There's no reason to entirely wedge the gpu, for the minimal deadlock
bugfix we only need to unbreak/decouple the atomic commit from the gpu
reset. The simplest way to fix that is by replacing the
unconditional fence wait a the top of commit_tail by a wait which
completes either when the fences are done (normal case, or when a
reset doesn't need to touch the display state). Or when the gpu reset
needs to force-unblock all pending modeset states.
The lesser source of deadlocks is when we try to pin a new framebuffer
and run into a stall. There's a bunch of places this can happen, like
eviction, changing the caching mode, acquiring a fence on older
platforms. And we can't just break the depency loop and keep going,
the only way would be to break out and restart. But the problem with
that approach is that we must stall for the reset to complete before
we grab any locks, and with the atomic infrastructure that's a bit
tricky. The only place is the ioctl code, and we don't want to insert
code into e.g. the BUSY ioctl. Hence for that problem just create a
critical section, and if any code is in there, wedge the GPU. For the
steady-state this should never be a problem.
Note that in both cases TDR itself keeps working, so from a userspace
pov this trickery isn't observable. Users themselvs might spot a short
glitch while the rendering is catching up again, but that's still
better than pre-TDR where we've thrown away all the rendering,
including innocent batches. Also, this fixes the regression TDR
introduced of making gpu resets deadlock-prone when we do need to
touch the display.
One thing I noticed is that gpu_error.flags seems to use both our own
wait-queue in gpu_error.wait_queue, and the generic wait_on_bit
facilities. Not entirely sure why this inconsistency exists, I just
picked one style.
A possible future avenue could be to insert the gpu reset in-between
ongoing modeset changes, which would avoid the momentary glitch. But
that's a lot more work to implement in the atomic commit machinery,
and given that we only need this for pre-g4x hw, of questionable
utility just for the sake of polishing gpu reset even more on those
old boxes. It might be useful for other features though.
v2: Rebase onto 4.13 with a s/wait_queue_t/struct wait_queue_entry/.
v3: Really emabarrassing fixup, I checked the wrong bit and broke the
unbreak/wakeup logic.
v4: Also handle deadlocks in pin_to_display.
v5: Review from Michel:
- Fixup the BUILD_BUG_ON
- Don't forget about the overlay
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> (v2)
Cc: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170808080828.23650-3-daniel.vetter@ffwll.ch
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
The motivation behind this new interface is expose at runtime the
creation of new OA configs which can be used as part of the i915 perf
open interface. This will enable the kernel to learn new configs which
may be experimental, or otherwise not part of the core set currently
available through the i915 perf interface.
v2: Drop DRM_ERROR for userspace errors (Matthew)
Add padding to userspace structure (Matthew)
s/guid/uuid/ (Matthew)
v3: Use u32 instead of int to iterate through registers (Matthew)
v4: Lock access to dynamic config list (Lionel)
v5: by Matthew:
Fix uninitialized error values
Fix incorrect unwiding when opening perf stream
Use kmalloc_array() to store register
Use uuid_is_valid() to valid config uuids
Declare ioctls as write only
Check padding members are set to 0
by Lionel:
Return ENOENT rather than EINVAL when trying to remove non
existing config
v6: by Chris:
Use ref counts for OA configs
Store UUID in drm_i915_perf_oa_config rather then using pointer
Shuffle fields of drm_i915_perf_oa_config to avoid padding
v7: by Chris
Rename uapi pointers fields to end with '_ptr'
v8: by Andrzej, Marek, Sebastian
Update register whitelisting
by Lionel
Add more register names for documentation
Allow configuration programming in non-paranoid mode
Add support for value filter for a couple of registers already
programmed in other part of the kernel
v9: Documentation fix (Lionel)
Allow writing WAIT_FOR_RC6_EXIT only on Gen8+ (Andrzej)
v10: Perform read access_ok() on register pointers (Lionel)
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Andrzej Datczuk <andrzej.datczuk@intel.com>
Reviewed-by: Andrzej Datczuk <andrzej.datczuk@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170803165812.2373-2-lionel.g.landwerlin@intel.com
In the following commit we'll introduce loadable userspace
configs. This change reworks how configurations are handled in the
perf driver and retains only the test configurations in kernel space.
