1# Scheduler VM expectations 2 3Hafnium requires there to be a special 'primary' or 'scheduler' VM which is 4responsible for scheduling the other VMs. There are some particular expectations 5on this VM that are required for the rest of the system to function normally. 6 7## Scheduling 8 9The scheduler VM is responsible for scheduling the vCPUs of all the other VMs. 10It should request information about the VMs in the system using the 11`FFA_PARTITION_INFO_GET` function, and then schedule their vCPUs as it wishes. 12The recommended way of doing this is to create a kernel thread for each vCPU, 13which will repeatedly run that vCPU by calling `FFA_RUN`. 14 15`FFA_RUN` will return one of several possible functions, which must be handled 16as follows: 17 18### `FFA_INTERRUPT` 19 20The vCPU has been preempted but still has work to do. If the scheduling quantum 21has not expired, the scheduler MUST call `FFA_RUN` on the vCPU to allow it to 22continue. 23 24If `w1` is non-zero, then Hafnium would like `FFA_RUN` to be called on the vCPU 25specified there. The scheduler MUST either wake the vCPU in question up if it is 26blocked, or preempt and re-run it if it is already running somewhere. This gives 27Hafnium a chance to update any CPU state which might have changed. The scheduler 28should call `FFA_RUN` again on the sending VM as usual. 29 30### `FFA_YIELD` 31 32The vCPU has voluntarily yielded the CPU. The scheduler SHOULD take a scheduling 33decision to give cycles to those that need them but MUST call `FFA_RUN` on the 34vCPU at a later point. 35 36### `FFA_MSG_WAIT` 37 38The vCPU is blocked waiting for a message. The scheduler MUST take it off the 39run queue and not call `FFA_RUN` on the vCPU until it has either: 40 41* injected an interrupt 42* sent it a message 43* received `FFA_INTERRUPT` for it from another vCPU 44* the timeout provided in `w2` is not `FFA_SLEEP_INDEFINITE` and the 45 specified duration has expired. 46 47### `FFA_MSG_SEND` 48 49A message has been sent by the vCPU. If the recipient is the scheduler VM itself 50then it can handle it as it pleases. Otherwise the scheduler MUST run a vCPU 51from the recipient VM and priority SHOULD be given to those vCPUs that are 52waiting for a message. The scheduler should call `FFA_RUN` again on the sending 53VM as usual. 54 55### `FFA_RX_RELEASE` 56 57The vCPU has made the mailbox writable and there are pending waiters. The 58scheduler MUST call `hf_mailbox_waiter_get()` repeatedly and notify all waiters 59by injecting an `HF_MAILBOX_WRITABLE_INTID` interrupt. The scheduler should call 60`FFA_RUN` again on the sending VM as usual. 61 62### `HF_FFA_RUN_WAIT_FOR_INTERRUPT` 63 64_This is a Hafnium-specific function not part of the FF-A standard._ 65 66The vCPU is blocked waiting for an interrupt. The scheduler MUST take it off the 67run queue and not call `FFA_RUN` on the vCPU until it has either: 68 69* injected an interrupt 70* received `FFA_INTERRUPT` for it from another vCPU 71* the timeout provided in `w2` is not `FFA_SLEEP_INDEFINITE` and the 72 specified duration has expired. 73 74### `FFA_ERROR` 75 76#### `FFA_ABORTED` 77 78The vCPU has aborted triggering the whole VM to abort. The scheduler MUST treat 79this the same as `FFA_INTERRUPT` for all the other vCPUs of the VM. For this 80vCPU the scheduler SHOULD either never call `FFA_RUN` on the vCPU again, or treat 81it the same as `HF_FFA_RUN_WAIT_FOR_INTERRUPT`. 82 83#### Any other error code 84 85This should not happen if the scheduler VM has called `FFA_RUN` correctly, but 86in case there is some other error it should be logged. The scheduler SHOULD 87either try again or suspend the vCPU indefinitely. 88 89## Interrupt handling 90 91The scheduler VM is responsible for handling all hardware interrupts. Many of 92these will be intended for the scheduler VM itself and it can handle them as 93usual. However, it must also: 94 95* Enable, handle and ignore interrupts for the non-secure hypervisor physical 96 timer (PPI 10, IRQ 26). 97* Forward interrupts intended for secondary VMs to an appropriate vCPU of the 98 VM by calling `hf_interrupt_inject` and then running the vCPU as usual with 99 `FFA_RUN`. (If the vCPU is already running at the time that 100 `hf_interrupt_inject` is called then it must be preempted and run again so 101 that Hafnium can inject the interrupt.) 102