drm/i915: Add WABB blit for Wa_16018031267 / Wa_16018063123

Apply WABB blit for Wa_16018031267 / Wa_16018063123.

v3: drop unused enum definition
v4: move selftest to separate patch, use wa only on BCS0.
v5: fixed selftest caller to context_wabb

Signed-off-by: Nirmoy Das <nirmoy.das@intel.com>
Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Signed-off-by: Andrzej Hajda <andrzej.hajda@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231026-wabb-v6-2-4aa7d55d0a8a@intel.com

drm/i915/lrc: User PXP contexts requires runalone bit in lrc

On Meteorlake onwards, HW specs require that all user contexts that
run on render or compute engines and require PXP must enforce
run-alone bit in lrc. Add this enforcement for protected contexts.

Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com>
Reviewed-by: Vivaik Balasubrawmanian <vivaik.balasubrawmanian@intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230917211933.1407559-4-alan.previn.teres.alexis@intel.com

drm/i915: Add Wa_18022495364

Invalidate instruction and State cache bit using INDIRECT_CTX on
every gpu context switch for gen12.
The goal of this workaround is to actually perform an explicit
invalidation of that cache (by re-writing the register) during every GPU
context switch, which is accomplished via a "workaround batchbuffer"
that's attached to the context via INDIRECT_CTX. (Matt Roper)

Please refer [1] for more reviews and comment on the same patch

[1] https://patchwork.freedesktop.org/series/123377/

v2:
- Remove extra parentheses from the condition (Lucas)
- Align spacing and new line (Lucas)

v3:
- Fix commit message.

v4:
- Only Gen12 changes are kept and Remove DG2+ condition (Matt Roper)
- Fix the commit message for r-b (Matt Roper)
- Rename the register bit in define

v5:
- Move out this workaround from golden context init (Matt Roper)
- Use INDIRECT_CTX to set bit on each GPU context switch (Matt Roper)

v6:
- Change IP Version base condition for Gen12 (Matt Roper)
- Made imperative form of commit version messages (Suraj)
- s/Added/Add in patch header (Suraj)

v7:
- In version descriptions s/Ropper/Roper (Matt Atwood)

BSpec: 11354
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Suraj Kandpal <suraj.kandpal@intel.com>
Cc: Matt Atwood <matthew.s.atwood@intel.com>
Signed-off-by: Dnyaneshwar Bhadane <dnyaneshwar.bhadane@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230914202000.1069884-1-dnyaneshwar.bhadane@intel.com

drm/i915/gt: Prevent error pointer dereference

Move the check for "if (IS_ERR(obj))" in front of the call to
i915_gem_object_set_cache_coherency() which dereferences "obj".
Otherwise it will lead to a crash.

Fixes: 43aa755eae2c ("drm/i915/mtl: Update cache coherency setting for context structure")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/455b2279-2e08-4d00-9784-be56d8ee42e3@moroto.mountain

drm/i915: Eliminate IS_MTL_GRAPHICS_STEP

Several workarounds are guarded by IS_MTL_GRAPHICS_STEP. However none
of these workarounds are actually tied to MTL as a platform; they only
relate to the Xe_LPG graphics IP, regardless of what platform it appears
in. At the moment MTL is the only platform that uses Xe_LPG with IP
versions 12.70 and 12.71, but we can't count on this being true in the
future. Switch these to use a new IS_GFX_GT_IP_STEP() macro instead
that is purely based on IP version. IS_GFX_GT_IP_STEP() is also
GT-based rather than device-based, which will help prevent mistakes
where we accidentally try to apply Xe_LPG graphics workarounds to the
Xe_LPM+ media GT and vice-versa.

v2:
- Switch to a more generic and shorter IS_GT_IP_STEP macro that can be
used for both graphics and media IP (and any other kind of GTs that
show up in the future).
v3:
- Switch back to long-form IS_GFX_GT_IP_STEP macro. (Jani)
- Move macro to intel_gt.h. (Andi)
v4:
- Build IS_GFX_GT_IP_STEP on top of IS_GFX_GT_IP_RANGE and
IS_GRAPHICS_STEP building blocks and name the parameters from/until
rather than begin/fixed. (Jani)
- Fix usage examples in comment.
v5:
- Tweak comment on macro. (Gustavo)

Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821180619.650007-15-matthew.d.roper@intel.com

drm/i915/dg2: Drop pre-production GT workarounds

DG2 first production steppings were C0 (for DG2-G10), B1 (for DG2-G11),
and A1 (for DG2-G12). Several workarounds that apply onto to
pre-production hardware can be dropped. Furthermore, several
workarounds that apply to all production steppings can have their
conditions simplified to no longer check the GT stepping.

v2:
- Keep Wa_16011777198 in place for now; it will be removed separately
in a follow-up patch to keep review easier.

Bspec: 44477
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Matt Atwood <matthew.s.atwood@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230816214201.534095-10-matthew.d.roper@intel.com

drm/i915/gt: Prevent error pointer dereference

Move the check for "if (IS_ERR(obj))" in front of the call to
i915_gem_object_set_cache_coherency() which dereferences "obj".
Otherwise it will lead to a crash.

Fixes: 43aa755eae2c ("drm/i915/mtl: Update cache coherency setting for context structure")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/455b2279-2e08-4d00-9784-be56d8ee42e3@moroto.mountain
(cherry picked from commit c92ec50822fb84306d951520d81919328421acbd)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Apply workaround 22016122933 correctly

WA_22016122933 was recently applied to all MeteorLake engines, which is
simultaneously too broad (should only apply to Media engines) and too
specific (should apply to all platforms that use the same media engine
as MeteorLake). Correct this in cases where coherency settings are
modified.

There were also two additional places where the workaround was applied
unconditionally. The change was confirmed as necessary for all
platforms, so the workaround label was removed.

Suggested-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Acked-by: Fei Yang <fei.yang@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230801153242.2445478-4-jonathan.cavitt@intel.com
Link: https://patchwork.freedesktop.org/patch/msgid/20230807121957.598420-4-andi.shyti@linux.intel.com

drm/i915: Make i915_coherent_map_type GT-centric

Refactor i915_coherent_map_type to be GT-centric rather than
device-centric. Each GT may require different coherency
handling due to hardware workarounds.

Since the function now takes a GT instead of the i915, the function is
renamed and moved to the gt folder.

Suggested-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Acked-by: Fei Yang <fei.yang@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Acked-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230801153242.2445478-3-jonathan.cavitt@intel.com
Link: https://patchwork.freedesktop.org/patch/msgid/20230807121957.598420-3-andi.shyti@linux.intel.com

drm/i915/gt: Support aux invalidation on all engines

Perform some refactoring with the purpose of keeping in one
single place all the operations around the aux table
invalidation.

With this refactoring add more engines where the invalidation
should be performed.

Fixes: 972282c4cf24 ("drm/i915/gen12: Add aux table invalidate for all engines")
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Jonathan Cavitt <jonathan.cavitt@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: <stable@vger.kernel.org> # v5.8+
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230725001950.1014671-8-andi.shyti@linux.intel.com

drm/i915/gt: Cleanup aux invalidation registers

Fix the 'NV' definition postfix that is supposed to be INV.

Take the chance to also order properly the registers based on
their address and call the GEN12_GFX_CCS_AUX_INV address as
GEN12_CCS_AUX_INV like all the other similar registers.

Remove also VD1, VD3 and VE1 registers that don't exist and add
BCS0 and CCS0.

Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.8+
Reviewed-by: Nirmoy Das <nirmoy.das@intel.com>
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230725001950.1014671-2-andi.shyti@linux.intel.com

drm/i915/mtl: Update cache coherency setting for context structure

As context structure is shared memory for CPU/GPU, Wa_22016122933 is
needed for this memory block as well.

Signed-off-by: Zhanjun Dong <zhanjun.dong@intel.com>
CC: Fei Yang <fei.yang@intel.com>
Reviewed-by: Fei Yang <fei.yang@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230706174704.177929-1-zhanjun.dong@intel.com

drm/i915/gt: Support aux invalidation on all engines

Perform some refactoring with the purpose of keeping in one
single place all the operations around the aux table
invalidation.

With this refactoring add more engines where the invalidation
should be performed.

Fixes: 972282c4cf24 ("drm/i915/gen12: Add aux table invalidate for all engines")
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Jonathan Cavitt <jonathan.cavitt@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: <stable@vger.kernel.org> # v5.8+
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230725001950.1014671-8-andi.shyti@linux.intel.com
(cherry picked from commit 76ff7789d6e63d1a10b3b58f5c70b2e640c7a880)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

drm/i915/gt: Cleanup aux invalidation registers

Fix the 'NV' definition postfix that is supposed to be INV.

Take the chance to also order properly the registers based on
their address and call the GEN12_GFX_CCS_AUX_INV address as
GEN12_CCS_AUX_INV like all the other similar registers.

Remove also VD1, VD3 and VE1 registers that don't exist and add
BCS0 and CCS0.

Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.8+
Reviewed-by: Nirmoy Das <nirmoy.das@intel.com>
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230725001950.1014671-2-andi.shyti@linux.intel.com
(cherry picked from commit 2f0b927d3ca3440445975ebde27f3df1c3ed6f76)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

drm/i915/mtl: Extend Wa_16014892111 to MTL A-step

Like DG2, MTL a-step hardware is subject to Wa_16014892111 which
requires that any changes made to the DRAW_WATERMARK register be
done via an INDIRECT_CTX batch buffer rather than through a regular
context workaround.

The bspec gives the same non-default recommended tuning value
for DRAW_WATERMARK as DG2, so we can re-use the INDIRECT_CTX code
to apply that tuning setting on A-step hardware.

Application of the tuning setting on B-step and later does not
need INDIRECT_CTX handling and is already done in
mtl_ctx_workarounds_init() as usual.

v2: Limit the WA for A-step
v3: Update the commit message.
v4: Reorder platform checks and update commit message.

Bspec: 68331
Cc: Haridhar Kalvala <haridhar.kalvala@intel.com>
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230517233111.297542-2-radhakrishna.sripada@intel.com

drm/i915/mtl: Fix bcs default context

Commit 0d0e7d1eea9e ("drm/i915/mtl: Define engine context layouts")
added the engine context for Meteor Lake. In a second revision of the
patch it was believed the xcs offsets were wrong due to a tagging
issue in the spec. The first version was actually correct, as shown
by the intel_lrc_live_selftests/live_lrc_layout test:

i915: Running gt_lrc
i915: Running intel_lrc_live_selftests/live_lrc_layout
bcs0: LRI command mismatch at dword 1, expected 1108101d found 11081019
[drm:drm_helper_probe_single_connector_modes [drm_kms_helper]] [CONNECTOR:236:DP-1] disconnected
bcs0: HW register image:
[0000] 00000000 1108101d 00022244 ffff0008 00022034 00000088 00022030 00000088
...
bcs0: SW register image:
[0000] 00000000 11081019 00022244 00090009 00022034 00000000 00022030 00000000

The difference in the 2 additional dwords (0x1d vs 0x19) are the offsets
0x120 / 0x124 that are indeed part of the context image.

Bspec: 45585

Fixes: 0d0e7d1eea9e ("drm/i915/mtl: Define engine context layouts")
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230111235531.3353815-2-radhakrishna.sripada@intel.com

drm/i915: Move/adjust register definitions related to Wa_22011450934

The implementation of Wa_22011450934 introduced three new register
definitions in i915_reg.h that didn't get moved to the GT/engine
register headers when all the other registers moved; let's move them to
the appropriate headers and tidy up their definitions now for
consistency:

- STATE_ACK_DEBUG is moved to the engine register header and converted
to a parameterized definition; the workaround only needs the RCS
instance to be programmed, but there are instances on other engines
that could be used by other workarounds in the future.

- The two CULLBIT registers move to the GT register header. Since
they belong to MMIO ranges that became MCR starting with Xe_HP,
their definitions should be defined as MCR_REG() and use an Xe_HP
prefix to keep the register semantics clear.

Note that the MCR definition is just for consistency and to prevent
accidental misuse if other workarounds related to these registers show
up in the future. There's no functional change to today's driver since
the workaround that references these registers only accesses them via
MI_LRR engine instructions. Engine-initiated register accesses do not
utilize the same steering controls as CPU-initiated accesses; they
use a different steering control register (0x20CC) which is initialized
to a non-terminated DSS target by pre-OS firmware and never changed
thereafter (i915 does not touch it and userspace does not have
permission to change that register).

Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230117202627.4134579-1-matthew.d.roper@intel.com

drm/i915/mtl: Fix bcs default context

Commit 0d0e7d1eea9e ("drm/i915/mtl: Define engine context layouts")
added the engine context for Meteor Lake. In a second revision of the
patch it was believed the xcs offsets were wrong due to a tagging
issue in the spec. The first version was actually correct, as shown
by the intel_lrc_live_selftests/live_lrc_layout test:

i915: Running gt_lrc
i915: Running intel_lrc_live_selftests/live_lrc_layout
bcs0: LRI command mismatch at dword 1, expected 1108101d found 11081019
[drm:drm_helper_probe_single_connector_modes [drm_kms_helper]] [CONNECTOR:236:DP-1] disconnected
bcs0: HW register image:
[0000] 00000000 1108101d 00022244 ffff0008 00022034 00000088 00022030 00000088
...
bcs0: SW register image:
[0000] 00000000 11081019 00022244 00090009 00022034 00000000 00022030 00000000

The difference in the 2 additional dwords (0x1d vs 0x19) are the offsets
0x120 / 0x124 that are indeed part of the context image.

Bspec: 45585

Fixes: 0d0e7d1eea9e ("drm/i915/mtl: Define engine context layouts")
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230111235531.3353815-2-radhakrishna.sripada@intel.com
(cherry picked from commit ca54a9a32da0f0ef7e5cbcd111b66f3c9d78b7d2)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Document function to decode register state context

It's not obvious how the encode/decode of the per platform tables is
done. Document it so while adding tables for new platforms people can be
confident they right things is being done.

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220930050903.3479619-3-lucas.demarchi@intel.com

drm/i915: Fix __gen125_emit_bb_start() without WA

ce->wa_bb_page is allocated only for graphics version 12. However
__gen125_emit_bb_start() is used for any graphics version >= 12.50. For
the currently supported platforms this is not an issue, but for future
ones there's a mismatch causing the jump to
`wa_offset + DG2_PREDICATE_RESULT_BB` to be invalid since wa_offset is
not correct.

As in other places in the driver, check for graphics version "greater or
equal" to future-proof the support for new platforms.

Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220930050903.3479619-2-lucas.demarchi@intel.com

drm/i915/mtl: Define engine context layouts

The part of the media and blitter engine contexts that we care about for
setting up an initial state on MTL are nearly similar to DG2 (and PVC).
The difference being PRT_BB_STATE being replaced with NOP.

For render/compute engines, the part of the context images are nearly
the same, although the layout had a very slight change --- one POSH
register was removed and the placement of some LRI/noops adjusted
slightly to compensate.

v2:
- Dg2, mtl xcs offsets slightly vary. Use a separate offsets array(Bala)
- Add missing nop in xcs offsets(Bala)
v3:
- Fix the spacing for nop in xcs offset(MattR)
v4:
- Fix rcs register offset(MattR)
v4.1:
- Fix commit message(Lucas)

Bspec: 46261, 46260, 45585
Cc: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>
Cc: Licas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220928155511.2379663-1-radhakrishna.sripada@intel.com

drm/i915: Noop lrc_init_wa_ctx() on recent/future platforms

Except for graphics version 8 and 9, nothing is done in
lrc_init_wa_ctx(). Assume this won't be needed on future platforms as
well and remove the warning.

Note that this function is not called for anything below version 8 since
those don't use either guc or execlist, i.e. HAS_EXECLISTS() is false.

Reviewed-by: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220907230841.1703574-1-lucas.demarchi@intel.com

drm/i915/gt: Explicitly clear BB_OFFSET for new contexts

Even though the initial protocontext we load onto HW has the register
cleared, by the time we save it into the default image, BB_OFFSET has
had the enable bit set. Reclear BB_OFFSET for each new context.

Testcase: igt/i915_selftests/gt_lrc

v2:
Extend it for gen8.
v3:
BB_OFFSET is recorded per engine from Gen9 onwards

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Reviewed-by: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Signed-off-by: Karolina Drobnik <karolina.drobnik@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/37c67abb3303852f06a570a4360addf52bf941c1.1663081418.git.karolina.drobnik@intel.com

drm/i915/mtl: Add gsi_offset when emitting aux table invalidation

The aux table invalidation registers are a bit unique --- they're
engine-centric registers that reside in the GSI register space rather
than within the engines' regular MMIO ranges. That means that when
issuing invalidation on engines in the standalone media GT, the GSI
offset must be added to the regular MMIO offset for the invalidation
registers.

Cc: Aravind Iddamsetty <aravind.iddamsetty@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Aravind Iddamsetty <aravind.iddamsetty@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220906234934.3655440-12-matthew.d.roper@intel.com
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/dg2: Incorporate Wa_16014892111 into DRAW_WATERMARK tuning

Although register tuning settings are generally implemented via the
workaround infrastructure, it turns out that the DRAW_WATERMARK register
is not properly saved/restored by hardware around power events (i.e.,
RC6 entry) so updates to the value cannot be applied in the usual
manner. New workaround Wa_16014892111 informs us that any tuning
updates to this register must instead be applied via an INDIRECT_CTX
batch buffer. This will ensure that the necessary value is re-applied
when a context begins running, even if an RC6 entry had wiped the
register back to hardware defaults since the last context ran.

Fixes: 6dc85721df74 ("drm/i915/dg2: Add additional tuning settings")
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/6642
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220823202449.83727-1-matthew.d.roper@intel.com

drm/i915/gt: Clear SET_PREDICATE_RESULT prior to executing the ring

Userspace may leave predication enabled upon return from the batch
buffer, which has the consequent of preventing all operation from the
ring from being executed, including all the synchronisation, coherency
control, arbitration and user signaling. This is more than just a local
gpu hang in one client, as the user has the ability to prevent the
kernel from applying critical workarounds and can cause a full GT reset.

We could simply execute MI_SET_PREDICATE upon return from the user
batch, but this has the repercussion of modifying the user's context
state. Instead, we opt to execute a fixup batch which by mixing
predicated operations can determine the state of the
SET_PREDICATE_RESULT register and restore it prior to the next userspace
batch. This allows us to protect the kernel's ring without changing the
uABI.

Suggested-by: Zbigniew Kempczynski <zbigniew.kempczynski@intel.com>
Signed-off-by: Chris Wilson <chris.p.wilson@intel.com>
Cc: Zbigniew Kempczynski <zbigniew.kempczynski@intel.com>
Cc: Thomas Hellstrom <thomas.hellstrom@intel.com>
Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220425152317.4275-4-ramalingam.c@intel.com

drm/i915: Track context current active time

Track context active (on hardware) status together with the start
timestamp.

This will be used to provide better granularity of context
runtime reporting in conjunction with already tracked pphwsp accumulated
runtime.

The latter is only updated on context save so does not give us visibility
to any currently executing work.

As part of the patch the existing runtime tracking data is moved under the
new ce->stats member and updated under the seqlock. This provides the
ability to atomically read out accumulated plus active runtime.

v2:
* Rename and make __intel_context_get_active_time unlocked.

v3:
* Use GRAPHICS_VER.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Aravind Iddamsetty <aravind.iddamsetty@intel.com> # v1
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220401142205.3123159-6-tvrtko.ursulin@linux.intel.com

drm/i915: avoid concurrent writes to aux_inv

GPU hangs have been observed when multiple engines write to the
same aux_inv register at the same time. To avoid this each engine
should only invalidate its own auxiliary table. The function
gen12_emit_flush_xcs() currently invalidate the auxiliary table for
all engines because the rq->engine is not necessarily the engine
eventually carrying out the request, and potentially the engine
could even be a virtual one (with engine->instance being -1).
With the MMIO remap feature, we can actually set bit 17 of MI_LRI
instruction and let the hardware to figure out the local aux_inv
register at runtime to avoid invalidating auxiliary table for all
engines.

Bspec: 45728

v2: Invalidate AUX table for indirect context as well.

Cc: Stuart Summers <stuart.summers@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Chris Wilson <chris.p.wilson@intel.com>
Signed-off-by: Fei Yang <fei.yang@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220328171650.1900674-1-fei.yang@intel.com

drm/i915/xehp: Add compute workarounds

Additional workarounds are required once we start exposing CCS engines.

Note that we have a number of workarounds that update registers in the
shared render/compute reset domain. Historically we've just added such
registers to the RCS engine's workaround list. But going forward we
should be more careful to place such workarounds on a wa_list for an
engine that definitely exists and is not fused off (e.g., a platform
with no RCS would never apply the RCS wa_list). We'll keep
rcs_engine_wa_init() focused on RCS-specific workarounds that only need
to be applied if the RCS engine is present. A separate
general_render_compute_wa_init() function will be used to define
workarounds that touch registers in the shared render/compute reset
domain and that we need to apply regardless of what render and/or
compute engines actually exist. Any workarounds defined in this new
function will internally be added to the first present RCS or CCS
engine's workaround list to ensure they get applied (and only get
applied once rather than being needlessly re-applied several times).

Co-author: Srinivasan Shanmugam
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220301231549.1817978-13-matthew.d.roper@intel.com

drm/i915/xehp: CCS should use RCS setup functions

The compute engine handles the same commands the render engine can
(except 3D pipeline), so it makes sense that CCS is more similar to RCS
than non-render engines.

The CCS context state (lrc) is also similar to the render one, so reuse
it. Note that the compute engine has its own CTX_R_PWR_CLK_STATE
register.

In order to avoid having multiple RCS && CCS checks, add the following
engine flag:
- I915_ENGINE_HAS_RCS_REG_STATE - use the render (larger) reg state ctx.

BSpec: 46260
Original-author: Michel Thierry
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220301231549.1817978-6-matthew.d.roper@intel.com

drm/i915/lrc: replace include with forward declarations

Prefer forward declarations over includes if possible.

