the counter
snapshots written by 2 MI_RECORD_PERF_COUNT commands, rather than the
more complex scheme we currently have in the GL driver, using 2
MI_RECORD_PERF_COUNT commands and doing some post processing on the
stream of OA reports, coming from the global OA buffer, to remove any
unrelated deltas in between the 2 MI_RECORD_PERF_COUNT.
Disabling preemption only apply to a single context with which want to
query performance counters for and is considered a privileged
operation, by default protected by CAP_SYS_ADMIN. It is possible to
enable it for a normal user by disabling the paranoid stream setting.
v2: Store preemption setting in intel_context (Chris)
v3: Use priorities to avoid preemption rather than the HW mechanism
v4: Just modify the port priority reporting function
v5: Add nopreempt flag on gem context and always flag requests
appropriately, regarless of OA reconfiguration.
Link: https://gitlab.freedesktop.org/mesa/mesa/merge_requests/932
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@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/20191014201404.22468-4-chris@chris-wilson.co.uk
Introduce a new perf_ioctl command to change the OA configuration of the
active stream. This allows the OA stream to be reconfigured between
batch buffers, giving greater flexibility in sampling. We inject a
request into the OA context to reconfigure the stream asynchronously on
the GPU in between and ordered with execbuffer calls.
Original patch for dynamic reconfiguration by Lionel Landwerlin.
Link: https://gitlab.freedesktop.org/mesa/mesa/merge_requests/932
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191014201404.22468-3-chris@chris-wilson.co.uk
Listing configurations at the moment is supported only through sysfs.
This might cause issues for applications wanting to list
configurations from a container where sysfs isn't available.
This change adds a way to query the number of configurations and their
content through the i915 query uAPI.
v2: Fix sparse warnings (Lionel)
Add support to query configuration using uuid (Lionel)
v3: Fix some inconsistency in uapi header (Lionel)
Fix unlocking when not locked issue (Lionel)
Add debug messages (Lionel)
v4: Fix missing unlock (Dan)
v5: Drop lock when copying config content to userspace (Chris)
v6: Drop lock when copying config list to userspace (Chris)
Fix deadlock when calling i915_perf_get_oa_config() under
perf.metrics_lock (Lionel)
Add i915_oa_config_get() (Chris)
Link: https://gitlab.freedesktop.org/mesa/mesa/merge_requests/932
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@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/20191014201404.22468-2-chris@chris-wilson.co.uk
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@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/20191014201404.22468-1-chris@chris-wilson.co.uk
Gen12 has dual-subslices (DSS), which compared to gen11 subslices have
some duplicated resources/paths. Although DSS behave similarly to 2
subslices, instead of splitting this and presenting userspace with bits
not directly representative of hardware resources, present userspace
with a subslice_mask made up of DSS bits instead.
v2: GEM_BUG_ON on mask size (Lionel)
Bspec: 29547
Bspec: 12247
Cc: Kelvin Gardiner <kelvin.gardiner@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
CC: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com> #v1
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Signed-off-by: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Sudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: Stuart Summers <stuart.summers@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190913075137.18476-2-chris@chris-wilson.co.uk
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Expose whether or not we support the PMU software tracking in our
scheduler capabilities, so userspace can query at runtime.
v2: Use I915_SCHEDULER_CAP_ENGINE_BUSY_STATS for a less ambiguous
capability name.
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/20190703143702.11339-1-chris@chris-wilson.co.uk
Engine discovery query allows userspace to enumerate engines, probe their
configuration features, all without needing to maintain the internal PCI
ID based database.
A new query for the generic i915 query ioctl is added named
DRM_I915_QUERY_ENGINE_INFO, together with accompanying structure
drm_i915_query_engine_info. The address of latter should be passed to the
kernel in the query.data_ptr field, and should be large enough for the
kernel to fill out all known engines as struct drm_i915_engine_info
elements trailing the query.
As with other queries, setting the item query length to zero allows
userspace to query minimum required buffer size.
Enumerated engines have common type mask which can be used to query all
hardware engines, versus engines userspace can submit to using the execbuf
uAPI.
Engines also have capabilities which are per engine class namespace of
bits describing features not present on all engine instances.
v2:
* Fixed HEVC assignment.
* Reorder some fields, rename type to flags, increase width. (Lionel)
* No need to allocate temporary storage if we do it engine by engine.
(Lionel)
v3:
* Describe engine flags and mark mbz fields. (Lionel)
* HEVC only applies to VCS.
v4:
* Squash SFC flag into main patch.
* Tidy some comments.
v5:
* Add uabi_ prefix to engine capabilities. (Chris Wilson)
* Report exact size of engine info array. (Chris Wilson)
* Drop the engine flags. (Joonas Lahtinen)
* Added some more reserved fields.
* Move flags after class/instance.
v6:
* Do not check engine info array was zeroed by userspace but zero the
unused fields for them instead.
v7:
* Simplify length calculation loop. (Lionel Landwerlin)
v8:
* Remove MBZ comments where not applicable.
* Rename ABI flags to match engine class define naming.
