Hi Thomas / Kevin,
I did verify Thomas Petazzoni's patch - [PATCH] omap: prcm: switch to a chained IRQ handler mechanism, and I have below questions or comments.
1. I see for each WKUP_ST or IO_ST interrupt the _prcm_int_handle_wakeup handler is getting called 2 times which impacts on performance.
printk("irq:%d,%d\n",irq,c); just before returning from the handler shows.
[ 221.966308] irq wkst:377,2
[ 221.968597] irq wkst:377,0
I see, the code checking the below warning is removed, won't it be good to retain this check ?
WARN(c == 0, "prcm: WARNING: PRCM indicated "
"MPU wakeup but no wakeup sources "
"are marked\n");
Also need to address the corner case issue, for which I submitted the patch fix.
[ 222.002563] irq wkst:368,3
[ 222.004913] irq iost:377,0
Regards
Gowda
________________________________________
From: ext Kevin Hilman [khilman@xxxxxxxxxxxxxxxxxxx]
Sent: Friday, November 19, 2010 6:36 PM
To: Gowda Madhusudhan.1 (EXT-Elektrobit/Helsinki)
Cc: linux-omap@xxxxxxxxxxxxxxx; paul@xxxxxxxxx
Subject: Re: [PATCH v2] OMAP3: PM: PRCM interrupt: Fix warning "MPU wakeup but no wakeup sources"
Madhusudhan Gowda <ext-madhusudhan.1.gowda@xxxxxxxxx> writes:
> A corner case where prcm_interrupt handler is handling the WKST_WKUP and
> before acknowledging the wakeup sources if an IO Pad wakeup ST_IO is
> indicated then hits the below warning since the wakeup sources are already
> cleared.
> WARN(c == 0, "prcm: WARNING: PRCM indicated "
> "MPU wakeup but no wakeup sources "
> "are marked\n");
>
> Since the above warning condition is only valid if the prcm_interrupt
> handler is called but no wakeup sources are marked in first iteration.
>
> The patch fixes this corner case.
>
> Updated after Paul Walmsley's "only handle selected PRCM interrupts" patch.
Can you have a look at the recent work by Thomas Petazzoni:
[PATCH] omap: prcm: switch to a chained IRQ handler mechanism
where the PRCM IRQ handler is broken up to see if this problem still
exists? I suspect the problem is gone as each type of interrupt is
separated out, but should be verified.
Kevin
>
> Signed-off-by: Madhusudhan Gowda <ext-madhusudhan.1.gowda@xxxxxxxxx>
> ---
> arch/arm/mach-omap2/pm34xx.c | 7 +++++--
> 1 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm/mach-omap2/pm34xx.c b/arch/arm/mach-omap2/pm34xx.c
> index 75c0cd1..2ed3662 100644
> --- a/arch/arm/mach-omap2/pm34xx.c
> +++ b/arch/arm/mach-omap2/pm34xx.c
> @@ -266,6 +266,7 @@ static irqreturn_t prcm_interrupt_handler (int irq, void *dev_id)
> {
> u32 irqenable_mpu, irqstatus_mpu;
> int c = 0;
> + int ct = 0;
>
> irqenable_mpu = prm_read_mod_reg(OCP_MOD,
> OMAP3_PRM_IRQENABLE_MPU_OFFSET);
> @@ -277,13 +278,15 @@ static irqreturn_t prcm_interrupt_handler (int irq, void *dev_id)
> if (irqstatus_mpu & (OMAP3430_WKUP_ST_MASK |
> OMAP3430_IO_ST_MASK)) {
> c = _prcm_int_handle_wakeup();
> + ct++;
>
> /*
> * Is the MPU PRCM interrupt handler racing with the
> * IVA2 PRCM interrupt handler ?
> */
> - WARN(c == 0, "prcm: WARNING: PRCM indicated MPU wakeup "
> - "but no wakeup sources are marked\n");
> + WARN(!c && (ct == 1), "prcm: WARNING: PRCM indicated "
> + "MPU wakeup but no wakeup sources "
> + "are marked\n");
> } else {
> /* XXX we need to expand our PRCM interrupt handler */
> WARN(1, "prcm: WARNING: PRCM interrupt received, but "
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
