On 08/29/11 06:39, malas wrote: > > Dear Jonathan Cameron, > > I last reported that after creating device node and opening it, read() call was failing. > I got to fix this. > > The reason was this: > 4 functions are exported from ad7793 module to iio-ring module as ring setup ops: > namely: > .preenable = &ad7793_ring_preenable, > .postenable = &iio_triggered_ring_postenable, > .predisable = &iio_triggered_ring_predisable, > .postdisable = &ad7793_ring_postdisable, > > In the iio-ring modules "iio_store_ring_enable" function is invoked from user app,which > inturn invokes the func "ad7793_ring_preenable" and subsequently driver's postenable > function. > > The "iio_triggered_ring_postenable" function then attaches the poll function to the ring > trigger soft interrupt. > > Since continuous mode was set and enabled the configured GPIO interrupt for capturing > data ready signal (in ad7793_ring_preenable) as interrupt, iio_trigger_poll() invoked on > interrupt (from drivers data rdy interrupt handler) failed to raise soft interrupt and > as a consequence driver's ad7793_trigger_handler() was not invoked when data ready. > > I patched as follows in ad7793.c: > > 1. Moved irq enabling from ad7793_ring_preenable to ad7793_ring_postenable > 2. Defined ad7793_ring_setup_ops ".postenable" as ad7793_ring_postenable > 3. In ad7793_ring_postenable, invoked &iio_triggered_ring_postenable and then enabled > the data ready signalling GPIO interrupt. Hmm.. That will fix the issue. Normally I'd expect to see the enable actually happening in the trigger_set_state callback rather than there, but this device shares the interrupt line with on of the spi signal lines so is messing with irq's a lot more than normal so I guess that doesn't work? > > Then, from user app after opening the device, Poll() for rcv data presence was > successful and read() call returned non-zero size . Could print the data from ring buffer. > > Sincere thanks to you and Michael for guiding me through. > > Take care. > > regards > Mala > > > > > -- To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