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220214173810.2108975-2-jani.nikula@intel.com

drm/i915/lrc: move lrc_get_runtime() to intel_lrc.c

Move the static inline next to the only caller.

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220214173810.2108975-1-jani.nikula@intel.com

drm/i915/dg2: Add Wa_22011450934

An indirect ctx wabb is implemented as per Wa_22011450934 to avoid rcs
restore hang during context restore of a preempted context in GPGPU mode

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
cc: Chris Wilson <chris.p.wilson@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220128185209.18077-2-ramalingam.c@intel.com

drm/i915: Move GT registers to their own header file

This is a huge, chaotic mass of registers copied over as-is without any
real cleanup. We'll come back and organize these better, align on
consistent coding style, remove dead code, etc. in separate patches
later that will be easier to review.

v2:
- Add missing include in intel_pxp_irq.c
v3:
- Correct a few indentation errors (Lucas)
- Minor conflict resolution

Cc: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220127234334.4016964-6-matthew.d.roper@intel.com

drm/i915/gt: Move engine registers to their own header

Let's continue breaking up and cleaning up the massive i915_reg.h file
by moving all registers that are defined in relation to an engine base
to their own header.

There are probably a bunch of other "engine registers" that we haven't
moved yet (especially those that belong to the render engine in the
0x2??? range), but this is a relatively straightforward first step.

Cc: Jani Nikula <jani.nikula@linux.intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220111051600.3429104-8-matthew.d.roper@intel.com

drm/i915/execlists: Weak parallel submission support for execlists

A weak implementation of parallel submission (multi-bb execbuf IOCTL) for
execlists. Doing as little as possible to support this interface for
execlists - basically just passing submit fences between each request
generated and virtual engines are not allowed. This is on par with what
is there for the existing (hopefully soon deprecated) bonding interface.

We perma-pin these execlists contexts to align with GuC implementation.

v2:
(John Harrison)
- Drop siblings array as num_siblings must be 1
v3:
(John Harrison)
- Drop single submission
v4:
(John Harrison)
- Actually drop single submission
- Use IS_ERR check on return value from intel_context_create
- Set last request to NULL on unpin

Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211222223532.28698-1-matthew.brost@intel.com

drm/i915/dg2: Add Wa_16013000631

Invalidate IC cache through pipe control command as part of the ctx
restore flow through indirect ctx pointer.

v2:
- Move pipe control from xcs indirect context to the rcs indirect
context. We'll eventually need this on the CCS engines too, but
support for those hasn't landed yet.

Cc: Chris Wilson <chris.p.wilson@intel.com>
Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211116174818.2128062-5-matthew.d.roper@intel.com

drm/i915/guc: Add multi-lrc context registration

Add multi-lrc context registration H2G. In addition a workqueue and
process descriptor are setup during multi-lrc context registration as
these data structures are needed for multi-lrc submission.

v2:
(John Harrison)
- Move GuC specific fields into sub-struct
- Clean up WQ defines
- Add comment explaining math to derive WQ / PD address
v3:
(John Harrison)
- Add PARENT_SCRATCH_SIZE define
- Update comment explaining multi-lrc register
v4:
(John Harrison)
- Move PARENT_SCRATCH_SIZE to common file

Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211014172005.27155-9-matthew.brost@intel.com

drm/i915: Don't back up pinned LMEM context images and rings during suspend

Pinned context images are now reset during resume. Don't back them up,
and assuming that rings can be assumed empty at suspend, don't back them
up either.

Introduce a new object flag, I915_BO_ALLOC_PM_VOLATILE meaning that an
object is allowed to lose its content on suspend.

v3:
- Slight documentation clarification (Matthew Auld)

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210922062527.865433-7-thomas.hellstrom@linux.intel.com

drm/i915/dg2: Add new LRI reg offsets

New LRI register offsets were introduced for DG2, this patch adds
those extra registers, and create new register table for setting offsets
to compare with HW generated context image - especially for gt_lrc test.
Also updates general purpose register with scratch offset for DG2, in
order to use it for live_lrc_fixed selftest.

Cc: Chris P Wilson <chris.p.wilson@intel.com>
Cc: Prathap Kumar Valsan <prathap.kumar.valsan@intel.com>
Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Caz Yokoyama <caz.yokoyama@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210805163647.801064-8-matthew.d.roper@intel.com

drm/i915/gt: remove GRAPHICS_VER == 10

Replace all remaining handling of GRAPHICS_VER {==,>=} 10 with
{==,>=} 11. With the removal of CNL, there is no platform with graphics
version equals 10.

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210728220326.1578242-5-lucas.demarchi@intel.com

drm/i915/xehp: New engine context offsets

The layout of some engine contexts has changed on Xe_HP. Define the new
offsets.

Bspec: 45585, 46256
Signed-off-by: Prathap Kumar Valsan <prathap.kumar.valsan@intel.com>
Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Venkata Ramana Nayana <venkata.ramana.nayana@intel.com>
Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Matt Atwood <matthew.s.atwood@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210721223043.834562-10-matthew.d.roper@intel.com

drm/i915/xehp: Handle new device context ID format

Xe_HP changes the format of the context ID from past platforms.

Signed-off-by: Stuart Summers <stuart.summers@intel.com>
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Matt Atwood <matthew.s.atwood@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210721223043.834562-9-matthew.d.roper@intel.com

drm/i915: Stop storing the ring size in the ring pointer (v3)

Previously, we were storing the ring size in the ring pointer before it
was actually allocated. We would then guard setting the ring size on
checking for CONTEXT_ALLOC_BIT. This is error-prone at best and really
only saves us a few bytes on something that already burns at least 4K.
Instead, this patch adds a new ring_size field and makes everything use
that.

v2 (Daniel Vetter):
- Replace 512 * SZ_4K with SZ_2M

v2 (Jason Ekstrand):
- Rebase on top of page migration code

Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210708154835.528166-3-jason@jlekstrand.net

drm/i915/gt: replace IS_GEN and friends with GRAPHICS_VER

This was done by the following semantic patch:

@@ expression i915; @@
- INTEL_GEN(i915)
+ GRAPHICS_VER(i915)

@@ expression i915; expression E; @@
- INTEL_GEN(i915) >= E
+ GRAPHICS_VER(i915) >= E

@@ expression dev_priv; expression E; @@
- !IS_GEN(dev_priv, E)
+ GRAPHICS_VER(dev_priv) != E

@@ expression dev_priv; expression E; @@
- IS_GEN(dev_priv, E)
+ GRAPHICS_VER(dev_priv) == E

@@
expression dev_priv;
expression from, until;
@@
- IS_GEN_RANGE(dev_priv, from, until)
+ IS_GRAPHICS_VER(dev_priv, from, until)

@def@
expression E;
identifier id =~ "^gen$";
@@
- id = GRAPHICS_VER(E)
+ ver = GRAPHICS_VER(E)

@@
identifier def.id;
@@
- id
+ ver

It also takes care of renaming the variable we assign to GRAPHICS_VER()
so to use "ver" rather than "gen".

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210605155356.4183026-2-lucas.demarchi@intel.com

drm/i915: Update the helper to set correct mapping

Determine the possible coherent map type based on object location,
and if target has llc or if user requires an always coherent
mapping.

Cc: Matthew Auld <matthew.auld@intel.com>
Cc: CQ Tang <cq.tang@intel.com>
Suggested-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Venkata Sandeep Dhanalakota <venkata.s.dhanalakota@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210427085417.120246-2-matthew.auld@intel.com

drm/i915: allocate context from LMEM

Prefer allocating the context from LMEM on dgfx.

Based on a patch from Michel Thierry.

v2: flatten the chain

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/20210127131417.393872-6-matthew.auld@intel.com
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

drm/i915: Make lrc_init_wa_ctx compatible with ww locking, v3.

Make creation separate from pinning, in order to take the lock only
once, and pin the mapping with the lock held.

Changes since v1:
- Rebase on top of upstream changes.
Changes since v2:
- Fully clear wa_ctx on error.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-27-maarten.lankhorst@linux.intel.com

drm/i915/gt: Prune inlines

Remove all the manual inlines from non-critical sections in gt/

add/remove: 2/0 grow/shrink: 0/3 up/down: 762/-1473 (-711)
Function old new delta
mi_set_context.isra - 602 +602
write_dma_entry - 160 +160
__set_pd_entry 214 69 -145
clear_pd_entry 190 42 -148
ring_request_alloc 2021 841 -1180
Total: Before=1605086, After=1604375, chg -0.04%

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210113152224.29794-1-chris@chris-wilson.co.uk

drm/i915/gt: Remove unused function 'dword_in_page'

>> drivers/gpu/drm/i915/gt/intel_lrc.c:17:28: error: unused function 'dword_in_page' [-Werror,-Wunused-function]
static inline unsigned int dword_in_page(void *addr)

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210109163455.28466-1-chris@chris-wilson.co.uk

drm/i915/gt: Exercise lrc_wa_ctx initialisation failure

Inject a fault into lrc_init_wa_ctx() to ensure that we can tolerate a
failure to construct the workarounds.

v2: Avoid mentioning an error for fault-injection, other CI will
complain about the dmesg spam.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210109114453.27798-1-chris@chris-wilson.co.uk

drm/i915/gt: Prevent use of engine->wa_ctx after error

On error we unpin and free the wa_ctx.vma, but do not clear any of the
derived flags. During lrc_init, we look at the flags and attempt to
dereference the wa_ctx.vma if they are set. To protect the error path
where we try to limp along without the wa_ctx, make sure we clear those
flags!

Reported-by: Matt Roper <matthew.d.roper@intel.com>
Fixes: 604a8f6f1e33 ("drm/i915/lrc: Only enable per-context and per-bb buffers if set")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v4.15+
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210108204026.20682-1-chris@chris-wilson.co.uk

drm/i915: Populate logical context during first pin.

This allows us to remove pin_map from state allocation, which saves
us a few retry loops. We won't need this until first pin, anyway.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20201231170405.22843-1-chris@chris-wilson.co.uk

drm/i915/gt: Track all timelines created using the HWSP

We assume that the contents of the HWSP are lost across suspend, and so
upon resume we must restore critical values such as the timeline seqno.
Keep track of every timeline allocated that uses the HWSP as its storage
and so we can then reset all seqno values by walking that list.

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/20201222104242.10993-1-chris@chris-wilson.co.uk

drm/i915/gt: Split logical ring contexts from execlist submission

Split the definition, construction and updating of the Logical Ring
Context from the execlist submission interface. The LRC is used by the
HW, irrespective of our different submission backends.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201219020343.22681-1-chris@chris-wilson.co.uk

drm/i915/gt: Clear the execlists timers upon reset

Across a reset, we stop the engine but not the timers. This leaves a
window where the timers have inconsistent state with the engine, but
should only result in a spurious timeout. As we cancel the outstanding
events, also cancel their timers.

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/20201204151234.19729-4-chris@chris-wilson.co.uk

drm/i915/gt: Cancel the preemption timeout on responding to it

We currently presume that the engine reset is successful, cancelling the
expired preemption timer in the process. However, engine resets can
fail, leaving the timeout still pending and we will then respond to the
timeout again next time the tasklet fires. What we want is for the
failed engine reset to be promoted to a full device reset, which is
kicked by the heartbeat once the engine stops processing events.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1168
Fixes: 3a7a92aba8fb ("drm/i915/execlists: Force preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201204151234.19729-2-chris@chris-wilson.co.uk

drm/i915/gt: Ignore repeated attempts to suspend request flow across reset

Before reseting the engine, we suspend the execution of the guilty
request, so that we can continue execution with a new context while we
slowly compress the captured error state for the guilty context. However,
if the reset fails, we will promptly attempt to reset the same request
again, and discover the ongoing capture. Ignore the second attempt to
suspend and capture the same request.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1168
Fixes: 32ff621fd744 ("drm/i915/gt: Allow temporary suspension of inflight requests")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v5.7+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201204151234.19729-1-chris@chris-wilson.co.uk

drm/i915/gt: Check for a completed last request once

Pull the repeated check for the last active request being completed to a
single spot, when deciding whether or not execlist preemption is
required.

In doing so, we remove the tasklet kick, introduced with the completion
checks in commit 35f3fd8182ba ("drm/i915/execlists: Workaround switching
back to a completed context"), if we find the request was completed but
have not yet seen the corresponding CS event. This was devolving into a
busy spin of the tasklet while we waited for the event as the delivery
was not as instantaneous as expected. Under load this is sufficient to
exhaust the tasklet softirq timeslice, and force ksoftirqd. Quite
noticeable overhead for no apparent improvement in latency.

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/20201126140407.31952-2-chris@chris-wilson.co.uk

drm/i915/gt: Decouple completed requests on unwind

Since the introduction of preempt-to-busy, requests can complete in the
background, even while they are not on the engine->active.requests list.
As such, the engine->active.request list itself is not in strict
retirement order, and we have to scan the entire list while unwinding to
not miss any. However, if the request is completed we currently leave it
on the list [until retirement], but we could just as simply remove it
and stop treating it as active. We would only have to then traverse it
once while unwinding in quick succession.

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/20201126140407.31952-1-chris@chris-wilson.co.uk

drm/i915/gt: Free stale request on destroying the virtual engine

Since preempt-to-busy, we may unsubmit a request while it is still on
the HW and completes asynchronously. That means it may be retired and in
the process destroy the virtual engine (as the user has closed their
context), but that engine may still be holding onto the unsubmitted
compelted request. Therefore we need to potentially cleanup the old
request on destroying the virtual engine. We also have to keep the
virtual_engine alive until after the sibling's execlists_dequeue() have
finished peeking into the virtual engines, for which we serialise with
RCU.

v2: Be paranoid and flush the tasklet as well.
v3: And flush the tasklet before the engines, as the tasklet may
re-attach an rb_node after our removal from the siblings.

Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine")
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/20201123113717.20500-4-chris@chris-wilson.co.uk

drm/i915: Lift waiter/signaler iterators

Lift the list iteration defines for traversing the signaler/waiter lists
into i915_scheduler.h for reuse.

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/20201119165616.10834-5-chris@chris-wilson.co.uk

drm/i915/gt: Update request status flags for debug pretty-printer

We plan to expand upon the number of available statuses for when we
pretty-print the requests along the timelines, and so need a new set of
flags. We have settled upon:

Unready [U]
- initial status after being submitted, the request is not
ready for execution as it is waiting for external fences

Ready [R]
- all fences the request was waiting on have been signaled,
and the request is now ready for execution and will be
in a backend queue

- a ready request may still need to wait on semaphores
[internal fences]

Ready/virtual [V]
- same as ready, but queued over multiple backends

Executing [E]
- the request has been transferred from the backend queue and
submitted for execution on HW

- a completed request may still be regarded as executing, its
status may not be updated until it is retired and removed
from the lists

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/20201119165616.10834-3-chris@chris-wilson.co.uk

drm/i915: Lift i915_request_show()

Extract i915_request_show for reuse in other request chain pretty
printers.

For a bonus point, quietly change the seqno format from %llx to %lld to
match everywhere else.

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/20201119165616.10834-2-chris@chris-wilson.co.uk

drm/i915/gt: Remember to free the virtual breadcrumbs

Since we allocate some breadcrumbs for the virtual engine, and the
virtual engine has a custom destructor, we also need to free the
breadcrumbs after use.

Fixes: b3786b29379c ("drm/i915/gt: Distinguish the virtual breadcrumbs from the irq breadcrumbs")
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/20201118133839.1783-1-chris@chris-wilson.co.uk

drm/i915/gt: Ignore dt==0 for reporting underflows

The presumption was that some time would always elapse between recording
the start and the finish of a context switch. This turns out to be a
regular occurrence and emitting a debug statement superfluous.

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/20201117113103.21480-4-chris@chris-wilson.co.uk

drm/i915/gt: Prevent use of engine->wa_ctx after error

On error we unpin and free the wa_ctx.vma, but do not clear any of the
derived flags. During lrc_init, we look at the flags and attempt to
dereference the wa_ctx.vma if they are set. To protect the error path
where we try to limp along without the wa_ctx, make sure we clear those
flags!

Reported-by: Matt Roper <matthew.d.roper@intel.com>
Fixes: 604a8f6f1e33 ("drm/i915/lrc: Only enable per-context and per-bb buffers if set")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v4.15+
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210108204026.20682-1-chris@chris-wilson.co.uk
(cherry-picked from 5b4dc95cf7f573e927fbbd406ebe54225d41b9b2)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210118095332.458813-1-chris@chris-wilson.co.uk

drm/i915/gt: Cancel the preemption timeout on responding to it

We currently presume that the engine reset is successful, cancelling the
expired preemption timer in the process. However, engine resets can
fail, leaving the timeout still pending and we will then respond to the
timeout again next time the tasklet fires. What we want is for the
failed engine reset to be promoted to a full device reset, which is
kicked by the heartbeat once the engine stops processing events.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1168
Fixes: 3a7a92aba8fb ("drm/i915/execlists: Force preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201204151234.19729-2-chris@chris-wilson.co.uk
(cherry picked from commit d997e240ceecb4f732611985d3a939ad1bfc1893)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Ignore repeated attempts to suspend request flow across reset

Before reseting the engine, we suspend the execution of the guilty
request, so that we can continue execution with a new context while we
slowly compress the captured error state for the guilty context. However,
if the reset fails, we will promptly attempt to reset the same request
again, and discover the ongoing capture. Ignore the second attempt to
suspend and capture the same request.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1168
Fixes: 32ff621fd744 ("drm/i915/gt: Allow temporary suspension of inflight requests")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v5.7+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201204151234.19729-1-chris@chris-wilson.co.uk
(cherry picked from commit b969540500bce60cf1cdfff5464388af32b9a553)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Free stale request on destroying the virtual engine

Since preempt-to-busy, we may unsubmit a request while it is still on
the HW and completes asynchronously. That means it may be retired and in
the process destroy the virtual engine (as the user has closed their
context), but that engine may still be holding onto the unsubmitted
compelted request. Therefore we need to potentially cleanup the old
request on destroying the virtual engine. We also have to keep the
virtual_engine alive until after the sibling's execlists_dequeue() have
finished peeking into the virtual engines, for which we serialise with
RCU.

v2: Be paranoid and flush the tasklet as well.
v3: And flush the tasklet before the engines, as the tasklet may
re-attach an rb_node after our removal from the siblings.

Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine")
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/20201123113717.20500-4-chris@chris-wilson.co.uk
(cherry picked from commit 46eecfccb4c2b0f258adbafb2e53ca3b822cd663)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Remember to free the virtual breadcrumbs

Since we allocate some breadcrumbs for the virtual engine, and the
virtual engine has a custom destructor, we also need to free the
breadcrumbs after use.

Fixes: b3786b29379c ("drm/i915/gt: Distinguish the virtual breadcrumbs from the irq breadcrumbs")
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/20201118133839.1783-1-chris@chris-wilson.co.uk
(cherry picked from commit 45e50f48b7907e650cfbbc7879abfe3a0c419c73)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915: Improve record of hung engines in error state

Between events which trigger engine and GPU resets and capturing the error
state we lose information on which engine triggered the reset. Improve
this by passing in the hung engine mask down to error capture.

Result is that the list of engines in user visible "GPU HANG: ecode
<gen>:<engines>:<ecode>, <process>" is now a list of hanging and not just
active engines. Most importantly the displayed process is now the one
which was actually hung.

v2:
* Stub prototype. (Chris)

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/20201104134743.916027-1-tvrtko.ursulin@linux.intel.com

drm/i915/gt: Flush xcs before tgl breadcrumbs

In a simple test case that writes to scratch and then busy-waits for the
batch to be signaled, we observe that the signal is before the write is
posted. That is bad news.

Splitting the flush + write_dword into two separate flush_dw prevents
the issue from being reproduced, we can presume the post-sync op is not
so post-sync.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/216
Testcase: igt/gem_exec_fence/parallel
Testcase: igt/i915_selftest/live/gt_timelines
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: stable@vger.kernel.org
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201102221057.29626-2-chris@chris-wilson.co.uk
(cherry picked from commit 09212e81e5450743e5b06b27c4e344e4c45b630d)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Use the local HWSP offset during submission

We wrap the timeline on construction of the next request, but there may
still be requests in flight that have not yet finalized the breadcrumb.
(The breadcrumb is delayed as we need engine-local offsets, and for the
virtual engine that is not known until execution.) As such, by the time
we write to the timeline's HWSP offset it may have changed, and we
should use the value we preserved in the request instead.

Though the window is small and infrequent (at full flow we can expect a
timeline's seqno to wrap once every 30 minutes), the impact of writing
the old seqno into the new HWSP is severe: the old requests are never
completed, and the new requests are completed before they are even
submitted.

Fixes: ebece7539242 ("drm/i915: Keep timeline HWSP allocated until idle across the system")
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>
Cc: <stable@vger.kernel.org> # v5.2+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201022064127.10159-1-chris@chris-wilson.co.uk
(cherry picked from commit c10f6019d0b2dc8a6a62b55459f3ada5bc4e5e1a)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Flush xcs before tgl breadcrumbs

In a simple test case that writes to scratch and then busy-waits for the
batch to be signaled, we observe that the signal is before the write is
posted. That is bad news.

Splitting the flush + write_dword into two separate flush_dw prevents
the issue from being reproduced, we can presume the post-sync op is not
so post-sync.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/216
Testcase: igt/gem_exec_fence/parallel
Testcase: igt/i915_selftest/live/gt_timelines
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: stable@vger.kernel.org
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201102221057.29626-2-chris@chris-wilson.co.uk

drm/i915/gt: Use the local HWSP offset during submission

We wrap the timeline on construction of the next request, but there may
still be requests in flight that have not yet finalized the breadcrumb.
(The breadcrumb is delayed as we need engine-local offsets, and for the
virtual engine that is not known until execution.) As such, by the time
we write to the timeline's HWSP offset it may have changed, and we
should use the value we preserved in the request instead.

Though the window is small and infrequent (at full flow we can expect a
timeline's seqno to wrap once every 30 minutes), the impact of writing
the old seqno into the new HWSP is severe: the old requests are never
completed, and the new requests are completed before they are even
submitted.