We now store the test config in dev_priv and resolve the id only once
when opening the perf stream. The OA config is then handled through a
pointer to the structure holding the configuration details.
v2: Rework how test configs are handled (Lionel)
v3: Use u32 to hold number of register (Matthew)
v4: Removed unused dev_priv->perf.oa.current_config variable (Matthew)
v5: Lock device when accessing exclusive_stream (Lionel)
v6: Ensure OACTXCONTROL is always reprogrammed (Lionel)
v7: Switch a couple of index variable from int to u32 (Matthew)
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/20170803165812.2373-3-lionel.g.landwerlin@intel.com
The pattern of a power well backing a set of fuses whose initialization
we need to wait for during power well enabling is common to all GEN9+
platforms. Adding support for this to the HSW power well enable helper
allows us to use the HSW/BDW power well code for GEN9+ as well in a
follow-up patch.
v2:
- Use an enum for power gates instead of raw numbers. (Ville)
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170711204236.5618-6-imre.deak@intel.com
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The pattern of a power well backing a set of pipe IRQ or VGA
functionality applies to all HSW+ platforms. Using power well attributes
instead of platform checks to decide whether to init/reset pipe IRQs and
VGA correspondingly is cleaner and it allows us to unify the HSW/BDW and
GEN9+ power well code in follow-up patches.
Also use u8 for pipe_mask in related helpers to match the type in the
power well struct.
v2:
- Use u8 instead of u32 for irq_pipe_mask. (Ville)
v3:
- Use u8 for pipe_mask in related helpers too for clarity.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170712155413.29839-1-imre.deak@intel.com
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Atm, the power well IDs are defined in separate platform specific enums,
which isn't ideal for the following reasons:
- the IDs are used by helpers like lookup_power_well() in a platform
independent way
- the always-on power well is used by multiple platforms and so needs
now separate IDs, although these IDs refer to the same thing
To make things more consistent use a single enum instead of the two
separate ones, listing the IDs per platform (or set of very similar
platforms like all GEN9/10). Replace the separate always-on power
well IDs with a single ID.
While at it also add a note clarifying the distinction between regular
power wells that follow a common programming pattern and custom ones
that are programmed in some other way. The IDs for regular power wells
need to stay fixed, since they also define the request and state HW flag
positions in their corresponding power well control register(s).
v2:
- Add comment about id to req,status bit mapping to the enum. (Rodrigo)
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170711204236.5618-1-imre.deak@intel.com
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Resync with upstream to avoid git getting too badly confused. Also, we
have a conflict with the drm_vblank_cleanup removal, which cannot be
resolved by simply taking our side. Bake that in properly.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
I need this to be able to apply the deferred fbdev setup patches, I
need the relevant prep work that landed through the drm-intel tree.
Also squash in conflict fixup from Laurent Pinchart.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch make naming of fixed-point wrappers consistent
operation_<any_post_operation>_<1st operand>_<2nd operand>
also shorten the name for fixed_16_16 to fixed16
s/u32_to_fixed_16_16/u32_to_fixed16
s/fixed_16_16_to_u32/fixed16_to_u32
s/fixed_16_16_to_u32_round_up/fixed16_to_u32_round_up
s/min_fixed_16_16/min_fixed16
s/max_fixed_16_16/max_fixed16
s/mul_u32_fixed_16_16/mul_u32_fixed16
s/fixed_16_16_div/div_fixed16
Changes Since V1:
- Split the patch in more logical patches (Maarten)
Changes Since V2:
- Rebase
Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170705143154.32132-4-mahesh1.kumar@intel.com
Resync with the main drm-next pull request for 4.13. What we really
need is to fully resync with pending drm-misc, but that's not yet
possible due to the still ongoing merge window.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Once a client has requested a waitboost, we keep that waitboost active
until all clients are no longer waiting. This is because we don't
distinguish which waiter deserves the boost. However, with the advent of
fence signaling, the signaler threads appear as waiters to the RPS
interrupt handler. So instead of using a single boolean to track when to
keep the waitboost active, use a counter of all outstanding waitboosted
requests.
