The "mmio" writes into vgpu registers are simple memory traps from the
guest into the host. We do not need to assert in the guest that the
device is awake for the io as we do not write to the device itself.
However, over time we have refactored all the mmio accessors with the
result that the vgpu reuses the gen2 accessors and so inherits the
assert for runtime-pm of the native device. The assert though has
actually been there since commit 3be0bf5acc ("drm/i915: Create vGPU
specific MMIO operations to reduce traps").
References: 3be0bf5acc ("drm/i915: Create vGPU specific MMIO operations to reduce traps")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Yan Zhao <yan.y.zhao@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Reviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20200811092532.13753-1-chris@chris-wilson.co.uk
Underruns happens when plane height + y offset is not a modulo of 4
when FBC is enabled. It happens when scanline is at vactive - 10 but
that is not feasible to do from the software side so here completely
disabling FBC when height + y offset matches to avoid visual glitches.
Specification says that it only affects TGL display C stepping and
newer but to simply the check and as TGL is already in final costumers
hands, pre-production display stepping A and B was also included.
BSpec: 52887 ICL
BSpec: 52888 EHL/JSL
BSpec: 52890/55378 TGL
BSpec: 53508 DG1
BSpec: 53273 RKL
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201019175609.28715-1-jose.souza@intel.com
On Tigerlake, we are seeing a repeat of commit d8f5053117 ("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: d8f5053117 ("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 233c1ae3c8)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
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: 8ab3a3812a ("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 bb65548e3c)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
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 233c1ae3c8 ("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 6ca7217dff)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
In commit 7994672309 ("drm/i915: Assume 100% brightness when not in
DPCD control mode"), we fixed the brightness level when DPCD control was
not active to max brightness. This is as good as we can guess since most
backlights go on full when uncontrolled.
However in doing so we changed the semantics of the initial
'backlight.enabled' value. At least on Pixelbooks, they were relying
on the brightness level in DP_EDP_BACKLIGHT_BRIGHTNESS_MSB to be 0 on
boot such that enabled would be false. This causes the device to be
enabled when the brightness is set. Without this, brightness control
doesn't work. So by changing brightness to max, we also flipped enabled
to be true on boot.
To fix this, make enabled a function of brightness and backlight control
mechanism.
Fixes: 7994672309 ("drm/i915: Assume 100% brightness when not in DPCD control mode")
Cc: Lyude Paul <lyude@redhat.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Cc: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Kevin Chowski <chowski@chromium.org>>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Signed-off-by: Lyude Paul <lyude@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200918002845.32766-1-sean@poorly.run
(cherry picked from commit 4ade8f31c2)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
If guest fills non-priv bb on ApolloLake/Broxton as Mesa i965 does in:
717e7539124d (i965: Use a WC map and memcpy for the batch instead of pw-)
Due to the missing flush of bb filled by VM vCPU, host GPU hangs on
executing these MI_BATCH_BUFFER.
Temporarily workaround this by setting SNOOP bit for PAT3 used by PPGTT
PML4 PTE: PAT(0) PCD(1) PWT(1).
The performance is still expected to be low, will need further improvement.
Acked-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Colin Xu <colin.xu@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20201012045231.226748-1-colin.xu@intel.com
Guest driver may reset HWSP to 0 as init value during D3->D0:
The full sequence is:
- Boot ->D0
- Update HWSP
- D0->D3
- ...In D3 state...
- D3->D0
- DMLR reset.
- Set engine HWSP to 0.
- Set engine ring mode to 0.
- Set engine HWSP to correct value.
- Set engine ring mode to correct value.
Ring mode is masked register so set 0 won't take effect.
However HWPS addr 0 is considered as invalid GGTT address which will
report error like:
gvt: vgpu 1: write invalid HWSP address, reg:0x2080, value:0x0
gvt: vgpu 1: fail to emulate MMIO write 00002080 len 4
Detected your guest driver doesn't support GVT-g.
Now vgpu 2 will enter failsafe mode.
Zero out HWSP addr is considered as a valid setting from device driver
so don't treat it as invalid HWSP addr.
V2:
Treat HWSP addr 0 as valid. (zhenyu)
V3:
Change patch title.
Reviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Colin Xu <colin.xu@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20200911065239.147789-1-colin.xu@intel.com
Pull rdma updates from Jason Gunthorpe:
"A usual cycle for RDMA with a typical mix of driver and core subsystem
updates:
- Driver minor changes and bug fixes for mlx5, efa, rxe, vmw_pvrdma,
hns, usnic, qib, qedr, cxgb4, hns, bnxt_re
- Various rtrs fixes and updates
- Bug fix for mlx4 CM emulation for virtualization scenarios where
MRA wasn't working right
- Use tracepoints instead of pr_debug in the CM code
- Scrub the locking in ucma and cma to close more syzkaller bugs
- Use tasklet_setup in the subsystem
- Revert the idea that 'destroy' operations are not allowed to fail
at the driver level. This proved unworkable from a HW perspective.
- Revise how the umem API works so drivers make fewer mistakes using
it
- XRC support for qedr
- Convert uverbs objects RWQ and MW to new the allocation scheme
- Large queue entry sizes for hns
- Use hmm_range_fault() for mlx5 On Demand Paging
- uverbs APIs to inspect the GID table instead of sysfs
- Move some of the RDMA code for building large page SGLs into
lib/scatterlist"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (191 commits)
RDMA/ucma: Fix use after free in destroy id flow
RDMA/rxe: Handle skb_clone() failure in rxe_recv.c
RDMA/rxe: Move the definitions for rxe_av.network_type to uAPI
RDMA: Explicitly pass in the dma_device to ib_register_device
lib/scatterlist: Do not limit max_segment to PAGE_ALIGNED values
IB/mlx4: Convert rej_tmout radix-tree to XArray
RDMA/rxe: Fix bug rejecting all multicast packets
RDMA/rxe: Fix skb lifetime in rxe_rcv_mcast_pkt()
RDMA/rxe: Remove duplicate entries in struct rxe_mr
IB/hfi,rdmavt,qib,opa_vnic: Update MAINTAINERS
IB/rdmavt: Fix sizeof mismatch
MAINTAINERS: CISCO VIC LOW LATENCY NIC DRIVER
RDMA/bnxt_re: Fix sizeof mismatch for allocation of pbl_tbl.
RDMA/bnxt_re: Use rdma_umem_for_each_dma_block()
RDMA/umem: Move to allocate SG table from pages
lib/scatterlist: Add support in dynamic allocation of SG table from pages
tools/testing/scatterlist: Show errors in human readable form
tools/testing/scatterlist: Rejuvenate bit-rotten test
RDMA/ipoib: Set rtnl_link_ops for ipoib interfaces
RDMA/uverbs: Expose the new GID query API to user space
...
Add a helper so we can set per asic default values. Also,
the module parameter is currently clamped to 8, but clamp it
per asic just in case some asics have different limits in the
future. Enable the option on gfx6,7 as well for consistency.
Acked-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
