It has been many years since the last confirmed sighting (and fix) of an
RC6 related bug (usually a system hang). Remove the parameter to stop
users from setting dangerous values, as they often set it during triage
and end up disabling the entire runtime pm instead (the option is not a
fine scalpel!).
Furthermore, it allows users to set known dangerous values which were
intended for testing and not for production use. For testing, we can
always patch in the required setting without having to expose ourselves
to random abuse.
v2: Fixup NEEDS_WaRsDisableCoarsePowerGating fumble, and document the
lack of ilk support better.
v3: Clear intel_info->rc6p if we don't support rc6 itself.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171201113030.18360-1-chris@chris-wilson.co.uk
assert_rpm_wakelock_held is triggered from i915_pmic_bus_access_notifier
even though it gets unregistered on (runtime) suspend, this is caused
by a race happening under the following circumstances:
intel_runtime_pm_put does:
atomic_dec(&dev_priv->pm.wakeref_count);
pm_runtime_mark_last_busy(kdev);
pm_runtime_put_autosuspend(kdev);
And pm_runtime_put_autosuspend calls intel_runtime_suspend from
a workqueue, so there is ample of time between the atomic_dec() and
intel_runtime_suspend() unregistering the notifier. If the notifier
gets called in this windowd assert_rpm_wakelock_held falsely triggers
(at this point we're not runtime-suspended yet).
This commit adds disable_rpm_wakeref_asserts and
enable_rpm_wakeref_asserts calls around the
intel_uncore_forcewake_get(FORCEWAKE_ALL) call in
i915_pmic_bus_access_notifier fixing the false-positive WARN_ON.
Changes in v2:
-Reword comment explaining why disabling the wakeref asserts is
ok and necessary
Cc: stable@vger.kernel.org
Reported-by: FKr <bugs-freedesktop@ubermail.me>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110150301.9601-2-hdegoede@redhat.com
intel_uncore_forcewake_reset() does forcewake puts and gets as such
we need to make sure that no-one tries to access the PUNIT->PMIC bus
(on systems where this bus is shared) while it runs, otherwise bad
things happen.
Normally this is taken care of by the i915_pmic_bus_access_notifier()
which does an intel_uncore_forcewake_get(FORCEWAKE_ALL) when some other
driver tries to access the PMIC bus, so that later forcewake gets are
no-ops (for the duration of the bus access).
But intel_uncore_forcewake_reset gets called in 3 cases:
1) Before registering the pmic_bus_access_notifier
2) After unregistering the pmic_bus_access_notifier
3) To reset forcewake state on a GPU reset
In all 3 cases the i915_pmic_bus_access_notifier() protection is
insufficient.
This commit fixes this race by calling iosf_mbi_punit_acquire() before
calling intel_uncore_forcewake_reset(). In the case where it is called
directly after unregistering the pmic_bus_access_notifier, we need to
hold the punit-lock over both calls to avoid a race where
intel_uncore_fw_release_timer() may execute between the 2 calls.
Reviewed-by: Imre Deak <imre.deak@intel.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171019111620.26761-3-hdegoede@redhat.com
There is a possibility on gen9 hardware to miss the forcewake ack
message. The recommended workaround is to use another free
bit and toggle it until original bit is successfully acknowledged.
Some future gen9 revs might or might not fix the underlying issue but
using fallback forcewake bit dance can be considered as harmless:
without the ack timeout we never reach the fallback bit forcewake.
Thus as of now we adopt a blanket approach for all gen9 and leave
the bypassing the fallback bit approach for future patches if
corresponding hw revisions do appear.
Commit 83e3337204 ("drm/i915: Increase maximum polling time to 50ms
for forcewake request/clear ack") did increase the forcewake timeout.
If the issue was a delayed ack, future work could include finding
a suitable timeout value both for primary ack and reserve toggle
to reduce the worst case latency.
v2: use bit 15, naming, comment (Chris), only wait fallback ack
v3: fix return on fallback, backoff after fallback write (Chris)
v4: udelay on first pass, grammar (Chris)
v4: s/reserve/fallback
References: HSDES #1604254524
References: https://bugs.freedesktop.org/show_bug.cgi?id=102051
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20171102094836.2506-1-mika.kuoppala@linux.intel.com
This patch adds per engine reset and recovery (TDR) support when GuC is
used to submit workloads to GPU.
In the case of i915 directly submission to ELSP, driver manages hang
detection, recovery and resubmission. With GuC submission these tasks
are shared between driver and GuC. i915 is still responsible for detecting
a hang, and when it does it only requests GuC to reset that Engine. GuC
internally manages acquiring forcewake and idling the engine before
resetting it.
Once the reset is successful, i915 takes over again and handles the
resubmission. The scheduler in i915 knows which requests are pending so
after resetting a engine, pending workloads/requests are resubmitted
again.
v2: s/i915_guc_request_engine_reset/i915_guc_reset_engine/ to match the
non-guc function names.
v3: Removed debug message about engine restarting from which request,
since the new baseline do it regardless of submission mode. (Chris)
v4: Rebase.
v5: Do not pass unnecessary reporting flags to the fw (Jeff);
tasklet_schedule(&execlists->irq_tasklet) handles the resubmit; rebase.
v6: Rename the existing reset engine function and share a similar
interface between guc and non-guc paths (Chris).
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20171031225309.10888-1-michel.thierry@intel.com
Reviewed-by: Jeff McGee <jeff.mcgee@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Forcewake is not affected by the engine reset on gen6+. Indeed the
reason why we added intel_uncore_forcewake_reset() to
gen6_reset_engines() was to keep the bookkeeping intact because the
reset did not touch the forcewake bit (yet we cancelled the forcewake
consumers)! This was done in commit 521198a2e7:
Author: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Fri Aug 23 16:52:30 2013 +0300
drm/i915: sanitize forcewake registers on reset
In reset we try to restore the forcewake state to
pre reset state, using forcewake_count. The reset
doesn't seem to clear the forcewake bits so we
get warn on forcewake ack register not clearing.