At this point, I have removed all vestiges of the rate limiting on
clients. Whilst this means that compositors should remain more fluid,
it also means that boosts are more prevalent. See commit b29c19b645
("drm/i915: Boost RPS frequency for CPU stalls") for a longer discussion
on the pros and cons of both approaches.
A drawback of this implementation is that it requires constant request
submission to keep the waitboost trimmed (as it is now cancelled when the
request is completed). This will be fine for a busy system, but near
idle the boosts may be kept for longer than desired (effectively tens of
vblanks worstcase) and there is a reliance on rc6 instead.
v2: Remove defunct rps.client_lock
Reported-by: Michał Winiarski <michal.winiarski@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170628123548.9236-1-chris@chris-wilson.co.uk
Make the code less confusiong by always using the top 9 bits of the
LPC bridge device ID to detect the PCH type. We need to add a bit of
new code for WPT, and we need to adjust the KBP ID as well. All the
other pre-CNP IDs are fine as is.
The virtualization cases I think are fine. These P2X and P3X IDs
actually just look like the old PIIX4 and PIIX3 IDs to me. Not sure
why they're not called PIIX3/4 though. The qemu one has a comment
saying the full ID is 0x2918 which is fine with 9 bits.
v2: Keep the CNP ID as 0xa300 (DK)
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170621174944.23306-1-ville.syrjala@linux.intel.com
Final pile of features for 4.13
New uabi:
- batch bo in first slot, for faster execbuf assembly in userspace
(Chris Wilson)
- (sub)slice getparam, needed for mesa perf support (Robert Bragg)
First pile of patches for cnl/cfl support, maintained by Rodrigo but
with lots of contributions from others. Still incomplete since public
review still ongoing.
Features/refactoring:
- Make execbuf faster (Chris Wilson), a pile of series to make execbuf
buffer handling have fewer passes, use less list walking, postpone
more work to async workers and shuffle buffers less, all to make the
common case much faster (in some cases at least).
- cold boot support for glk dsi (Madhav Chauhan)
- Clean up pipe A quirk and related old platform hacks (Ville)
- perf sampling support for kbl/glk (Lionel)
- perf cleanups (Robert Bragg)
- wire atomic state to backlight code, to avoid pipe lookup hacks
(Maarten)
- reduce request waiting latency/overhead to remove the spinning and
associated cpu cycle wasting (Chris)
- fix 90/270 rotation wm computation (Ville)
- new ddb allocation algo for skl (Kumar Mahesh)
- fix regression due to system suspend optimiazatino (Imre)
- the usual pile of small cleanups and refactors all over
GVT updates contained in this tag:
- optimization for per-VM mmio save/restore (Changbin)
- optimization for mmio hash table (Changbin)
- scheduler optimization with event (Ping)
- vGPU reset refinement (Fred)
- other misc refactor and cleanups, etc.
* tag 'drm-intel-next-2017-06-19' of git://anongit.freedesktop.org/git/drm-intel: (170 commits)
drm/i915: Update DRIVER_DATE to 20170619
drm/i915/cfl: Introduce Coffee Lake workarounds.
drm/i915: Store 9 bits of PCI Device ID for platforms with a LP PCH
drm/i915: Stash a pointer to the obj's resv in the vma
drm/i915: Async GPU relocation processing
drm/i915: Allow execbuffer to use the first object as the batch
drm/i915: Wait upon userptr get-user-pages within execbuffer
drm/i915: First try the previous execbuffer location
drm/i915: Store a persistent reference for an object in the execbuffer cache
drm/i915: Eliminate lots of iterations over the execobjects array
drm/i915: Disable EXEC_OBJECT_ASYNC when doing relocations
drm/i915: Pass vma to relocate entry
drm/i915: Store a direct lookup from object handle to vma
drm/i915: Fix retrieval of hangcheck stats
drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
drm/i915: Mark CPU cache as dirty on every transition for CPU writes
drm/i915: Make i915_vma_destroy() static
drm/i915: Actually attach the tv_format property to the SDVO connector
Revert "drm/i915/skl: New ddb allocation algorithm"
drm/i915/glk: Add cold boot sequence for GLK DSI
...
