[PATCH] sdio: optimized SDIO IRQ handling for single function

[PATCH] sdio: optimized SDIO IRQ handling for single function

[Per Forlin]

From: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>

If there is only 1 function registered, and IRQ:s are supported and
currently enabled, call the callback handler directly
without checking the CCCR registers.

Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
Signed-off-by: Per Forlin <per.forlin@linaro.org>
---
 drivers/mmc/core/sdio_irq.c |   14 ++++++++++++++
 1 files changed, 14 insertions(+), 0 deletions(-)

diff --git a/drivers/mmc/core/sdio_irq.c b/drivers/mmc/core/sdio_irq.c
index b300161..25291bf 100644
--- a/drivers/mmc/core/sdio_irq.c
+++ b/drivers/mmc/core/sdio_irq.c
@@ -32,6 +32,20 @@ static int process_sdio_pending_irqs(struct mmc_card *card)
 	int i, ret, count;
 	unsigned char pending;
 
+	/*
+	 * If there is only 1 function registered
+	 * call irq directly without checking the CCCR registers.
+	 */
+	if ((card->host->caps & MMC_CAP_SDIO_IRQ) &&
+	    card->host->sdio_irqs && (card->sdio_funcs == 1))
+		for (i = 0; i < SDIO_MAX_FUNCS; i++) {
+			struct sdio_func *func = card->sdio_func[i];
+			if (func && func->irq_handler) {
+				func->irq_handler(func);
+				return 1;
+			}
+		}
+
 	ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_INTx, 0, &pending);
 	if (ret) {
 		printk(KERN_DEBUG "%s: error %d reading SDIO_CCCR_INTx\n",
-- 
1.7.4.1

[PATCH] sdio: optimized SDIO IRQ handling for single function

[Nicolas Pitre]

On Tue, 3 May 2011, Per Forlin wrote:

> From: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
> 
> If there is only 1 function registered, and IRQ:s are supported and
> currently enabled, call the callback handler directly
> without checking the CCCR registers.
> 
> Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
> Signed-off-by: Per Forlin <per.forlin@linaro.org>

Acked-by: Nicolas Pitre <nicolas.pitre@linaro.org>

> ---
>  drivers/mmc/core/sdio_irq.c |   14 ++++++++++++++
>  1 files changed, 14 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/mmc/core/sdio_irq.c b/drivers/mmc/core/sdio_irq.c
> index b300161..25291bf 100644
> --- a/drivers/mmc/core/sdio_irq.c
> +++ b/drivers/mmc/core/sdio_irq.c
> @@ -32,6 +32,20 @@ static int process_sdio_pending_irqs(struct mmc_card *card)
>  	int i, ret, count;
>  	unsigned char pending;
>  
> +	/*
> +	 * If there is only 1 function registered
> +	 * call irq directly without checking the CCCR registers.
> +	 */
> +	if ((card->host->caps & MMC_CAP_SDIO_IRQ) &&
> +	    card->host->sdio_irqs && (card->sdio_funcs == 1))
> +		for (i = 0; i < SDIO_MAX_FUNCS; i++) {
> +			struct sdio_func *func = card->sdio_func[i];
> +			if (func && func->irq_handler) {
> +				func->irq_handler(func);
> +				return 1;
> +			}
> +		}
> +
>  	ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_INTx, 0, &pending);
>  	if (ret) {
>  		printk(KERN_DEBUG "%s: error %d reading SDIO_CCCR_INTx\n",
> -- 
> 1.7.4.1
> 
> 
> _______________________________________________
> linaro-dev mailing list
> linaro-dev at lists.linaro.org
> http://lists.linaro.org/mailman/listinfo/linaro-dev
> 

[PATCH] sdio: optimized SDIO IRQ handling for single function

[Per Forlin]

On 4 May 2011 05:40, Nicolas Pitre <nicolas.pitre@linaro.org> wrote:
> On Tue, 3 May 2011, Per Forlin wrote:
>
>> From: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
>>
>> If there is only 1 function registered, and IRQ:s are supported and
>> currently enabled, call the callback handler directly
>> without checking the CCCR registers.
>>
>> Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
>> Signed-off-by: Per Forlin <per.forlin@linaro.org>
>
> Acked-by: Nicolas Pitre <nicolas.pitre@linaro.org>
>
I am working o a patch version 2 after offline discussion with Ulf Hansson.
Instead of adding this code here.
Add sdio_single_func member in mmc_card. Set and reset this function
in sdio_claim_irq and sdio_release_irq.
process_sdio_pending_irqs would only check if sdio_single_func is !=
null and call it.