That futzing of the forcewake bookkeeping was dropped in commit
0294ae7b44 ("drm/i915: Consolidate forcewake resetting to a single
function"), but it did not make the realisation that the remaining
intel_uncore_forcewake_reset() was redundant.
The new danger with using intel_uncore_forcewake_reset() with per-engine
resets is that the driver and hw are still in an active state as we
perform the reset. We may be using the forcewake to read protected
registers elsewhere and those results may be clobbered by the concurrent
dropping of forcewake.
Reported-by: Michel Thierry <michel.thierry@intel.com>
Fixes: 142bc7d99b ("drm/i915: Modify error handler for per engine hang recovery")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170817173229.20324-1-chris@chris-wilson.co.uk
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
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
Currently the timer is armed for 1ms after the first use and is killed
immediately, dropping the forcewake as early as possible. However, for
very frequent operations the forcewake dance has a large impact on
latency and keeping the timer alive until we are idle is preferred. To
achieve this, if we call intel_uncore_forcewake_get whilst the timer is
alive (repeated use), then set a flag to restart the timer on expiry
rather than drop the forcewake usage count. The timer is racy, the
consequence of the race is to expire the timer earlier than is now
desired but does not impact on correct behaviour. The offset the race
slightly, we set the active flag again on intel_uncore_forcewake_put.
The effect should be to reduce the jitter of reacquiring the fw every
1ms on a busy system. However, the cost is to keep the timer alive for
an extra 1ms on a nearly idle system. We chose to incur the jitter
previously to keep the timer off for as much as possible.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170526132209.14640-1-chris@chris-wilson.co.uk
ELK seems to very picky about the preconditions to reset.
Evidence on Eaglelake (8086:2e12 (rev 03)) shows that it does
not like if reset occurs when there is active ring.
Ville found out that there is workaround with name
'WaMediaResetMainRingCleanup' which suggests that we need to
cleanup rings before resetting. It is unclear what cleanup
exactly means but evidence shows that stopping the ring
does have an effect on reset reliability. This patch makes
reset successful on hangs induced by chained batches (the igt ones).
Note that if the hang is inside a shader, it is possible
that our attempts to stop the ring achieves anything.
v2: zero ctl,head,tail also. bug ref. use driver debugs (Chris)
v3: specify platform on testcases, comment tidyup (Chris)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100942
Testcase: igt/gem_busy/*-hang #elk
Testcase: igt/gem_ringfill/hang-* #elk
Suggested-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tomi Sarvela <tomi.p.sarvela@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170519091340.21439-1-mika.kuoppala@intel.com
Replace the handcrafter loop when checking for fifo slots
with atomic wait for. This brings this wait in line with
the other waits on register access. We also get a readable
timeout constraint, so make it to fail after 10ms.
Chris suggested that we should fail silently as the fifo debug
handler, now attached to unclaimed mmio handling, will take care of the
possible errors at later stage.
Note that the decision to wait was changed so that we avoid
allocating the first reserved entry. Nor do we reduce the count
if we fail the wait, removing the possiblity to wrap the
count if the hw fifo returned zero.
v2: remove unclaimed check on timeout (Chris)
v3: use void return (Chris)
References: https://bugs.freedesktop.org/show_bug.cgi?id=100247
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/1491493182-31540-1-git-send-email-mika.kuoppala@intel.com
Remove the per-mmio checking of the FIFO debug register into the common
conditional mmio debug handling. Based on patch from Chris Wilson.
v2: postpone warn on fifodbg for unclaimed reg debugs
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Looks like intel_guc_reset had the ability to sleep under the
uncore spinlock since forever but it wasn't detected until the
recent changes annotated the wait for register with might_sleep.
I have fixed it by removing holding of the uncore spinlock over
the call to gen6_hw_domain_reset, since I do not see that is
really needed. But there is always a possibility I am missing
some nasty detail so please double check.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Acked-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Baytrail PMIC vs. PMU race fixes from Hans de Goede
This time the right version (v4), with the compile fix.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Listen for PMIC bus access notifications and get FORCEWAKE_ALL while
the bus is accessed to avoid needing to do any forcewakes, which need
PMIC bus access, while the PMIC bus is busy:
This fixes errors like these showing up in dmesg, usually followed
by a gfx or system freeze:
[drm:fw_domains_get [i915]] *ERROR* render: timed out waiting for forcewake ack request.
[drm:fw_domains_get [i915]] *MEDIA* render: timed out waiting for forcewake ack request.
i2c_designware 808622C1:06: punit semaphore timed out, resetting
i2c_designware 808622C1:06: PUNIT SEM: 2
i2c_designware 808622C1:06: couldn't acquire bus ownership
Downside of this approach is that it causes wakeups whenever the PMIC
bus is accessed. Unfortunately we cannot simply wait for the PMIC bus
to go idle when we hit a race, as forcewakes may be done from interrupt
handlers where we cannot sleep to wait for the i2c PMIC bus access to
finish.
Note that the notifications and thus the wakeups will only happen on
baytrail / cherrytrail devices using PMICs with a shared i2c bus for
P-Unit and host PMIC access (i2c busses with a _SEM method in their
APCI node), e.g. an axp288 PMIC.
I plan to write some patches for drivers accessing the PMIC bus to
limit their bus accesses to a bare minimum (e.g. cache registers, do not
update battery level more often then 4 times a minute), to limit the
amount of wakeups.
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=155241
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Tested-by: tagorereddy <tagore.chandan@gmail.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
[danvet: Wiggle in conflicts.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
