1 config DRM_I915_USERFAULT_AUTOSUSPEND
2 int "Runtime autosuspend delay for userspace GGTT mmaps (ms)"
3 default 250 # milliseconds
5 On runtime suspend, as we suspend the device, we have to revoke
6 userspace GGTT mmaps and force userspace to take a pagefault on
7 their next access. The revocation and subsequent recreation of
8 the GGTT mmap can be very slow and so we impose a small hysteris
9 that complements the runtime-pm autosuspend and provides a lower
10 floor on the autosuspend delay.
12 May be 0 to disable the extra delay and solely use the device level
13 runtime pm autosuspend delay tunable.
15 config DRM_I915_HEARTBEAT_INTERVAL
16 int "Interval between heartbeat pulses (ms)"
17 default 2500 # milliseconds
19 The driver sends a periodic heartbeat down all active engines to
20 check the health of the GPU and undertake regular house-keeping of
21 internal driver state.
23 This is adjustable via
24 /sys/class/drm/card?/engine/*/heartbeat_interval_ms
26 May be 0 to disable heartbeats and therefore disable automatic GPU
29 config DRM_I915_PREEMPT_TIMEOUT
30 int "Preempt timeout (ms, jiffy granularity)"
31 default 640 # milliseconds
33 How long to wait (in milliseconds) for a preemption event to occur
34 when submitting a new context via execlists. If the current context
35 does not hit an arbitration point and yield to HW before the timer
36 expires, the HW will be reset to allow the more important context
39 This is adjustable via
40 /sys/class/drm/card?/engine/*/preempt_timeout_ms
42 May be 0 to disable the timeout.
44 config DRM_I915_MAX_REQUEST_BUSYWAIT
45 int "Busywait for request completion limit (ns)"
46 default 8000 # nanoseconds
48 Before sleeping waiting for a request (GPU operation) to complete,
49 we may spend some time polling for its completion. As the IRQ may
50 take a non-negligible time to setup, we do a short spin first to
51 check if the request will complete in the time it would have taken
52 us to enable the interrupt.
54 This is adjustable via
55 /sys/class/drm/card?/engine/*/max_busywait_duration_ns
57 May be 0 to disable the initial spin. In practice, we estimate
58 the cost of enabling the interrupt (if currently disabled) to be
61 config DRM_I915_STOP_TIMEOUT
62 int "How long to wait for an engine to quiesce gracefully before reset (ms)"
63 default 100 # milliseconds
65 By stopping submission and sleeping for a short time before resetting
66 the GPU, we allow the innocent contexts also on the system to quiesce.
67 It is then less likely for a hanging context to cause collateral
68 damage as the system is reset in order to recover. The corollary is
69 that the reset itself may take longer and so be more disruptive to
70 interactive or low latency workloads.
72 This is adjustable via
73 /sys/class/drm/card?/engine/*/stop_timeout_ms
75 config DRM_I915_TIMESLICE_DURATION
76 int "Scheduling quantum for userspace batches (ms, jiffy granularity)"
77 default 1 # milliseconds
79 When two user batches of equal priority are executing, we will
80 alternate execution of each batch to ensure forward progress of
81 all users. This is necessary in some cases where there may be
82 an implicit dependency between those batches that requires
83 concurrent execution in order for them to proceed, e.g. they
84 interact with each other via userspace semaphores. Each context
85 is scheduled for execution for the timeslice duration, before
86 switching to the next context.
88 This is adjustable via
89 /sys/class/drm/card?/engine/*/timeslice_duration_ms
91 May be 0 to disable timeslicing.