* Rename SFC ABI flag to reflect it applies to VCS and VECS.
* SFC is wired to even _logical_ engine instances.
* SFC applies to VCS and VECS.
* HEVC is present on all instances on Gen11. (Tony)
* Simplify length calculation even more. (Chris Wilson)
* Move info_ptr assigment closer to loop for clarity. (Chris Wilson)
* Use vdbox_sfc_access from runtime info.
* Rebase for RUNTIME_INFO.
* Refactor for lower indentation.
* Rename uAPI class/instance to engine_class/instance to avoid C++
keyword.
v9:
* Rebase for s/num_rings/num_engines/ in RUNTIME_INFO.
v10:
* Use new copy_query_item.
v11:
* Consolidate with struct i915_engine_class_instnace.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Jon Bloomfield <jon.bloomfield@intel.com>
Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tony Ye <tony.ye@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> # v7
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20190522090054.6007-1-tvrtko.ursulin@linux.intel.com
There is a desire to split a task onto two engines and have them run at
the same time, e.g. scanline interleaving to spread the workload evenly.
Through the use of the out-fence from the first execbuf, we can
coordinate secondary execbuf to only become ready simultaneously with
the first, so that with all things idle the second execbufs are executed
in parallel with the first. The key difference here between the new
EXEC_FENCE_SUBMIT and the existing EXEC_FENCE_IN is that the in-fence
waits for the completion of the first request (so that all of its
rendering results are visible to the second execbuf, the more common
userspace fence requirement).
Since we only have a single input fence slot, userspace cannot mix an
in-fence and a submit-fence. It has to use one or the other! This is not
such a harsh requirement, since by virtue of the submit-fence, the
secondary execbuf inherit all of the dependencies from the first
request, and for the application the dependencies should be common
between the primary and secondary execbuf.
Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Testcase: igt/gem_exec_fence/parallel
Link: https://github.com/intel/media-driver/pull/546
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/20190521211134.16117-10-chris@chris-wilson.co.uk
Some users require that when a master batch is executed on one particular
engine, a companion batch is run simultaneously on a specific slave
engine. For this purpose, we introduce virtual engine bonding, allowing
maps of master:slaves to be constructed to constrain which physical
engines a virtual engine may select given a fence on a master engine.
For the moment, we continue to ignore the issue of preemption deferring
the master request for later. Ideally, we would like to then also remove
the slave and run something else rather than have it stall the pipeline.
With load balancing, we should be able to move workload around it, but
there is a similar stall on the master pipeline while it may wait for
the slave to be executed. At the cost of more latency for the bonded
request, it may be interesting to launch both on their engines in
lockstep. (Bubbles abound.)
Opens: Also what about bonding an engine as its own master? It doesn't
break anything internally, so allow the silliness.
v2: Emancipate the bonds
v3: Couple in delayed scheduling for the selftests
v4: Handle invalid mutually exclusive bonding
v5: Mention what the uapi does
v6: s/nbond/num_bonds/
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/20190521211134.16117-9-chris@chris-wilson.co.uk
Having allowed the user to define a set of engines that they will want
to only use, we go one step further and allow them to bind those engines
into a single virtual instance. Submitting a batch to the virtual engine
will then forward it to any one of the set in a manner as best to
distribute load. The virtual engine has a single timeline across all
engines (it operates as a single queue), so it is not able to concurrently
run batches across multiple engines by itself; that is left up to the user
to submit multiple concurrent batches to multiple queues. Multiple users
will be load balanced across the system.
The mechanism used for load balancing in this patch is a late greedy
balancer. When a request is ready for execution, it is added to each
engine's queue, and when an engine is ready for its next request it
claims it from the virtual engine. The first engine to do so, wins, i.e.
the request is executed at the earliest opportunity (idle moment) in the
system.
As not all HW is created equal, the user is still able to skip the
virtual engine and execute the batch on a specific engine, all within the
same queue. It will then be executed in order on the correct engine,
with execution on other virtual engines being moved away due to the load
detection.
A couple of areas for potential improvement left!
- The virtual engine always take priority over equal-priority tasks.
Mostly broken up by applying FQ_CODEL rules for prioritising new clients,
and hopefully the virtual and real engines are not then congested (i.e.
all work is via virtual engines, or all work is to the real engine).
- We require the breadcrumb irq around every virtual engine request. For
normal engines, we eliminate the need for the slow round trip via
interrupt by using the submit fence and queueing in order. For virtual
engines, we have to allow any job to transfer to a new ring, and cannot
coalesce the submissions, so require the completion fence instead,
forcing the persistent use of interrupts.
- We only drip feed single requests through each virtual engine and onto
the physical engines, even if there was enough work to fill all ELSP,
leaving small stalls with an idle CS event at the end of every request.
Could we be greedy and fill both slots? Being lazy is virtuous for load
distribution on less-than-full workloads though.
