Gcc report warning as follows:
drivers/irqchip/irq-imx-intmux.c:316:29: warning:
variable 'irqchip_data' set but not used [-Wunused-but-set-variable]
316 | struct intmux_irqchip_data irqchip_data;
| ^~~~~~~~~~~~
irqdata regs is stored to this variable on the stack in
imx_intmux_runtime_suspend(), which means a nop. this commit
fix to save regs to the right place.
Fixes: bb403111e017 ("irqchip/imx-intmux: Implement intmux runtime power management")
Reported-by: Hulk Robot <[email protected]>
Signed-off-by: Wei Yongjun <[email protected]>
---
drivers/irqchip/irq-imx-intmux.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/irqchip/irq-imx-intmux.c b/drivers/irqchip/irq-imx-intmux.c
index 4c9e40d193d6..e35b7b09c3ab 100644
--- a/drivers/irqchip/irq-imx-intmux.c
+++ b/drivers/irqchip/irq-imx-intmux.c
@@ -313,12 +313,12 @@ static int imx_intmux_remove(struct platform_device *pdev)
static int imx_intmux_runtime_suspend(struct device *dev)
{
struct intmux_data *data = dev_get_drvdata(dev);
- struct intmux_irqchip_data irqchip_data;
+ struct intmux_irqchip_data *irqchip_data;
int i;
for (i = 0; i < data->channum; i++) {
- irqchip_data = data->irqchip_data[i];
- irqchip_data.saved_reg = readl_relaxed(data->regs + CHANIER(i));
+ irqchip_data = &data->irqchip_data[i];
+ irqchip_data->saved_reg = readl_relaxed(data->regs + CHANIER(i));
}
clk_disable_unprepare(data->ipg_clk);
@@ -329,7 +329,7 @@ static int imx_intmux_runtime_suspend(struct device *dev)
static int imx_intmux_runtime_resume(struct device *dev)
{
struct intmux_data *data = dev_get_drvdata(dev);
- struct intmux_irqchip_data irqchip_data;
+ struct intmux_irqchip_data *irqchip_data;
int ret, i;
ret = clk_prepare_enable(data->ipg_clk);
@@ -339,8 +339,8 @@ static int imx_intmux_runtime_resume(struct device *dev)
}
for (i = 0; i < data->channum; i++) {
- irqchip_data = data->irqchip_data[i];
- writel_relaxed(irqchip_data.saved_reg, data->regs + CHANIER(i));
+ irqchip_data = &data->irqchip_data[i];
+ writel_relaxed(irqchip_data->saved_reg, data->regs + CHANIER(i));
}
return 0;
On 2020-07-29 16:58, Wei Yongjun wrote:
> Gcc report warning as follows:
>
> drivers/irqchip/irq-imx-intmux.c:316:29: warning:
> variable 'irqchip_data' set but not used [-Wunused-but-set-variable]
> 316 | struct intmux_irqchip_data irqchip_data;
> | ^~~~~~~~~~~~
>
> irqdata regs is stored to this variable on the stack in
> imx_intmux_runtime_suspend(), which means a nop. this commit
> fix to save regs to the right place.
>
> Fixes: bb403111e017 ("irqchip/imx-intmux: Implement intmux runtime
> power management")
> Reported-by: Hulk Robot <[email protected]>
> Signed-off-by: Wei Yongjun <[email protected]>
> ---
> drivers/irqchip/irq-imx-intmux.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/irqchip/irq-imx-intmux.c
> b/drivers/irqchip/irq-imx-intmux.c
> index 4c9e40d193d6..e35b7b09c3ab 100644
> --- a/drivers/irqchip/irq-imx-intmux.c
> +++ b/drivers/irqchip/irq-imx-intmux.c
> @@ -313,12 +313,12 @@ static int imx_intmux_remove(struct
> platform_device *pdev)
> static int imx_intmux_runtime_suspend(struct device *dev)
> {
> struct intmux_data *data = dev_get_drvdata(dev);
> - struct intmux_irqchip_data irqchip_data;
> + struct intmux_irqchip_data *irqchip_data;
> int i;
>
> for (i = 0; i < data->channum; i++) {
> - irqchip_data = data->irqchip_data[i];
> - irqchip_data.saved_reg = readl_relaxed(data->regs + CHANIER(i));
> + irqchip_data = &data->irqchip_data[i];
> + irqchip_data->saved_reg = readl_relaxed(data->regs + CHANIER(i));
> }
>
> clk_disable_unprepare(data->ipg_clk);
> @@ -329,7 +329,7 @@ static int imx_intmux_runtime_suspend(struct device
> *dev)
> static int imx_intmux_runtime_resume(struct device *dev)
> {
> struct intmux_data *data = dev_get_drvdata(dev);
> - struct intmux_irqchip_data irqchip_data;
> + struct intmux_irqchip_data *irqchip_data;
> int ret, i;
>
> ret = clk_prepare_enable(data->ipg_clk);
> @@ -339,8 +339,8 @@ static int imx_intmux_runtime_resume(struct device
> *dev)
> }
>
> for (i = 0; i < data->channum; i++) {
> - irqchip_data = data->irqchip_data[i];
> - writel_relaxed(irqchip_data.saved_reg, data->regs + CHANIER(i));
> + irqchip_data = &data->irqchip_data[i];
> + writel_relaxed(irqchip_data->saved_reg, data->regs + CHANIER(i));
> }
>
> return 0;
Amazing. Thanks for fixing this.
