accel/ivpu: Drop IRQF_ONESHOT to allow IPC IRQ threading on PREEMPT_RT
Karol Wachowski
maintainerPATCH 1/1 accel/ivpu: Drop IRQF_ONESHOT to allow IPC IRQ threading on PREEMPT_RT
The IPC RX hardirq handler matches consumers under a spinlock and allocates rx_msg buffers. On PREEMPT_RT these spinlocks become sleeping locks and the allocation may sleep, neither of which is allowed in true hardirq context, resulting in "sleeping function called from invalid context" splats.
IRQF_ONESHOT makes genirq keep the primary handler in hardirq even when forced threading is enabled, so on PREEMPT_RT the handler cannot be threaded. Drop the flag so the primary handler is threaded on PREEMPT_RT and the IPC RX path runs in a context where sleeping is allowed. On the MSI interrupt chip (IRQCHIP_ONESHOT_SAFE) the flag was stripped anyway, so non-RT behaviour is unchanged.
--- drivers/accel/ivpu/ivpu_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/accel/ivpu/ivpu_drv.c +1 -1
@@ -623,7 +623,7 @@ static int ivpu_irq_init(struct ivpu_device *vdev) vdev->irq =pci_irq_vector (pdev,0 ); ret =devm_request_threaded_irq (vdev->drm.dev, vdev->irq, ivpu_hw_irq_handler, - ivpu_ipc_irq_thread_handler,IRQF_NO_AUTOEN |IRQF_ONESHOT , + ivpu_ipc_irq_thread_handler,IRQF_NO_AUTOEN ,DRIVER_NAME , vdev);if (ret)ivpu_err (vdev,"Failed to request an IRQ %d\n" , ret);
Signature · 1 line
2.43.0
Quoted 27 lines · Karol Wachowski
The IPC RX hardirq handler matches consumers under a spinlock and allocates
rx_msgbuffers. OnPREEMPT_RTthese spinlocks become sleeping locks and the allocation may sleep, neither of which is allowed in true hardirq context, resulting in "sleeping function called from invalid context" splats.
IRQF_ONESHOTmakes genirq keep the primary handler in hardirq even when forced threading is enabled, so onPREEMPT_RTthe handler cannot be threaded. Drop the flag so the primary handler is threaded onPREEMPT_RTand the IPC RX path runs in a context where sleeping is allowed. On the MSI interrupt chip (IRQCHIP_ONESHOT_SAFE) the flag was stripped anyway, so non-RT behaviour is unchanged.Fixes 85c9cc2d25f8 ("accel/ivpu: Use threaded IRQ for IPC callback processing")Cc Andrzej Kacprowski <[email protected]>Cc Karol Wachowski <[email protected]>Signed-off-by Karol Wachowski <[email protected]>--- drivers/accel/ivpu/ivpu_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/accel/ivpu/ivpu_drv.c +0
-0
drivers/accel/ivpu/ivpu_drv.c static int ivpu_irq_init(struct ivpu_device *vdev) · line 629
Quoted 8 lines
drivers/accel/ivpu/ivpu_drv.c +1
-1@@ -623,7 +623,7 @@ static int ivpu_irq_init(struct ivpu_device *vdev) vdev->irq =pci_irq_vector (pdev,0 ); ret =devm_request_threaded_irq (vdev->drm.dev, vdev->irq, ivpu_hw_irq_handler, - ivpu_ipc_irq_thread_handler,IRQF_NO_AUTOEN |IRQF_ONESHOT , + ivpu_ipc_irq_thread_handler,IRQF_NO_AUTOEN ,DRIVER_NAME , vdev);if (ret)ivpu_err (vdev,"Failed to request an IRQ %d\n" , ret);