Other areas of improvement are more general, such as reducing lock
contention, reducing dispatch overhead, looking at direct submission
rather than bouncing around tasklets etc.
sseu: Lift the restriction to allow sseu to be reconfigured on virtual
engines composed of RENDER_CLASS (rcs).
v2: macroize check_user_mbz()
v3: Cancel virtual engines on wedging
v4: Commence commenting
v5: Replace 64b sibling_mask with a list of class:instance
v6: Drop the one-element array in the uabi
v7: Assert it is an virtual engine in to_virtual_engine()
v8: Skip over holes in [class][inst] so we can selftest with (vcs0, vcs2)
Link: https://github.com/intel/media-driver/pull/283
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/20190521211134.16117-6-chris@chris-wilson.co.uk
A usecase arose out of handling context recovery in mesa, whereby they
wish to recreate a context with fresh logical state but preserving all
other details of the original. Currently, they create a new context and
iterate over which bits they want to copy across, but it would much more
convenient if they were able to just pass in a target context to clone
during creation. This essentially extends the setparam during creation
to pull the details from a target context instead of the user supplied
parameters.
The ideal here is that we don't expose control over anything more than
can be obtained via CONTEXT_PARAM. That is userspace retains explicit
control over all features, and this api is just convenience.
For example, you could replace
struct context_param p = { .param = CONTEXT_PARAM_VM };
param.ctx_id = old_id;
gem_context_get_param(&p.param);
new_id = gem_context_create();
param.ctx_id = new_id;
gem_context_set_param(&p.param);
gem_vm_destroy(param.value); /* drop the ref to VM_ID handle */
with
struct create_ext_param p = {
{ .name = CONTEXT_CREATE_CLONE },
.clone_id = old_id,
.flags = CLONE_FLAGS_VM
}
new_id = gem_context_create_ext(&p);
and not have to worry about stray namespace pollution etc.
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/20190521211134.16117-5-chris@chris-wilson.co.uk
The SINGLE_TIMELINE flag can be used to create a context such that all
engine instances within that context share a common timeline. This can
be useful for mixing operations between real and virtual engines, or
when using a composite context for a single client API context.
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/20190521211134.16117-4-chris@chris-wilson.co.uk
Allow the user to specify a local engine index (as opposed to
class:index) that they can use to refer to a preset engine inside the
ctx->engine[] array defined by an earlier I915_CONTEXT_PARAM_ENGINES.
This will be useful for setting SSEU parameters on virtual engines that
are local to the context and do not have a valid global class:instance
lookup.
Note that due to the ambiguity in using class:instance with
ctx->engines[], if a user supplied engine map is active the user must
specify the engine to alter by its index into the ctx->engines[].
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/20190521211134.16117-3-chris@chris-wilson.co.uk
Over the last few years, we have debated how to extend the user API to
support an increase in the number of engines, that may be sparse and
even be heterogeneous within a class (not all video decoders created
equal). We settled on using (class, instance) tuples to identify a
specific engine, with an API for the user to construct a map of engines
to capabilities. Into this picture, we then add a challenge of virtual
engines; one user engine that maps behind the scenes to any number of
physical engines. To keep it general, we want the user to have full
control over that mapping. To that end, we allow the user to constrain a
context to define the set of engines that it can access, order fully
controlled by the user via (class, instance). With such precise control
in context setup, we can continue to use the existing execbuf uABI of
specifying a single index; only now it doesn't automagically map onto
the engines, it uses the user defined engine map from the context.
v2: Fixup freeing of local on success of get_engines()
v3: Allow empty engines[]
v4: s/nengine/num_engines/
v5: Replace 64 limit on num_engines with a note that execbuf is
currently limited to only using the first 64 engines.
v6: Actually use the engines_mutex to guard the ctx->engines.
Testcase: igt/gem_ctx_engines
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/20190521211134.16117-2-chris@chris-wilson.co.uk
Having hid the partially exposed new ABI from the PR, put it back again
for completion of context recovery. A significant part of context
recovery is the ability to reuse as much of the old context as is
feasible (to avoid expensive reconstruction). The biggest chunk kept
hidden at the moment is fine-control over the ctx->ppgtt (the GPU page
tables and associated translation tables and kernel maps), so make
control over the ctx->ppgtt explicit.
This allows userspace to create and share virtual memory address spaces
(within the limits of a single fd) between contexts they own, along with
the ability to query the contexts for the vm state.
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/20190521211134.16117-1-chris@chris-wilson.co.uk
SSEU reprogramming of the context introduced the notion of engine class
and instance for a forwards compatible method of describing any engine
beyond the old execbuf interface. We wish to adopt this class:instance
description for more interfaces, so pull it out into a separate type for
userspace convenience.
Fixes: e46c2e99f6 ("drm/i915: Expose RPCS (SSEU) configuration to userspace (Gen11 only)")
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>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Tony Ye <tony.ye@intel.com>
Cc: Andi Shyti <andi@etezian.org>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Acked-by: Tony Ye <tony.ye@intel.com>
Reviewed-by: Andi Shyti <andi@etezian.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20190412071416.30097-1-chris@chris-wilson.co.uk
The intent was to expose these as part of the means to perform full
context recovery (though not the SINGLE_TIMELINE, that is for later and
just sucked as collateral damage). As that requires a couple more
patches to complete the series, roll back the earlier chunks of ABI for
an intervening PR. We keep all the internals intact and under selftests.