Fixes: ebece7539242 ("drm/i915: Keep timeline HWSP allocated until idle across the system")
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>
Cc: <stable@vger.kernel.org> # v5.2+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201022064127.10159-1-chris@chris-wilson.co.uk

drm/i915/gt: Wait for CSB entries on Tigerlake

On Tigerlake, we are seeing a repeat of commit d8f505311717 ("drm/i915/icl:
Forcibly evict stale csb entries") where, presumably, due to a missing
Global Observation Point synchronisation, the write pointer of the CSB
ringbuffer is updated _prior_ to the contents of the ringbuffer. That is
we see the GPU report more context-switch entries for us to parse, but
those entries have not been written, leading us to process stale events,
and eventually report a hung GPU.

However, this effect appears to be much more severe than we previously
saw on Icelake (though it might be best if we try the same approach
there as well and measure), and Bruce suggested the good idea of resetting
the CSB entry after use so that we can detect when it has been updated by
the GPU. By instrumenting how long that may be, we can set a reliable
upper bound for how long we should wait for:

513 late, avg of 61 retries (590 ns), max of 1061 retries (10099 ns)

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2045
References: d8f505311717 ("drm/i915/icl: Forcibly evict stale csb entries")
References: HSDES#22011327657, HSDES#1508287568
Suggested-by: Bruce Chang <yu.bruce.chang@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: stable@vger.kernel.org # v5.4
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-2-chris@chris-wilson.co.uk
(cherry picked from commit 233c1ae3c83f21046c6c4083da904163ece8f110)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Widen CSB pointer to u64 for the parsers

A CSB entry is 64b, and it is simpler for us to treat it as an array of
64b entries than as an array of pairs of 32b entries.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-1-chris@chris-wilson.co.uk
(cherry picked from commit f24a44e52fbc9881fc5f3bcef536831a15a439f3)
(cherry picked from commit 3d4dbe0e0f0d04ebcea917b7279586817da8cf46)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Undo forced context restores after trivial preemptions

We may try to preempt the currently executing request, only to find that
after unravelling all the dependencies that the original executing
context is still the earliest in the topological sort and re-submitted
back to HW (if we do detect some change in the ELSP that requires
re-submission). However, due to the way we check for wrap-around during
the unravelling, we mark any context that has been submitted just once
(i.e. with the rq->wa_tail set, but the ring->tail earlier) as
potentially wrapping and requiring a forced restore on resubmission.
This was expected to be not a problem, as it was anticipated that most
unwinding for preemption would result in a context switch and the few
that did not would be lost in the noise. It did not take long for
someone to find one particular workload where the cost of those extra
context restores was measurable.

However, since we know the wa_tail is of fixed size, and we know that a
request must be larger than the wa_tail itself, we can safely maintain
the check for request wrapping and check against a slightly future point
in the ring that includes an expected wa_tail. (That is if the
ring->tail is already set to rq->wa_tail, including another 8 bytes in
the check does not invalidate the incremental wrap detection.)

Fixes: 8ab3a3812aa9 ("drm/i915/gt: Incrementally check for rewinding")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Ramalingam C <ramalingam.c@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201002083425.4605-1-chris@chris-wilson.co.uk
(cherry picked from commit bb65548e3c6e299175a9e8c3e24b2b9577656a5d)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Delay execlist processing for tgl

When running gem_exec_nop, it floods the system with many requests (with
the goal of userspace submitting faster than the HW can process a single
empty batch). This causes the driver to continually resubmit new
requests onto the end of an active context, a flood of lite-restore
preemptions. If we time this just right, Tigerlake hangs.

Inserting a small delay between the processing of CS events and
submitting the next context, prevents the hang. Naturally it does not
occur with debugging enabled. The suspicion then is that this is related
to the issues with the CS event buffer, and inserting an mmio read of
the CS pointer status appears to be very successful in preventing the
hang. Other registers, or uncached reads, or plain mb, do not prevent
the hang, suggesting that register is key -- but that the hang can be
prevented by a simple udelay, suggests it is just a timing issue like
that encountered by commit 233c1ae3c83f ("drm/i915/gt: Wait for CSB
entries on Tigerlake"). Also note that the hang is not prevented by
applying CTX_DESC_FORCE_RESTORE, or by inserting a delay on the GPU
between requests.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: stable@vger.kernel.org
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201015195023.32346-1-chris@chris-wilson.co.uk
(cherry picked from commit 6ca7217dffaf1abba91558e67a2efb655ac91405)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Confirm the context survives execution

Repeat our sanitychecks from before execution to after execution. One
expects that if we were to see these, the gpu would already be on fire,
but the timing may be informative.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201015190816.31763-1-chris@chris-wilson.co.uk

drm/i915/gt: Undo forced context restores after trivial preemptions

We may try to preempt the currently executing request, only to find that
after unravelling all the dependencies that the original executing
context is still the earliest in the topological sort and re-submitted
back to HW (if we do detect some change in the ELSP that requires
re-submission). However, due to the way we check for wrap-around during
the unravelling, we mark any context that has been submitted just once
(i.e. with the rq->wa_tail set, but the ring->tail earlier) as
potentially wrapping and requiring a forced restore on resubmission.
This was expected to be not a problem, as it was anticipated that most
unwinding for preemption would result in a context switch and the few
that did not would be lost in the noise. It did not take long for
someone to find one particular workload where the cost of those extra
context restores was measurable.

However, since we know the wa_tail is of fixed size, and we know that a
request must be larger than the wa_tail itself, we can safely maintain
the check for request wrapping and check against a slightly future point
in the ring that includes an expected wa_tail. (That is if the
ring->tail is already set to rq->wa_tail, including another 8 bytes in
the check does not invalidate the incremental wrap detection.)

Fixes: 8ab3a3812aa9 ("drm/i915/gt: Incrementally check for rewinding")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Ramalingam C <ramalingam.c@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201002083425.4605-1-chris@chris-wilson.co.uk

drm/i915/gt: Delay execlist processing for tgl

When running gem_exec_nop, it floods the system with many requests (with
the goal of userspace submitting faster than the HW can process a single
empty batch). This causes the driver to continually resubmit new
requests onto the end of an active context, a flood of lite-restore
preemptions. If we time this just right, Tigerlake hangs.

Inserting a small delay between the processing of CS events and
submitting the next context, prevents the hang. Naturally it does not
occur with debugging enabled. The suspicion then is that this is related
to the issues with the CS event buffer, and inserting an mmio read of
the CS pointer status appears to be very successful in preventing the
hang. Other registers, or uncached reads, or plain mb, do not prevent
the hang, suggesting that register is key -- but that the hang can be
prevented by a simple udelay, suggests it is just a timing issue like
that encountered by commit 233c1ae3c83f ("drm/i915/gt: Wait for CSB
entries on Tigerlake"). Also note that the hang is not prevented by
applying CTX_DESC_FORCE_RESTORE, or by inserting a delay on the GPU
between requests.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: stable@vger.kernel.org
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201015195023.32346-1-chris@chris-wilson.co.uk

drm/i915/gt: Signal cancelled requests

After marking the requests on an engine as cancelled upon wedging, send
any signals for their completions.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200930163253.2789-1-chris@chris-wilson.co.uk

drm/i915/gt: Remove defunct intel_virtual_engine_get_sibling()

As the last user was eliminated in commit e21fecdcde40 ("drm/i915/gt:
Distinguish the virtual breadcrumbs from the irq breadcrumbs"), we can
remove the function. One less implementation detail creeping beyond its
scope.

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/20200826132811.17577-16-chris@chris-wilson.co.uk

drm/i915: Be wary of data races when reading the active execlists

To implement preempt-to-busy (and so efficient timeslicing and best utilization
of the hardware submission ports) we let the GPU run asynchronously in respect
to the ELSP submission queue. This created challenges in keeping and accessing
the driver state mirroring the asynchronous GPU execution.

The latest occurence of this was spotted by KCSAN:

[ 1413.563200] BUG: KCSAN: data-race in __await_execution+0x217/0x370 [i915]
[ 1413.563221]
[ 1413.563236] race at unknown origin, with read to 0xffff88885bb6c478 of 8 bytes by task 9654 on cpu 1:
[ 1413.563548] __await_execution+0x217/0x370 [i915]
[ 1413.563891] i915_request_await_dma_fence+0x4eb/0x6a0 [i915]
[ 1413.564235] i915_request_await_object+0x421/0x490 [i915]
[ 1413.564577] i915_gem_do_execbuffer+0x29b7/0x3c40 [i915]
[ 1413.564967] i915_gem_execbuffer2_ioctl+0x22f/0x5c0 [i915]
[ 1413.564998] drm_ioctl_kernel+0x156/0x1b0
[ 1413.565022] drm_ioctl+0x2ff/0x480
[ 1413.565046] __x64_sys_ioctl+0x87/0xd0
[ 1413.565069] do_syscall_64+0x4d/0x80
[ 1413.565094] entry_SYSCALL_64_after_hwframe+0x44/0xa9

To complicate matters, we have to both avoid the read tearing of *active and
avoid any write tearing as perform the pending[] -> inflight[] promotion of the
execlists.

This is because we cannot rely on the memcpy doing u64 aligned copies on all
kernels/platforms and so we opt to open-code it with explicit WRITE_ONCE
annotations to satisfy KCSAN.

v2: When in doubt, write the same comment again.
v3: Expanded commit message.

Fixes: b55230e5e800 ("drm/i915: Check for awaits on still currently executing requests")
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/20200716142207.13003-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
[Joonas: Rebased and reordered into drm-intel-gt-next branch]
[Joonas: Added expanded commit message from Tvrtko and Chris]
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
(cherry picked from commit b4d9145b0154f8c71dafc2db5fd445f1f3db9426)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/gt: Use a mmio read of the CSB in case of failure

If we find the GPU didn't update the CSB within 50us, we currently fail
and eventually reset the GPU. Lets report the value from the mmio space
as a last resort, it may just stave off an unnecessary GPU reset.

References: HSDES#22011327657
Suggested-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-4-chris@chris-wilson.co.uk

drm/i915/gt: Apply the CSB w/a for all

Since we expect to inline the csb_parse() routines, the w/a for the
stale CSB data on Tigerlake will be pulled into process_csb(), and so we
might as well simply reuse the logic for all, and so will hopefully
avoid any strange behaviour on Icelake that was not covered by our
previous w/a.

References: d8f505311717 ("drm/i915/icl: Forcibly evict stale csb entries")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-3-chris@chris-wilson.co.uk

drm/i915/gt: Wait for CSB entries on Tigerlake

On Tigerlake, we are seeing a repeat of commit d8f505311717 ("drm/i915/icl:
Forcibly evict stale csb entries") where, presumably, due to a missing
Global Observation Point synchronisation, the write pointer of the CSB
ringbuffer is updated _prior_ to the contents of the ringbuffer. That is
we see the GPU report more context-switch entries for us to parse, but
those entries have not been written, leading us to process stale events,
and eventually report a hung GPU.

However, this effect appears to be much more severe than we previously
saw on Icelake (though it might be best if we try the same approach
there as well and measure), and Bruce suggested the good idea of resetting
the CSB entry after use so that we can detect when it has been updated by
the GPU. By instrumenting how long that may be, we can set a reliable
upper bound for how long we should wait for:

513 late, avg of 61 retries (590 ns), max of 1061 retries (10099 ns)

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2045
References: d8f505311717 ("drm/i915/icl: Forcibly evict stale csb entries")
References: HSDES#22011327657, HSDES#1508287568
Suggested-by: Bruce Chang <yu.bruce.chang@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: stable@vger.kernel.org # v5.4
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-2-chris@chris-wilson.co.uk

drm/i915/gt: Widen CSB pointer to u64 for the parsers

A CSB entry is 64b, and it is simpler for us to treat it as an array of
64b entries than as an array of pairs of 32b entries.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-1-chris@chris-wilson.co.uk

drm/i915: Be wary of data races when reading the active execlists

To implement preempt-to-busy (and so efficient timeslicing and best utilization
of the hardware submission ports) we let the GPU run asynchronously in respect
to the ELSP submission queue. This created challenges in keeping and accessing
the driver state mirroring the asynchronous GPU execution.

The latest occurence of this was spotted by KCSAN:

[ 1413.563200] BUG: KCSAN: data-race in __await_execution+0x217/0x370 [i915]
[ 1413.563221]
[ 1413.563236] race at unknown origin, with read to 0xffff88885bb6c478 of 8 bytes by task 9654 on cpu 1:
[ 1413.563548] __await_execution+0x217/0x370 [i915]
[ 1413.563891] i915_request_await_dma_fence+0x4eb/0x6a0 [i915]
[ 1413.564235] i915_request_await_object+0x421/0x490 [i915]
[ 1413.564577] i915_gem_do_execbuffer+0x29b7/0x3c40 [i915]
[ 1413.564967] i915_gem_execbuffer2_ioctl+0x22f/0x5c0 [i915]
[ 1413.564998] drm_ioctl_kernel+0x156/0x1b0
[ 1413.565022] drm_ioctl+0x2ff/0x480
[ 1413.565046] __x64_sys_ioctl+0x87/0xd0
[ 1413.565069] do_syscall_64+0x4d/0x80
[ 1413.565094] entry_SYSCALL_64_after_hwframe+0x44/0xa9

To complicate matters, we have to both avoid the read tearing of *active and
avoid any write tearing as perform the pending[] -> inflight[] promotion of the
execlists.

This is because we cannot rely on the memcpy doing u64 aligned copies on all
kernels/platforms and so we opt to open-code it with explicit WRITE_ONCE
annotations to satisfy KCSAN.

v2: When in doubt, write the same comment again.
v3: Expanded commit message.

Fixes: b55230e5e800 ("drm/i915: Check for awaits on still currently executing requests")
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/20200716142207.13003-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
[Joonas: Rebased and reordered into drm-intel-gt-next branch]
[Joonas: Added expanded commit message from Tvrtko and Chris]
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915: Make sure execbuffer always passes ww state to i915_vma_pin.

As a preparation step for full object locking and wait/wound handling
during pin and object mapping, ensure that we always pass the ww context
in i915_gem_execbuffer.c to i915_vma_pin, use lockdep to ensure this
happens.

This also requires changing the order of eb_parse slightly, to ensure
we pass ww at a point where we could still handle -EDEADLK safely.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200819140904.1708856-15-maarten.lankhorst@linux.intel.com
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915: Rework intel_context pinning to do everything outside of pin_mutex

Instead of doing everything inside of pin_mutex, we move all pinning
outside. Because i915_active has its own reference counting and
pinning is also having the same issues vs mutexes, we make sure
everything is pinned first, so the pinning in i915_active only needs
to bump refcounts. This allows us to take pin refcounts correctly
all the time.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200819140904.1708856-14-maarten.lankhorst@linux.intel.com
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Distinguish the virtual breadcrumbs from the irq breadcrumbs

On the virtual engines, we only use the intel_breadcrumbs for tracking
signaling of stale breadcrumbs from the irq_workers. They do not have
any associated interrupt handling, active requests are passed to a
physical engine and associated breadcrumb interrupt handler. This causes
issues for us as we need to ensure that we do not actually try and
enable interrupts and the powermanagement required for them on the
virtual engine, as they will never be disabled. Instead, let's
specify the physical engine used for interrupt handler on a particular
breadcrumb.

v2: Drop b->irq_armed = true mocking for no interrupt HW

Fixes: 4fe6abb8f513 ("drm/i915/gt: Ignore irq enabling on the virtual 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/20200731154834.8378-4-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Only transfer the virtual context to the new engine if active

One more complication of preempt-to-busy with respect to the virtual
engine is that we may have retired the last request along the virtual
engine at the same time as preparing to submit the completed request to
a new engine. That submit will be shortcircuited, but not before we have
updated the context with the new register offsets and marked the virtual
engine as bound to the new engine (by calling swap on ve->siblings[]).
As we may have just retired the completed request, we may also be in the
middle of calling virtual_context_exit() to turn off the power management
associated with the virtual engine, and that in turn walks the
ve->siblings[]. If we happen to call swap() on the array as we walk, we
will call intel_engine_pm_put() twice on the same engine.

In this patch, we prevent this by only updating the bound engine after a
successful submission which weeds out the already completed requests.

Alternatively, we could walk a non-volatile array for the pm, such as
using the engine->mask. The small advantage to performing the update
after the submit is that we then only have to do a swap for active
requests.

Fixes: 22b7a426bbe1 ("drm/i915/execlists: Preempt-to-busy")
References: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine"
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: "Nayana, Venkata Ramana" <venkata.ramana.nayana@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200731154834.8378-3-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Replace intel_engine_transfer_stale_breadcrumbs

After staring at the breadcrumb enabling/cancellation and coming to the
conclusion that the cause of the mysterious stale breadcrumbs must the
act of submitting a completed requests, we can then redirect those
completed requests onto a dedicated signaled_list at the time of
construction and so eliminate intel_engine_transfer_stale_breadcrumbs().

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/20200731154834.8378-2-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915: Remove requirement for holding i915_request.lock for breadcrumbs

Since the breadcrumb enabling/cancelling itself is serialised by the
breadcrumbs.irq_lock, with a bit of care we can remove the outer
serialisation with i915_request.lock for concurrent
dma_fence_enable_signaling(). This has the important side-effect of
eliminating the nested i915_request.lock within request submission.

The challenge in serialisation is around the unsubmission where we take
an active request that wants a breadcrumb on the signaling engine and
put it to sleep. We do not want a concurrent
dma_fence_enable_signaling() to attach a breadcrumb as we unsubmit, so
we must mark the request as no longer active before serialising with the
concurrent enable-signaling.

On retire, we serialise with the concurrent enable-signaling, but
instead of clearing ACTIVE, we mark it as SIGNALED.

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/20200731154834.8378-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
[Joonas: Rebased and reordered into drm-intel-gt-next branch]
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Support multiple pinned timelines

We may need to allocate more than one pinned context/timeline for each
engine which can utilise the per-engine HWSP, so we need to give each
a different offset within it.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200730183906.25422-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Disable preparser around xcs invalidations on tgl

Unlike rcs where we have conclusive evidence from our selftesting that
disabling the preparser before performing the TLB invalidate and
relocations does impact upon the GPU execution, the evidence for the
same requirement on xcs is much more circumstantial. Let's apply the
preparser disable between batches as we invalidate the TLB as a dose of
healthy paranoia, just in case.

References: https://gitlab.freedesktop.org/drm/intel/-/issues/2169
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200728152110.830-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/kbl: Fix revision ID checks

We usually assume that increasing PCI device revision ID's translates to
newer steppings; macros like IS_KBL_REVID() that we use rely on this
behavior. Unfortunately this turns out to not be true on KBL; the
newer device 2 revision ID's sometimes go backward to older steppings.
The situation is further complicated by different GT and display
steppings associated with each revision ID.

Let's work around this by providing a table to map the revision ID to
specific GT and display steppings, and then perform our comparisons on
the mapped values.

v2:
- Move the kbl_revids[] array to intel_workarounds.c to avoid compiler
warnings about an unused variable in files that don't call the
macros (kernel test robot).

Bspec: 18329
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200811032105.2819370-1-matthew.d.roper@intel.com
Reviewed-by: Swathi Dhanavanthri <swathi.dhanavanthri@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

treewide: Remove uninitialized_var() usage

Using uninitialized_var() is dangerous as it papers over real bugs[1]
(or can in the future), and suppresses unrelated compiler warnings
(e.g. "unused variable"). If the compiler thinks it is uninitialized,
either simply initialize the variable or make compiler changes.

In preparation for removing[2] the[3] macro[4], remove all remaining
needless uses with the following script:

git grep '\buninitialized_var\b' | cut -d: -f1 | sort -u | \
xargs perl -pi -e \
's/\buninitialized_var\(([^\)]+)\)/\1/g;
s:\s*/\* (GCC be quiet|to make compiler happy) \*/$::g;'

drivers/video/fbdev/riva/riva_hw.c was manually tweaked to avoid
pathological white-space.

No outstanding warnings were found building allmodconfig with GCC 9.3.0
for x86_64, i386, arm64, arm, powerpc, powerpc64le, s390x, mips, sparc64,
alpha, and m68k.

[1] https://lore.kernel.org/lkml/20200603174714.192027-1-glider@google.com/
[2] https://lore.kernel.org/lkml/CA+55aFw+Vbj0i=1TGqCR5vQkCzWJ0QxK6CernOU6eedsudAixw@mail.gmail.com/
[3] https://lore.kernel.org/lkml/CA+55aFwgbgqhbp1fkxvRKEpzyR5J8n1vKT1VZdz9knmPuXhOeg@mail.gmail.com/
[4] https://lore.kernel.org/lkml/CA+55aFz2500WfbKXAx8s67wrm9=yVJu65TpLgN_ybYNv0VEOKA@mail.gmail.com/

Reviewed-by: Leon Romanovsky <leonro@mellanox.com> # drivers/infiniband and mlx4/mlx5
Acked-by: Jason Gunthorpe <jgg@mellanox.com> # IB
Acked-by: Kalle Valo <kvalo@codeaurora.org> # wireless drivers
Reviewed-by: Chao Yu <yuchao0@huawei.com> # erofs
Signed-off-by: Kees Cook <keescook@chromium.org>

drm/i915/gt: Only swap to a random sibling once upon creation

The danger in switching at random upon intel_context_pin is that the
context may still actually be inflight, as it will not be scheduled out
until a context switch after it is complete -- that may be a long time
after we do a final intel_context_unpin.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2118
Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.3+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200713160549.17344-1-chris@chris-wilson.co.uk
(cherry picked from commit 90a987205c6cf74116a102ed446d22d92cdaf915)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/gt: Ignore irq enabling on the virtual engines

We do not use the virtual engines for interrupts (they have physical
components), but we do use them to decouple the fence signaling during
submission. Currently, when we submit a completed request, we try to
enable the interrupt handler for the virtual engine, but we never disarm
it. A quick fix is then to mark the irq as enabled, and it will then
remain enabled -- and this prevents us from waking the device and never
letting it sleep again.

