Re: [PATCH] driver core: check bus->match without holding device lock

[Ming Lei <tom.leiming@gmail.com>]

2009/3/25 Guennadi Liakhovetski <g.liakhovetski@gmx.de>:
> Hi,
>
> this patch breaks soc-camera because of this change in behaviour:
>
> On Tue, 20 Jan 2009, tom.leiming@gmail.com wrote:
>
>> From: Ming Lei <tom.leiming@gmail.com>
>>
>> This patch moves bus->match out from driver_probe_device and
>> does not hold device lock to check the match between a device
>> and a driver.
>>
>> The idea has been verified by the commit 6cd495860901,
>> which leads to a faster boot. But the commit 6cd495860901 has
>> the following drawbacks: 1),only does the quick check in
>> the path of __driver_attach->driver_probe_device, not in other
>> paths; 2),for a matched device and driver, check the same match
>> twice. It is a waste of cpu ,especially for some drivers with long
>> device id table (eg. usb-storage driver).
>>
>> This patch adds a helper of driver_match_device to check the match
>> in all paths, and testes the match only once.
>>
>> Signed-off-by: Ming Lei <tom.leiming@gmail.com>
>> ---
>>  drivers/base/bus.c     |    8 +++++++-
>>  drivers/base/dd.c      |   19 +++++++------------
>>  include/linux/device.h |    6 ++++++
>>  3 files changed, 20 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/base/bus.c b/drivers/base/bus.c
>> index 83f32b8..c570d16 100644
>> --- a/drivers/base/bus.c
>> +++ b/drivers/base/bus.c
>> @@ -199,6 +199,12 @@ static ssize_t driver_bind(struct device_driver *drv,
>>
>>       dev = bus_find_device_by_name(bus, NULL, buf);
>>       if (dev && dev->driver == NULL) {
>> +
>> +             if (!driver_match_device(drv, dev)) {
>> +                     err = 0;
>> +                     goto not_match;
>> +             }
>> +
>>               if (dev->parent)        /* Needed for USB */
>>                       down(&dev->parent->sem);
>>               down(&dev->sem);
>> @@ -206,7 +212,7 @@ static ssize_t driver_bind(struct device_driver *drv,
>>               up(&dev->sem);
>>               if (dev->parent)
>>                       up(&dev->parent->sem);
>> -
>> +not_match:
>>               if (err > 0) {
>>                       /* success */
>>                       err = count;
>> diff --git a/drivers/base/dd.c b/drivers/base/dd.c
>> index 315bed8..61f32db 100644
>> --- a/drivers/base/dd.c
>> +++ b/drivers/base/dd.c
>> @@ -172,14 +172,8 @@ int driver_probe_done(void)
>>   * @drv: driver to bind a device to
>>   * @dev: device to try to bind to the driver
>>   *
>> - * First, we call the bus's match function, if one present, which should
>> - * compare the device IDs the driver supports with the device IDs of the
>> - * device. Note we don't do this ourselves because we don't know the
>> - * format of the ID structures, nor what is to be considered a match and
>> - * what is not.
>> - *
>> - * This function returns 1 if a match is found, -ENODEV if the device is
>> - * not registered, and 0 otherwise.
>> + * This function returns -ENODEV if the device is not registered,
>> + * and 0 otherwise.
>>   *
>>   * This function must be called with @dev->sem held.  When called for a
>>   * USB interface, @dev->parent->sem must be held as well.
>> @@ -190,21 +184,22 @@ int driver_probe_device(struct device_driver *drv, struct device *dev)
>>
>>       if (!device_is_registered(dev))
>>               return -ENODEV;
>> -     if (drv->bus->match && !drv->bus->match(dev, drv))
>> -             goto done;
>
> Previously, if no .match() was specified, the normal probing has been
> called.
>
>>
>>       pr_debug("bus: '%s': %s: matched device %s with driver %s\n",
>>                drv->bus->name, __func__, dev_name(dev), drv->name);
>>
>>       ret = really_probe(dev, drv);
>>
>> -done:
>>       return ret;
>>  }
>>
>>  static int __device_attach(struct device_driver *drv, void *data)
>>  {
>>       struct device *dev = data;
>> +
>> +     if (!driver_match_device(drv, dev))
>> +             return 0;
>> +
>
> Now, without .match() no probing is done. Is this an intended change and
> soc-camera has to be fixed or is this a bug?

It is not a driver-core bug, and soc-camera should be fixed.

>
>>       return driver_probe_device(drv, dev);
>>  }
>>
>> @@ -257,7 +252,7 @@ static int __driver_attach(struct device *dev, void *data)
>>        * is an error.
>>        */
>>
>> -     if (drv->bus->match && !drv->bus->match(dev, drv))
>> +     if (!driver_match_device(drv, dev))
>>               return 0;
>>
>>       if (dev->parent)        /* Needed for USB */
>> diff --git a/include/linux/device.h b/include/linux/device.h
>> index 45e5b19..3c61315 100644
>> --- a/include/linux/device.h
>> +++ b/include/linux/device.h
>> @@ -466,6 +466,12 @@ static inline int device_is_registered(struct device *dev)
>>       return dev->kobj.state_in_sysfs;
>>  }
>>
>> +static inline int driver_match_device(struct device_driver *drv,
>> +                                   struct device *dev)
>> +{
>> +     return drv->bus->match && drv->bus->match(dev, drv);
>> +}
>> +
>>  void driver_init(void);
>>
>>  /*
>> --
>> 1.6.0
>>
>> --
>> 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/
>>
>
> Thanks
> Guennadi
> ---
> Guennadi Liakhovetski, Ph.D.
> Freelance Open-Source Software Developer
>



-- 
Lei Ming