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/20190327105814.14694-1-chris@chris-wilson.co.uk
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Previously, our view has been always to run the engines independently
within a context. (Multiple engines happened before we had contexts and
timelines, so they always operated independently and that behaviour
persisted into contexts.) However, at the user level the context often
represents a single timeline (e.g. GL contexts) and userspace must
ensure that the individual engines are serialised to present that
ordering to the client (or forgot about this detail entirely and hope no
one notices - a fair ploy if the client can only directly control one
engine themselves ;)
In the next patch, we will want to construct a set of engines that
operate as one, that have a single timeline interwoven between them, to
present a single virtual engine to the user. (They submit to the virtual
engine, then we decide which engine to execute on based.)
To that end, we want to be able to create contexts which have a single
timeline (fence context) shared between all engines, rather than multiple
timelines.
v2: Move the specialised timeline ordering to its own function.
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/20190322092325.5883-4-chris@chris-wilson.co.uk
It can be useful to have a single ioctl to create a context with all
the initial parameters instead of a series of create + setparam + setparam
ioctls. This extension to create context allows any of the parameters
to be passed in as a linked list to be applied to the newly constructed
context.
v2: Make a local copy of user setparam (Tvrtko)
v3: Use flags to detect availability of extension interface
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/20190322092325.5883-3-chris@chris-wilson.co.uk
In preparation to making the ppGTT binding for a context explicit (to
facilitate reusing the same ppGTT between different contexts), allow the
user to create and destroy named ppGTT.
v2: Replace global barrier for swapping over the ppgtt and tlbs with a
local context barrier (Tvrtko)
v3: serialise with struct_mutex; it's lazy but required dammit
v4: Rewrite igt_ctx_shared_exec to be more different (aimed to be more
similarly, turned out different!)
v5: Fix up test unwind for aliasing-ppgtt (snb)
v6: Tighten language for uapi struct drm_i915_gem_vm_control.
v7: Patch the context image for runtime ppgtt switching!
Testcase: igt/gem_vm_create
Testcase: igt/gem_ctx_param/vm
Testcase: igt/gem_ctx_clone/vm
Testcase: igt/gem_ctx_shared
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/20190322092325.5883-2-chris@chris-wilson.co.uk
An idea for extending uABI inspired by Vulkan's extension chains.
Instead of expanding the data struct for each ioctl every time we need
to add a new feature, define an extension chain instead. As we add
optional interfaces to control the ioctl, we define a new extension
struct that can be linked into the ioctl data only when required by the
user. The key advantage being able to ignore large control structs for
optional interfaces/extensions, while being able to process them in a
consistent manner.
In comparison to other extensible ioctls, the key difference is the
use of a linked chain of extension structs vs an array of tagged
pointers. For example,
struct drm_amdgpu_cs_chunk {
__u32 chunk_id;
__u32 length_dw;
__u64 chunk_data;
};
struct drm_amdgpu_cs_in {
__u32 ctx_id;
__u32 bo_list_handle;
__u32 num_chunks;
__u32 _pad;
__u64 chunks;
};
allows userspace to pass in array of pointers to extension structs, but
must therefore keep constructing that array along side the command stream.
In dynamic situations like that, a linked list is preferred and does not
similar from extra cache line misses as the extension structs themselves
must still be loaded separate to the chunks array.
v2: Apply the tail call optimisation directly to nip the worry of stack
overflow in the bud.
v3: Defend against recursion.
v4: Fixup local types to match new uabi
Opens:
- do we include the result as an out-field in each chain?
struct i915_user_extension {
__u64 next_extension;
__u64 name;
__s32 result;
__u32 mbz; /* reserved for future use */
};
* Undecided, so provision some room for future expansion.
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>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190322092325.5883-1-chris@chris-wilson.co.uk
As we allow per-context engine allows the legacy concept of
I915_EXEC_RING no longer applies universally. We are still exposing the
unrelated exec-id in GEM_BUSY, so transition this ioctl (once more
slightly changing its ABI, but no one cares) over to only reporting the
uabi-class (not instance as we can not foreseeably fit those into the
small bitmask).
The only user of the extended ring information from GEM_BUSY is ddx/sna,
which tries to use the non-rcs business information to guide which
engine to use for subsequent operations on foreign bo. All that matters
for it is the decision between rcs and !rcs, so it is unaffected by the
change in higher bits.
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/20190305162643.20243-1-chris@chris-wilson.co.uk
This was supposed to be a mask of all known rings, but it is being used
by execbuffer to filter out invalid rings, and so is instead mapping high
unused values onto valid rings. Instead of a mask of all known rings,
we need it to be the mask of all possible rings.
Fixes: 549f736582 ("drm/i915: Enable SandyBridge blitter ring")
Fixes: de1add3605 ("drm/i915: Decouple execbuf uAPI from internal implementation")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v4.6+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190301140404.26690-21-chris@chris-wilson.co.uk
Having introduced per-context seqno, we now have a means to identity
progress across the system without feel of rollback as befell the
global_seqno. That is we can program a MI_SEMAPHORE_WAIT operation in
advance of submission safe in the knowledge that our target seqno and
address is stable.