Johakim: I guess this was never tested, was it?
M.
--
Jazz is not dead. It just smells funny...
> -----Original Message-----
> From: Marc Zyngier <[email protected]>
> Sent: 2020??7??30?? 1:00
> To: Wei Yongjun <[email protected]>
> Cc: Hulk Robot <[email protected]>; Thomas Gleixner <[email protected]>;
> Jason Cooper <[email protected]>; Shawn Guo <[email protected]>;
> Sascha Hauer <[email protected]>; Joakim Zhang
> <[email protected]>; [email protected];
> [email protected]
> Subject: Re: [PATCH -next] irqchip/imx-intmux: Fix irqdata regs save in
> imx_intmux_runtime_suspend()
>
> On 2020-07-29 16:58, Wei Yongjun wrote:
> > Gcc report warning as follows:
> >
> > drivers/irqchip/irq-imx-intmux.c:316:29: warning:
> > variable 'irqchip_data' set but not used [-Wunused-but-set-variable]
> > 316 | struct intmux_irqchip_data irqchip_data;
> > | ^~~~~~~~~~~~
> >
> > irqdata regs is stored to this variable on the stack in
> > imx_intmux_runtime_suspend(), which means a nop. this commit fix to
> > save regs to the right place.
> >
> > Fixes: bb403111e017 ("irqchip/imx-intmux: Implement intmux runtime
> > power management")
> > Reported-by: Hulk Robot <[email protected]>
> > Signed-off-by: Wei Yongjun <[email protected]>
> > ---
> > drivers/irqchip/irq-imx-intmux.c | 12 ++++++------
> > 1 file changed, 6 insertions(+), 6 deletions(-)
[...]
> Amazing. Thanks for fixing this.
>
> Johakim: I guess this was never tested, was it?
Hi Marc,
Firstly, thanks Wei Yongjun for fixing this.
I do the intmux function test for each version.
At my side, I test intmux with CAN device. This intmux in M4 core subsystem, devices located in M4 core subsystem will route irqs into intmux.
CAN transceiver is controlled by an I/O expander which is lined to I2C from M4 core subsystem. And CAN device is located in A35 core subsystem.
After implementing runtime pm, I test CAN work fine when do suspend/resume test and I check the runtime status. I guess intmux register's state will not lost in M4 core subsystem when system suspended.
However, this is a silly mistake, I am so sorry for this.
Best Regards,
Joakim Zhang
> M.
> --
> Jazz is not dead. It just smells funny...
On Wed, 29 Jul 2020 23:58:49 +0800, Wei Yongjun wrote:
> Gcc report warning as follows:
>
> drivers/irqchip/irq-imx-intmux.c:316:29: warning:
> variable 'irqchip_data' set but not used [-Wunused-but-set-variable]
> 316 | struct intmux_irqchip_data irqchip_data;
> | ^~~~~~~~~~~~
>
> [...]
Applied to irq/irqchip-next, thanks!
[1/1] irqchip/imx-intmux: Fix irqdata regs save in imx_intmux_runtime_suspend()
commit: 5b6570bbb26f1dfad5663d99351820865e8c5c7e
Cheers,
M.
--
Without deviation from the norm, progress is not possible.