This change implements support for per-engine reset as an initial, less
intrusive hang recovery option to be attempted before falling back to the
legacy full GPU reset recovery mode if necessary. This is only supported
from Gen8 onwards.
Hangchecker determines which engines are hung and invokes error handler to
recover from it. Error handler schedules recovery for each of those engines
that are hung. The recovery procedure is as follows,
- identifies the request that caused the hang and it is dropped
- force engine to idle: this is done by issuing a reset request
- reset the engine
- re-init the engine to resume submissions.
If engine reset fails then we fall back to heavy weight full gpu reset
which resets all engines and reinitiazes complete state of HW and SW.
v2: Rebase.
v3: s/*engine_reset*/*reset_engine*/; freeze engine and irqs before
calling i915_gem_reset_engine (Chris).
v4: Rebase, modify i915_gem_reset_prepare to use a ring mask and
reuse the function for reset_engine.
v5: intel_reset_engine_start/cancel instead of request/unrequest_reset.
v6: Clean up reset_engine function to not require mutex, i.e. no need to call
revoke/restore_fences and _retire_requests (Chris).
v7: Remove leftovers from v5, i.e. no need to disable irq, hold
forcewake or wakeup the handoff bit (Chris).
v8: engine_retire_requests should be (and it was) static; explain that
we have to re-init the engine after reset, which is why the init_hw call
is needed; check reset-in-progress flag (Chris).
v9: Rebase, include code to pass the active request to gem_reset_engine
(as it is already done in full reset). Remove unnecessary
intel_reset_engine_start/cancel, these are executed as part of the
reset.
v10: Rebase, use the right I915_RESET_ENGINE flag.
v11: Fixup to call reset_finish_engine even on error.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170615201828.23144-6-michel.thierry@intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620095751.13127-6-chris@chris-wilson.co.uk
This is a preparatory patch which modifies error handler to do per engine
hang recovery. The actual patch which implements this sequence follows
later in the series. The aim is to prepare existing recovery function to
adapt to this new function where applicable (which fails at this point
because core implementation is lacking) and continue recovery using legacy
full gpu reset.
A helper function is also added to query the availability of engine
reset. A subsequent patch will add the capability to query which type
of reset is present (engine -> full -> no-reset) via the get-param
ioctl.
It has been decided that the error events that are used to notify user of
reset will only be sent in case if full chip reset. In case of just
single (or multiple) engine resets, userspace won't be notified by these
events.
Note that this implementation of engine reset is for i915 directly
submitting to the ELSP, where the driver manages the hang detection,
recovery and resubmission. With GuC submission these tasks are shared
between driver and firmware; i915 will still responsible for detecting a
hang, and when it does it will have to request GuC to reset that Engine and
remind the firmware about the outstanding submissions. This will be
added in different patch.
v2: rebase, advertise engine reset availability in platform definition,
add note about GuC submission.
v3: s/*engine_reset*/*reset_engine*/. (Chris)
Handle reset as 2 level resets, by first going to engine only and fall
backing to full/chip reset as needed, i.e. reset_engine will need the
struct_mutex.
v4: Pass the engine mask to i915_reset. (Chris)
v5: Rebase, update selftests.
v6: Rebase, prepare for mutex-less reset engine.
v7: Pass reset_engine mask as a function parameter, and iterate over the
engine mask for reset_engine. (Chris)
v8: Use i915.reset >=2 in has_reset_engine; remove redundant reset
logging; add a reset-engine-in-progress flag to prevent concurrent
resets, and avoid dual purposing of reset-backoff. (Chris)
v9: Support reset of different engines in parallel (Chris)
v10: Handle reset-engine flag locking better (Chris)
v11: Squash in reporting of per-engine-reset availability.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Ian Lister <ian.lister@intel.com>
Signed-off-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170615201828.23144-4-michel.thierry@intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620095751.13127-5-chris@chris-wilson.co.uk
If we move the actual cleanup of the context to a worker, we can allow
the final free to be called from any context and avoid undue latency in
the caller.