Fixes: f8db4d051b5e ("drm/i915: Initialise breadcrumb lists on the virtual engine")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200711203236.12330-1-chris@chris-wilson.co.uk
(cherry picked from commit 4fe6abb8f51355224808ab02a9febf65d184c40b)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/gt: Only swap to a random sibling once upon creation

The danger in switching at random upon intel_context_pin is that the
context may still actually be inflight, as it will not be scheduled out
until a context switch after it is complete -- that may be a long time
after we do a final intel_context_unpin.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2118
Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.3+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200713160549.17344-1-chris@chris-wilson.co.uk

drm/i915/gt: Ignore irq enabling on the virtual engines

We do not use the virtual engines for interrupts (they have physical
components), but we do use them to decouple the fence signaling during
submission. Currently, when we submit a completed request, we try to
enable the interrupt handler for the virtual engine, but we never disarm
it. A quick fix is then to mark the irq as enabled, and it will then
remain enabled -- and this prevents us from waking the device and never
letting it sleep again.

Fixes: f8db4d051b5e ("drm/i915: Initialise breadcrumb lists on the virtual engine")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200711203236.12330-1-chris@chris-wilson.co.uk

drm/i915/gt: Always reset the engine, even if inactive, on execlists failure

If something has gone awry with the CSB processing, we need to pause,
unwind and restart the request submission and event processing. However,
currently we skip the engine reset if we raise an error but discover no
active context, in the mistaken belief that it was merely a glitch in
the matrix. The glitches are real enough, and we do need to unwind even
if the engine appears idle (as it has gone permanently idle!) The
simplest way to unwind and recover is simply do the engine reset, which
should be very fast and _safe_ as nothing is active.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200711091349.28865-1-chris@chris-wilson.co.uk

drm/i915/gt: Be defensive in the face of false CS events

If the HW throws a curve ball and reports either en event before it is
possible, or just a completely impossible event, we have to grin and
bear it. The first few events, we will likely not notice as we would be
expecting some event, but as soon as we stop expecting an event and yet
they still keep coming, then we enter into undefined state territory.
In which case, bail out, stop processing the events, and reset the
engine and our set of queued requests to recover.

The sporadic hangs and warnings will continue to plague CI, but at least
system stability should not be compromised.

v2: Commentary and force the reset-on-error.
v3: Customised user facing message for forced resets from internal errors.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2045
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/20200710133125.30194-1-chris@chris-wilson.co.uk

drm/i915: Release shortlived maps of longlived objects

Some objects we map once during their construction, and then never
access their mappings again, even if they are kept around for the
duration of the driver. Keeping those pages mapped, often vmapped, is
therefore wasteful and we should release the maps as soon as we no
longer need them.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200708173748.32734-3-chris@chris-wilson.co.uk

drm/i915/gt: Replace opencoded i915_gem_object_pin_map()

As we have a pin_map interface, that knows how to flush the data to the
device, use it. The only downside is that we keep the kmap around, as
once acquired we keep the mapping cached until the object's backing
store is released.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200708173748.32734-2-chris@chris-wilson.co.uk

drm/i915/sseu: Move sseu_info under gt_info

SSEUs are a GT capability, so track them under gt_info.

Signed-off-by: Venkata Sandeep Dhanalakota <venkata.s.dhanalakota@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Andi Shyti <andi.shyti@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/20200708003952.21831-8-daniele.ceraolospurio@intel.com

drm/i915/gt: Incrementally check for rewinding

In commit 5ba32c7be81e ("drm/i915/execlists: Always force a context
reload when rewinding RING_TAIL"), we placed the check for rewinding a
context on actually submitting the next request in that context. This
was so that we only had to check once, and could do so with precision
avoiding as many forced restores as possible. For example, to ensure
that we can resubmit the same request a couple of times, we include a
small wa_tail such that on the next submission, the ring->tail will
appear to move forwards when resubmitting the same request. This is very
common as it will happen for every lite-restore to fill the second port
after a context switch.

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

Fixes: 5ba32c7be81e ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Bruce Chang <yu.bruce.chang@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200609151723.12971-1-chris@chris-wilson.co.uk
(cherry picked from commit e36ba817fa966f81fb1c8d16f3721b5a644b2fa9)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Prevent timeslicing into unpreemptable requests

We have a I915_REQUEST_NOPREEMPT flag that we set when we must prevent
the HW from preempting during the course of this request. We need to
honour this flag and protect the HW even if we have a heartbeat request,
or other maximum priority barrier, pending. As such, restrict the
timeslicing check to avoid preempting into the topmost priority band,
leaving the unpreemptable requests in blissful peace running
uninterrupted on the HW.

v2: Set the I915_PRIORITY_BARRIER to be less than
I915_PRIORITY_UNPREEMPTABLE so that we never submit a request
(heartbeat or barrier) that can legitimately preempt the current
non-premptable request.

Fixes: 2a98f4e65bba ("drm/i915: add infrastructure to hold off preemption on a request")
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/20200527162418.24755-1-chris@chris-wilson.co.uk
(cherry picked from commit b72f02d78e4f257761ed003444ae52083f962076)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/gt: Incorporate the virtual engine into timeslicing

It was quite the oversight to only factor in the normal queue to decide
the timeslicing switch priority. By leaving out the next virtual request
from the priority decision, we would not timeslice the current engine if
there was an available virtual request.

Testcase: igt/gem_exec_balancer/sliced
Fixes: 3df2deed411e ("drm/i915/execlists: Enable timeslice on partial virtual engine dequeue")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200519132046.22443-3-chris@chris-wilson.co.uk
(cherry picked from commit 6ad249ba59badc7ff157d4db1f835748f0e2c9b6)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

drm/i915/execlists: Lift opportunistic process_csb to before engine lock

Since the process_csb() does not require us to hold the
engine->active.lock, we can move the opportunistic flush before
direction submission to outside of the lock.

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/20200612221113.9129-1-chris@chris-wilson.co.uk

drm/i915/gt: Incrementally check for rewinding

In commit 5ba32c7be81e ("drm/i915/execlists: Always force a context
reload when rewinding RING_TAIL"), we placed the check for rewinding a
context on actually submitting the next request in that context. This
was so that we only had to check once, and could do so with precision
avoiding as many forced restores as possible. For example, to ensure
that we can resubmit the same request a couple of times, we include a
small wa_tail such that on the next submission, the ring->tail will
appear to move forwards when resubmitting the same request. This is very
common as it will happen for every lite-restore to fill the second port
after a context switch.

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

Fixes: 5ba32c7be81e ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Bruce Chang <yu.bruce.chang@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200609151723.12971-1-chris@chris-wilson.co.uk

drm/i915: Adjust the sentinel assert to match implementation

Sentinels are supposed to be last requests in the elsp queue, not the
only one, so adjust the assert accordingly.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200607222108.14401-1-chris@chris-wilson.co.uk
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

drm/i915/gt: Always check to enable timeslicing if not submitting

We may choose not to submit for a number of reasons, yet not fill both
ELSP. In which case we must start timeslicing (there will be no ACK
event on which to hook the start) if the queue would benefit from the
currently active context being evicted.

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/20200605122334.2798-2-chris@chris-wilson.co.uk

drm/i915/gt: Set timeslicing priority from queue

If we only submit the first port, leaving the second empty yet have
ready requests pending in the queue, use that to set the timeslicing
priority (i.e. the priority at which we will decided to enabling
timeslicing and evict the currently active context if the queue is of
equal priority after its quantum expired).

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/20200605122334.2798-1-chris@chris-wilson.co.uk

drm/i915/gt: Track if an engine requires forcewake w/a

Sometimes an engine might need to keep forcewake active while it is busy
submitting requests for a particular workaround. Track such nuisance
with engine->fw_domain.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Venkata Sandeep Dhanalakota <venkata.s.dhanalakota@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200604153145.21068-1-chris@chris-wilson.co.uk

drm/i915: Drop i915_request.i915 backpointer

We infrequently use the direct i915 backpointer from the i915_request,
so do we really need to waste the space in the struct for it? 8 bytes
from the most frequently allocated struct vs an 3 bytes and pointer
chasing in using rq->engine->i915?

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

drm/i915/gt: Start timeslice on partial submission

We may choose to only submit ELSP[0], even though we have sufficient
requests to fill the whole ELSP. Normally, we only start timeslicing if
we fill more than one port, but in this case we need to start
timeslicing for the queue that we choose not to submit.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200528205727.20309-1-chris@chris-wilson.co.uk

drm/i915/gt: Prevent timeslicing into unpreemptable requests

We have a I915_REQUEST_NOPREEMPT flag that we set when we must prevent
the HW from preempting during the course of this request. We need to
honour this flag and protect the HW even if we have a heartbeat request,
or other maximum priority barrier, pending. As such, restrict the
timeslicing check to avoid preempting into the topmost priority band,
leaving the unpreemptable requests in blissful peace running
uninterrupted on the HW.

v2: Set the I915_PRIORITY_BARRIER to be less than
I915_PRIORITY_UNPREEMPTABLE so that we never submit a request
(heartbeat or barrier) that can legitimately preempt the current
non-premptable request.

Fixes: 2a98f4e65bba ("drm/i915: add infrastructure to hold off preemption on a request")
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/20200527162418.24755-1-chris@chris-wilson.co.uk

drm/i915/execlists: Shortcircuit queue_prio() for no internal levels

If there are no internal levels and the user priority-shift is zero, we
can help the compiler eliminate some dead code:

Function old new delta
start_timeslice 169 154 -15
__execlists_submission_tasklet 4696 4659 -37

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/20200525075347.582-4-chris@chris-wilson.co.uk

drm/i915/gt: Incorporate the virtual engine into timeslicing

It was quite the oversight to only factor in the normal queue to decide
the timeslicing switch priority. By leaving out the next virtual request
from the priority decision, we would not timeslice the current engine if
there was an available virtual request.

Testcase: igt/gem_exec_balancer/sliced
Fixes: 3df2deed411e ("drm/i915/execlists: Enable timeslice on partial virtual engine dequeue")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200519132046.22443-3-chris@chris-wilson.co.uk

drm/i915/gt: Kick virtual siblings on timeslice out

If we decide to timeslice out the current virtual request, we will
unsubmit it while it is still busy (ve->context.inflight == sibling[0]).
If the virtual tasklet and then the other sibling tasklets run before we
completely schedule out the active virtual request for the preemption,
those other tasklets will see that the virtul request is still inflight
on sibling[0] and leave it be. Therefore when we finally schedule-out
the virtual request and if we see that we have passed it back to the
virtual engine, reschedule the virtual tasklet so that it may be
resubmitted on any of the siblings.

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/20200519132046.22443-2-chris@chris-wilson.co.uk

drm/i915/gt: Reuse the tasklet priority for virtual as their siblings

In order to keep all the tasklets in the same execution lists and so
fifo ordered, be consistent and use the same priority for all.

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/20200518081440.17948-3-chris@chris-wilson.co.uk

drm/i915/gt: Transfer old virtual breadcrumbs to irq_worker

The second try at staging the transfer of the breadcrumb. In part one,
we realised we could not simply move to the second engine as we were
only holding the breadcrumb lock on the first. So in commit 6c81e21a4742
("drm/i915/gt: Stage the transfer of the virtual breadcrumb"), we
removed it from the first engine and marked up this request to reattach
the signaling on the new engine. However, this failed to take into
account that we only attach the breadcrumb if the new request is added
at the start of the queue, which if we are transferring, it is because
we know there to be a request to be signaled (and hence we would not be
attached).

In this attempt, we try to transfer the completed requests to the
irq_worker on its rq->engine->breadcrumbs. This preserves the coupling
between the rq and its breadcrumbs, so that
i915_request_cancel_breadcrumb() does not attempt to manipulate the list
under the wrong lock.

v2: Code sharing is fun.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1862
Fixes: 6c81e21a4742 ("drm/i915/gt: Stage the transfer of the virtual breadcrumb")
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/20200513074809.18194-1-chris@chris-wilson.co.uk

drm/i915: Drop no-semaphore boosting

Now that we have fast timeslicing on semaphores, we no longer need to
prioritise none-semaphore work as we will yield any work blocked on a
semaphore to the next in the queue. Previously with no timeslicing,
blocking on the semaphore caused extremely bad scheduling with multiple
clients utilising multiple rings. Now, there is no impact and we can
remove the complication.

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/20200513173504.28322-1-chris@chris-wilson.co.uk

drm/i915: Mark the addition of the initial-breadcrumb in the request

The initial-breadcrumb is used to mark the end of the awaiting and the
beginning of the user payload. We verify that we do not start the user
payload before all signaler are completed, checking our semaphore setup
by looking for the initial breadcrumb being written too early. We also
want to ensure that we do not add semaphore waits after we have already
closed the semaphore section, an issue for later deferred waits.

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/20200513165937.9508-2-chris@chris-wilson.co.uk

drm/i915/gt: Reset execlists registers before HWSP

Upon gt resume, we first poison then sanitize the engine. However, our
testing shows that gen9 will very rarely retain the poisoned value from
the HWSP mappings of the execlists status registers. This suggests that
it is reading back from the HWSP, so rejig the register reset.

v2: Maybe RING_CONTEXT_STATUS_PTR is write masked. It is.

References: https://gitlab.freedesktop.org/drm/intel/-/issues/1812
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200513100120.11617-1-chris@chris-wilson.co.uk

drm/i915: Mark concurrent submissions with a weak-dependency

We recorded the dependencies for WAIT_FOR_SUBMIT in order that we could
correctly perform priority inheritance from the parallel branches to the
common trunk. However, for the purpose of timeslicing and reset
handling, the dependency is weak -- as we the pair of requests are
allowed to run in parallel and not in strict succession.

The real significance though is that this allows us to rearrange
groups of WAIT_FOR_SUBMIT linked requests along the single engine, and
so can resolve user level inter-batch scheduling dependencies from user
semaphores.

Fixes: c81471f5e95c ("drm/i915: Copy across scheduler behaviour flags across submit fences")
Testcase: igt/gem_exec_fence/submit
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.6+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200507155109.8892-1-chris@chris-wilson.co.uk
(cherry picked from commit 6b6cd2ebd8d071e55998e32b648bb8081f7f02bb)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Restore Cherryview back to full-ppgtt

This reverts commit 0b718ba1e884f64dce27c19311dd2859b87e56b9.

There are still some residual issues with asynchronous binding and
execution, but since commit 92581f9fb99c ("drm/i915: Immediately execute
the fenced work") we prefer not to use asynchronous binds, and the
remaining issues do not seem restricted to Cherryview [at least the ones
seen over a few dozen CI runs, less frequent issues are sure to be
discovered!]

These issues seem to be mitigated, if not eliminated entirely, by the
previous commit 84eac0c65940 ("drm/i915/gt: Force pte cacheline to main
memory").

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200510102431.21959-3-chris@chris-wilson.co.uk

drm/i915/gt: Mark up the racy read of execlists->context_tag

Since we are using bitops on context_tag to allow us to reserve and
release inflight tags concurrently, the scan for the next bit is
intentionally racy.

[ 516.446854] BUG: KCSAN: data-race in execlists_schedule_in.isra.0 [i915] / execlists_schedule_out [i915]
[ 516.446874]
[ 516.446886] write (marked) to 0xffff8881f7644048 of 8 bytes by interrupt on cpu 2:
[ 516.447076] execlists_schedule_out+0x538/0x6a0 [i915]
[ 516.447263] process_csb+0x10b/0x3d0 [i915]
[ 516.447449] execlists_submission_tasklet+0x30/0x170 [i915]
[ 516.447468] tasklet_action_common.isra.0+0x42/0x90
[ 516.447484] __do_softirq+0xc8/0x206
[ 516.447498] irq_exit+0xcd/0xe0
[ 516.447516] do_IRQ+0x44/0xc0
[ 516.447535] ret_from_intr+0x0/0x1c
[ 516.447550] cpuidle_enter_state+0x199/0x400
[ 516.447572] cpuidle_enter+0x50/0x90
[ 516.447587] do_idle+0x197/0x1e0
[ 516.447600] cpu_startup_entry+0x14/0x20
[ 516.447619] start_secondary+0xf9/0x130
[ 516.447643] secondary_startup_64+0xa4/0xb0
[ 516.447655]
[ 516.447671] read to 0xffff8881f7644048 of 8 bytes by task 460 on cpu 1:
[ 516.447863] execlists_schedule_in.isra.0+0x3cf/0x5a0 [i915]
[ 516.448064] execlists_dequeue+0xf8f/0x1690 [i915]
[ 516.448252] __execlists_submission_tasklet+0x48/0x60 [i915]
[ 516.448440] execlists_submit_request+0x2e2/0x310 [i915]
[ 516.448634] submit_notify+0x8f/0xc8 [i915]
[ 516.448820] __i915_sw_fence_complete+0x61/0x420 [i915]
[ 516.449005] i915_sw_fence_complete+0x58/0x80 [i915]
[ 516.449208] i915_sw_fence_commit+0x16/0x20 [i915]
[ 516.449399] __i915_request_queue+0x60/0x70 [i915]
[ 516.449590] i915_gem_do_execbuffer+0x33f1/0x4a00 [i915]
[ 516.449782] i915_gem_execbuffer2_ioctl+0x2a2/0x550 [i915]
[ 516.449800] drm_ioctl_kernel+0xe9/0x130
[ 516.449814] drm_ioctl+0x27d/0x45e
[ 516.449827] ksys_ioctl+0x89/0xb0
[ 516.449842] __x64_sys_ioctl+0x42/0x60
[ 516.449864] do_syscall_64+0x6e/0x2c0
[ 516.449878] entry_SYSCALL_64_after_hwframe+0x44/0xa9

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/20200511075722.13483-1-chris@chris-wilson.co.uk

drm/i915: Replace zero-length array with flexible-array

The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:

struct foo {
int stuff;
struct boo array[];
};

By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.

Also, notice that, dynamic memory allocations won't be affected by
this change:

"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]

sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.

This issue was found with the help of Coccinelle.

[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour")

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
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/20200507185408.GA14561@embeddedor

drm/i915/gt: Improve precision on defer_request assert

The kernel_context does not use initial-breadcrumbs, so when we ask if
its requests have started we do so by comparing against the completion
seqno of the previous request. This is very imprecise, not precise
enough for the defer_request assertion.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1847
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/20200508104220.9872-1-chris@chris-wilson.co.uk

drm/i915/gen12: Add aux table invalidate for all engines

All engines, exception being blitter as it does not
care about the form, can access compressed surfaces.

So we need to add forced aux table invalidates
for those engines.

v2: virtual instance masking (Chris)
v3: bug on if not found (Chris)

References: d248b371f747 ("drm/i915/gen12: Invalidate aux table entries forcibly")
References bspec#43904, hsdes#1809175790
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chuansheng Liu <chuansheng.liu@intel.com>
Cc: Rafael Antognolli <rafael.antognolli@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-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/20200507142045.8668-1-mika.kuoppala@linux.intel.com

drm/i915: Remove wait priority boosting

Upon waiting a request (when asked), we gave that request a small
priority boost, not enough for it to cause preemption, but enough for it
to be scheduled next before all equals. We also used that bit to give
new clients a small priority boost, similar to FQ_CODEL, such that we
favoured short interactive tasks ahead of long running streams.

However, this is causing lots of complications with timeslicing where we
both want to honour the boost and yet ignore it. Those complications
cause unexpected user behaviour (tasks not being timesliced and run
concurrently as epxected), and the easiest way to resolve that is to
remove the boost. Hopefully, we can find a compromise again if we need
to, but in theory timeslicing itself and future more advanced schedulers
should give us the interactivity boost we seek.

Testcase: igt/gem_exec_schedule/lateslice
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/20200507152338.7452-3-chris@chris-wilson.co.uk

drm/i915: Mark concurrent submissions with a weak-dependency

We recorded the dependencies for WAIT_FOR_SUBMIT in order that we could
correctly perform priority inheritance from the parallel branches to the
common trunk. However, for the purpose of timeslicing and reset
handling, the dependency is weak -- as we the pair of requests are
allowed to run in parallel and not in strict succession.

The real significance though is that this allows us to rearrange
groups of WAIT_FOR_SUBMIT linked requests along the single engine, and
so can resolve user level inter-batch scheduling dependencies from user
semaphores.

Fixes: c81471f5e95c ("drm/i915: Copy across scheduler behaviour flags across submit fences")
Testcase: igt/gem_exec_fence/submit
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.6+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200507155109.8892-1-chris@chris-wilson.co.uk

drm/i915/gen12: Invalidate aux table entries forcibly

Aux table invalidation can fail on update. So
next access may cause memory access to be into stale entry.

Proposed workaround is to invalidate entries between
all batchbuffers.

v2: correct register address (Yang)
v3: respect the order (Chris)

References bspec#43904, hsdes#1809175790
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chuansheng Liu <chuansheng.liu@intel.com>
Cc: Rafael Antognolli <rafael.antognolli@intel.com>
Cc: Yang A Shi <yang.a.shi@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-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/20200506165310.1239-1-mika.kuoppala@linux.intel.com

drm/i915/gen12: Flush L3

Flush TDL,L3 and EUs

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200506144734.29297-3-mika.kuoppala@linux.intel.com

drm/i915/gen12: Fix HDC pipeline flush

HDC pipeline flush is bit on the first dword of
the PIPE_CONTROL, not the second. Make it so.

v2: function naming (Chris)

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200506144734.29297-2-mika.kuoppala@linux.intel.com

Revert "drm/i915/tgl: Include ro parts of l3 to invalidate"

This reverts commit 62037ffff229b7d94f1db5ef8d2e2ec819832ef3.

L3 ro cache invalidation is part of the dword0 of pipe
control. Also it is not relevant to this gen.