However, since we are telling the GPU to busy-spin on the target address
until it matches the signaling seqno, we only want to do so when we are
sure that busy-spin will be completed quickly. To achieve this we only
submit the request to HW once the signaler is itself executing (modulo
preemption causing us to wait longer), and we only do so for default and
above priority requests (so that idle priority tasks never themselves
hog the GPU waiting for others).
As might be reasonably expected, HW semaphores excel in inter-engine
synchronisation microbenchmarks (where the 3x reduced latency / increased
throughput more than offset the power cost of spinning on a second ring)
and have significant improvement (can be up to ~10%, most see no change)
for single clients that utilize multiple engines (typically media players
and transcoders), without regressing multiple clients that can saturate
the system or changing the power envelope dramatically.
v3: Drop the older NEQ branch, now we pin the signaler's HWSP anyway.
v4: Tell the world and include it as part of scheduler caps.
Testcase: igt/gem_exec_whisper
Testcase: igt/benchmarks/gem_wsim
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/20190301170901.8340-3-chris@chris-wilson.co.uk
We don't want to pre-reserve any holes in our uAPI for that is a sign of
nefarious and hidden activity. Add a reminder about our uAPI
expectations to encourage good practice when adding new defines/enums.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190218094628.13522-1-chris@chris-wilson.co.uk
Some clients, such as mesa, may only emit minimal incremental batches
that rely on the logical context state from previous batches. They know
that recovery is impossible after a hang as their required GPU state is
lost, and that each in flight and subsequent batch will hang (resetting
the context image back to default perpetuating the problem).
To avoid getting into the state in the first place, we can allow clients
to opt out of automatic recovery and elect to ban any guilty context
following a hang. This prevents the continual stream of hangs and allows
the client to recreate their context and rebuild the state from scratch.
v2: Prefer calling it recoverable rather than unrecoverable.
References: https://lists.freedesktop.org/archives/mesa-dev/2019-February/215431.html
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org> # for mesa
Link: https://patchwork.freedesktop.org/patch/msgid/20190218105821.17293-1-chris@chris-wilson.co.uk
We want to allow userspace to reconfigure the subslice configuration on a
per context basis.
This is required for the functional requirement of shutting down non-VME
enabled sub-slices on Gen11 parts.
To do so, we expose a context parameter to allow adjustment of the RPCS
register stored within the context image (and currently not accessible via
LRI).
If the context is adjusted before first use or whilst idle, the adjustment
is for "free"; otherwise if the context is active we queue a request to do
so (using the kernel context), following all other activity by that
context, which is also marked as barrier for all following submission
against the same context.
Since the overhead of device re-configuration during context switching can
be significant, especially in multi-context workloads, we limit this new
uAPI to only support the Gen11 VME use case. In this use case either the
device is fully enabled, and exactly one slice and half of the subslices
are enabled.
Example usage:
struct drm_i915_gem_context_param_sseu sseu = { };
struct drm_i915_gem_context_param arg = {
.param = I915_CONTEXT_PARAM_SSEU,
.ctx_id = gem_context_create(fd),
.size = sizeof(sseu),
.value = to_user_pointer(&sseu)
};
/* Query device defaults. */
gem_context_get_param(fd, &arg);
/* Set VME configuration on a 1x6x8 part. */
sseu.slice_mask = 0x1;
sseu.subslice_mask = 0xe0;
gem_context_set_param(fd, &arg);
v2: Fix offset of CTX_R_PWR_CLK_STATE in intel_lr_context_set_sseu()
(Lionel)
v3: Add ability to program this per engine (Chris)
v4: Move most get_sseu() into i915_gem_context.c (Lionel)
v5: Validate sseu configuration against the device's capabilities (Lionel)
v6: Change context powergating settings through MI_SDM on kernel context
(Chris)
v7: Synchronize the requests following a powergating setting change using
a global dependency (Chris)
Iterate timelines through dev_priv.gt.active_rings (Tvrtko)
Disable RPCS configuration setting for non capable users
(Lionel/Tvrtko)
v8: s/union intel_sseu/struct intel_sseu/ (Lionel)
s/dev_priv/i915/ (Tvrtko)
Change uapi class/instance fields to u16 (Tvrtko)
Bump mask fields to 64bits (Lionel)
Don't return EPERM when dynamic sseu is disabled (Tvrtko)
v9: Import context image into kernel context's ppgtt only when
reconfiguring powergated slice/subslices (Chris)
Use aliasing ppgtt when needed (Michel)
Tvrtko Ursulin:
v10:
* Update for upstream changes.
* Request submit needs a RPM reference.
* Reject on !FULL_PPGTT for simplicity.
* Pull out get/set param to helpers for readability and less indent.
* Use i915_request_await_dma_fence in add_global_barrier to skip waits
on the same timeline and avoid GEM_BUG_ON.
