lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Sun, 13 Jun 2010 19:44:56 +0200
From:	Geert Uytterhoeven <geert@...ux-m68k.org>
To:	Wan ZongShun <mcuos.com@...il.com>
Cc:	rtc-linux@...glegroups.com,
	Andrew Morton <akpm@...ux-foundation.org>,
	Alessandro Zummo <a.zummo@...ertech.it>,
	"Linux/m68k" <linux-m68k@...r.kernel.org>,
	Linux Kernel Development <linux-kernel@...r.kernel.org>
Subject: Re: [rtc-linux] [PATCH] rtc: rp5c01 - Add NVRAM support

On Sun, Jun 13, 2010 at 15:41, Wan ZongShun <mcuos.com@...il.com> wrote:
> Minior comment below:
>
> 2010/6/13 Geert Uytterhoeven <geert@...ux-m68k.org>:
>> The Ricoh RP5C01 RTC contains 26 x 4 bits of NVRAM.
>> Provide access to it via a sysfs "nvram" attribute file.
>>
>> Signed-off-by: Geert Uytterhoeven <geert@...ux-m68k.org>
>> --
>> Question: Is a spinlock in priv the right kind of locking?
>>          Should I use e.g. rtc-device.ops_lock instead?
>
> Firstly, this 'Question' text should not put here, please put all
> texts in front of '---'.

I explicitly put it there, as the question was meant for the reviewers only,
and I don't want it to be part of the final checkin comment.

> Secondly, You can not use rtc-device.ops_lock here, if so, it would
> arouse dead lock,
> before the ' rp5c01_read_time()' was called by RTC subsystem upper
> API, rtc-device.ops_lock
> has been required sucessfully, If continue to require the
> rtc-device.ops_lock in ' rp5c01_read_time()',
> the 'lock' will cannot be get, dead lock occurs.

Sorry, my question was not that correctly formulated...
I meant whether it's better to take rtc_device.ops_lock in the nvram
access functions
to synchronize with the RTC access functions (which already take that mutex),
instead of adding a spinlock to priv.

Thanks for your comments!

