Add support to partially execute a slice which is resized to zero.
Executing a zero size slice in a BO should mean that there is no DMA
transfers involved but you should still configure doorbell and semaphores.
For example consider a BO of size 18K and it is sliced into 3 6K slices
and user calls partial execute ioctl with resize as 10K.
slice 0 - size is 6k and offset is 0, so resize of 10K will not cut short
this slice hence we send the entire slice for execution.
slice 1 - size is 6k and offset is 6k, so resize of 10K will cut short this
slice and only the first 4k should be DMA along with configuring
doorbell and semaphores.
slice 2 - size is 6k and offset is 12k, so resize of 10k will cut short
this slice and no DMA transfer would be involved but we should
would configure doorbell and semaphores.
This change begs to change the behavior of 0 resize. Currently, 0 resize
partial execute ioctl behaves exactly like execute ioctl i.e. no resize.
After this patch all the slice in BO should behave exactly like slice 2 in
above example.
Refactor copy_partial_exec_reqs() to make it more readable and less
complex.
Signed-off-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231027164330.11978-1-quic_jhugo@quicinc.com
Current wrapper is right-sized to the message being transferred;
however, this is smaller than the structure defining message wrappers
since the trailing element is a union of message/transfer headers of
various sizes (8 and 32 bytes on 32-bit system where issue was
reported). Using the smaller header with a small message
(wire_trans_dma_xfer is 24 bytes including header) ends up being smaller
than a wrapper with the larger header. There are no accesses outside of
the defined size, however they are possible if the larger union member
is referenced.
Abort messages are outside of hot-path and changing the wrapper struct
would require a larger rewrite, so having the memory allocated to the
message be 8 bytes too big is acceptable.
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202310182253.bcb9JcyJ-lkp@intel.com/
Signed-off-by: Carl Vanderlip <quic_carlv@quicinc.com>
Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231027180810.4873-1-quic_jhugo@quicinc.com
The simple-framebuffer device tree bindings document the power-domains
property, so make sure that simplefb supports it. This ensures that the
power domains remain enabled as long as simplefb is active.
v2: - remove unnecessary call to simplefb_detach_genpds() since that's
already done automatically by devres
- fix crash if power-domains property is missing in DT
Signed-off-by: Thierry Reding <treding@nvidia.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231101172017.3872242-3-thierry.reding@gmail.com
Rather than call free_job and run_job in same work item have a dedicated
work item for each. This aligns with the design and intended use of work
queues.
v2:
- Test for DMA_FENCE_FLAG_TIMESTAMP_BIT before setting
timestamp in free_job() work item (Danilo)
v3:
- Drop forward dec of drm_sched_select_entity (Boris)
- Return in drm_sched_run_job_work if entity NULL (Boris)
v4:
- Replace dequeue with peek and invert logic (Luben)
- Wrap to 100 lines (Luben)
- Update comments for *_queue / *_queue_if_ready functions (Luben)
v5:
- Drop peek argument, blindly reinit idle (Luben)
- s/drm_sched_free_job_queue_if_ready/drm_sched_free_job_queue_if_done (Luben)
- Update work_run_job & work_free_job kernel doc (Luben)
v6:
- Do not move drm_sched_select_entity in file (Luben)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20231031032439.1558703-4-matthew.brost@intel.com
Reviewed-by: Luben Tuikov <ltuikov89@gmail.com>
Signed-off-by: Luben Tuikov <ltuikov89@gmail.com>
In Xe, the new Intel GPU driver, a choice has made to have a 1 to 1
mapping between a drm_gpu_scheduler and drm_sched_entity. At first this
seems a bit odd but let us explain the reasoning below.
1. In Xe the submission order from multiple drm_sched_entity is not
guaranteed to be the same completion even if targeting the same hardware
engine. This is because in Xe we have a firmware scheduler, the GuC,
which allowed to reorder, timeslice, and preempt submissions. If a using
shared drm_gpu_scheduler across multiple drm_sched_entity, the TDR falls
apart as the TDR expects submission order == completion order. Using a
dedicated drm_gpu_scheduler per drm_sched_entity solve this problem.
2. In Xe submissions are done via programming a ring buffer (circular
buffer), a drm_gpu_scheduler provides a limit on number of jobs, if the
limit of number jobs is set to RING_SIZE / MAX_SIZE_PER_JOB we get flow
control on the ring for free.
A problem with this design is currently a drm_gpu_scheduler uses a
kthread for submission / job cleanup. This doesn't scale if a large
number of drm_gpu_scheduler are used. To work around the scaling issue,
use a worker rather than kthread for submission / job cleanup.
v2:
- (Rob Clark) Fix msm build
- Pass in run work queue
v3:
- (Boris) don't have loop in worker
v4:
- (Tvrtko) break out submit ready, stop, start helpers into own patch
v5:
- (Boris) default to ordered work queue
v6:
- (Luben / checkpatch) fix alignment in msm_ringbuffer.c
- (Luben) s/drm_sched_submit_queue/drm_sched_wqueue_enqueue
- (Luben) Update comment for drm_sched_wqueue_enqueue
- (Luben) Positive check for submit_wq in drm_sched_init
- (Luben) s/alloc_submit_wq/own_submit_wq
v7:
- (Luben) s/drm_sched_wqueue_enqueue/drm_sched_run_job_queue
v8:
- (Luben) Adjust var names / comments
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://lore.kernel.org/r/20231031032439.1558703-3-matthew.brost@intel.com
Signed-off-by: Luben Tuikov <ltuikov89@gmail.com>
When a fence signals there is a very small race window where the timestamp
isn't updated yet. sync_file solves this by busy waiting for the
timestamp to appear, but on other ocassions didn't handled this
correctly.