v2: Negotiate handling the delayed contexts free by flushing the
workqueue before calling i915_gem_context_fini() and performing the final
free of the kernel context directly
v3: Flush deferred frees before new context allocations
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620110547.15947-2-chris@chris-wilson.co.uk
This simply hides the EAGAIN caused by userptr when userspace causes
resource contention. However, it is quite beneficial with highly
contended userptr users as we avoid repeating the setup costs and
kernel-user context switches.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
The major scaling bottleneck in execbuffer is the processing of the
execobjects. Creating an auxiliary list is inefficient when compared to
using the execobject array we already have allocated.
Reservation is then split into phases. As we lookup up the VMA, we
try and bind it back into active location. Only if that fails, do we add
it to the unbound list for phase 2. In phase 2, we try and add all those
objects that could not fit into their previous location, with fallback
to retrying all objects and evicting the VM in case of severe
fragmentation. (This is the same as before, except that phase 1 is now
done inline with looking up the VMA to avoid an iteration over the
execobject array. In the ideal case, we eliminate the separate reservation
phase). During the reservation phase, we only evict from the VM between
passes (rather than currently as we try to fit every new VMA). In
testing with Unreal Engine's Atlantis demo which stresses the eviction
logic on gen7 class hardware, this speed up the framerate by a factor of
2.
The second loop amalgamation is between move_to_gpu and move_to_active.
As we always submit the request, even if incomplete, we can use the
current request to track active VMA as we perform the flushes and
synchronisation required.
The next big advancement is to avoid copying back to the user any
execobjects and relocations that are not changed.
v2: Add a Theory of Operation spiel.
v3: Fall back to slow relocations in preparation for flushing userptrs.
v4: Document struct members, factor out eb_validate_vma(), add a few
more comments to explain some magic and hide other magic behind macros.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
The advent of full-ppgtt lead to an extra indirection between the object
and its binding. That extra indirection has a noticeable impact on how
fast we can convert from the user handles to our internal vma for
execbuffer. In order to bypass the extra indirection, we use a
resizable hashtable to jump from the object to the per-ctx vma.
rhashtable was considered but we don't need the online resizing feature
and the extra complexity proved to undermine its usefulness. Instead, we
simply reallocate the hastable on demand in a background task and
serialize it before iterating.
In non-full-ppgtt modes, multiple files and multiple contexts can share
the same vma. This leads to having multiple possible handle->vma links,
so we only use the first to establish the fast path. The majority of
buffers are not shared and so we should still be able to realise
speedups with multiple clients.
v2: Prettier names, more magic.
v3: Many style tweaks, most notably hiding the misuse of execobj[].rsvd2
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
In earlier iterations of the i915-perf driver we had a number of
callbacks/hooks from other parts of the i915 driver to e.g. notify us
when a legacy context was pinned and these could run asynchronously with
respect to the stream file operations and might also run in atomic
context.
dev_priv->perf.hook_lock had been for serialising access to state needed
within these callbacks, but as the code has evolved some of the hooks
have gone away or are implemented to avoid needing to lock any state.
The remaining use of this lock was actually redundant considering how
the gen7 oacontrol state used to be updated as part of a context pin
hook.
Signed-off-by: Robert Bragg <robert@sixbynine.org>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
An oa_exponent_to_ns() utility and per-gen timebase constants where
recently removed when updating the tail pointer race condition WA, and
this restores those so we can update the _PROP_OA_EXPONENT validation
done in read_properties_unlocked() to not assume we have a 12.5MHz
timebase as we did for Haswell.
Accordingly the oa_sample_rate_hard_limit value that's referenced by
proc_dointvec_minmax defining the absolute limit for the OA sampling
frequency is now initialized to (timestamp_frequency / 2) instead of the
6.25MHz constant for Haswell.
v2:
Specify frequency of 19.2MHz for BXT (Ville)
Initialize oa_sample_rate_hard_limit per-gen too (Lionel)
Signed-off-by: Robert Bragg <robert@sixbynine.org>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Enables access to OA unit metrics for BDW, CHV, SKL and BXT which all
share (more-or-less) the same OA unit design.