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200506144734.29297-1-mika.kuoppala@linux.intel.com

drm/i915/execlists: Track inflight CCID

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

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

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

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796
Fixes: 2935ed5339c4 ("drm/i915: Remove logical HW ID")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-2-chris@chris-wilson.co.uk
(cherry picked from commit 5c4a53e3b1cbc38d0906e382f1037290658759bb)
(cherry picked from commit 134711240307d5586ae8e828d2699db70a8b74f2)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/execlists: Avoid reusing the same logical CCID

The bspec is confusing on the nature of the upper 32bits of the LRC
descriptor. Once upon a time, it said that it uses the upper 32b to
decide if it should perform a lite-restore, and so we must ensure that
each unique context submitted to HW is given a unique CCID [for the
duration of it being on the HW]. Currently, this is achieved by using
a small circular tag, and assigning every context submitted to HW a
new id. However, this tag is being cleared on repinning an inflight
context such that we end up re-using the 0 tag for multiple contexts.

To avoid accidentally clearing the CCID in the upper 32bits of the LRC
descriptor, split the descriptor into two dwords so we can update the
GGTT address separately from the CCID.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796
Fixes: 2935ed5339c4 ("drm/i915: Remove logical HW ID")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-1-chris@chris-wilson.co.uk
(cherry picked from commit 2632f174a2e1a5fd40a70404fa8ccfd0b1f79ebd)
(cherry picked from commit a4b70fcc587860f4b972f68217d8ebebe295ec15)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Yield the timeslice if caught waiting on a user semaphore

If we find ourselves waiting on a MI_SEMAPHORE_WAIT, either within the
user batch or in our own preamble, the engine raises a
GT_WAIT_ON_SEMAPHORE interrupt. We can unmask that interrupt and so
respond to a semaphore wait by yielding the timeslice, if we have
another context to yield to!

The only real complication is that the interrupt is only generated for
the start of the semaphore wait, and is asynchronous to our
process_csb() -- that is, we may not have registered the timeslice before
we see the interrupt. To ensure we don't miss a potential semaphore
blocking forward progress (e.g. selftests/live_timeslice_preempt) we mark
the interrupt and apply it to the next timeslice regardless of whether it
was active at the time.

v2: We use semaphores in preempt-to-busy, within the timeslicing
implementation itself! Ergo, when we do insert a preemption due to an
expired timeslice, the new context may start with the missed semaphore
flagged by the retired context and be yielded, ad infinitum. To avoid
this, read the context id at the time of the semaphore interrupt and
only yield if that context is still active.

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200407130811.17321-1-chris@chris-wilson.co.uk
(cherry picked from commit c4e8ba7390346a77ffe33ec3f210bc62e0b6c8c6)
(cherry picked from commit cd60e4ac4738a6921592c4f7baf87f9a3499f0e2)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Stop holding onto the pinned_default_state

As we only restore the default context state upon banning a context, we
only need enough of the state to run the ring and nothing more. That is
we only need our bare protocontext.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Andi Shyti <andi.shyti@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200504180745.15645-1-chris@chris-wilson.co.uk

drm/i915/execlists: Record the active CCID from before reset

If we cannot trust the reset will flush out the CS event queue such that
process_csb() reports an accurate view of HW, we will need to search the
active and pending contexts to determine which was actually running at
the time we issued the reset.

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/20200505084629.31365-1-chris@chris-wilson.co.uk

drm/i915/gt: Small tidy of gen8+ breadcrumb emission

Use a local to shrink a line under 80 columns, and refactor the common
emit_xcs_breadcrumb() wrapper of ggtt-write.

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/20200504180507.6017-1-chris@chris-wilson.co.uk

drm/i915/gt: Make timeslicing an explicit engine property

In order to allow userspace to rely on timeslicing to reorder their
batches, we must support preemption of those user batches. Declare
timeslicing as an explicit property that is a combination of having the
kernel support and HW support.

Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
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/20200501122249.12417-1-chris@chris-wilson.co.uk
(cherry picked from commit a211da9c771bf97395a3ced83a3aa383372b13a7)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Make timeslicing an explicit engine property

In order to allow userspace to rely on timeslicing to reorder their
batches, we must support preemption of those user batches. Declare
timeslicing as an explicit property that is a combination of having the
kernel support and HW support.

Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
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/20200501122249.12417-1-chris@chris-wilson.co.uk

drm/i915/gt: Always enable busy-stats for execlists

In the near future, we will utilize the busy-stats on each engine to
approximate the C0 cycles of each, and use that as an input to a manual
RPS mechanism. That entails having busy-stats always enabled and so we
can remove the enable/disable routines and simplify the pmu setup. As a
consequence of always having the stats enabled, we can also show the
current active time via sysfs/engine/xcs/active_time_ns.

Signed-off-by: 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/20200429205446.3259-1-chris@chris-wilson.co.uk

drm/i915/gt: Keep a no-frills swappable copy of the default context state

We need to keep the default context state around to instantiate new
contexts (aka golden rendercontext), and we also keep it pinned while
the engine is active so that we can quickly reset a hanging context.
However, the default contexts are large enough to merit keeping in
swappable memory as opposed to kernel memory, so we store them inside
shmemfs. Currently, we use the normal GEM objects to create the default
context image, but we can throw away all but the shmemfs file.

This greatly simplifies the tricky power management code which wants to
run underneath the normal GT locking, and we definitely do not want to
use any high level objects that may appear to recurse back into the GT.
Though perhaps the primary advantage of the complex GEM object is that
we aggressively cache the mapping, but here we are recreating the
vm_area everytime time we unpark. At the worst, we add a lightweight
cache, but first find a microbenchmark that is impacted.

Having started to create some utility functions to make working with
shmemfs objects easier, we can start putting them to wider use, where
GEM objects are overkill, such as storing persistent error state.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Ramalingam C <ramalingam.c@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200429172429.6054-1-chris@chris-wilson.co.uk

drm/i915/execlists: Verify we don't submit two identical CCIDs

Check that we do not submit two contexts into ELSP with the same CCID
[upper portion of the descriptor].

References: https://gitlab.freedesktop.org/drm/intel/-/issues/1793
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/20200428184751.11257-3-chris@chris-wilson.co.uk

drm/i915/execlists: Track inflight CCID

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

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

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

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796
Fixes: 2935ed5339c4 ("drm/i915: Remove logical HW ID")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-2-chris@chris-wilson.co.uk

drm/i915/execlists: Avoid reusing the same logical CCID

The bspec is confusing on the nature of the upper 32bits of the LRC
descriptor. Once upon a time, it said that it uses the upper 32b to
decide if it should perform a lite-restore, and so we must ensure that
each unique context submitted to HW is given a unique CCID [for the
duration of it being on the HW]. Currently, this is achieved by using
a small circular tag, and assigning every context submitted to HW a
new id. However, this tag is being cleared on repinning an inflight
context such that we end up re-using the 0 tag for multiple contexts.

To avoid accidentally clearing the CCID in the upper 32bits of the LRC
descriptor, split the descriptor into two dwords so we can update the
GGTT address separately from the CCID.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1796
Fixes: 2935ed5339c4 ("drm/i915: Remove logical HW ID")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200428184751.11257-1-chris@chris-wilson.co.uk

drm/i915/gt: Sanitize GT first

We see that if the HW doesn't actually sleep, the HW may eat the poison
we set in its write-only HWSP during sanitize:

intel_gt_resume.part.8: 0000:00:02.0
__gt_unpark: 0000:00:02.0
gt_sanitize: 0000:00:02.0 force:yes
process_csb: 0000:00:02.0 vcs0: cs-irq head=5, tail=90
process_csb: 0000:00:02.0 vcs0: csb[0]: status=0x5a5a5a5a:0x5a5a5a5a
assert_pending_valid: Nothing pending for promotion!

The CS TAIL pointer should have been reset by reset_csb_pointers(), so
in this case it is likely that we have read back from the CPU cache and
so we must clflush our control over that page. In doing so, push the
sanitisation to the start of the GT sequence so that our poisoning is
assuredly before we start talking to the HW.

References: https://gitlab.freedesktop.org/drm/intel/-/issues/1794
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/20200427084000.10999-1-chris@chris-wilson.co.uk

drm/i915/execlists: Check preempt-timeout target before submit_ports

We evaluate *active, which is a pointer into execlists->inflight[]
during dequeue to decide how long a preempt-timeout we need to apply.
However, as soon as we do the submit_ports, the HW may send its ACK
interrupt causing us to promote execlists->pending[] tp
execlists->inflight[], overwriting the value of *active. We know *active
is only stable until we submit (as we only submit when there is no
pending promotion).

[ 16.102328] BUG: KCSAN: data-race in execlists_dequeue+0x1449/0x1600 [i915]
[ 16.102356]
[ 16.102375] race at unknown origin, with read to 0xffff8881e9500488 of 8 bytes by task 429 on cpu 1:
[ 16.102780] execlists_dequeue+0x1449/0x1600 [i915]
[ 16.103160] __execlists_submission_tasklet+0x48/0x60 [i915]
[ 16.103540] execlists_submit_request+0x38e/0x3c0 [i915]
[ 16.103940] submit_notify+0x8f/0xc0 [i915]
[ 16.104308] __i915_sw_fence_complete+0x61/0x420 [i915]
[ 16.104683] i915_sw_fence_complete+0x58/0x80 [i915]
[ 16.105054] i915_sw_fence_commit+0x16/0x20 [i915]
[ 16.105457] __i915_request_queue+0x60/0x70 [i915]
[ 16.105843] i915_gem_do_execbuffer+0x2d6b/0x4230 [i915]
[ 16.106227] i915_gem_execbuffer2_ioctl+0x2b0/0x580 [i915]
[ 16.106257] drm_ioctl_kernel+0xe9/0x130
[ 16.106279] drm_ioctl+0x27d/0x45e
[ 16.106311] ksys_ioctl+0x89/0xb0
[ 16.106336] __x64_sys_ioctl+0x42/0x60
[ 16.106370] do_syscall_64+0x6e/0x2c0
[ 16.106397] entry_SYSCALL_64_after_hwframe+0x44/0xa9

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/20200426094231.21995-1-chris@chris-wilson.co.uk

drm/i915: Use indirect ctx bb to mend CMD_BUF_CCTL

Use indirect ctx bb to load cmd buffer control value
from context image to avoid corruption.

v2: add to lrc layout (Chris)
v3: end to a cacheline (Chris)
v4: add to lrc fixed (Chris)
v5: value in offset+1

Testcase: igt/i915_selftest/gt_lrc
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-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/20200424230632.30333-1-mika.kuoppala@linux.intel.com

drm/i915: Add per ctx batchbuffer wa for timestamp

Restoration of a previous timestamp can collide
with updating the timestamp, causing a value corruption.

Combat this issue by using indirect ctx bb to
modify the context image during restoring process.

We can preload value into scratch register. From which
we then do the actual write with LRR. LRR is faster and
thus less error prone as probability of race drops.

v2: tidying (Chris)
v3: lrr for all engines
v4: grp
v5: reg bit
v6: wa_bb_offset, virtual engines (Chris)

References: HSDES#16010904313
Testcase: igt/i915_selftest/gt_lrc
Suggested-by: Joseph Koston <joseph.koston@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-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/20200424230546.30271-1-mika.kuoppala@linux.intel.com

drm/i915: Add engine scratch register to live_lrc_fixed

General purpose registers are per engine and
in a fixed location. Add to live_lrc_fixed.

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200424214841.28076-1-mika.kuoppala@linux.intel.com

drm/i915: Make define for lrc state offset

More often than not, we need a byte offset into lrc
register state from the start of the hw state. Make it so.

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200423182355.21837-3-mika.kuoppala@linux.intel.com

drm/i915/selftests: Add context batchbuffers registers to live_lrc_fixed

Add per ctx bb and indirect ctx bb register locations to live_lrc_fixed
for verification.

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200423224159.22078-1-chris@chris-wilson.co.uk

drm/i915/gt: Carefully order virtual_submission_tasklet

During the virtual engine's submission tasklet, we take the request and
insert into the submission queue on each of our siblings. This seems
quite simply, and so no problems with ordering. However, the sibling
execlists' submission tasklets may run concurrently with the virtual
engine's tasklet, submitting the request to HW before the virtual
finishes its task of telling all the siblings. If this happens, the
sibling tasklet may *reorder* the ve->sibling[] array that the virtual
engine tasklet is processing. This can *only* reorder within the
elements already processed by the virtual engine, nevertheless the
race is detected by KCSAN:

[ 185.580014] BUG: KCSAN: data-race in execlists_dequeue [i915] / virtual_submission_tasklet [i915]
[ 185.580054]
[ 185.580076] write to 0xffff8881f1919860 of 8 bytes by interrupt on cpu 2:
[ 185.580553] execlists_dequeue+0x6ad/0x1600 [i915]
[ 185.581044] __execlists_submission_tasklet+0x48/0x60 [i915]
[ 185.581517] execlists_submission_tasklet+0xd3/0x170 [i915]
[ 185.581554] tasklet_action_common.isra.0+0x42/0x90
[ 185.581585] __do_softirq+0xc8/0x206
[ 185.581613] run_ksoftirqd+0x15/0x20
[ 185.581641] smpboot_thread_fn+0x15a/0x270
[ 185.581669] kthread+0x19a/0x1e0
[ 185.581695] ret_from_fork+0x1f/0x30
[ 185.581717]
[ 185.581736] read to 0xffff8881f1919860 of 8 bytes by interrupt on cpu 0:
[ 185.582231] virtual_submission_tasklet+0x10e/0x5c0 [i915]
[ 185.582265] tasklet_action_common.isra.0+0x42/0x90
[ 185.582291] __do_softirq+0xc8/0x206
[ 185.582315] run_ksoftirqd+0x15/0x20
[ 185.582340] smpboot_thread_fn+0x15a/0x270
[ 185.582368] kthread+0x19a/0x1e0
[ 185.582395] ret_from_fork+0x1f/0x30
[ 185.582417]

We can prevent this race by checking for the ve->request after looking
up the sibling array.

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/20200423115315.26825-1-chris@chris-wilson.co.uk

drm/i915/execlists: Drop request-before-CS assertion

When we migrated to execlists, one of the conditions we wanted to test
for was whether the breadcrumb seqno was being written before the
breadcumb interrupt was delivered. This was following on from issues
observed on previous generations which were not so strongly ordered. With
the removal of the missed interrupt detection, we have not reliable
means of detecting the out-of-order seqno/interrupt but instead tried to
assert that the relationship between the CS event interrupt and the
breadwrite should be strongly ordered. However, Icelake proves it is
possible for the HW implementation to forget about minor little details
such as write ordering and so the order between *processing* the CS
event and the breadcrumb is unreliable.

Remove the unreliable assertion, but leave a debug telltale in case we
have reason to suspect.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1658
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200422141749.28709-1-chris@chris-wilson.co.uk

drm/i915/gt: Poison residual state [HWSP] across resume.

Since we may lose the content of any buffer when we relinquish control
of the system (e.g. suspend/resume), we have to be careful not to rely
on regaining control. A good method to detect when we might be using
garbage is by always injecting that garbage prior to first use on
load/resume/etc.

v2: Drop sanitize callback on cleanup
v3: Move seqno reset to timeline enter, so we reset all timelines.
However, this is done on every activation during runtime and not reset.
The similar level of paranoia we apply to correcting context state after
a period of inactivity.

Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Venkata Ramana Nayana <venkata.ramana.nayana@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200421092504.7416-1-chris@chris-wilson.co.uk

drm/i915/gt: Scrub execlists state on resume

Before we resume, we reset the HW so we restart from a known good state.
However, as a part of the reset process, we drain our pending CS event
queue -- and if we are resuming that does not correspond to internal
state. On setup, we are scrubbing the CS pointers, but alas only on
setup.

Apply the sanitization not just to setup, but to all resumes.

Reported-by: Venkata Ramana Nayana <venkata.ramana.nayana@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Venkata Ramana Nayana <venkata.ramana.nayana@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200416114117.3460-1-chris@chris-wilson.co.uk

drm/i915/gt: prefer struct drm_device based logging

Prefer struct drm_device based logging over struct device based logging.

No functional changes.

Cc: Wambui Karuga <wambui.karugax@gmail.com>
Reviewed-by: Wambui Karuga <wambui.karugax@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200402114819.17232-16-jani.nikula@intel.com

drm/i915/gt: Yield the timeslice if caught waiting on a user semaphore

If we find ourselves waiting on a MI_SEMAPHORE_WAIT, either within the
user batch or in our own preamble, the engine raises a
GT_WAIT_ON_SEMAPHORE interrupt. We can unmask that interrupt and so
respond to a semaphore wait by yielding the timeslice, if we have
another context to yield to!

The only real complication is that the interrupt is only generated for
the start of the semaphore wait, and is asynchronous to our
process_csb() -- that is, we may not have registered the timeslice before
we see the interrupt. To ensure we don't miss a potential semaphore
blocking forward progress (e.g. selftests/live_timeslice_preempt) we mark
the interrupt and apply it to the next timeslice regardless of whether it
was active at the time.

v2: We use semaphores in preempt-to-busy, within the timeslicing
implementation itself! Ergo, when we do insert a preemption due to an
expired timeslice, the new context may start with the missed semaphore
flagged by the retired context and be yielded, ad infinitum. To avoid
this, read the context id at the time of the semaphore interrupt and
only yield if that context is still active.

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200407130811.17321-1-chris@chris-wilson.co.uk

drm/i915/gt: Free request pool from virtual engines

While extremely unlikely to be populated, we could capture a request on
the virtual engine which we should free along with the virtual engine.

Fixes: 43acd6516ca9 ("drm/i915: Keep a per-engine request pool")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200403203303.10903-1-chris@chris-wilson.co.uk

drm/i915/execlists: Peek at the next submission for error interrupts

If we receive the error interrupt before the CS interrupt, we may find
ourselves without an active request to reset, skipping the GPU reset.
All because the attempt to reset was too early.

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/20200401110435.30389-1-chris@chris-wilson.co.uk

drm/i915/execlists: Pause CS flow before reset

Since we may be attempting to reset an active engine, we try to freeze
it in place before resetting -- to be on the safe side. We can go one
step further if we are using the CS flow semaphore to prevent the
context switching into the next.

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/20200331091459.29179-2-chris@chris-wilson.co.uk

drm/i915/gt: Include a few tracek for timeslicing

Add a few telltales to see when timeslicing is being enabled.

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/20200331120502.14713-1-chris@chris-wilson.co.uk

drm/i915/execlists: Double check breadcrumb before crying foul

process_csb: 0000:00:02.0 bcs0: cs-irq head=4, tail=5
process_csb: 0000:00:02.0 bcs0: csb[5]: status=0x00008002:0x60000020
trace_ports: 0000:00:02.0 bcs0: preempted { ff84:45154! prio 2 }
trace_ports: 0000:00:02.0 bcs0: promote { ff84:45155* prio 2 }
trace_ports: 0000:00:02.0 bcs0: submit { ff84:45156 prio 2 }

process_csb: 0000:00:02.0 bcs0: cs-irq head=5, tail=6
process_csb: 0000:00:02.0 bcs0: csb[6]: status=0x00000018:0x60000020
trace_ports: 0000:00:02.0 bcs0: completed { ff84:45155* prio 2 }
process_csb: 0000:00:02.0 bcs0: ring:{start:0x00178000, head:0928, tail:0928, ctl:00000000, mode:00000200}
process_csb: 0000:00:02.0 bcs0: rq:{start:00178000, head:08b0, tail:08f0, seqno:ff84:45155, hwsp:45156},
process_csb: 0000:00:02.0 bcs0: ctx:{start:00178000, head:e000928, tail:0928},
process_csb: GEM_BUG_ON("context completed before request")

In this sequence, we can see that although we have submitted the next
request [ff84:45156] to HW (via ELSP[]) it has not yet reported the
lite-restore. Instead, we see the completion event of the currently
active request [ff84:45155] but at the time of processing that event,
the breadcrumb has not yet been written. Though by the time we do print
out the debug info, the seqno write of ff84:45156 has landed!

Therefore there is a serialisation problem between the seqno writes and
CS events, not just between the CS buffer and its head/tail pointers as
previously observed on Icelake.

This is not a huge problem, as we don't strictly rely on the breadcrumb
to determine HW activity, but it may indicate that interrupt delivery is
before the seqno write, aka bringing back the plague of missed
interrupts from yesteryear. However, there is no indication of this
wider problem, so let's just flush the seqno read before reporting an
error. If it persists after the fresh read we can worry again.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200330234318.30638-1-chris@chris-wilson.co.uk

drm/i915/execlists: Explicitly reset both reg and context runtime

Upon a GPU reset, we copy the default context image over top of the
guilty image. This will rollback the CTX_TIMESTAMP register to before
our value of ce->runtime.last. Reset both back to 0 so that we do not
encounter an underflow on the next schedule out after resume.

This should not be a huge issue in practice, as hangs should be rare in
correct code.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200330125827.5804-1-chris@chris-wilson.co.uk

drm/i915/execlists: Include priority info in trace_ports

Add some extra information into trace_ports to help with reviewing
correctness.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200330113137.24425-1-chris@chris-wilson.co.uk

drm/i915/execlists: Workaround switching back to a completed context

In what seems remarkably similar to the w/a required to not reload an
idle context with HEAD==TAIL, it appears we must prevent the HW from
switching to an idle context in ELSP[1], while simultaneously trying to
preempt the HW to run another context and a continuation of the idle
context (which is no longer idle).

