On 26.05.2021 13:47:42, Torin Cooper-Bennun wrote: > Hi Marc and list, > > This series is a follow-up to the RFC here: > https://lore.kernel.org/linux-can/20210514121946.2344901-1-torin@xxxxxxxxxxxxxxxxxx/ > > In this series we enable M_CAN-based devices to implement their own > device-specific interrupt handling, and add such handling for tcan4x5x. > > In 1, we replace the clear_interrupts() m_can_ops callback with > handle_interrupts(), which returns irqreturn_t and has an extra > parameter, clear_only, which is used if it isn't necessary to handle > device interrupts, only clear them. > > In 2, we use the new infrastructure in m_can_isr(). If M_CAN core > interrupts are handled, we still only clear device interrupts. > Otherwise, we try to handle any pending device interrupts. > > In 3-5, we clean up interrupt enabling and clearing in tcan4x5x, and > handle device interrupts appropriately. We specifically look for fatal > errors arising from transceiver or power, and SPI errors, which are not > necessarily fatal, but are dangerous! > > TCAN4550 shutdown is attempted by setting the device into standby mode. > There is probably a better way, but I understand we are limited by being > in the ISR context. Not exactly. The tcan's ISR runs in a threaded context, so you can basically do normal SPI or regmap transactions, shut down clocks and regulators, etc... Marc -- Pengutronix e.K. | Marc Kleine-Budde | Embedded Linux | https://www.pengutronix.de | Vertretung West/Dortmund | Phone: +49-231-2826-924 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
Attachment:
signature.asc
Description: PGP signature