* No need to explicitly assign a NULL pointer to engine in legacy mode.
* No need to move gen8_make_rpcs up.
* Factored out global barrier as prep patch.
* Allow to only CAP_SYS_ADMIN if !Gen11.
v11:
* Remove engine vfunc in favour of local helper. (Chris Wilson)
* Stop retiring requests before updates since it is not needed
(Chris Wilson)
* Implement direct CPU update path for idle contexts. (Chris Wilson)
* Left side dependency needs only be on the same context timeline.
(Chris Wilson)
* It is sufficient to order the timeline. (Chris Wilson)
* Reject !RCS configuration attempts with -ENODEV for now.
v12:
* Rebase for make_rpcs.
v13:
* Centralize SSEU normalization to make_rpcs.
* Type width checking (uAPI <-> implementation).
* Gen11 restrictions uAPI checks.
* Gen11 subslice count differences handling.
Chris Wilson:
* args->size handling fixes.
* Update context image from GGTT.
* Postpone context image update to pinning.
* Use i915_gem_active_raw instead of last_request_on_engine.
v14:
* Add activity tracker on intel_context to fix the lifetime issues
and simplify the code. (Chris Wilson)
v15:
* Fix context pin leak if no space in ring by simplifying the
context pinning sequence.
v16:
* Rebase for context get/set param locking changes.
* Just -ENODEV on !Gen11. (Joonas)
v17:
* Fix one Gen11 subslice enablement rule.
* Handle error from i915_sw_fence_await_sw_fence_gfp. (Chris Wilson)
v18:
* Update commit message. (Joonas)
* Restrict uAPI to VME use case. (Joonas)
v19:
* Rebase.
v20:
* Rebase for ce->active_tracker.
v21:
* Rebase for IS_GEN changes.
v22:
* Reserve uAPI for flags straight away. (Chris Wilson)
v23:
* Rebase for RUNTIME_INFO.
v24:
* Added some headline docs for the uapi usage. (Joonas/Chris)
v25:
* Renamed class/instance to engine_class/engine_instance to avoid clash
with C++ keyword. (Tony Ye)
v26:
* Rebased for runtime pm api changes.
v27:
* Rebased for intel_context_init.
* Wrap commit msg to 75.
v28:
(Chris Wilson)
* Use i915_gem_ggtt.
* Use i915_request_await_dma_fence to show a better example.
v29:
* i915_timeline_set_barrier can now fail. (Chris Wilson)
v30:
* Capture some acks.
v31:
* Drop the WARN_ON from use controllable paths. (Chris Wilson)
* Use overflows_type for all checks.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100899
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107634
Issue: https://github.com/intel/media-driver/issues/267
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Zhipeng Gong <zhipeng.gong@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tony Ye <tony.ye@intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: Timo Aaltonen <timo.aaltonen@canonical.com>
Acked-by: Takashi Iwai <tiwai@suse.de>
Acked-by: Stéphane Marchesin <marcheu@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20190205095032.22673-4-tvrtko.ursulin@linux.intel.com
The way our hardware is designed doesn't seem to let us use the
MI_RECORD_PERF_COUNT command without setting up a circular buffer.
In the case where the user didn't request OA reports to be available
through the i915 perf stream, we can set the OA buffer to the minimum
size to avoid consuming memory which won't be used by the driver.
v2: Simplify oa buffer size exponent selection (Chris)
Reuse vma size field (Lionel)
v3: Restrict size opening parameter to values supported by HW (Chris)
v4: Drop out of date comment (Matt)
Add debug message when buffer size is rejected (Matt)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20181023100707.31738-5-lionel.g.landwerlin@intel.com
Now that we are confident in providing full-ppgtt where supported,
remove the ability to override the context isolation.
v2: Remove faked aliasing-ppgtt for testing as it no longer is accepted.
v3: s/USES/HAS/ to match usage and reject attempts to load the module on
old GVT-g setups that do not provide support for full-ppgtt.
v4: Insulate ABI ppGTT values from our internal enum (later plans
involve moving ppGTT depth out of the enum, thus potentially breaking
ABI unless we document the current values).
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>
Acked-by: Zhi Wang <zhi.a.wang@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180926201222.5643-1-chris@chris-wilson.co.uk
Not all chipsets have an internal buffer delaying the visibility of
writes via the GGTT being visible by other physical paths, but we use a
very heavy workaround for all. We only need to apply that workarounds to
the chipsets we know suffer from the delay and the resulting coherency
issue.
Similarly, the same inconsistent coherency fouls up our ABI promise that
a write into a mmap_gtt is immediately visible to others. Since the HW
has made that a lie, let userspace know when that contract is broken.
(Not that userspace would want to use mmap_gtt on those chipsets for
other performance reasons...)
Testcase: igt/drv_selftest/live_coherency
Testcase: igt/gem_mmap_gtt/coherency
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100587
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Tomasz Lis <tomasz.lis@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180720101910.11153-1-chris@chris-wilson.co.uk
With the introduction of asymmetric slices in CNL, we cannot rely on
the previous SUBSLICE_MASK getparam to tell userspace what subslices
are available. Here we introduce a more detailed way of querying the
Gen's GPU topology that doesn't aggregate numbers.