We can achieve this by preventing the context from completing while we
reload a new ELSP (by applying ring_set_paused(1) across the whole of
dequeue), except this eventually fails due to a lite-restore into a
waiting semaphore does not generate an ACK. Instead, we try to avoid
making the GPU do anything too challenging and not submit a new ELSP
while the interrupts + CSB events appear to have fallen behind the
completed contexts. We expect it to catch up shortly so we queue another
tasklet execution and hope for the best.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1501
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200327201433.21864-1-chris@chris-wilson.co.uk

drm/i915/gt: Stage the transfer of the virtual breadcrumb

We move the virtual breadcrumb from one physical engine to the next, if
the next virtual request is scheduled on a new physical engine. Since
the virtual context can only be in one signal queue, we need it to track
the current physical engine for the new breadcrumbs. However, to move
the list we need both breadcrumb locks -- and since we cannot take both
at the same time (unless we are careful and always ensure consistent
ordering) stage the movement of the signaler via the current virtual
request.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1510
Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine")
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/20200325130059.30600-1-chris@chris-wilson.co.uk
(cherry picked from commit 6c81e21a4742385c00713137c6fdcade0412e93c)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

drm/i915/gt: Stage the transfer of the virtual breadcrumb

We move the virtual breadcrumb from one physical engine to the next, if
the next virtual request is scheduled on a new physical engine. Since
the virtual context can only be in one signal queue, we need it to track
the current physical engine for the new breadcrumbs. However, to move
the list we need both breadcrumb locks -- and since we cannot take both
at the same time (unless we are careful and always ensure consistent
ordering) stage the movement of the signaler via the current virtual
request.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1510
Fixes: 6d06779e8672 ("drm/i915: Load balancing across a virtual engine")
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/20200325130059.30600-1-chris@chris-wilson.co.uk

drm/i915/execlists: Pull tasklet interrupt-bh local to direct submission

We dropped calling process_csb prior to handling direct submission in
order to avoid the nesting of spinlocks and lift process_csb() and the
majority of the tasklet out of irq-off. However, we do want to avoid
ksoftirqd latency in the fast path, so try and pull the interrupt-bh
local to direct submission if we can acquire the tasklet's lock.

v2: Document the read of pending[0] from outside the tasklet with
READ_ONCE.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200325120227.8044-1-chris@chris-wilson.co.uk

drm/i915/execlists: Drop setting sibling priority hint on virtual engines

We set the priority hint on execlists to avoid executing the tasklet for
when we know that there will be no change in execution order. However,
as we set it from the virtual engine for all siblings, but only one
physical engine may respond, we leave the hint set on the others
stopping direct submission that could take place.

If we do not set the hint, we may attempt direct submission even if we
don't expect to submit. If we set the hint, we may not do any submission
until the tasklet is run (and sometimes we may park the engine before
that has had a chance). Ergo there's only a minor ill-effect on mixed
virtual/physical engine workloads where we may try and fail to do direct
submission more often than required. (Pure virtual / engine workloads
will have redundant tasklet execution suppressed as normal.)

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1522
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/20200325101358.12231-1-chris@chris-wilson.co.uk

drm/i915: Rely on direct submission to the queue

Drop the pretense of kicking the tasklet (used only for the defunct guc
submission backend, it should just take ownership of the submit!) and so
remove the bh-kicking from around submission.

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/20200323092841.22240-5-chris@chris-wilson.co.uk

drm/i915/lrc: convert to struct drm_device based logging macros.

Convert various instances of the printk based drm logging macros to the
struct drm_device based logging macros.

Note that this converts DRM_DEBUG_DRIVER() to drm_dbg() but does not
convert DRM_DEBUG() due to the lack of an analogous drm_device based
macro.

References: https://lists.freedesktop.org/archives/dri-devel/2020-January/253381.html
Signed-off-by: Wambui Karuga <wambui.karugax@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200314183344.17603-3-wambui.karugax@gmail.com

Revert "drm/i915/tgl: Add extra hdc flush workaround"

This reverts commit 36a6b5d964d995b536b1925ec42052ee40ba92c4.

The commit takes care Wa_1604544889 which was fixed on a0 stepping based on
a0 replan. So no SW workaround is required on any stepping now.

Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Caz Yokoyama <caz.yokoyama@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Fixes: 36a6b5d964d9 ("drm/i915/tgl: Add extra hdc flush workaround")
Link: https://patchwork.freedesktop.org/patch/msgid/1c751032ce79c80c5485cae315f1a9904ce07cac.1583359940.git.caz.yokoyama@intel.com
(cherry picked from commit 175c4d9b3b9a60b4ea0b8cd034011808c6a03b05)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/execlists: Track active elements during dequeue

Record the initial active element we use when building the next ELSP
submission, so that we can compare against it latter to see if there's
no change.

Fixes: 44d0a9c05bc0 ("drm/i915/execlists: Skip redundant resubmission")
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/20200311092624.10012-2-chris@chris-wilson.co.uk
(cherry picked from commit 60ef5b7ac6a131f09d287a5f156c878c2c926a30)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Revert "drm/i915/tgl: Add extra hdc flush workaround"

This reverts commit 36a6b5d964d995b536b1925ec42052ee40ba92c4.

The commit takes care Wa_1604544889 which was fixed on a0 stepping based on
a0 replan. So no SW workaround is required on any stepping now.

Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Caz Yokoyama <caz.yokoyama@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Fixes: 36a6b5d964d9 ("drm/i915/tgl: Add extra hdc flush workaround")
Link: https://patchwork.freedesktop.org/patch/msgid/1c751032ce79c80c5485cae315f1a9904ce07cac.1583359940.git.caz.yokoyama@intel.com

drm/i915/execlists: Enable timeslice on partial virtual engine dequeue

If we stop filling the ELSP due to an incompatible virtual engine
request, check if we should enable the timeslice on behalf of the queue.

This fixes the case where we are inspecting the last->next element when
we know that the last element is the last request in the execution queue,
and so decided we did not need to enable timeslicing despite the intent
to do so!

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200306113012.3184606-1-chris@chris-wilson.co.uk
(cherry picked from commit 3df2deed411e0f1b7312baf0139aab8bba4c0410)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/execlists: Track active elements during dequeue

Record the initial active element we use when building the next ELSP
submission, so that we can compare against it latter to see if there's
no change.

Fixes: 44d0a9c05bc0 ("drm/i915/execlists: Skip redundant resubmission")
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/20200311092624.10012-2-chris@chris-wilson.co.uk

drm/i915/execlists: Mark up data-races in virtual engines

The virtual engine passes tokens back and forth to its backing physical
engines.

[ 57.372993] BUG: KCSAN: data-race in execlists_dequeue [i915] / virtual_submission_tasklet [i915]
[ 57.373012]
[ 57.373023] write to 0xffff8881f47324c0 of 4 bytes by interrupt on cpu 2:
[ 57.373241] execlists_dequeue+0x6fa/0x2150 [i915]
[ 57.373458] __execlists_submission_tasklet+0x48/0x60 [i915]
[ 57.373677] execlists_submission_tasklet+0xd3/0x170 [i915]
[ 57.373694] tasklet_action_common.isra.0+0x42/0xa0
[ 57.373709] __do_softirq+0xd7/0x2cd
[ 57.373723] irq_exit+0xbe/0xe0
[ 57.373735] do_IRQ+0x51/0x100
[ 57.373748] ret_from_intr+0x0/0x1c
[ 57.373963] engine_retire+0x89/0xe0 [i915]
[ 57.373977] process_one_work+0x3b1/0x690
[ 57.373990] worker_thread+0x80/0x670
[ 57.374004] kthread+0x19a/0x1e0
[ 57.374017] ret_from_fork+0x1f/0x30
[ 57.374027]
[ 57.374038] read to 0xffff8881f47324c0 of 4 bytes by interrupt on cpu 3:
[ 57.374256] virtual_submission_tasklet+0x27/0x5a0 [i915]
[ 57.374273] tasklet_action_common.isra.0+0x42/0xa0
[ 57.374288] __do_softirq+0xd7/0x2cd
[ 57.374302] run_ksoftirqd+0x15/0x20
[ 57.374315] smpboot_thread_fn+0x1ab/0x300
[ 57.374329] kthread+0x19a/0x1e0
[ 57.374342] ret_from_fork+0x1f/0x30

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/20200310141320.24149-3-chris@chris-wilson.co.uk

drm/i915/gt: Defend against concurrent updates to execlists->active

[ 206.875637] BUG: KCSAN: data-race in __i915_schedule+0x7fc/0x930 [i915]
[ 206.875654]
[ 206.875666] race at unknown origin, with read to 0xffff8881f7644480 of 8 bytes by task 703 on cpu 3:
[ 206.875901] __i915_schedule+0x7fc/0x930 [i915]
[ 206.876130] __bump_priority+0x63/0x80 [i915]
[ 206.876361] __i915_sched_node_add_dependency+0x258/0x300 [i915]
[ 206.876593] i915_sched_node_add_dependency+0x50/0xa0 [i915]
[ 206.876824] i915_request_await_dma_fence+0x1da/0x530 [i915]
[ 206.877057] i915_request_await_object+0x2fe/0x470 [i915]
[ 206.877287] i915_gem_do_execbuffer+0x45dc/0x4c20 [i915]
[ 206.877517] i915_gem_execbuffer2_ioctl+0x2c3/0x580 [i915]
[ 206.877535] drm_ioctl_kernel+0xe4/0x120
[ 206.877549] drm_ioctl+0x297/0x4c7
[ 206.877563] ksys_ioctl+0x89/0xb0
[ 206.877577] __x64_sys_ioctl+0x42/0x60
[ 206.877591] do_syscall_64+0x6e/0x2c0
[ 206.877606] entry_SYSCALL_64_after_hwframe+0x44/0xa9

v2: Be safe and include mb

References: https://gitlab.freedesktop.org/drm/intel/issues/1318
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200309170540.10332-1-chris@chris-wilson.co.uk

drm/i915/execlsts: Mark up racy inspection of current i915_request priority

[ 120.176548] BUG: KCSAN: data-race in __i915_schedule [i915] / effective_prio [i915]
[ 120.176566]
[ 120.176577] write to 0xffff8881e35e6540 of 4 bytes by task 730 on cpu 3:
[ 120.176792] __i915_schedule+0x63e/0x920 [i915]
[ 120.177007] __bump_priority+0x63/0x80 [i915]
[ 120.177220] __i915_sched_node_add_dependency+0x258/0x300 [i915]
[ 120.177438] i915_sched_node_add_dependency+0x50/0xa0 [i915]
[ 120.177654] i915_request_await_dma_fence+0x1da/0x530 [i915]
[ 120.177867] i915_request_await_object+0x2fe/0x470 [i915]
[ 120.178081] i915_gem_do_execbuffer+0x45dc/0x4c20 [i915]
[ 120.178292] i915_gem_execbuffer2_ioctl+0x2c3/0x580 [i915]
[ 120.178309] drm_ioctl_kernel+0xe4/0x120
[ 120.178322] drm_ioctl+0x297/0x4c7
[ 120.178335] ksys_ioctl+0x89/0xb0
[ 120.178348] __x64_sys_ioctl+0x42/0x60
[ 120.178361] do_syscall_64+0x6e/0x2c0
[ 120.178375] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 120.178387]
[ 120.178397] read to 0xffff8881e35e6540 of 4 bytes by interrupt on cpu 2:
[ 120.178606] effective_prio+0x25/0xc0 [i915]
[ 120.178812] process_csb+0xe8b/0x10a0 [i915]
[ 120.179021] execlists_submission_tasklet+0x30/0x170 [i915]
[ 120.179038] tasklet_action_common.isra.0+0x42/0xa0
[ 120.179053] __do_softirq+0xd7/0x2cd
[ 120.179066] irq_exit+0xbe/0xe0
[ 120.179078] do_IRQ+0x51/0x100
[ 120.179090] ret_from_intr+0x0/0x1c
[ 120.179104] cpuidle_enter_state+0x1b8/0x5d0
[ 120.179117] cpuidle_enter+0x50/0x90
[ 120.179131] do_idle+0x1a1/0x1f0
[ 120.179145] cpu_startup_entry+0x14/0x16
[ 120.179158] start_secondary+0x120/0x180
[ 120.179172] secondary_startup_64+0xa4/0xb0

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/20200309110934.868-5-chris@chris-wilson.co.uk

drm/i915/execlists: Mark up read of i915_request.fence.flags

[ 145.927961] BUG: KCSAN: data-race in can_merge_rq [i915] / signal_irq_work [i915]
[ 145.927980]
[ 145.927992] write (marked) to 0xffff8881e513fab0 of 8 bytes by interrupt on cpu 2:
[ 145.928250] signal_irq_work+0x134/0x640 [i915]
[ 145.928268] irq_work_run_list+0xd7/0x120
[ 145.928283] irq_work_run+0x1d/0x50
[ 145.928300] smp_irq_work_interrupt+0x21/0x30
[ 145.928328] irq_work_interrupt+0xf/0x20
[ 145.928356] _raw_spin_unlock_irqrestore+0x34/0x40
[ 145.928596] execlists_submission_tasklet+0xde/0x170 [i915]
[ 145.928616] tasklet_action_common.isra.0+0x42/0xa0
[ 145.928632] __do_softirq+0xd7/0x2cd
[ 145.928646] irq_exit+0xbe/0xe0
[ 145.928665] do_IRQ+0x51/0x100
[ 145.928684] ret_from_intr+0x0/0x1c
[ 145.928699] schedule+0x0/0xb0
[ 145.928719] worker_thread+0x194/0x670
[ 145.928743] kthread+0x19a/0x1e0
[ 145.928765] ret_from_fork+0x1f/0x30
[ 145.928784]
[ 145.928796] read to 0xffff8881e513fab0 of 8 bytes by task 738 on cpu 1:
[ 145.929046] can_merge_rq+0xb1/0x100 [i915]
[ 145.929282] __execlists_submission_tasklet+0x866/0x25a0 [i915]
[ 145.929518] execlists_submit_request+0x2a4/0x2b0 [i915]
[ 145.929758] submit_notify+0x8f/0xc0 [i915]
[ 145.929989] __i915_sw_fence_complete+0x5d/0x3e0 [i915]
[ 145.930221] i915_sw_fence_complete+0x58/0x80 [i915]
[ 145.930453] i915_sw_fence_commit+0x16/0x20 [i915]
[ 145.930698] __i915_request_queue+0x60/0x70 [i915]
[ 145.930935] i915_gem_do_execbuffer+0x3997/0x4c20 [i915]
[ 145.931175] i915_gem_execbuffer2_ioctl+0x2c3/0x580 [i915]
[ 145.931194] drm_ioctl_kernel+0xe4/0x120
[ 145.931208] drm_ioctl+0x297/0x4c7
[ 145.931222] ksys_ioctl+0x89/0xb0
[ 145.931238] __x64_sys_ioctl+0x42/0x60
[ 145.931260] do_syscall_64+0x6e/0x2c0
[ 145.931275] entry_SYSCALL_64_after_hwframe+0x44/0xa9

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/20200309110934.868-4-chris@chris-wilson.co.uk

drm/i915/gt: Mark up racy check of last list element

[ 25.025543] BUG: KCSAN: data-race in __i915_request_create [i915] / process_csb [i915]
[ 25.025561]
[ 25.025573] write (marked) to 0xffff8881e85c1620 of 8 bytes by task 696 on cpu 1:
[ 25.025789] __i915_request_create+0x54b/0x5d0 [i915]
[ 25.026001] i915_request_create+0xcc/0x150 [i915]
[ 25.026218] i915_gem_do_execbuffer+0x2f70/0x4c20 [i915]
[ 25.026428] i915_gem_execbuffer2_ioctl+0x2c3/0x580 [i915]
[ 25.026445] drm_ioctl_kernel+0xe4/0x120
[ 25.026459] drm_ioctl+0x297/0x4c7
[ 25.026472] ksys_ioctl+0x89/0xb0
[ 25.026484] __x64_sys_ioctl+0x42/0x60
[ 25.026497] do_syscall_64+0x6e/0x2c0
[ 25.026510] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 25.026522]
[ 25.026532] read to 0xffff8881e85c1620 of 8 bytes by interrupt on cpu 2:
[ 25.026742] process_csb+0x8d6/0x1070 [i915]
[ 25.026949] execlists_submission_tasklet+0x30/0x170 [i915]
[ 25.026969] tasklet_action_common.isra.0+0x42/0xa0
[ 25.026984] __do_softirq+0xd7/0x2cd
[ 25.026997] irq_exit+0xbe/0xe0
[ 25.027009] do_IRQ+0x51/0x100
[ 25.027021] ret_from_intr+0x0/0x1c
[ 25.027033] poll_idle+0x3e/0x13b
[ 25.027047] cpuidle_enter_state+0x189/0x5d0
[ 25.027060] cpuidle_enter+0x50/0x90
[ 25.027074] do_idle+0x1a1/0x1f0
[ 25.027086] cpu_startup_entry+0x14/0x16
[ 25.027100] start_secondary+0x120/0x180
[ 25.027116] secondary_startup_64+0xa4/0xb0

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/20200309110934.868-2-chris@chris-wilson.co.uk

drm/i915/execlists: Mark up the racy access to switch_priority_hint

[ 7534.150687] BUG: KCSAN: data-race in __execlists_submission_tasklet [i915] / process_csb [i915]
[ 7534.150706]
[ 7534.150717] write to 0xffff8881f1bc24b4 of 4 bytes by task 24404 on cpu 3:
[ 7534.150925] __execlists_submission_tasklet+0x1158/0x2780 [i915]
[ 7534.151133] execlists_submit_request+0x2e8/0x2f0 [i915]
[ 7534.151348] submit_notify+0x8f/0xc0 [i915]
[ 7534.151549] __i915_sw_fence_complete+0x5d/0x3e0 [i915]
[ 7534.151753] i915_sw_fence_complete+0x58/0x80 [i915]
[ 7534.151963] i915_sw_fence_commit+0x16/0x20 [i915]
[ 7534.152179] __i915_request_queue+0x60/0x70 [i915]
[ 7534.152388] i915_gem_do_execbuffer+0x3997/0x4c20 [i915]
[ 7534.152598] i915_gem_execbuffer2_ioctl+0x2c3/0x580 [i915]
[ 7534.152615] drm_ioctl_kernel+0xe4/0x120
[ 7534.152629] drm_ioctl+0x297/0x4c7
[ 7534.152642] ksys_ioctl+0x89/0xb0
[ 7534.152654] __x64_sys_ioctl+0x42/0x60
[ 7534.152667] do_syscall_64+0x6e/0x2c0
[ 7534.152681] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 7534.152693]
[ 7534.152703] read to 0xffff8881f1bc24b4 of 4 bytes by interrupt on cpu 2:
[ 7534.152914] process_csb+0xe7c/0x10a0 [i915]
[ 7534.153120] execlists_submission_tasklet+0x30/0x170 [i915]
[ 7534.153138] tasklet_action_common.isra.0+0x42/0xa0
[ 7534.153153] __do_softirq+0xd7/0x2cd
[ 7534.153166] run_ksoftirqd+0x15/0x20
[ 7534.153180] smpboot_thread_fn+0x1ab/0x300
[ 7534.153194] kthread+0x19a/0x1e0
[ 7534.153207] ret_from_fork+0x1f/0x30

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/20200309144249.10309-1-chris@chris-wilson.co.uk

drm/i915/execlists: Enable timeslice on partial virtual engine dequeue

If we stop filling the ELSP due to an incompatible virtual engine
request, check if we should enable the timeslice on behalf of the queue.

This fixes the case where we are inspecting the last->next element when
we know that the last element is the last request in the execution queue,
and so decided we did not need to enable timeslicing despite the intent
to do so!

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200306113012.3184606-1-chris@chris-wilson.co.uk

drm/i915: Assert requests within a context are submitted in order

Check the flow of requests into the hardware to verify that are
submitted in order along their timeline.

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/20200306071614.2846708-1-chris@chris-wilson.co.uk

drm/i915/execlists: Show the "switch priority hint" in dumps

Show the timeslicing priority hint in engine dumps to aide debugging.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200305135843.2760512-1-chris@chris-wilson.co.uk

drm/i915/gt: Propagate change in error status to children on unhold

As we release the head requests back into the queue, propagate any
change in error status that may have occurred while the requests were
temporarily suspended.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1277
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200304121849.2448028-2-chris@chris-wilson.co.uk

drm/i915: Apply i915_request_skip() on submission

Trying to use i915_request_skip() prior to i915_request_add() causes us
to try and fill the ring upto request->postfix, which has not yet been
set, and so may cause us to memset() past the end of the ring.

Instead of skipping the request immediately, just flag the error on the
request (only accepting the first fatal error we see) and then clear the
request upon submission.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200304121849.2448028-1-chris@chris-wilson.co.uk

drm/i915/execlists: Check the sentinel is alone in the ELSP

We only use sentinel requests for "preempt-to-idle" passes, so assert
that they are the only request in a new submission.

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/20200302085812.4172450-12-chris@chris-wilson.co.uk

drm/i915/gt: Reset queue_priority_hint after wedging

An odd and highly unlikely path caught us out. On delayed submission
(due to an asynchronous reset handler), we poked the priority_hint and
kicked the tasklet. However, we had already marked the device as wedged
and swapped out the tasklet for a no-op. The result was that we never
cleared the priority hint and became upset when we later checked.

