From: Grant Likely <grant.likely@secretlab.ca>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
Andrew Morton <akpm@linux-foundation.org>,
Gregory Bean <gbean@codeaurora.org>,
Jean Delvare <khali@linux-fr.org>,
Lennert Buytenhek <buytenh@secretlab.ca>
Subject: Re: [patch] drivers: gpio: sx150x: Do not access I2C from mask/unmask functions
Date: Tue, 15 Feb 2011 19:48:26 -0700 [thread overview]
Message-ID: <20110216024826.GL28005@angua.secretlab.ca> (raw)
In-Reply-To: <alpine.LFD.2.00.1102141807000.26192@localhost6.localdomain6>
On Mon, Feb 14, 2011 at 06:08:51PM +0100, Thomas Gleixner wrote:
> On Mon, 14 Feb 2011, Thomas Gleixner wrote:
>
> Ooops. quilt refresh should be mandatory for quilt mail. One day I'll fix that.
>
> ------->
> Subject: drivers: gpio: sx150x: Do not access I2C from mask/unmask functions
> From: Thomas Gleixner <tglx@linutronix.de>
> Date: Mon, 14 Feb 2011 16:58:11 +0100
>
> irq_chip->irq_mask/unmask are called with interrupts disabled and
> irq_desc->lock held. So we cannot access i2c from this context. That's
> what irq_bus_sync_unlock() is for.
>
> Store the masked information in the chip data structure and update the
> i2c bus from the irq_bus_sync_unlock() callback.
>
> This does not need a while(pending) loop because the update to this is
> always serialized via the bus lock, so we never have more than one pin
> update pending.
>
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Gregory Bean <gbean@codeaurora.org>
> Cc: Jean Delvare <khali@linux-fr.org>
> Cc: Lennert Buytenhek <buytenh@secretlab.ca>
Applied, thanks.
g.
> ---
> drivers/gpio/sx150x.c | 54 ++++++++++++++++++++++++++++++++++----------------
> 1 file changed, 37 insertions(+), 17 deletions(-)
>
> Index: linux-2.6-tip/drivers/gpio/sx150x.c
> ===================================================================
> --- linux-2.6-tip.orig/drivers/gpio/sx150x.c
> +++ linux-2.6-tip/drivers/gpio/sx150x.c
> @@ -25,6 +25,8 @@
> #include <linux/workqueue.h>
> #include <linux/i2c/sx150x.h>
>
> +#define NO_UPDATE_PENDING -1
> +
> struct sx150x_device_data {
> u8 reg_pullup;
> u8 reg_pulldn;
> @@ -47,8 +49,11 @@ struct sx150x_chip {
> const struct sx150x_device_data *dev_cfg;
> int irq_summary;
> int irq_base;
> + int irq_update;
> u32 irq_sense;
> - unsigned long irq_set_type_pending;
> + u32 irq_masked;
> + u32 dev_sense;
> + u32 dev_masked;
> struct irq_chip irq_chip;
> struct mutex lock;
> };
> @@ -312,9 +317,8 @@ static void sx150x_irq_mask(struct irq_d
>
> chip = container_of(ic, struct sx150x_chip, irq_chip);
> n = d->irq - chip->irq_base;
> -
> - sx150x_write_cfg(chip, n, 1, chip->dev_cfg->reg_irq_mask, 1);
> - sx150x_write_cfg(chip, n, 2, chip->dev_cfg->reg_sense, 0);
> + chip->irq_masked |= (1 << n);
> + chip->irq_update = n;
> }
>
> static void sx150x_irq_unmask(struct irq_data *d)
> @@ -326,9 +330,8 @@ static void sx150x_irq_unmask(struct irq
> chip = container_of(ic, struct sx150x_chip, irq_chip);
> n = d->irq - chip->irq_base;
>
> - sx150x_write_cfg(chip, n, 1, chip->dev_cfg->reg_irq_mask, 0);
> - sx150x_write_cfg(chip, n, 2, chip->dev_cfg->reg_sense,
> - chip->irq_sense >> (n * 2));
> + chip->irq_masked &= ~(1 << n);
> + chip->irq_update = n;
> }
>
> static int sx150x_irq_set_type(struct irq_data *d, unsigned int flow_type)
> @@ -350,7 +353,7 @@ static int sx150x_irq_set_type(struct ir
>
> chip->irq_sense &= ~(3UL << (n * 2));
> chip->irq_sense |= val << (n * 2);
> - chip->irq_set_type_pending |= BIT(n);
> + chip->irq_update = n;
> return 0;
> }
>
> @@ -404,15 +407,29 @@ static void sx150x_irq_bus_sync_unlock(s
>
> chip = container_of(ic, struct sx150x_chip, irq_chip);
>
> - while (chip->irq_set_type_pending) {
> - n = __ffs(chip->irq_set_type_pending);
> - chip->irq_set_type_pending &= ~BIT(n);
> - if (!(irq_to_desc(n + chip->irq_base)->status & IRQ_MASKED))
> - sx150x_write_cfg(chip, n, 2,
> - chip->dev_cfg->reg_sense,
> - chip->irq_sense >> (n * 2));
> - }
> + if (chip->irq_update == NO_UPDATE_PENDING)
> + goto out;
>
> + n = chip->irq_update;
> + chip->irq_update = NO_UPDATE_PENDING;
> +
> + /* Avoid updates if nothing changed */
> + if (chip->dev_sense == chip->irq_sense &&
> + chip->dev_sense == chip->irq_masked)
> + goto out;
> +
> + chip->dev_sense = chip->irq_sense;
> + chip->dev_masked = chip->irq_masked;
> +
> + if (chip->irq_masked & (1 << n)) {
> + sx150x_write_cfg(chip, n, 1, chip->dev_cfg->reg_irq_mask, 1);
> + sx150x_write_cfg(chip, n, 2, chip->dev_cfg->reg_sense, 0);
> + } else {
> + sx150x_write_cfg(chip, n, 1, chip->dev_cfg->reg_irq_mask, 0);
> + sx150x_write_cfg(chip, n, 2, chip->dev_cfg->reg_sense,
> + chip->irq_sense >> (n * 2));
> + }
> +out:
> mutex_unlock(&chip->lock);
> }
>
> @@ -445,8 +462,11 @@ static void sx150x_init_chip(struct sx15
> chip->irq_chip.irq_bus_sync_unlock = sx150x_irq_bus_sync_unlock;
> chip->irq_summary = -1;
> chip->irq_base = -1;
> + chip->irq_masked = ~0;
> chip->irq_sense = 0;
> - chip->irq_set_type_pending = 0;
> + chip->dev_masked = ~0;
> + chip->dev_sense = 0;
> + chip->irq_update = NO_UPDATE_PENDING;
> }
>
> static int sx150x_init_io(struct sx150x_chip *chip, u8 base, u16 cfg)
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
next prev parent reply other threads:[~2011-02-16 2:48 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-02-14 16:48 [patch] drivers: gpio: sx150x: Do not access I2C from mask/unmask functions Thomas Gleixner
2011-02-14 17:08 ` Thomas Gleixner
2011-02-16 2:48 ` Grant Likely [this message]
2011-02-14 19:45 ` Gregory Bean
2011-02-14 19:58 ` Thomas Gleixner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20110216024826.GL28005@angua.secretlab.ca \
--to=grant.likely@secretlab.ca \
--cc=akpm@linux-foundation.org \
--cc=buytenh@secretlab.ca \
--cc=gbean@codeaurora.org \
--cc=khali@linux-fr.org \
--cc=linux-kernel@vger.kernel.org \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.