2015-11-25 18:20:09

by Saurabh Sengar

Subject: [PATCH] drivers/rtc/rtc-tps65910.c: Add IRQF_ONESHOT to interrupt flags

If no primary handler is specified for threaded_irq then a
default one is assigned which always returns IRQ_WAKE_THREAD.
This handler requires the IRQF_ONESHOT, because the source of
interrupt is not disabled

Signed-off-by: Saurabh Sengar <[email protected]>
---
drivers/rtc/rtc-tps65910.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/rtc/rtc-tps65910.c b/drivers/rtc/rtc-tps65910.c
index f42aa2b..9afafde 100644
--- a/drivers/rtc/rtc-tps65910.c
+++ b/drivers/rtc/rtc-tps65910.c
@@ -268,7 +268,8 @@ static int tps65910_rtc_probe(struct platform_device *pdev)
}

ret = devm_request_threaded_irq(&pdev->dev, irq, NULL,
- tps65910_rtc_interrupt, IRQF_TRIGGER_LOW | IRQF_EARLY_RESUME,
+ tps65910_rtc_interrupt,
+ IRQF_TRIGGER_LOW | IRQF_EARLY_RESUME | IRQF_ONESHOT,
dev_name(&pdev->dev), &pdev->dev);
if (ret < 0) {
dev_err(&pdev->dev, "IRQ is not free.\n");
--
1.9.1

2015-11-25 18:42:06

by Alexandre Belloni

Subject: Re: [PATCH] drivers/rtc/rtc-tps65910.c: Add IRQF_ONESHOT to interrupt flags

Hi,

On 25/11/2015 at 23:49:57 +0530, Saurabh Sengar wrote :
> If no primary handler is specified for threaded_irq then a
> default one is assigned which always returns IRQ_WAKE_THREAD.
> This handler requires the IRQF_ONESHOT, because the source of
> interrupt is not disabled
>

Are you sure this is necessary? The parent irqchip is the mfd and it
sets IRQF_ONESHOT.

Does that solve a real problem or is it a false positive coming from a
static analysis tool?


--
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com