<0> [574.303565] i915_sel-6278 2.... 481822445us : __i915_subtests: Running intel_execlists_live_selftests/live_error_interrupt
<0> [574.303565] i915_sel-6278 2.... 481822472us : __engine_unpark: 0000:00:02.0 rcs0:
<0> [574.303565] i915_sel-6278 2.... 481822491us : __gt_unpark: 0000:00:02.0
<0> [574.303565] i915_sel-6278 2.... 481823220us : execlists_context_reset: 0000:00:02.0 rcs0: context:f4ee reset
<0> [574.303565] i915_sel-6278 2.... 481824830us : __intel_context_active: 0000:00:02.0 rcs0: context:f51b active
<0> [574.303565] i915_sel-6278 2.... 481825258us : __intel_context_do_pin: 0000:00:02.0 rcs0: context:f51b pin ring:{start:00006000, head:0000, tail:0000}
<0> [574.303565] i915_sel-6278 2.... 481825311us : __i915_request_commit: 0000:00:02.0 rcs0: fence f51b:2, current 0
<0> [574.303565] i915_sel-6278 2d..1 481825347us : __i915_request_submit: 0000:00:02.0 rcs0: fence f51b:2, current 0
<0> [574.303565] i915_sel-6278 2d..1 481825363us : trace_ports: 0000:00:02.0 rcs0: submit { f51b:2, 0:0 }
<0> [574.303565] i915_sel-6278 2.... 481826809us : __intel_context_active: 0000:00:02.0 rcs0: context:f51c active
<0> [574.303565] <idle>-0 7d.h2 481827326us : cs_irq_handler: 0000:00:02.0 rcs0: CS error: 1
<0> [574.303565] <idle>-0 7..s1 481827377us : process_csb: 0000:00:02.0 rcs0: cs-irq head=3, tail=4
<0> [574.303565] <idle>-0 7..s1 481827379us : process_csb: 0000:00:02.0 rcs0: csb[4]: status=0x10000001:0x00000000
<0> [574.305593] <idle>-0 7..s1 481827385us : trace_ports: 0000:00:02.0 rcs0: promote { f51b:2*, 0:0 }
<0> [574.305611] <idle>-0 7..s1 481828179us : execlists_reset: 0000:00:02.0 rcs0: reset for CS error
<0> [574.305611] i915_sel-6278 2.... 481828284us : __intel_context_do_pin: 0000:00:02.0 rcs0: context:f51c pin ring:{start:00007000, head:0000, tail:0000}
<0> [574.305611] i915_sel-6278 2.... 481828345us : __i915_request_commit: 0000:00:02.0 rcs0: fence f51c:2, current 0
<0> [574.305611] <idle>-0 7dNs2 481847823us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence f51b:2, current 1
<0> [574.305611] <idle>-0 7dNs2 481847857us : execlists_hold: 0000:00:02.0 rcs0: fence f51b:2, current 1 on hold
<0> [574.305611] <idle>-0 7.Ns1 481847863us : intel_engine_reset: 0000:00:02.0 rcs0: flags=4
<0> [574.305611] <idle>-0 7.Ns1 481847945us : execlists_reset_prepare: 0000:00:02.0 rcs0: depth<-1
<0> [574.305611] <idle>-0 7.Ns1 481847946us : intel_engine_stop_cs: 0000:00:02.0 rcs0:
<0> [574.305611] <idle>-0 7.Ns1 538584284us : intel_engine_stop_cs: 0000:00:02.0 rcs0: timed out on STOP_RING -> IDLE
<0> [574.305611] <idle>-0 7.Ns1 538584347us : __intel_gt_reset: 0000:00:02.0 engine_mask=1
<0> [574.305611] <idle>-0 7.Ns1 538584406us : execlists_reset_rewind: 0000:00:02.0 rcs0:
<0> [574.305611] <idle>-0 7dNs2 538585050us : __i915_request_reset: 0000:00:02.0 rcs0: fence f51b:2, current 1 guilty? yes
<0> [574.305611] <idle>-0 7dNs2 538585063us : __execlists_reset: 0000:00:02.0 rcs0: replay {head:0000, tail:0068}
<0> [574.306565] <idle>-0 7.Ns1 538588457us : intel_engine_cancel_stop_cs: 0000:00:02.0 rcs0:
<0> [574.306565] <idle>-0 7dNs2 538588462us : __i915_request_submit: 0000:00:02.0 rcs0: fence f51c:2, current 0
<0> [574.306565] <idle>-0 7dNs2 538588471us : trace_ports: 0000:00:02.0 rcs0: submit { f51c:2, 0:0 }
<0> [574.306565] <idle>-0 7.Ns1 538588474us : execlists_reset_finish: 0000:00:02.0 rcs0: depth->1
<0> [574.306565] kworker/-202 2.... 538588755us : i915_request_retire: 0000:00:02.0 rcs0: fence f51c:2, current 2
<0> [574.306565] ksoftirq-46 7..s. 538588773us : process_csb: 0000:00:02.0 rcs0: cs-irq head=11, tail=1
<0> [574.306565] ksoftirq-46 7..s. 538588774us : process_csb: 0000:00:02.0 rcs0: csb[0]: status=0x10000001:0x00000000
<0> [574.306565] ksoftirq-46 7..s. 538588776us : trace_ports: 0000:00:02.0 rcs0: promote { f51c:2!, 0:0 }
<0> [574.306565] ksoftirq-46 7..s. 538588778us : process_csb: 0000:00:02.0 rcs0: csb[1]: status=0x10000018:0x00000020
<0> [574.306565] ksoftirq-46 7..s. 538588779us : trace_ports: 0000:00:02.0 rcs0: completed { f51c:2!, 0:0 }
<0> [574.306565] kworker/-202 2.... 538588826us : intel_context_unpin: 0000:00:02.0 rcs0: context:f51c unpin
<0> [574.306565] i915_sel-6278 6.... 538589663us : __intel_gt_set_wedged.part.32: 0000:00:02.0 start
<0> [574.306565] i915_sel-6278 6.... 538589667us : execlists_reset_prepare: 0000:00:02.0 rcs0: depth<-0
<0> [574.306565] i915_sel-6278 6.... 538589710us : intel_engine_stop_cs: 0000:00:02.0 rcs0:
<0> [574.306565] i915_sel-6278 6.... 538589732us : execlists_reset_prepare: 0000:00:02.0 bcs0: depth<-0
<0> [574.307591] i915_sel-6278 6.... 538589733us : intel_engine_stop_cs: 0000:00:02.0 bcs0:
<0> [574.307591] i915_sel-6278 6.... 538589757us : execlists_reset_prepare: 0000:00:02.0 vcs0: depth<-0
<0> [574.307591] i915_sel-6278 6.... 538589758us : intel_engine_stop_cs: 0000:00:02.0 vcs0:
<0> [574.307591] i915_sel-6278 6.... 538589771us : execlists_reset_prepare: 0000:00:02.0 vcs1: depth<-0
<0> [574.307591] i915_sel-6278 6.... 538589772us : intel_engine_stop_cs: 0000:00:02.0 vcs1:
<0> [574.307591] i915_sel-6278 6.... 538589778us : execlists_reset_prepare: 0000:00:02.0 vecs0: depth<-0
<0> [574.307591] i915_sel-6278 6.... 538589780us : intel_engine_stop_cs: 0000:00:02.0 vecs0:
<0> [574.307591] i915_sel-6278 6.... 538589786us : __intel_gt_reset: 0000:00:02.0 engine_mask=ff
<0> [574.307591] i915_sel-6278 6.... 538591175us : execlists_reset_cancel: 0000:00:02.0 rcs0:
<0> [574.307591] i915_sel-6278 6.... 538591970us : execlists_reset_cancel: 0000:00:02.0 bcs0:
<0> [574.307591] i915_sel-6278 6.... 538591982us : execlists_reset_cancel: 0000:00:02.0 vcs0:
<0> [574.307591] i915_sel-6278 6.... 538591996us : execlists_reset_cancel: 0000:00:02.0 vcs1:
<0> [574.307591] i915_sel-6278 6.... 538592759us : execlists_reset_cancel: 0000:00:02.0 vecs0:
<0> [574.307591] i915_sel-6278 6.... 538592977us : execlists_reset_finish: 0000:00:02.0 rcs0: depth->0
<0> [574.307591] i915_sel-6278 6.N.. 538592996us : execlists_reset_finish: 0000:00:02.0 bcs0: depth->0
<0> [574.307591] i915_sel-6278 6.N.. 538593023us : execlists_reset_finish: 0000:00:02.0 vcs0: depth->0
<0> [574.307591] i915_sel-6278 6.N.. 538593037us : execlists_reset_finish: 0000:00:02.0 vcs1: depth->0
<0> [574.307591] i915_sel-6278 6.N.. 538593051us : execlists_reset_finish: 0000:00:02.0 vecs0: depth->0
<0> [574.307591] i915_sel-6278 6.... 538593407us : __intel_gt_set_wedged.part.32: 0000:00:02.0 end
<0> [574.307591] kworker/-210 7d..1 551958381us : execlists_unhold: 0000:00:02.0 rcs0: fence f51b:2, current 2 hold release
<0> [574.307591] i915_sel-6278 0.... 559490788us : i915_request_retire: 0000:00:02.0 rcs0: fence f51b:2, current 2
<0> [574.307591] i915_sel-6278 0.... 559490793us : intel_context_unpin: 0000:00:02.0 rcs0: context:f51b unpin
<0> [574.307591] i915_sel-6278 0.... 559490798us : __engine_park: 0000:00:02.0 rcs0: parked
<0> [574.307591] i915_sel-6278 0.... 559490982us : __intel_context_retire: 0000:00:02.0 rcs0: context:f51c retire runtime: { total:30004ns, avg:30004ns }
<0> [574.307591] i915_sel-6278 0.... 559491372us : __engine_park: __engine_park:261 GEM_BUG_ON(engine->execlists.queue_priority_hint != (-((int)(~0U >> 1)) - 1))

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/20200227085723.1961649-9-chris@chris-wilson.co.uk

drm/i915/gt: Check engine-is-awake on reset later

As we drop the engine-pm on retiring, that may happen while there are
still CS events in the buffer. As such we cannot assert the engine is
still active on reset, until we know that the current request is still
in flight.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1338
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200227204727.2009346-1-chris@chris-wilson.co.uk

drm/i915: Allow userspace to specify ringsize on construction

No good reason why we must always use a static ringsize, so let
userspace select one during construction.

Link: https://github.com/intel/compute-runtime/pull/261
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Steve Carbonari <steven.carbonari@intel.com>
Reviewed-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200225192206.1107336-2-chris@chris-wilson.co.uk

drm/i915/gt: Protect signaler walk with RCU

While we know that the waiters cannot disappear as we walk our list
(only that they might be added), the same cannot be said for our
signalers as they may be completed by the HW and retired as we process
this request. Ergo we need to use rcu to protect the list iteration and
remember to mark up the list_del_rcu.

v2: Mark the deps as safe-for-rcu

Fixes: 793c22617367 ("drm/i915/gt: Protect execlists_hold/unhold from new waiters")
Fixes: 32ff621fd744 ("drm/i915/gt: Allow temporary suspension of inflight requests")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200220075025.1539375-1-chris@chris-wilson.co.uk

drm/i915/gt: Avoid resetting ring->head outside of its timeline mutex

We manipulate ring->head while active in i915_request_retire underneath
the timeline manipulation. We cannot rely on a stable ring->head outside
of the timeline->mutex, in particular while setting up the context for
resume and reset.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1126
Fixes: 0881954965e3 ("drm/i915: Introduce intel_context.pin_mutex for pin management")
Fixes: e5dadff4b093 ("drm/i915: Protect request retirement with timeline->mutex")
References: f3c0efc9fe7a ("drm/i915/execlists: Leave resetting ring to intel_ring")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211120131.958949-1-chris@chris-wilson.co.uk
(cherry picked from commit 42827350f75c56d0fe9f15d8425a1390528958b6)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/execlists: Always force a context reload when rewinding RING_TAIL

If we rewind the RING_TAIL on a context, due to a preemption event, we
must force the context restore for the RING_TAIL update to be properly
handled. Rather than note which preemption events may cause us to rewind
the tail, compare the new request's tail with the previously submitted
RING_TAIL, as it turns out that timeslicing was causing unexpected
rewinds.

<idle>-0 0d.s2 1280851190us : __execlists_submission_tasklet: 0000:00:02.0 rcs0: expired last=130:4698, prio=3, hint=3
<idle>-0 0d.s2 1280851192us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 66:119966, current 119964
<idle>-0 0d.s2 1280851195us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 130:4698, current 4695
<idle>-0 0d.s2 1280851198us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 130:4696, current 4695
^---- Note we unwind 2 requests from the same context

<idle>-0 0d.s2 1280851208us : __i915_request_submit: 0000:00:02.0 rcs0: fence 130:4696, current 4695
<idle>-0 0d.s2 1280851213us : __i915_request_submit: 0000:00:02.0 rcs0: fence 134:1508, current 1506
^---- But to apply the new timeslice, we have to replay the first request
before the new client can start -- the unexpected RING_TAIL rewind

<idle>-0 0d.s2 1280851219us : trace_ports: 0000:00:02.0 rcs0: submit { 130:4696*, 134:1508 }
synmark2-5425 2..s. 1280851239us : process_csb: 0000:00:02.0 rcs0: cs-irq head=5, tail=0
synmark2-5425 2..s. 1280851240us : process_csb: 0000:00:02.0 rcs0: csb[0]: status=0x00008002:0x00000000
^---- Preemption event for the ELSP update; note the lite-restore

synmark2-5425 2..s. 1280851243us : trace_ports: 0000:00:02.0 rcs0: preempted { 130:4698, 66:119966 }
synmark2-5425 2..s. 1280851246us : trace_ports: 0000:00:02.0 rcs0: promote { 130:4696*, 134:1508 }
synmark2-5425 2.... 1280851462us : __i915_request_commit: 0000:00:02.0 rcs0: fence 130:4700, current 4695
synmark2-5425 2.... 1280852111us : __i915_request_commit: 0000:00:02.0 rcs0: fence 130:4702, current 4695
synmark2-5425 2.Ns1 1280852296us : process_csb: 0000:00:02.0 rcs0: cs-irq head=0, tail=2
synmark2-5425 2.Ns1 1280852297us : process_csb: 0000:00:02.0 rcs0: csb[1]: status=0x00000814:0x00000000
synmark2-5425 2.Ns1 1280852299us : trace_ports: 0000:00:02.0 rcs0: completed { 130:4696!, 134:1508 }
synmark2-5425 2.Ns1 1280852301us : process_csb: 0000:00:02.0 rcs0: csb[2]: status=0x00000818:0x00000040
synmark2-5425 2.Ns1 1280852302us : trace_ports: 0000:00:02.0 rcs0: completed { 134:1508, 0:0 }
synmark2-5425 2.Ns1 1280852313us : process_csb: process_csb:2336 GEM_BUG_ON(!i915_request_completed(*execlists->active) && !reset_in_progress(execlists))

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Referenecs: 82c69bf58650 ("drm/i915/gt: Detect if we miss WaIdleLiteRestore")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Link: https://patchwork.freedesktop.org/patch/msgid/20200207211452.2860634-1-chris@chris-wilson.co.uk
(cherry picked from commit 5ba32c7be81e53ea8a27190b0f6be98e6c6779af)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/gt: Protect defer_request() from new waiters

Mika spotted

<4>[17436.705441] general protection fault: 0000 [#1] PREEMPT SMP PTI
<4>[17436.705447] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 5.5.0+ #1
<4>[17436.705449] Hardware name: System manufacturer System Product Name/Z170M-PLUS, BIOS 3805 05/16/2018
<4>[17436.705512] RIP: 0010:__execlists_submission_tasklet+0xc4d/0x16e0 [i915]
<4>[17436.705516] Code: c5 4c 8d 60 e0 75 17 e9 8c 07 00 00 49 8b 44 24 20 49 39 c5 4c 8d 60 e0 0f 84 7a 07 00 00 49 8b 5c 24 08 49 8b 87 80 00 00 00 <48> 39 83 d8 fe ff ff 75 d9 48 8b 83 88 fe ff ff a8 01 0f 84 b6 05
<4>[17436.705518] RSP: 0018:ffffc9000012ce80 EFLAGS: 00010083
<4>[17436.705521] RAX: ffff88822ae42000 RBX: 5a5a5a5a5a5a5a5a RCX: dead000000000122
<4>[17436.705523] RDX: ffff88822ae42588 RSI: ffff8881e32a7908 RDI: ffff8881c429fd48
<4>[17436.705525] RBP: ffffc9000012cf00 R08: ffff88822ae42588 R09: 00000000fffffffe
<4>[17436.705527] R10: ffff8881c429fb80 R11: 00000000a677cf08 R12: ffff8881c42a0aa8
<4>[17436.705529] R13: ffff8881c429fd38 R14: ffff88822ae42588 R15: ffff8881c429fb80
<4>[17436.705532] FS: 0000000000000000(0000) GS:ffff88822ed00000(0000) knlGS:0000000000000000
<4>[17436.705534] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
<4>[17436.705536] CR2: 00007f858c76d000 CR3: 0000000005610003 CR4: 00000000003606e0
<4>[17436.705538] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
<4>[17436.705540] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
<4>[17436.705542] Call Trace:
<4>[17436.705545] <IRQ>
<4>[17436.705603] execlists_submission_tasklet+0xc0/0x130 [i915]

which is us consuming a partially initialised new waiter in
defer_requests(). We can prevent this by initialising the i915_dependency
prior to making it visible, and since we are using a concurrent
list_add/iterator mark them up to the compiler.

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200206204915.2636606-2-chris@chris-wilson.co.uk
(cherry picked from commit f14f27b1663269a81ed62d3961fe70250a1a0623)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/gt: Rearrange code to silence compiler

Without selftests enabled, I915_SELFTEST_ONLY becomes a dummy,
generating a bare '0'. This causes the compiler to complain about a
useless line, and while we could use I915_SELFTEST_DECLARE instead, it
is a bit messier. Move the selftest-only code to a helper and make that
conditional on having selftests enabled.

Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200217095835.599827-1-chris@chris-wilson.co.uk

drm/i915: Track hw reported context runtime

GPU saves accumulated context runtime (in CS timestamp units) in PPHWSP
which will be useful for us in cases when we are not able to track context
busyness ourselves (like with GuC). Keep a copy of this in struct
intel_context from where it can be easily read even if the context is not
pinned.

v2:
(Chris)
* Do not store pphwsp address in intel_context.
* Log CS wrap-around.
* Simplify calculation by relying on integer wraparound.
v3:
* Include total/avg in traces and error state for debugging

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@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/20200216133620.394962-1-chris@chris-wilson.co.uk

drm/i915/gt: Suppress warnings for unused debugging locals

With debugging turned off, we have to tell the compiler not to warn
about the unused debug locals.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200213081217.3107410-1-chris@chris-wilson.co.uk

drm/i915/gt: Expand bad CS completion event debug

Show the ring/request/context state if we see what we believe is an
early CS completion.

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/20200211230944.1203098-1-chris@chris-wilson.co.uk

drm/i915/execlists: Reclaim the hanging virtual request

If we encounter a hang on a virtual engine, as we process the hang the
request may already have been moved back to the virtual engine (we are
processing the hang on the physical engine). We need to reclaim the
request from the virtual engine so that the locking is consistent and
local to the real engine on which we will hold the request for error
state capturing.

v2: Pull the reclamation into execlists_hold() and assert that cannot be
called from outside of the reset (i.e. with the tasklet disabled).
v3: Added selftest
v4: Drop the reference owned by the virtual engine

Fixes: ad18ba7b5eeb ("drm/i915/execlists: Offline error capture")
Testcase: igt/gem_exec_balancer/hang
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200122140243.495621-2-chris@chris-wilson.co.uk
(cherry picked from commit 989df3a7bd2abe566521e61d1aebf603eb013b7f)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/execlists: Take a reference while capturing the guilty request

Thanks to preempt-to-busy, we leave the request on the HW as we submit
the preemption request. This means that the request may complete at any
moment as we process HW events, and in particular the request may be
retired as we are planning to capture it for a preemption timeout.

Be more careful while obtaining the request to capture after a
preemption timeout, and check to see if it completed before we were able
to put it on the on-hold list. If we do see it did complete just before
we capture the request, proclaim the preemption-timeout a false positive
and pardon the reset as we should hit an arbitration point momentarily
and so be able to process the preemption.

Note that even after we move the request to be on hold it may be retired
(as the reset to stop the HW comes after), so we do require to hold our
own reference as we work on the request for capture (and all of the
peeking at state within the request needs to be carefully protected).

Fixes: c3f1ed90e6ff ("drm/i915/gt: Allow temporary suspension of inflight requests")
Closes: https://gitlab.freedesktop.org/drm/intel/issues/997
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/20200122140243.495621-1-chris@chris-wilson.co.uk
(cherry picked from commit 4ba5c086a1d8e38d6927967ae1a3271a6ab7a927)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/execlists: Offline error capture

Currently, we skip error capture upon forced preemption. We apply forced
preemption when there is a higher priority request that should be
running but is being blocked, and we skip inline error capture so that
the preemption request is not further delayed by a user controlled
capture -- extending the denial of service.

However, preemption reset is also used for heartbeats and regular GPU
hangs. By skipping the error capture, we remove the ability to debug GPU
hangs.

In order to capture the error without delaying the preemption request
further, we can do an out-of-line capture by removing the guilty request
from the execution queue and scheduling a worker to dump that request.
When removing a request, we need to remove the entire context and all
descendants from the execution queue, so that they do not jump past.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/738
Fixes: 3a7a92aba8fb ("drm/i915/execlists: Force preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116184754.2860848-3-chris@chris-wilson.co.uk
(cherry picked from commit 748317386afb235e11616098d2c7772e49776b58)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/gt: Allow temporary suspension of inflight requests

In order to support out-of-line error capture, we need to remove the
active request from HW and put it to one side while a worker compresses
and stores all the details associated with that request. (As that
compression may take an arbitrary user-controlled amount of time, we
want to let the engine continue running on other workloads while the
hanging request is dumped.) Not only do we need to remove the active
request, but we also have to remove its context and all requests that
were dependent on it (both in flight, queued and future submission).

Finally once the capture is complete, we need to be able to resubmit the
request and its dependents and allow them to execute.

v2: Replace stack recursion with a simple list.
v3: Check all the parents, not just the first, when searching for a
stuck ancestor!

References: https://gitlab.freedesktop.org/drm/intel/issues/738
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/20200116184754.2860848-2-chris@chris-wilson.co.uk
(cherry picked from commit 32ff621fd74496f0c33644125fb69ff175859b1f)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915: Keep track of request among the scheduling lists

If we keep track of when the i915_request.sched.link is on the HW
runlist, or in the priority queue we can simplify our interactions with
the request (such as during rescheduling). This also simplifies the next
patch where we introduce a new in-between list, for requests that are
ready but neither on the run list or in the queue.

v2: Update i915_sched_node.link explanation for current usage where it
is a link on both the queue and on the runlists.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116184754.2860848-1-chris@chris-wilson.co.uk
(cherry picked from commit 672c368f9398042b629740cc9816e8e939eff2db)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915: Disable use of hwsp_cacheline for kernel_context

Currently on execlists, we use a local hwsp for the kernel_context,
rather than the engine's HWSP, as this is the default for execlists.
However, seqno wrap requires allocating a new HWSP cacheline, and may
require pinning a new HWSP page in the GGTT. This operation requiring
pinning in the GGTT is not allowed within the kernel_context timeline,
as doing so may require re-entering the kernel_context in order to evict
from the GGTT. As we want to avoid requiring a new HWSP for the
kernel_context, we can use the permanently pinned engine's HWSP instead.
However to do so we must prevent the use of semaphores reading the
kernel_context's HWSP, as the use of semaphores do not support rollover
onto the same cacheline. Fortunately, the kernel_context is mostly
isolated, so unlikely to give benefit to semaphores.

