Re: [PATCH 2/5] pinctrl: keembay: Propagate errors from optional IRQ lookup
Andy Shevchenko <[email protected]>
| Newsgroups | org.kernel.vger.linux-gpio,org.infradead.lists.linux-arm-kernel,org.kernel.vger.linux-kernel |
|---|---|
| Organization | Intel Finland Oy - BIC 0357606-4 - c/o Alberga Business Park, 6 krs, Bertel Jungin Aukio 5, 02600 Espoo |
| Message-ID | <[email protected]> |
On Tue, Aug 11, 2026 at 09:48:01AM +0700, Bui Duc Phuc wrote: > > While this looks okay, have you considered actually having the > > platform_get_irq_optional() to be optional in this sense? > > As I understand it, if we decide to use platform_get_irq_optional(), > then -ENXIO should be handled as the expected case where the IRQ is > not available. If platform_get_irq_optional() was optional, it would return 0 instead of ENXIO (and avoid printing any message in such a case). > If we do not want to handle -ENXIO as an optional IRQ, > then I think we should use platform_get_irq() instead, > which matches its intended semantics: > https://elixir.bootlin.com/linux/v7.2-rc6/source/drivers/base/platform.c#L301 > ------------------------------- > ret = platform_get_irq_optional(dev, num); > if (ret < 0) > return dev_err_probe(&dev->dev, ret, > "IRQ index %u not found\n", num); > ------------------------------- > > In my opinion, we should not use an optional API if the IRQ is not > actually intended to be optional. Right. But does it suit for this driver? -- With Best Regards, Andy Shevchenko