On Wed, 22 May 2019 at 20:09, Mark Brown <broonie@xxxxxxxxxx> wrote: > > On Wed, May 22, 2019 at 05:27:23PM +0900, Masahisa Kojima wrote: > > On Wed, 22 May 2019 at 03:16, Mark Brown <broonie@xxxxxxxxxx> wrote: > > > On Tue, May 21, 2019 at 08:59:58PM +0900, Masahisa Kojima wrote: > > > > > + ret = devm_request_irq(&pdev->dev, rx_irq, sq_spi_rx_handler, > > > > + 0, "synquacer-spi-rx", sspi); > > > > + ret = devm_request_irq(&pdev->dev, tx_irq, sq_spi_tx_handler, > > > > + 0, "synquacer-spi-tx", sspi); > > > > The code looked awfully like we depend on having interrupts? > > > I"m not sure I correctly understand what this comment means, > > should driver assume the case interrupt is not available? > > Do I need to support both interrupt and polling handling? > > If the driver requires interrupts it should not just ignore errors when > it requests interrupts. I'm sorry, I misunderstood. Yes, interrupt is required. I will handle errors properly.
