From: Pierre-Louis Bossart pierre-louis.bossart@linux.intel.com
The interrupt handling in SoundWire requires software to re-read the interrupt status after clearing an interrupt. In case the interrupt is still outstanding, the code in bus.c will loop a number of times, however that loop is limited to the interrupts detected in the first read. This strategy helps meet SoundWire requirements without remaining forever in an interrupt handler.
Add a couple of comments to document this design.
Signed-off-by: Pierre-Louis Bossart pierre-louis.bossart@linux.intel.com Reviewed-by: Guennadi Liakhovetski guennadi.liakhovetski@linux.intel.com Signed-off-by: Bard Liao yung-chuan.liao@linux.intel.com --- drivers/soundwire/bus.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/soundwire/bus.c b/drivers/soundwire/bus.c index ffe4600fd95b..45131b9f5080 100644 --- a/drivers/soundwire/bus.c +++ b/drivers/soundwire/bus.c @@ -1334,6 +1334,7 @@ static int sdw_handle_dp0_interrupt(struct sdw_slave *slave, u8 *slave_status) "SDW_DP0_INT read failed:%d\n", status2); return status2; } + /* filter to limit loop to interrupts identified in the first status read */ status &= status2;
count++; @@ -1404,6 +1405,7 @@ static int sdw_handle_port_interrupt(struct sdw_slave *slave, "SDW_DPN_INT read failed:%d\n", status2); return status2; } + /* filter to limit loop to interrupts identified in the first status read */ status &= status2;
count++; @@ -1589,7 +1591,10 @@ static int sdw_handle_slave_alerts(struct sdw_slave *slave) sdca_cascade = ret & SDW_DP0_SDCA_CASCADE; }
- /* Make sure no interrupts are pending */ + /* + * Make sure no interrupts are pending, but filter to limit loop + * to interrupts identified in the first status read + */ buf &= _buf; buf2[0] &= _buf2[0]; buf2[1] &= _buf2[1];