Reported-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200210205722.794180-5-chris@chris-wilson.co.uk

drm/i915/gt: Avoid resetting ring->head outside of its timeline mutex

We manipulate ring->head while active in i915_request_retire underneath
the timeline manipulation. We cannot rely on a stable ring->head outside
of the timeline->mutex, in particular while setting up the context for
resume and reset.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1126
Fixes: 0881954965e3 ("drm/i915: Introduce intel_context.pin_mutex for pin management")
Fixes: e5dadff4b093 ("drm/i915: Protect request retirement with timeline->mutex")
References: f3c0efc9fe7a ("drm/i915/execlists: Leave resetting ring to intel_ring")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211120131.958949-1-chris@chris-wilson.co.uk

drm/i915/execlists: Leave resetting ring to intel_ring

We need to allow concurrent intel_context_unpin, which means avoiding
doing destructive operations like intel_ring_reset(). This was already
fixed for intel_ring_unpin() in commit 0725d9a31869 ("drm/i915/gt: Make
intel_ring_unpin() safe for concurrent pint"), but I overlooked that
execlists_context_unpin() also made the same mistake.

Reported-by: Matthew Brost <matthew.brost@intel.com>
Fixes: 841350223816 ("drm/i915/gt: Drop mutex serialisation between context pin/unpin")
References: 0725d9a31869 ("drm/i915/gt: Make intel_ring_unpin() safe for concurrent pint")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115175829.2761329-1-chris@chris-wilson.co.uk
(cherry picked from commit f3c0efc9fe7a4e61544034f525348a3aa86ac5aa)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/selftests: Drop live_preempt_hang

live_preempt_hang's use of hang injection has been superseded by
live_preempt_reset's use of an non-preemptible spinner. The latter does
not require intrusive hacks into the code.

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/20200209230838.361154-2-chris@chris-wilson.co.uk

drm/i915/gt: Use the BIT when checking the flags, not the index

In converting over to using set_bit()/test_bit(), when manually
inspecting the rq->fence.flags, we need to use BIT().

Fixes: e1c31fb5dde3 ("drm/i915: Merge i915_request.flags with i915_request.fence.flags")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115122509.2673075-1-chris@chris-wilson.co.uk
(cherry picked from commit 72ff2b8d5f2dcb09bfa37b902c23311eec426496)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/i915/execlists: Ignore tracek for nop process_csb

Recording the frequent inspection of CSB head/tail when there is
expected to be no update adds noise to the debug trace. (Not entirely
useless, but since we know the sequence of function calls, we can
surmise the function was called -- so redundant.)

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200209131922.180287-2-chris@chris-wilson.co.uk

drm/i915/gt: Fix hold/unhold recursion

In eliminating the recursion from walking the tree of signalers/waiters
for processing the hold/unhold operations, a crucial error crept in
where we looked at the parent request and not the list element when
processing the list.

Brown paper bag, much?

Closes: https://gitlab.freedesktop.org/drm/intel/issues/1166
Fixes: 32ff621fd744 ("drm/i915/gt: Allow temporary suspension of inflight requests")
Fixes: 748317386afb ("drm/i915/execlists: Offline error capture")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200209131922.180287-1-chris@chris-wilson.co.uk

drm/i915: Remove lite restore defines

We have switched from tail manipulation to forced context restore
to implement WaIdleLiteRestore. Remove the old defines and comments.

Note: we still do emit the WA tail, and use it as our first attempt to
avoid forcing a full-restore instead of a lite-restore, we just have a
much stronger backup mechanism for repeated preemptions.

References: f26a9e959a7b ("drm/i915/gt: Detect if we miss WaIdleLiteRestore")
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.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/20200203163312.15475-1-mika.kuoppala@linux.intel.com

drm/i915/execlists: Always force a context reload when rewinding RING_TAIL

If we rewind the RING_TAIL on a context, due to a preemption event, we
must force the context restore for the RING_TAIL update to be properly
handled. Rather than note which preemption events may cause us to rewind
the tail, compare the new request's tail with the previously submitted
RING_TAIL, as it turns out that timeslicing was causing unexpected
rewinds.

<idle>-0 0d.s2 1280851190us : __execlists_submission_tasklet: 0000:00:02.0 rcs0: expired last=130:4698, prio=3, hint=3
<idle>-0 0d.s2 1280851192us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 66:119966, current 119964
<idle>-0 0d.s2 1280851195us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 130:4698, current 4695
<idle>-0 0d.s2 1280851198us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence 130:4696, current 4695
^---- Note we unwind 2 requests from the same context

<idle>-0 0d.s2 1280851208us : __i915_request_submit: 0000:00:02.0 rcs0: fence 130:4696, current 4695
<idle>-0 0d.s2 1280851213us : __i915_request_submit: 0000:00:02.0 rcs0: fence 134:1508, current 1506
^---- But to apply the new timeslice, we have to replay the first request
before the new client can start -- the unexpected RING_TAIL rewind

<idle>-0 0d.s2 1280851219us : trace_ports: 0000:00:02.0 rcs0: submit { 130:4696*, 134:1508 }
synmark2-5425 2..s. 1280851239us : process_csb: 0000:00:02.0 rcs0: cs-irq head=5, tail=0
synmark2-5425 2..s. 1280851240us : process_csb: 0000:00:02.0 rcs0: csb[0]: status=0x00008002:0x00000000
^---- Preemption event for the ELSP update; note the lite-restore

synmark2-5425 2..s. 1280851243us : trace_ports: 0000:00:02.0 rcs0: preempted { 130:4698, 66:119966 }
synmark2-5425 2..s. 1280851246us : trace_ports: 0000:00:02.0 rcs0: promote { 130:4696*, 134:1508 }
synmark2-5425 2.... 1280851462us : __i915_request_commit: 0000:00:02.0 rcs0: fence 130:4700, current 4695
synmark2-5425 2.... 1280852111us : __i915_request_commit: 0000:00:02.0 rcs0: fence 130:4702, current 4695
synmark2-5425 2.Ns1 1280852296us : process_csb: 0000:00:02.0 rcs0: cs-irq head=0, tail=2
synmark2-5425 2.Ns1 1280852297us : process_csb: 0000:00:02.0 rcs0: csb[1]: status=0x00000814:0x00000000
synmark2-5425 2.Ns1 1280852299us : trace_ports: 0000:00:02.0 rcs0: completed { 130:4696!, 134:1508 }
synmark2-5425 2.Ns1 1280852301us : process_csb: 0000:00:02.0 rcs0: csb[2]: status=0x00000818:0x00000040
synmark2-5425 2.Ns1 1280852302us : trace_ports: 0000:00:02.0 rcs0: completed { 134:1508, 0:0 }
synmark2-5425 2.Ns1 1280852313us : process_csb: process_csb:2336 GEM_BUG_ON(!i915_request_completed(*execlists->active) && !reset_in_progress(execlists))

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Referenecs: 82c69bf58650 ("drm/i915/gt: Detect if we miss WaIdleLiteRestore")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Link: https://patchwork.freedesktop.org/patch/msgid/20200207211452.2860634-1-chris@chris-wilson.co.uk

drm/i915/gt: Protect execlists_hold/unhold from new waiters

As we may add new waiters to a request as it is being run, we need to
mark the list iteration as being safe for concurrent addition.

v2: Mika spotted that we used the same trick for signalers_list, so warn
the compiler about the lockless walk there as well.

Fixes: 32ff621fd744 ("drm/i915/gt: Allow temporary suspension of inflight requests")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200207110213.2734386-1-chris@chris-wilson.co.uk

drm/i915/gt: Protect defer_request() from new waiters

Mika spotted

<4>[17436.705441] general protection fault: 0000 [#1] PREEMPT SMP PTI
<4>[17436.705447] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 5.5.0+ #1
<4>[17436.705449] Hardware name: System manufacturer System Product Name/Z170M-PLUS, BIOS 3805 05/16/2018
<4>[17436.705512] RIP: 0010:__execlists_submission_tasklet+0xc4d/0x16e0 [i915]
<4>[17436.705516] Code: c5 4c 8d 60 e0 75 17 e9 8c 07 00 00 49 8b 44 24 20 49 39 c5 4c 8d 60 e0 0f 84 7a 07 00 00 49 8b 5c 24 08 49 8b 87 80 00 00 00 <48> 39 83 d8 fe ff ff 75 d9 48 8b 83 88 fe ff ff a8 01 0f 84 b6 05
<4>[17436.705518] RSP: 0018:ffffc9000012ce80 EFLAGS: 00010083
<4>[17436.705521] RAX: ffff88822ae42000 RBX: 5a5a5a5a5a5a5a5a RCX: dead000000000122
<4>[17436.705523] RDX: ffff88822ae42588 RSI: ffff8881e32a7908 RDI: ffff8881c429fd48
<4>[17436.705525] RBP: ffffc9000012cf00 R08: ffff88822ae42588 R09: 00000000fffffffe
<4>[17436.705527] R10: ffff8881c429fb80 R11: 00000000a677cf08 R12: ffff8881c42a0aa8
<4>[17436.705529] R13: ffff8881c429fd38 R14: ffff88822ae42588 R15: ffff8881c429fb80
<4>[17436.705532] FS: 0000000000000000(0000) GS:ffff88822ed00000(0000) knlGS:0000000000000000
<4>[17436.705534] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
<4>[17436.705536] CR2: 00007f858c76d000 CR3: 0000000005610003 CR4: 00000000003606e0
<4>[17436.705538] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
<4>[17436.705540] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
<4>[17436.705542] Call Trace:
<4>[17436.705545] <IRQ>
<4>[17436.705603] execlists_submission_tasklet+0xc0/0x130 [i915]

which is us consuming a partially initialised new waiter in
defer_requests(). We can prevent this by initialising the i915_dependency
prior to making it visible, and since we are using a concurrent
list_add/iterator mark them up to the compiler.

Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200206204915.2636606-2-chris@chris-wilson.co.uk

drm/i915: extract engine WA programming to common resume function

The workarounds are a common "feature" across gens and submission
mechanisms and we already call the other WA related functions from
common engine ones (<setup/cleanup>_common), so it makes sense to
do the same with WA application. Medium-term, This will help us
reduce the duplication once the GuC resume function is added, but short
term it will also allow us to use the workaround lists for pre-gen8
engine workarounds.

Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Matthew Brost <matthew.brost@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/20200131075716.2212299-2-chris@chris-wilson.co.uk

drm/i915: Use the async worker to avoid reclaim tainting the ggtt->mutex

On Braswell and Broxton (also known as Valleyview and Apollolake), we
need to serialise updates of the GGTT using the big stop_machine()
hammer. This has the side effect of appearing to lockdep as a possible
reclaim (since it uses the cpuhp mutex and that is tainted by per-cpu
allocations). However, we want to use vm->mutex (including ggtt->mutex)
from within the shrinker and so must avoid such possible taints. For this
purpose, we introduced the asynchronous vma binding and we can apply it
to the PIN_GLOBAL so long as take care to add the necessary waits for
the worker afterwards.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/211
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200130181710.2030251-3-chris@chris-wilson.co.uk

drm/i915/execlists: Ignore discrepancies in pending[] across resets

When we reset the engine, we first remove the guilty request from the
active list. If it so happens that there is a pending preemption event
to process before we handle the reset, when we inspect that event we
find ourselves a little confused as we have bent the rules slightly to
perform the reset.

Just ignore any discrepancies inside reset, we know we'll start again
from scratch afterwards.

<0>[ 536.940213] <idle>-0 6..s1 537441383us : execlists_reset: 0000:00:02.0 vcs0: reset for CS error
<0>[ 536.940213] i915_sel-7302 2d..1 537441386us : trace_ports: 0000:00:02.0 vcs0: submit { 10c59:2*, 10c5a:2 }
<0>[ 536.940213] <idle>-0 6d.s2 537471320us : __i915_request_unsubmit: 0000:00:02.0 vcs0: fence 10c59:2, current 1
<0>[ 536.940213] <idle>-0 6d.s2 537471321us : execlists_hold: 0000:00:02.0 vcs0: fence 10c59:2, current 1 on hold
<0>[ 536.940213] <idle>-0 6.Ns1 537471328us : intel_engine_reset: 0000:00:02.0 vcs0: flags=10
<0>[ 536.940213] <idle>-0 6.Ns1 537471421us : execlists_reset_prepare: 0000:00:02.0 vcs0: depth<-1
<0>[ 536.940213] <idle>-0 6.Ns1 537471422us : intel_engine_stop_cs: 0000:00:02.0 vcs0:
<0>[ 536.940213] <idle>-0 6.Ns1 537472424us : intel_engine_stop_cs: 0000:00:02.0 vcs0: timed out on STOP_RING -> IDLE
<0>[ 536.940213] <idle>-0 6.Ns1 537472429us : __intel_gt_reset: 0000:00:02.0 engine_mask=4
<0>[ 536.940213] <idle>-0 6.Ns1 537472442us : execlists_reset_rewind: 0000:00:02.0 vcs0:
<0>[ 536.940213] <idle>-0 6dNs2 537472443us : process_csb: 0000:00:02.0 vcs0: cs-irq head=4, tail=5
<0>[ 536.940213] <idle>-0 6dNs2 537472444us : process_csb: 0000:00:02.0 vcs0: csb[5]: status=0x00008002:0x20000060
<0>[ 536.940213] <idle>-0 6dNs2 537472464us : trace_ports: 0000:00:02.0 vcs0: preempted { 10c59:2*, 0:0 }
<0>[ 536.940213] <idle>-0 6dNs2 537472465us : trace_ports: 0000:00:02.0 vcs0: promote { 10c59:2*, 10c5a:2 }
<0>[ 536.940213] <idle>-0 6dNs2 537472706us : assert_pending_valid: assert_pending_valid:1417 GEM_BUG_ON(!i915_request_is_active(rq))

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/20200129165935.1266132-1-chris@chris-wilson.co.uk

drm/i915/gt: Hook up CS_MASTER_ERROR_INTERRUPT

Now that we have offline error capture and can reset an engine from
inside an atomic context while also preserving the GPU state for
post-mortem analysis, it is time to handle error interrupts thrown by
the command parser.

This provides a much, much faster mechanism for us to detect known
problems than using heartbeats/hangchecks, and also provides a mechanism
for when those are disabled. However, it is limited to problems the HW
can detect in the CS and so not a complete solution for detecting lockups.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128204318.4182039-2-chris@chris-wilson.co.uk

drm/i915/execlist: Mark up racy read of execlists->pending[0]

We write to execlists->pending[0] in process_csb() to acknowledge the
completion of the ESLP update, outside of the main spinlock. When we
check the current status of the previous submission in
__execlists_submission_tasklet() we should therefore use READ_ONCE() to
reflect and document the unsynchronized read.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128171614.3845825-1-chris@chris-wilson.co.uk

drm/i915/perf: Fix OA context id overlap with idle context id

Engine context pinned in perf OA was set to same context id as
the idle context. Set the context id to an unused value.

Clear the sw context id field in lrc descriptor before ORing with
ce->tag (Chris)

Closes: https://gitlab.freedesktop.org/drm/intel/issues/756
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@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/20200124013701.40609-1-umesh.nerlige.ramappa@intel.com

drm/i915/execlists: Reclaim the hanging virtual request

If we encounter a hang on a virtual engine, as we process the hang the
request may already have been moved back to the virtual engine (we are
processing the hang on the physical engine). We need to reclaim the
request from the virtual engine so that the locking is consistent and
local to the real engine on which we will hold the request for error
state capturing.

v2: Pull the reclamation into execlists_hold() and assert that cannot be
called from outside of the reset (i.e. with the tasklet disabled).
v3: Added selftest
v4: Drop the reference owned by the virtual engine

Fixes: 748317386afb ("drm/i915/execlists: Offline error capture")
Testcase: igt/gem_exec_balancer/hang
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200122140243.495621-2-chris@chris-wilson.co.uk

drm/i915/execlists: Take a reference while capturing the guilty request

Thanks to preempt-to-busy, we leave the request on the HW as we submit
the preemption request. This means that the request may complete at any
moment as we process HW events, and in particular the request may be
retired as we are planning to capture it for a preemption timeout.

Be more careful while obtaining the request to capture after a
preemption timeout, and check to see if it completed before we were able
to put it on the on-hold list. If we do see it did complete just before
we capture the request, proclaim the preemption-timeout a false positive
and pardon the reset as we should hit an arbitration point momentarily
and so be able to process the preemption.

Note that even after we move the request to be on hold it may be retired
(as the reset to stop the HW comes after), so we do require to hold our
own reference as we work on the request for capture (and all of the
peeking at state within the request needs to be carefully protected).

Fixes: 32ff621fd744 ("drm/i915/gt: Allow temporary suspension of inflight requests")
Closes: https://gitlab.freedesktop.org/drm/intel/issues/997
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/20200122140243.495621-1-chris@chris-wilson.co.uk

drm/i915/execlists: Offline error capture

Currently, we skip error capture upon forced preemption. We apply forced
preemption when there is a higher priority request that should be
running but is being blocked, and we skip inline error capture so that
the preemption request is not further delayed by a user controlled
capture -- extending the denial of service.

However, preemption reset is also used for heartbeats and regular GPU
hangs. By skipping the error capture, we remove the ability to debug GPU
hangs.

In order to capture the error without delaying the preemption request
further, we can do an out-of-line capture by removing the guilty request
from the execution queue and scheduling a worker to dump that request.
When removing a request, we need to remove the entire context and all
descendants from the execution queue, so that they do not jump past.

Closes: https://gitlab.freedesktop.org/drm/intel/issues/738
Fixes: 3a7a92aba8fb ("drm/i915/execlists: Force preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116184754.2860848-3-chris@chris-wilson.co.uk

drm/i915/gt: Allow temporary suspension of inflight requests

In order to support out-of-line error capture, we need to remove the
active request from HW and put it to one side while a worker compresses
and stores all the details associated with that request. (As that
compression may take an arbitrary user-controlled amount of time, we
want to let the engine continue running on other workloads while the
hanging request is dumped.) Not only do we need to remove the active
request, but we also have to remove its context and all requests that
were dependent on it (both in flight, queued and future submission).

Finally once the capture is complete, we need to be able to resubmit the
request and its dependents and allow them to execute.

v2: Replace stack recursion with a simple list.
v3: Check all the parents, not just the first, when searching for a
stuck ancestor!

References: https://gitlab.freedesktop.org/drm/intel/issues/738
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/20200116184754.2860848-2-chris@chris-wilson.co.uk

drm/i915: Keep track of request among the scheduling lists

If we keep track of when the i915_request.sched.link is on the HW
runlist, or in the priority queue we can simplify our interactions with
the request (such as during rescheduling). This also simplifies the next
patch where we introduce a new in-between list, for requests that are
ready but neither on the run list or in the queue.

v2: Update i915_sched_node.link explanation for current usage where it
is a link on both the queue and on the runlists.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116184754.2860848-1-chris@chris-wilson.co.uk

drm/i915/execlists: Leave resetting ring to intel_ring

We need to allow concurrent intel_context_unpin, which means avoiding
doing destructive operations like intel_ring_reset(). This was already
fixed for intel_ring_unpin() in commit 0725d9a31869 ("drm/i915/gt: Make
intel_ring_unpin() safe for concurrent pint"), but I overlooked that
execlists_context_unpin() also made the same mistake.

Reported-by: Matthew Brost <matthew.brost@intel.com>
Fixes: 841350223816 ("drm/i915/gt: Drop mutex serialisation between context pin/unpin")
References: 0725d9a31869 ("drm/i915/gt: Make intel_ring_unpin() safe for concurrent pint")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115175829.2761329-1-chris@chris-wilson.co.uk

drm/i915/gt: Use the BIT when checking the flags, not the index

In converting over to using set_bit()/test_bit(), when manually
inspecting the rq->fence.flags, we need to use BIT().

Fixes: e1c31fb5dde3 ("drm/i915: Merge i915_request.flags with i915_request.fence.flags")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115122509.2673075-1-chris@chris-wilson.co.uk

drm/i915/gt: Always reset the timeslice after a context switch

Currently, we reset the timer after a pre-eemption event. This has the
side-effect that the timeslice runs into the second context after the
first is completed after a normal promotion event, causing the second
context to be swapped out early and switched for a third context. To be
more fair, we want to reset the clock after promotion as well.

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/20200113214546.1990139-1-chris@chris-wilson.co.uk

drm/i915/gen9: Clear residual context state on context switch

Intel ID: PSIRT-TA-201910-001
CVEID: CVE-2019-14615

Intel GPU Hardware prior to Gen11 does not clear EU state
during a context switch. This can result in information
leakage between contexts.

For Gen8 and Gen9, hardware provides a mechanism for
fast cleardown of the EU state, by issuing a PIPE_CONTROL
with bit 27 set. We can use this in a context batch buffer
to explicitly cleardown the state on every context switch.

As this workaround is already in place for gen8, we can borrow
the code verbatim for Gen9.

Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Cc: Kumar Valsan Prathap <prathap.kumar.valsan@intel.com>
Cc: Chris Wilson <chris.p.wilson@intel.com>
Cc: Balestrieri Francesco <francesco.balestrieri@intel.com>
Cc: Bloomfield Jon <jon.bloomfield@intel.com>
Cc: Dutt Sudeep <sudeep.dutt@intel.com>