| Message ID | 20201215024542.18888-4-zev@bewilderbeest.net |
|---|---|
| State | New |
| Headers | show |
| Series | aspeed-video: extend spurious interrupt handling | expand |
On Tue, 15 Dec 2020 at 02:46, Zev Weiss <zev@bewilderbeest.net> wrote: > > This joins CAPTURE_COMPLETE and FRAME_COMPLETE in the set of interrupts > that have been seen asserted by the hardware even when disabled, leading > to the interrupt eventually getting disabled as described in commit > 65d270acb2d662c3346793663ac3a759eb4491b8. > > Signed-off-by: Zev Weiss <zev@bewilderbeest.net> I have less experience with this part of the chip, so I defer to Jae or Ryan for an ack. > --- > drivers/media/platform/aspeed-video.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/drivers/media/platform/aspeed-video.c b/drivers/media/platform/aspeed-video.c > index 218aae3be809..48c52bf91a1b 100644 > --- a/drivers/media/platform/aspeed-video.c > +++ b/drivers/media/platform/aspeed-video.c > @@ -564,7 +564,8 @@ static void aspeed_video_irq_res_change(struct aspeed_video *video, ulong delay) > * register. > */ > #define VE_SPURIOUS_IRQS \ > - (VE_INTERRUPT_CAPTURE_COMPLETE | VE_INTERRUPT_FRAME_COMPLETE) > + (VE_INTERRUPT_CAPTURE_COMPLETE | VE_INTERRUPT_FRAME_COMPLETE \ > + | VE_INTERRUPT_COMP_READY) > > static irqreturn_t aspeed_video_irq(int irq, void *arg) > { > -- > 2.29.2 >
On 12/21/2020 8:49 PM, Joel Stanley wrote: > On Tue, 15 Dec 2020 at 02:46, Zev Weiss <zev@bewilderbeest.net> wrote: >> >> This joins CAPTURE_COMPLETE and FRAME_COMPLETE in the set of interrupts >> that have been seen asserted by the hardware even when disabled, leading >> to the interrupt eventually getting disabled as described in commit >> 65d270acb2d662c3346793663ac3a759eb4491b8. >> >> Signed-off-by: Zev Weiss <zev@bewilderbeest.net> > > I have less experience with this part of the chip, so I defer to Jae > or Ryan for an ack. I didn't see unexpected VE_INTERRUPT_COMP_READY events in my system but it seems a fix good to have. Acked-by: Jae Hyun Yoo <jae.hyun.yoo@linux.intel.com> >> --- >> drivers/media/platform/aspeed-video.c | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/media/platform/aspeed-video.c b/drivers/media/platform/aspeed-video.c >> index 218aae3be809..48c52bf91a1b 100644 >> --- a/drivers/media/platform/aspeed-video.c >> +++ b/drivers/media/platform/aspeed-video.c >> @@ -564,7 +564,8 @@ static void aspeed_video_irq_res_change(struct aspeed_video *video, ulong delay) >> * register. >> */ >> #define VE_SPURIOUS_IRQS \ >> - (VE_INTERRUPT_CAPTURE_COMPLETE | VE_INTERRUPT_FRAME_COMPLETE) >> + (VE_INTERRUPT_CAPTURE_COMPLETE | VE_INTERRUPT_FRAME_COMPLETE \ >> + | VE_INTERRUPT_COMP_READY) >> >> static irqreturn_t aspeed_video_irq(int irq, void *arg) >> { >> -- >> 2.29.2 >>
diff --git a/drivers/media/platform/aspeed-video.c b/drivers/media/platform/aspeed-video.c index 218aae3be809..48c52bf91a1b 100644 --- a/drivers/media/platform/aspeed-video.c +++ b/drivers/media/platform/aspeed-video.c @@ -564,7 +564,8 @@ static void aspeed_video_irq_res_change(struct aspeed_video *video, ulong delay) * register. */ #define VE_SPURIOUS_IRQS \ - (VE_INTERRUPT_CAPTURE_COMPLETE | VE_INTERRUPT_FRAME_COMPLETE) + (VE_INTERRUPT_CAPTURE_COMPLETE | VE_INTERRUPT_FRAME_COMPLETE \ + | VE_INTERRUPT_COMP_READY) static irqreturn_t aspeed_video_irq(int irq, void *arg) {
This joins CAPTURE_COMPLETE and FRAME_COMPLETE in the set of interrupts that have been seen asserted by the hardware even when disabled, leading to the interrupt eventually getting disabled as described in commit 65d270acb2d662c3346793663ac3a759eb4491b8. Signed-off-by: Zev Weiss <zev@bewilderbeest.net> --- drivers/media/platform/aspeed-video.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)