Of particular note in comparison to Haswell: some OA unit HW config
state has become per-context state and as a consequence it is somewhat
more complicated to manage synchronous state changes from the cpu while
there's no guarantee of what context (if any) is currently actively
running on the gpu.
The periodic sampling frequency which can be particularly useful for
system-wide analysis (as opposed to command stream synchronised
MI_REPORT_PERF_COUNT commands) is perhaps the most surprising state to
have become per-context save and restored (while the OABUFFER
destination is still a shared, system-wide resource).
This support for gen8+ takes care to consider a number of timing
challenges involved in synchronously updating per-context state
primarily by programming all config state from the cpu and updating all
current and saved contexts synchronously while the OA unit is still
disabled.
The driver intentionally avoids depending on command streamer
programming to update OA state considering the lack of synchronization
between the automatic loading of OACTXCONTROL state (that includes the
periodic sampling state and enable state) on context restore and the
parsing of any general purpose BB the driver can control. I.e. this
implementation is careful to avoid the possibility of a context restore
temporarily enabling any out-of-date periodic sampling state. In
addition to the risk of transiently-out-of-date state being loaded
automatically; there are also internal HW latencies involved in the
loading of MUX configurations which would be difficult to account for
from the command streamer (and we only want to enable the unit when once
the MUX configuration is complete).
Since the Gen8+ OA unit design no longer supports clock gating the unit
off for a single given context (which effectively stopped any progress
of counters while any other context was running) and instead supports
tagging OA reports with a context ID for filtering on the CPU, it means
we can no longer hide the system-wide progress of counters from a
non-privileged application only interested in metrics for its own
context. Although we could theoretically try and subtract the progress
of other contexts before forwarding reports via read() we aren't in a
position to filter reports captured via MI_REPORT_PERF_COUNT commands.
As a result, for Gen8+, we always require the
dev.i915.perf_stream_paranoid to be unset for any access to OA metrics
if not root.
v5: Drain submitted requests when enabling metric set to ensure no
lite-restore erases the context image we just updated (Lionel)
v6: In addition to drain, switch to kernel context & update all
context in place (Chris)
v7: Add missing mutex_unlock() if switching to kernel context fails
(Matthew)
v8: Simplify OA period/flex-eu-counters programming by using the
batchbuffer instead of modifying ctx-image (Lionel)
v9: Back to updating the context image (due to erroneous testing,
batchbuffer programming the OA unit doesn't actually work)
(Lionel)
Pin context before updating context image (Chris)
Drop MMIO programming now that we switch to a kernel context with
right values in initial context image (Chris)
v10: Just pin_map the contexts we want to modify or let the
configuration happen on first use (Chris)
v11: Update kernel context OA config through the batchbuffer rather
than on the fly ctx-image update (Lionel)
v12: Rework OA context registers update again by swithing away from
user contexts and reconfiguring the kernel context through the
batchbuffer and updating all the other contexts' context image.
Also take care to lock slice/subslice configuration when OA is
on. (Lionel)
v13: Request rpcs updates on all engine when updating the OA config
(Lionel)
v14: Drop any kind of rpcs management now that we monitor sseu
configuration changes in a later patch (Lionel)
Remove usleep after programming the NOA configs on Gen8+, this
doesn't seem to be needed (Lionel)
v15: Respect coding style for block comments (Chris)
v16: Add missing i915_add_request() in case we fail to emit OA
configuration (Matthew)
Signed-off-by: Robert Bragg <robert@sixbynine.org>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com> \o/
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Adds a static OA unit, MUX, B Counter + Flex EU configurations for basic
render metrics on Broadwell, Cherryview, Skylake and Broxton. These are
auto generated from an XML description of metric sets, currently
maintained in gputop, ref:
https://github.com/rib/gputop
> gputop-data/oa-*.xml
> scripts/i915-perf-kernelgen.py
$ make -C gputop-data -f Makefile.xml WHITELIST=RenderBasic
v2: add newlines to debug messages + fix comment (Matthew Auld)
Signed-off-by: Robert Bragg <robert@sixbynine.org>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