This is essential for monitoring parts of the GPU with the OA unit,
because counters need to be normalized to the number of
EUs/subslices/slices. The current aggregated numbers like EU_TOTAL do
not gives us sufficient information.
The Mesa series making use of this API is :
https://patchwork.freedesktop.org/series/38795/
As a bonus we can draw representations of the GPU :
https://imgur.com/a/vuqpa
v2: Rename uapi struct s/_mask/_info/ (Tvrtko)
Report max_slice/subslice/eus_per_subslice rather than strides (Tvrtko)
Add uapi macros to read data from *_info structs (Tvrtko)
v3: Use !!(v & DRM_I915_BIT()) for uapi macros instead of custom shifts (Tvrtko)
v4: factorize query item writting (Tvrtko)
tweak uapi struct/define names (Tvrtko)
v5: Replace ALIGN() macro (Chris)
v6: Updated uapi comments (Tvrtko)
Moved flags != 0 checks into vfuncs (Tvrtko)
v7: Use access_ok() before copying anything, to avoid overflows (Chris)
Switch BUG_ON() to GEM_WARN_ON() (Tvrtko)
v8: Tweak uapi comments style to match the coding style (Lionel)
v9: Fix error in comment about computation of enabled subslice (Tvrtko)
v10: Fix/update comments in uAPI (Sagar)
v11: Drop drm_i915_query_(slice|subslice|eu)_info in favor of a single
drm_i915_query_topology_info (Joonas)
v12: Add subslice_stride/eu_stride in drm_i915_query_topology_info (Joonas)
v13: Fix comment in uAPI (Joonas)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180306122857.27317-7-lionel.g.landwerlin@intel.com
There are a number of information that are readable from hardware
registers and that we would like to make accessible to userspace. One
particular example is the topology of the execution units (how are
execution units grouped in subslices and slices and also which ones
have been fused off for die recovery).
At the moment the GET_PARAM ioctl covers some basic needs, but
generally is only able to return a single value for each defined
parameter. This is a bit problematic with topology descriptions which
are array/maps of available units.
This change introduces a new ioctl that can deal with requests to fill
structures of potentially variable lengths. The user is expected fill
a query with length fields set at 0 on the first call, the kernel then
sets the length fields to the their expected values. A second call to
the kernel with length fields at their expected values will trigger a
copy of the data to the pointed memory locations.
The scope of this uAPI is only to provide information to userspace,
not to allow configuration of the device.
v2: Simplify dispatcher code iteration (Tvrtko)
Tweak uapi drm_i915_query_item structure (Tvrtko)
v3: Rename pad fields into flags (Chris)
Return error on flags field != 0 (Chris)
Only copy length back to userspace in drm_i915_query_item (Chris)
v4: Use array of functions instead of switch (Chris)
v5: More comments in uapi (Tvrtko)
Return query item errors in length field (All)
v6: Tweak uapi comments style to match the coding style (Lionel)
v7: Add i915_query.h (Joonas)
v8: (Lionel) Change the behavior of the item iterator to report
invalid queries into the query item rather than stopping the
iteration. This enables userspace applications to query newer
items on older kernels and only have failure on the items that are
not supported.
v9: Edit copyright headers (Joonas)
v10: Typos & comments in uapi (Joonas)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180306122857.27317-6-lionel.g.landwerlin@intel.com
Deprecate the silly I915_SET_COLORKEY_NONE flag. The obvious
way to disable colorkey is to just set flags to 0, which is
exactly what the intel ddx has been doing all along.
Currently when userspace sets the flags to 0, we end up in a
funny state where colorkey is disabled, but various colorkey
vs. scaling checks still consider colorkey to be enabled, and
thus we don't allow plane scaling to kick in.
In case there is some other userspace out there that actually
uses this flag (unlikely as this is an i915 specific uapi)
we'll keep on accepting it.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180202204231.27905-1-ville.syrjala@linux.intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris has discovered that RC6, RC6p and RC6pp counters are mutually
exclusive, and even that on some SNB SKUs you get RC6p increasing, and on
the others RC6.
Furthermore RC6p and RC6pp were only present starting from GEN6 until,
GEN7, not including Haswell.
All this combined makes it questionable whether we need to reserve new ABI
for these counters. One idea was to just combine them all under the RC6
counter to simplify things for userspace. So that is what this patch does.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Suggested-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/20171124171331.17981-1-tvrtko.ursulin@linux.intel.com
We have agreed during the engine classes discussion that fields marked as
non-ABI are better left out altogether from uapi headers.
v2: Use a local define for maintanability. (Chris Wilson)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20171123100701.18430-1-tvrtko.ursulin@linux.intel.com
From: Chris Wilson <chris@chris-wilson.co.uk>
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
From: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
The first goal is to be able to measure GPU (and invidual ring) busyness
without having to poll registers from userspace. (Which not only incurs
holding the forcewake lock indefinitely, perturbing the system, but also
runs the risk of hanging the machine.) As an alternative we can use the
perf event counter interface to sample the ring registers periodically
and send those results to userspace.