Provide a dma_fence_timestamp() helper function for this and use it in
all appropriate cases.
Another alternative would be to grab the spinlock when that happens.
v2 by teddy: add a wait parameter to wait for the timestamp to show up, in case
the accurate timestamp is needed and/or the timestamp is not based on
ktime (e.g. hw timestamp)
v3 chk: drop the parameter again for unified handling
Signed-off-by: Yunxiang Li <Yunxiang.Li@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: 1774baa64f ("drm/scheduler: Change scheduled fence track v2")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
CC: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20230929104725.2358-1-christian.koenig@amd.com
Currently the VPU firmware prepares for D0i3 every time the VPU
is entering D0i2 Idle state. This is not optimal as we might not
enter D0i3 every time we enter D0i2 Idle and this preparation
is quite costly.
This optimization moves D0i3 preparation to a dedicated
message sent from the host driver only when the driver is about
to enter D0i3 - this reduces power consumption and latency for
certain workloads, for example audio workloads that submit
inference every 10 ms.
The VPU needs non zero time to enter IDLE state after responding to
D0i3 entry message. If the driver does not wait for the VPU to enter
IDLE state it could cause warm boot failures.
Signed-off-by: Andrzej Kacprowski <andrzej.kacprowski@linux.intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Signed-off-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231028133415.1169975-12-stanislaw.gruszka@linux.intel.com
The usage count of struct dev_pm_info is an implementation detail that
is only available if CONFIG_PM is enabled, so printing it in a debug message
causes a build failure in configurations without PM:
In file included from include/linux/device.h:15,
from include/linux/pci.h:37,
from drivers/accel/ivpu/ivpu_pm.c:8:
drivers/accel/ivpu/ivpu_pm.c: In function 'ivpu_rpm_get_if_active':
drivers/accel/ivpu/ivpu_pm.c:254:51: error: 'struct dev_pm_info' has no member named 'usage_count'
254 | atomic_read(&vdev->drm.dev->power.usage_count));
| ^
include/linux/dev_printk.h:129:48: note: in definition of macro 'dev_printk'
129 | _dev_printk(level, dev, fmt, ##__VA_ARGS__); \
| ^~~~~~~~~~~
drivers/accel/ivpu/ivpu_drv.h:75:17: note: in expansion of macro 'dev_dbg'
75 | dev_dbg((vdev)->drm.dev, "[%s] " fmt, #type, ##args); \
| ^~~~~~~
drivers/accel/ivpu/ivpu_pm.c:253:9: note: in expansion of macro 'ivpu_dbg'
253 | ivpu_dbg(vdev, RPM, "rpm_get_if_active count %d\n",
| ^~~~~~~~
The print message does not seem essential, so the easiest workaround is
to just remove it.
Fixes: c39dc15191 ("accel/ivpu: Read clock rate only if device is up")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Signed-off-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231027152633.528490-1-arnd@kernel.org
Several virtualization use-cases either don't support 32 MultiMSIs
(Xen/VMware) or have significant drawbacks to their use (KVM's vIOMMU,
which is required to support 32 MSI, needs to allocate an alternate
system memory space for each device using vIOMMU (e.g. 8GB VM mem and
2 cards => 8 + 2 * 8 = 24GB host memory required)). Support these
cases by enabling a 1 MSI fallback mode.
Whenever all 32 MSIs requested are not available, a second request for
a single MSI is made. Its success is the initiator of single MSI mode.
This mode causes all interrupts generated by the device to be directed
to the 0th MSI (firmware >=v1.10 will do this as a response to the PCIe
MSI capability configuration). Likewise, all interrupt handlers for the
device are registered to the 0th MSI.
Since the DBC interrupt handler checks if the DBC is in use or if
there is any pending changes, the 'spurious' interrupts are
disregarded. If there is work to be done, the standard threaded IRQ
handler is dispatched.
On every interrupt, the MHI handler wakes up its threaded interrupt
handler, and attempts to wake any waiters for MHI state events.
Performance is within +-0.6% for test cases that typify real world
use. Larger differences ([-4,+132]%, avg +47%) exist for very simple
tasks (e.g. addition) compiled for single NSPs. It is assumed that the
small work and many interrupts typically cause contention (e.g. 16 NSPs
vs 4 CPUs), as evidenced by the standard deviation between runs also
decreasing (r=-0.48 between delta(Performace_test) and
delta(StdDev_test/Avg_test))
Signed-off-by: Carl Vanderlip <quic_carlv@quicinc.com>
Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231016170036.5409-1-quic_jhugo@quicinc.com
Rather than doing this in the IP code for the SDMA paging
engine, move it up to the core device level init level.
This should fix the scheduler init ordering.
v2: drop extra parens
v3: drop SDMA helpers
v4: Added a Fixes tag because amdgpu dereferences an uninitialized
scheduler without this patch, and this patch fixes this. (Luben)
Tested-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://lore.kernel.org/r/20231025171928.3318505-1-alexander.deucher@amd.com
Acked-by: Christian König <christian.koenig@amd.com>
Fixes: 56e449603f ("drm/sched: Convert the GPU scheduler to variable number of run-queues")
Signed-off-by: Luben Tuikov <ltuikov89@gmail.com>