This will result in a bigger patch overall but the new code in
process_sdio_pending_irqs will be minimal.

>> ---
>> ?drivers/mmc/core/sdio_irq.c | ? 14 ++++++++++++++
>> ?1 files changed, 14 insertions(+), 0 deletions(-)
>>
>> diff --git a/drivers/mmc/core/sdio_irq.c b/drivers/mmc/core/sdio_irq.c
>> index b300161..25291bf 100644
>> --- a/drivers/mmc/core/sdio_irq.c
>> +++ b/drivers/mmc/core/sdio_irq.c
>> @@ -32,6 +32,20 @@ static int process_sdio_pending_irqs(struct mmc_card *card)
>> ? ? ? int i, ret, count;
>> ? ? ? unsigned char pending;
>>
>> + ? ? /*
>> + ? ? ?* If there is only 1 function registered
>> + ? ? ?* call irq directly without checking the CCCR registers.
>> + ? ? ?*/
>> + ? ? if ((card->host->caps & MMC_CAP_SDIO_IRQ) &&

>> + ? ? ? ? card->host->sdio_irqs && (card->sdio_funcs == 1))
>> + ? ? ? ? ? ? for (i = 0; i < SDIO_MAX_FUNCS; i++) {
Minor adjustments.
card->sdio_funcs may be > 1 but still only one irq is registered.
No need to iterate more than "sdio_funcs" number of elements.
+ card->host->sdio_irqs == 1)
+             for (i = 0; i < card->sdio_funcs; i++) {

>> + ? ? ? ? ? ? ? ? ? ? struct sdio_func *func = card->sdio_func[i];
>> + ? ? ? ? ? ? ? ? ? ? if (func && func->irq_handler) {
>> + ? ? ? ? ? ? ? ? ? ? ? ? ? ? func->irq_handler(func);
>> + ? ? ? ? ? ? ? ? ? ? ? ? ? ? return 1;
>> + ? ? ? ? ? ? ? ? ? ? }
>> + ? ? ? ? ? ? }
>> +
>> ? ? ? ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_INTx, 0, &pending);
>> ? ? ? if (ret) {
>> ? ? ? ? ? ? ? printk(KERN_DEBUG "%s: error %d reading SDIO_CCCR_INTx\n",
>> --
>> 1.7.4.1
>>
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev at lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>>
>

[PATCH] sdio: optimized SDIO IRQ handling for single function

[Nicolas Pitre]

On Wed, 4 May 2011, Per Forlin wrote:

> On 4 May 2011 05:40, Nicolas Pitre <nicolas.pitre@linaro.org> wrote:
> > On Tue, 3 May 2011, Per Forlin wrote:
> >
> >> From: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
> >>
> >> If there is only 1 function registered, and IRQ:s are supported and
> >> currently enabled, call the callback handler directly
> >> without checking the CCCR registers.
> >>
> >> Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
> >> Signed-off-by: Per Forlin <per.forlin@linaro.org>
> >
> > Acked-by: Nicolas Pitre <nicolas.pitre@linaro.org>
> >
> I am working o a patch version 2 after offline discussion with Ulf Hansson.
> Instead of adding this code here.
> Add sdio_single_func member in mmc_card. Set and reset this function
> in sdio_claim_irq and sdio_release_irq.
> process_sdio_pending_irqs would only check if sdio_single_func is !=
> null and call it.

Yes, that's what I was about to propose after thinking about it some 
more.

> This will result in a bigger patch overall but the new code in
> process_sdio_pending_irqs will be minimal.