>> diff --git a/drivers/rtc/rtc-rp5c01.c b/drivers/rtc/rtc-rp5c01.c
>> index a95f733..36eb661 100644
>> --- a/drivers/rtc/rtc-rp5c01.c
>> +++ b/drivers/rtc/rtc-rp5c01.c
>> @@ -63,6 +63,8 @@ enum {
>>  struct rp5c01_priv {
>>        u32 __iomem *regs;
>>        struct rtc_device *rtc;
>> +       spinlock_t lock;        /* against concurrent RTC/NVRAM access */
>> +       struct bin_attribute nvram_attr;
>>  };
>>
>>  static inline unsigned int rp5c01_read(struct rp5c01_priv *priv,
>> @@ -92,6 +94,7 @@ static int rp5c01_read_time(struct device *dev, struct rtc_time *tm)
>>  {
>>        struct rp5c01_priv *priv = dev_get_drvdata(dev);
>>
>> +       spin_lock_irq(&priv->lock);
>>        rp5c01_lock(priv);
>>
>>        tm->tm_sec  = rp5c01_read(priv, RP5C01_10_SECOND) * 10 +
>> @@ -111,6 +114,7 @@ static int rp5c01_read_time(struct device *dev, struct rtc_time *tm)
>>                tm->tm_year += 100;
>>
>>        rp5c01_unlock(priv);
>> +       spin_unlock_irq(&priv->lock);
>>
>>        return rtc_valid_tm(tm);
>>  }
>> @@ -119,6 +123,7 @@ static int rp5c01_set_time(struct device *dev, struct rtc_time *tm)
>>  {
>>        struct rp5c01_priv *priv = dev_get_drvdata(dev);
>>
>> +       spin_lock_irq(&priv->lock);
>>        rp5c01_lock(priv);
>>
>>        rp5c01_write(priv, tm->tm_sec / 10, RP5C01_10_SECOND);
>> @@ -139,6 +144,7 @@ static int rp5c01_set_time(struct device *dev, struct rtc_time *tm)
>>        rp5c01_write(priv, tm->tm_year % 10, RP5C01_1_YEAR);
>>
>>        rp5c01_unlock(priv);
>> +       spin_unlock_irq(&priv->lock);
>>        return 0;
>>  }
>>
>> @@ -147,6 +153,72 @@ static const struct rtc_class_ops rp5c01_rtc_ops = {
>>        .set_time       = rp5c01_set_time,
>>  };
>>
>> +
>> +/*
>> + * The NVRAM is organized as 2 blocks of 13 nibbles of 4 bits.
>> + * We provide access to them like AmigaOS does: the high nibble of each 8-bit
>> + * byte is stored in BLOCK10, the low nibble in BLOCK11.
>> + */
>> +
>> +static ssize_t rp5c01_nvram_read(struct file *filp, struct kobject *kobj,
>> +                                struct bin_attribute *bin_attr,
>> +                                char *buf, loff_t pos, size_t size)
>> +{
>> +       struct device *dev = container_of(kobj, struct device, kobj);
>> +       struct rp5c01_priv *priv = dev_get_drvdata(dev);
>> +       ssize_t count;
>> +
>> +       spin_lock_irq(&priv->lock);
>> +
>> +       for (count = 0; size > 0 && pos < RP5C01_MODE; count++, size--) {
>> +               u8 data;
>> +
>> +               rp5c01_write(priv,
>> +                            RP5C01_MODE_TIMER_EN | RP5C01_MODE_RAM_BLOCK10,
>> +                            RP5C01_MODE);
>> +               data = rp5c01_read(priv, pos) << 4;
>> +               rp5c01_write(priv,
>> +                            RP5C01_MODE_TIMER_EN | RP5C01_MODE_RAM_BLOCK11,
>> +                            RP5C01_MODE);
>> +               data |= rp5c01_read(priv, pos++);
>> +               rp5c01_write(priv, RP5C01_MODE_TIMER_EN | RP5C01_MODE_MODE01,
>> +                            RP5C01_MODE);
>> +               *buf++ = data;
>> +       }
>> +
>> +       spin_unlock_irq(&priv->lock);
>> +       return count;
>> +}
>> +
>> +static ssize_t rp5c01_nvram_write(struct file *filp, struct kobject *kobj,
>> +                                 struct bin_attribute *bin_attr,
>> +                                 char *buf, loff_t pos, size_t size)
>> +{
>> +       struct device *dev = container_of(kobj, struct device, kobj);
>> +       struct rp5c01_priv *priv = dev_get_drvdata(dev);
>> +       ssize_t count;
>> +
>> +       spin_lock_irq(&priv->lock);
>> +
>> +       for (count = 0; size > 0 && pos < RP5C01_MODE; count++, size--) {
>> +               u8 data = *buf++;
>> +
>> +               rp5c01_write(priv,
>> +                            RP5C01_MODE_TIMER_EN | RP5C01_MODE_RAM_BLOCK10,
>> +                            RP5C01_MODE);
>> +               rp5c01_write(priv, data >> 4, pos);
>> +               rp5c01_write(priv,
>> +                            RP5C01_MODE_TIMER_EN | RP5C01_MODE_RAM_BLOCK11,
>> +                            RP5C01_MODE);
>> +               rp5c01_write(priv, data & 0xf, pos++);
>> +               rp5c01_write(priv, RP5C01_MODE_TIMER_EN | RP5C01_MODE_MODE01,
>> +                            RP5C01_MODE);
>> +       }
>> +
>> +       spin_unlock_irq(&priv->lock);
>> +       return count;
>> +}
>> +
>>  static int __init rp5c01_rtc_probe(struct platform_device *dev)
>>  {
>>        struct resource *res;
>> @@ -168,6 +240,15 @@ static int __init rp5c01_rtc_probe(struct platform_device *dev)
>>                goto out_free_priv;
>>        }
>>
>> +       sysfs_bin_attr_init(&priv->nvram_attr);
>> +       priv->nvram_attr.attr.name = "nvram";
>> +       priv->nvram_attr.attr.mode = S_IRUGO | S_IWUSR;
>> +       priv->nvram_attr.read = rp5c01_nvram_read;
>> +       priv->nvram_attr.write = rp5c01_nvram_write;
>> +       priv->nvram_attr.size = RP5C01_MODE;
>> +
>> +       spin_lock_init(&priv->lock);
>> +
>>        rtc = rtc_device_register("rtc-rp5c01", &dev->dev, &rp5c01_rtc_ops,
>>                                  THIS_MODULE);
>>        if (IS_ERR(rtc)) {
>> @@ -177,8 +258,15 @@ static int __init rp5c01_rtc_probe(struct platform_device *dev)
>>
>>        priv->rtc = rtc;
>>        platform_set_drvdata(dev, priv);
>> +
>> +       error = sysfs_create_bin_file(&dev->dev.kobj, &priv->nvram_attr);
>> +       if (error)
>> +               goto out_unregister;
>> +
>>        return 0;
>>
>> +out_unregister:
>> +       rtc_device_unregister(rtc);
>>  out_unmap:
>>        iounmap(priv->regs);
>>  out_free_priv:
>> @@ -190,6 +278,7 @@ static int __exit rp5c01_rtc_remove(struct platform_device *dev)
>>  {
>>        struct rp5c01_priv *priv = platform_get_drvdata(dev);
>>
>> +       sysfs_remove_bin_file(&dev->dev.kobj, &priv->nvram_attr);
>>        rtc_device_unregister(priv->rtc);
>>        iounmap(priv->regs);
>>        kfree(priv);
>> --
>> 1.7.0.4

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...ux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