From mboxrd@z Thu Jan 1 00:00:00 1970 From: zhouqiao Date: Fri, 05 Jun 2015 11:16:28 +0000 Subject: Re: [PATCH] mfd: 88pm80x: refine irq bit operation Message-Id: <5571850C.3030801@marvell.com> List-Id: References: <1431688217-6960-1-git-send-email-zhouqiao@marvell.com> In-Reply-To: <1431688217-6960-1-git-send-email-zhouqiao@marvell.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: kernel-janitors@vger.kernel.org On 06/05/2015 06:15 PM, Lee Jones wrote: > On Fri, 15 May 2015, Qiao Zhou wrote: > >> Set_bit/clear_bit for wu_flag may be corrupted if irq > 5(or 6 for >> aarch64). The maximum irq number from 88pm80x chip series is 24. >> Here we refine the code to protect the potential memory corruption. >> >> Also change wu_flag to wakeup_flag for easier understanding. >> >> Reported-by: Dan Carpenter >> Signed-off-by: Qiao Zhou >> --- >> drivers/mfd/88pm80x.c | 4 ++-- >> include/linux/mfd/88pm80x.h | 19 ++++++++++++++++--- >> 2 files changed, 18 insertions(+), 5 deletions(-) >> >> diff --git a/drivers/mfd/88pm80x.c b/drivers/mfd/88pm80x.c >> index 5e72f65..e1d803a 100644 >> --- a/drivers/mfd/88pm80x.c >> +++ b/drivers/mfd/88pm80x.c >> @@ -136,7 +136,7 @@ static int pm80x_suspend(struct device *dev) >> struct i2c_client *client = container_of(dev, struct i2c_client, dev); >> struct pm80x_chip *chip = i2c_get_clientdata(client); >> >> - if (chip && chip->wu_flag) >> + if (chip && chip->wakeup_flag) >> if (device_may_wakeup(chip->dev)) >> enable_irq_wake(chip->irq); >> >> @@ -148,7 +148,7 @@ static int pm80x_resume(struct device *dev) >> struct i2c_client *client = container_of(dev, struct i2c_client, dev); >> struct pm80x_chip *chip = i2c_get_clientdata(client); >> >> - if (chip && chip->wu_flag) >> + if (chip && chip->wakeup_flag) >> if (device_may_wakeup(chip->dev)) >> disable_irq_wake(chip->irq); >> >> diff --git a/include/linux/mfd/88pm80x.h b/include/linux/mfd/88pm80x.h >> index 97cb283..1bca7eb 100644 >> --- a/include/linux/mfd/88pm80x.h >> +++ b/include/linux/mfd/88pm80x.h >> @@ -276,6 +276,9 @@ enum { >> #define PM805_EARPHONE_SETTING (0x29) >> #define PM805_AUTO_SEQ_SETTING (0x2A) >> >> +/* supported 24 sub-irq */ >> +#define PM80x_MAX_IRQ 24 > s/x/X/ will fix it. > >> struct pm80x_rtc_pdata { >> int vrtc; >> int rtc_wakeup; >> @@ -301,7 +304,7 @@ struct pm80x_chip { >> int type; >> int irq; >> int irq_mode; >> - unsigned long wu_flag; >> + unsigned long wakeup_flag; >> spinlock_t lock; >> }; >> >> @@ -348,8 +351,13 @@ static inline int pm80x_dev_suspend(struct device *dev) >> struct pm80x_chip *chip = dev_get_drvdata(pdev->dev.parent); >> int irq = platform_get_irq(pdev, 0); >> >> + if (irq < 0 || irq >= PM80x_MAX_IRQ) { >> + dev_err(dev, "Invalid irq %d\n", irq); > s/irq/IRQ/ will fix it. > >> + return 0; >> + } >> + >> if (device_may_wakeup(dev)) >> - set_bit((1 << irq), &chip->wu_flag); >> + set_bit(irq, &chip->wakeup_flag); >> >> return 0; >> } >> @@ -360,8 +368,13 @@ static inline int pm80x_dev_resume(struct device *dev) >> struct pm80x_chip *chip = dev_get_drvdata(pdev->dev.parent); >> int irq = platform_get_irq(pdev, 0); >> >> + if (irq < 0 || irq >= PM80x_MAX_IRQ) { >> + dev_err(dev, "Invalid irq %d\n", irq); >> + return 0; >> + } > Why would this have changed over a suspend/resume? Agree... The irq sanity check should be in driver probe, which should already be checked. I'll remove it if no concern. Thanks. > >> if (device_may_wakeup(dev)) >> - clear_bit((1 << irq), &chip->wu_flag); >> + clear_bit(irq, &chip->wakeup_flag); >> >> return 0; >> } -- Best Regards Qiao