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 May be 0 to disable heartbeats and therefore disable automatic GPU
26 config DRM_I915_PREEMPT_TIMEOUT
27 int "Preempt timeout (ms, jiffy granularity)"
28 default 640 # milliseconds
30 How long to wait (in milliseconds) for a preemption event to occur
31 when submitting a new context via execlists. If the current context
32 does not hit an arbitration point and yield to HW before the timer
33 expires, the HW will be reset to allow the more important context
36 May be 0 to disable the timeout.
38 config DRM_I915_SPIN_REQUEST
39 int "Busywait for request completion (us)"
40 default 5 # microseconds
42 Before sleeping waiting for a request (GPU operation) to complete,
43 we may spend some time polling for its completion. As the IRQ may
44 take a non-negligible time to setup, we do a short spin first to
45 check if the request will complete in the time it would have taken
46 us to enable the interrupt.
48 May be 0 to disable the initial spin. In practice, we estimate
49 the cost of enabling the interrupt (if currently disabled) to be
52 config DRM_I915_STOP_TIMEOUT
53 int "How long to wait for an engine to quiesce gracefully before reset (ms)"
54 default 100 # milliseconds
56 By stopping submission and sleeping for a short time before resetting
57 the GPU, we allow the innocent contexts also on the system to quiesce.
58 It is then less likely for a hanging context to cause collateral
59 damage as the system is reset in order to recover. The corollary is
60 that the reset itself may take longer and so be more disruptive to
61 interactive or low latency workloads.
63 config DRM_I915_TIMESLICE_DURATION
64 int "Scheduling quantum for userspace batches (ms, jiffy granularity)"
65 default 1 # milliseconds
67 When two user batches of equal priority are executing, we will
68 alternate execution of each batch to ensure forward progress of
69 all users. This is necessary in some cases where there may be
70 an implicit dependency between those batches that requires
71 concurrent execution in order for them to proceed, e.g. they
72 interact with each other via userspace semaphores. Each context
73 is scheduled for execution for the timeslice duration, before
74 switching to the next context.
76 May be 0 to disable timeslicing.