Functionality we are exporting to userspace is via the existing perf PMU
API and can be exercised via the existing tools. For example:
perf stat -a -e i915/rcs0-busy/ -I 1000
Will print the render engine busynnes once per second. All the performance
counters can be enumerated (perf list) and have their unit of measure
correctly reported in sysfs.
v1-v2 (Chris Wilson):
v2: Use a common timer for the ring sampling.
v3: (Tvrtko Ursulin)
* Decouple uAPI from i915 engine ids.
* Complete uAPI defines.
* Refactor some code to helpers for clarity.
* Skip sampling disabled engines.
* Expose counters in sysfs.
* Pass in fake regs to avoid null ptr deref in perf core.
* Convert to class/instance uAPI.
* Use shared driver code for rc6 residency, power and frequency.
v4: (Dmitry Rogozhkin)
* Register PMU with .task_ctx_nr=perf_invalid_context
* Expose cpumask for the PMU with the single CPU in the mask
* Properly support pmu->stop(): it should call pmu->read()
* Properly support pmu->del(): it should call stop(event, PERF_EF_UPDATE)
* Introduce refcounting of event subscriptions.
* Make pmu.busy_stats a refcounter to avoid busy stats going away
with some deleted event.
* Expose cpumask for i915 PMU to avoid multiple events creation of
the same type followed by counter aggregation by perf-stat.
* Track CPUs getting online/offline to migrate perf context. If (likely)
cpumask will initially set CPU0, CONFIG_BOOTPARAM_HOTPLUG_CPU0 will be
needed to see effect of CPU status tracking.
* End result is that only global events are supported and perf stat
works correctly.
* Deny perf driver level sampling - it is prohibited for uncore PMU.
v5: (Tvrtko Ursulin)
* Don't hardcode number of engine samplers.
* Rewrite event ref-counting for correctness and simplicity.
* Store initial counter value when starting already enabled events
to correctly report values to all listeners.
* Fix RC6 residency readout.
* Comments, GPL header.
v6:
* Add missing entry to v4 changelog.
* Fix accounting in CPU hotplug case by copying the approach from
arch/x86/events/intel/cstate.c. (Dmitry Rogozhkin)
v7:
* Log failure message only on failure.
* Remove CPU hotplug notification state on unregister.
v8:
* Fix error unwind on failed registration.
* Checkpatch cleanup.
v9:
* Drop the energy metric, it is available via intel_rapl_perf.
(Ville Syrjälä)
* Use HAS_RC6(p). (Chris Wilson)
* Handle unsupported non-engine events. (Dmitry Rogozhkin)
* Rebase for intel_rc6_residency_ns needing caller managed
runtime pm.
* Drop HAS_RC6 checks from the read callback since creating those
events will be rejected at init time already.
* Add counter units to sysfs so perf stat output is nicer.
* Cleanup the attribute tables for brevity and readability.
v10:
* Fixed queued accounting.
v11:
* Move intel_engine_lookup_user to intel_engine_cs.c
* Commit update. (Joonas Lahtinen)
v12:
* More accurate sampling. (Chris Wilson)
* Store and report frequency in MHz for better usability from
perf stat.
* Removed metrics: queued, interrupts, rc6 counters.
* Sample engine busyness based on seqno difference only
for less MMIO (and forcewake) on all platforms. (Chris Wilson)
v13:
* Comment spelling, use mul_u32_u32 to work around potential GCC
issue and somne code alignment changes. (Chris Wilson)
v14:
* Rebase.
v15:
* Rebase for RPS refactoring.
v16:
* Use the dynamic slot in the CPU hotplug state machine so that we are
free to setup our state as multi-instance. Previously we were re-using
the CPUHP_AP_PERF_X86_UNCORE_ONLINE slot which is neither used as
multi-instance, nor owned by our driver to start with.
* Register the CPU hotplug handlers after the PMU, otherwise the callback
will get called before the PMU is initialized which can end up in
perf_pmu_migrate_context with an un-initialized base.
* Added workaround for a probable bug in cpuhp core.
v17:
* Remove workaround for the cpuhp bug.
v18:
* Rebase for drm_i915_gem_engine_class getting upstream before us.
v19:
* Rebase. (trivial)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171121181852.16128-2-tvrtko.ursulin@linux.intel.com
We use to have this fixed per generation, but starting with CNL userspace
cannot tell just off the PCI ID. Let's make this information available. This
is particularly useful for performance monitoring where much of the
normalization work is done using those timestamps (this include pipeline
statistics in both GL & Vulkan as well as OA reports).
v2: Use variables for 24MHz/19.2MHz values (Ewelina)
Renamed function & coding style (Sagar)
v3: Fix frequency read on Broadwell (Sagar)
Fix missing divide by 4 on <= gen4 (Sagar)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Tested-by: Rafael Antognolli <rafael.antognolli@intel.com>
Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110190845.32574-7-lionel.g.landwerlin@intel.com