Something like this (untested) ?

diff --git a/drivers/mmc/core/sdio_irq.c b/drivers/mmc/core/sdio_irq.c
index b300161..4552727 100644
--- a/drivers/mmc/core/sdio_irq.c
+++ b/drivers/mmc/core/sdio_irq.c
@@ -32,6 +32,12 @@ static int process_sdio_pending_irqs(struct mmc_card *card)
 	int i, ret, count;
 	unsigned char pending;
 
+	if (card->sdio_single_irq) {
+		struct sdio_func *func = card->sdio_single_irq;
+		func->irq_handler(func);
+		return 1;
+	}
+
 	ret = mmc_io_rw_direct(card, 0, 0, SDIO_CCCR_INTx, 0, &pending);
 	if (ret) {
 		printk(KERN_DEBUG "%s: error %d reading SDIO_CCCR_INTx\n",
@@ -166,6 +172,7 @@ static int sdio_card_irq_get(struct mmc_card *card)
 			host->sdio_irqs--;
 			return err;
 		}
+		return 1;
 	}
 
 	return 0;
@@ -183,7 +190,7 @@ static int sdio_card_irq_put(struct mmc_card *card)
 		kthread_stop(host->sdio_irq_thread);
 	}
 
-	return 0;
+	return (host->sdio_irqs == 1);
 }
 
 /**
@@ -225,8 +232,12 @@ int sdio_claim_irq(struct sdio_func *func, sdio_irq_handler_t *handler)
 
 	func->irq_handler = handler;
 	ret = sdio_card_irq_get(func->card);
-	if (ret)
+	if (ret < 0) {
 		func->irq_handler = NULL;
+	} else if (ret == 1) {
+		card->sdio_single_irq = func;
+		ret = 0;
+	}
 
 	return ret;
 }
@@ -250,7 +261,17 @@ int sdio_release_irq(struct sdio_func *func)
 
 	if (func->irq_handler) {
 		func->irq_handler = NULL;
-		sdio_card_irq_put(func->card);
+		if (sdio_card_irq_put(func->card) == 1) {
+			int i;
+			for (i = 0, i < 7; i++) {
+				if (func->card->sdio_func[i].irq_handler) {
+					func->card->sdio_single_irq =
+						func->card->sdio_func[i];
+					break;
+				}
+			}
+		} else
+			func->card->sdio_single_irq = NULL;
 	}
 
 	ret = mmc_io_rw_direct(func->card, 0, 0, SDIO_CCCR_IENx, 0, &reg);
diff --git a/include/linux/mmc/card.h b/include/linux/mmc/card.h
index adb4888..fee3df3 100644
--- a/include/linux/mmc/card.h
+++ b/include/linux/mmc/card.h
@@ -145,6 +145,7 @@ struct mmc_card {
 	struct sdio_cccr	cccr;		/* common card info */
 	struct sdio_cis		cis;		/* common tuple info */
 	struct sdio_func	*sdio_func[SDIO_MAX_FUNCS]; /* SDIO functions (devices) */
+	struct sdio_func	*sdio_single_irq;		/* SDIO function when only one IRQ active */
 	unsigned		num_info;	/* number of info strings */
 	const char		**info;		/* info strings */
 	struct sdio_func_tuple	*tuples;	/* unknown common tuples */


Nicolas

[PATCH] sdio: optimized SDIO IRQ handling for single function

[Per Forlin]

On 4 May 2011 17:20, Nicolas Pitre <nicolas.pitre@linaro.org> wrote:
> On Wed, 4 May 2011, Per Forlin wrote:
>
>> On 4 May 2011 05:40, Nicolas Pitre <nicolas.pitre@linaro.org> wrote:
>> > On Tue, 3 May 2011, Per Forlin wrote:
>> >
>> >> From: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
>> >>
>> >> If there is only 1 function registered, and IRQ:s are supported and
>> >> currently enabled, call the callback handler directly
>> >> without checking the CCCR registers.
>> >>
>> >> Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
>> >> Signed-off-by: Per Forlin <per.forlin@linaro.org>
>> >
>> > Acked-by: Nicolas Pitre <nicolas.pitre@linaro.org>
>> >
>> I am working o a patch version 2 after offline discussion with Ulf Hansson.
>> Instead of adding this code here.
>> Add sdio_single_func member in mmc_card. Set and reset this function
>> in sdio_claim_irq and sdio_release_irq.
>> process_sdio_pending_irqs would only check if sdio_single_func is !=
>> null and call it.
>
> Yes, that's what I was about to propose after thinking about it some
> more.
>
>> This will result in a bigger patch overall but the new code in
>> process_sdio_pending_irqs will be minimal.
>
> Something like this (untested) ?
>
What I had in mind is similar. Please let me know what you think. I am
about to post "patch v2"

Regards,
Per