add_display_components() calls of_platform_populate, and we depopluate
on pdev remove, but not when probe fails. So if we get a probe deferral
in one of the components, we won't depopulate the platform. This causes
the core to keep references to devices which should be destroyed, which
causes issues when those same devices try to re-initialize on the next
probe attempt.
I think this is the reason we had issues with the gmu's device-managed
resources on deferral (worked around in commit 94e3a17f33a5).
Reviewed-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20190617201301.133275-3-sean@poorly.run
The 10nm pll driver didn't have any failure-path cleanup in register,
and the destroy function didn't unregister any of the hardware. This
patch adds both.
The reason things haven't been blowing up horribly is that msm_drv has a
reference count issue that keeps devices alive, so the destroy function
was never called. That will be fixed in a follow-up patch.
Reviewed-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20190617201301.133275-1-sean@poorly.run
This reverts commit ebc8c6f18322ad54275997a888ca1731d74b711f.
There are still missing corner cases with cursor interaction and these
fast plane updates on Picasso and Raven2 leading to endless PSTATE
warnings for typical desktop usage depending on the userspace.
This change should be reverted until these issues have been resolved.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110949
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: David Francis <david.francis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tonga sriov need to use smu to load firmware.
Remove sriov flag because the default return value is zero.
Signed-off-by: Jack Zhang <Jack.Zhang1@amd.com>
Reviewed-by: Trigger Huang <Trigger.Huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
add pmu attribute groups and structures for perf events.
add sysfs to track available df perfmon counters
fix overflow handling in perfmon counter reads.
v2: squash in fix (Alex)
Signed-off-by: Jonathan Kim <Jonathan.Kim@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
After clkmgr rework it gets initialized after resource pool.
The clkmgr is used in resource pool init for xgmi path.
That causes driver crash on Vega20 with xgmi due to NULL deref.
[How]
Move xgmi compensation code to dce121_clk_mgr_construct()
That also allows to make dce121_clock_patch_xgmi_ss_info()
internal static function.
Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add a folder structure to /sys/class/kfd/kfd/ called proc which contains
subfolders, each representing an active KFD process' PID, containing 1
file: pasid.
Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Under memory pressure, hmm_range_fault may return error code -ENOMEM
or -EBUSY, change pr_info to pr_debug to remove unnecessary kernel log
message because we will retry restore again.
Call get_user_pages_done if TTM get user pages failed will have
WARN_ONCE kernel calling stack dump log.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On 32-bit hosts mem->num_pages is 32-bits and can overflow
when shifted. Add a cast to avoid this.
(v2): Style fix.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If we have multiple contexts of equal priority pending execution,
activate a timer to demote the currently executing context in favour of
the next in the queue when that timeslice expires. This enforces
fairness between contexts (so long as they allow preemption -- forced
preemption, in the future, will kick those who do not obey) and allows
us to avoid userspace blocking forward progress with e.g. unbounded
MI_SEMAPHORE_WAIT.
For the starting point here, we use the jiffie as our timeslice so that
we should be reasonably efficient wrt frequent CPU wakeups.
Testcase: igt/gem_exec_scheduler/semaphore-resolve
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190620142052.19311-2-chris@chris-wilson.co.uk
When using a global seqno, we required a precise stop-the-workd event to
handle preemption and unwind the global seqno counter. To accomplish
this, we would preempt to a special out-of-band context and wait for the
machine to report that it was idle. Given an idle machine, we could very
precisely see which requests had completed and which we needed to feed
back into the run queue.
However, now that we have scrapped the global seqno, we no longer need
to precisely unwind the global counter and only track requests by their
per-context seqno. This allows us to loosely unwind inflight requests
while scheduling a preemption, with the enormous caveat that the
requests we put back on the run queue are still _inflight_ (until the
preemption request is complete). This makes request tracking much more
messy, as at any point then we can see a completed request that we
believe is not currently scheduled for execution. We also have to be
careful not to rewind RING_TAIL past RING_HEAD on preempting to the
running context, and for this we use a semaphore to prevent completion
of the request before continuing.
To accomplish this feat, we change how we track requests scheduled to
the HW. Instead of appending our requests onto a single list as we
submit, we track each submission to ELSP as its own block. Then upon
receiving the CS preemption event, we promote the pending block to the
inflight block (discarding what was previously being tracked). As normal
CS completion events arrive, we then remove stale entries from the
inflight tracker.
v2: Be a tinge paranoid and ensure we flush the write into the HWS page
for the GPU semaphore to pick in a timely fashion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190620142052.19311-1-chris@chris-wilson.co.uk
The interconnect API provides an interface for consumer drivers to
express their bandwidth needs in the SoC. This data is aggregated
and the on-chip interconnect hardware is configured to the most
appropriate power/performance profile.
Use the API to configure the interconnects and request bandwidth
between DDR and the display hardware (MDP port(s) and rotator
downscaler).
v2: update the path names to be consistent with dpu, handle the NULL
path case, updated commit msg from Georgi.
v3: split out icc setup into it's own function, and rework logic
slightly so no interconnect paths is not fatal.
Signed-off-by: Georgi Djakov <georgi.djakov@linaro.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-By: Jeffrey Hugo <jeffrey.l.hugo@gmail.com>
dpu_mdss_destroy() can get called not just from
msm_drm_uninit() but also from msm_drm_bind() in case
of any failures.
dpu_mdss_destroy() removes the icc voting by calling
icc_put. This could accidentally remove the voting
done by pm_runtime_enable.
To make the voting balanced add a minimum vote in
dpu_mdss_init() to avoid any unclocked access.
This change depends on the following patch which
introduces interconnect binding to MDSS driver:
https://patchwork.codeaurora.org/patch/708155/
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
Reviewed-by: Sean Paul <sean@poorly.run>
Signed-off-by: Rob Clark <robdclark@chromium.org>
The interconnect framework is designed to provide a
standard kernel interface to control the settings of
the interconnects on a SoC.
The interconnect API uses a consumer/provider-based model,
where the providers are the interconnect buses and the
consumers could be various drivers.
MDSS is one of the interconnect consumers which uses the
interconnect APIs to get the path between endpoints and
set its bandwidth requirement for the given interconnected
path.
Changes in v2:
- Remove error log and unnecessary check (Jordan Crouse)
Changes in v3:
- Code clean involving variable name change, removal
of extra paranthesis and variables (Matthias Kaehlcke)
Changes in v4:
- Add comments, spacings, tabs, proper port name
and icc macro (Georgi Djakov)
Changes in v5:
- Commit text and parenthesis alignment (Georgi Djakov)
Changes in v6:
- Change to new icc_set API's (Doug Anderson)
Changes in v7:
- Fixed a typo
Changes in v8:
- Handle the of_icc_get() returning NULL case. In practice
icc_set_bw() will gracefully handle the case of a NULL path,
but it's probably best for clarity to keep num_paths=0 in
this case.
Signed-off-by: Sravanthi Kollukuduru <skolluku@codeaurora.org>
Signed-off-by: Jayant Shekhar <jshekhar@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Acked-by: Georgi Djakov <georgi.djakov@linaro.org>
Reviewed-by: Sean Paul <sean@poorly.run>
Since the upstream interconnect bus framework has landed
upstream, the existing references of custom bus scaling
needs to be cleaned up.
Changes in v2:
- Fixed build error due to partial clean up
Changes in v3:
- Condense multiple lines into a single line (Sean Paul)
Changes in v4-v7:
- None
Signed-off-by: Sravanthi Kollukuduru <skolluku@codeaurora.org>
Signed-off-by: Jayant Shekhar <jshekhar@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Sean Paul <sean@poorly.run>
