[PATCH 01/15] media: atomisp: improve IRQ handling debug messages
From: Mauro Carvalho Chehab
Date: Wed Jun 03 2020 - 03:01:27 EST
When an IRQ is not handled, it is nice to know what's the
reason.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@xxxxxxxxxx>
---
drivers/staging/media/atomisp/pci/atomisp_cmd.c | 11 ++++++-----
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/drivers/staging/media/atomisp/pci/atomisp_cmd.c b/drivers/staging/media/atomisp/pci/atomisp_cmd.c
index d06173b62b99..875dd3ce6e48 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_cmd.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_cmd.c
@@ -566,7 +566,7 @@ irqreturn_t atomisp_isr(int irq, void *dev)
if (irq_infos & IA_CSS_IRQ_INFO_CSS_RECEIVER_SOF) {
dev_dbg_ratelimited(isp->dev,
- "irq:0x%x (IA_CSS_IRQ_INFO_CSS_RECEIVER_SOF)\n",
+ "irq:0x%x (SOF)\n",
irq_infos);
irq_infos &= ~IA_CSS_IRQ_INFO_CSS_RECEIVER_SOF;
}
@@ -591,14 +591,14 @@ irqreturn_t atomisp_isr(int irq, void *dev)
/* EOF Event does not have the css_pipe returned */
asd = __get_asd_from_port(isp, eof_event.event.port);
if (!asd) {
- dev_err(isp->dev, "%s:no subdev.event:%d",
+ dev_err(isp->dev, "%s: ISYS event, but no subdev.event:%d",
__func__, eof_event.event.type);
continue;
}
atomisp_eof_event(asd, eof_event.event.exp_id);
dev_dbg_ratelimited(isp->dev,
- "%s EOF exp_id %d, asd %d\n",
+ "%s ISYS event: EOF exp_id %d, asd %d\n",
__func__, eof_event.event.exp_id,
asd->index);
}
@@ -610,7 +610,7 @@ irqreturn_t atomisp_isr(int irq, void *dev)
spin_unlock_irqrestore(&isp->lock, flags);
- dev_dbg_ratelimited(isp->dev, "irq:0x%x\n", irq_infos);
+ dev_dbg_ratelimited(isp->dev, "irq:0x%x (unhandled)\n", irq_infos);
return IRQ_WAKE_THREAD;
@@ -618,7 +618,8 @@ irqreturn_t atomisp_isr(int irq, void *dev)
spin_unlock_irqrestore(&isp->lock, flags);
if (irq_infos)
- dev_dbg_ratelimited(isp->dev, "irq:0x%x\n", irq_infos);
+ dev_dbg_ratelimited(isp->dev, "irq:0x%x (ignored, as not streaming anymore)\n",
+ irq_infos);
return IRQ_HANDLED;
}
--
2.26.2