Hi Sujit,
Just a try to answer your question.
On Wed, Jun 17, 2009 at 1:32 PM, Sujit Karataparambil<sjt.kar@xxxxxxxxx> wrote:
> Neil,
>
> mddev is an static one time setup. Which should not have any problem
> with rcu, anyway.
> Any changes to the raid will require an rebuild using md.
>
mddev->raid_disks can change, when you add disks and also in the mean
while when you are updating the conf, the older conf might be being
used at various other places.
> Is this correct or wrong.
>
> Thanks,
> Sujit
>
> On Wed, Jun 17, 2009 at 12:16 PM, SandeepKsinha<sandeepksinha@xxxxxxxxx> wrote:
>> Here goes the updated patch.
>>
>>
>> commit 74da1595eb711b77969275070cda7516bac36f5e
>> Signed-off-by: Sandeep K Sinha <sandeepksinha@xxxxxxxxx>
>> Date: Sat Jun 6 20:49:37 2009 +0530
>> Due to the lack of memory ordering guarantees, we may have races around
>> mddev->conf. This patch addresses the same using rcu protection to avoid
>> such race conditions.
>>
>> diff --git a/drivers/md/linear.c b/drivers/md/linear.c
>> index 9ad6ec4..a56095c 100644
>> --- a/drivers/md/linear.c
>> +++ b/drivers/md/linear.c
>> @@ -28,9 +28,11 @@
>> static inline dev_info_t *which_dev(mddev_t *mddev, sector_t sector)
>> {
>> int lo, mid, hi;
>> - linear_conf_t *conf = mddev_to_conf(mddev);
>> + linear_conf_t *conf;
>>
>> + rcu_read_lock();
>> lo = 0;
>> + conf = rcu_dereference(mddev->private);
>> hi = mddev->raid_disks - 1;
>>
>> /*
>> @@ -45,7 +47,7 @@ static inline dev_info_t *which_dev(mddev_t *mddev,
>> sector_t sector)
>> else
>> lo = mid + 1;
>> }
>> -
>> + rcu_read_unlock();
>> return conf->disks + lo;
>> }
>>
>> @@ -86,36 +88,49 @@ static int linear_mergeable_bvec(struct request_queue *q,
>> static void linear_unplug(struct request_queue *q)
>> {
>> mddev_t *mddev = q->queuedata;
>> - linear_conf_t *conf = mddev_to_conf(mddev);
>> + linear_conf_t *conf;
>> int i;
>>
>> + rcu_read_lock();
>> + conf = rcu_dereference(mddev->private);
>> +
>> for (i=0; i < mddev->raid_disks; i++) {
>> struct request_queue *r_queue = bdev_get_queue(conf->disks[i].rdev->bdev);
>> blk_unplug(r_queue);
>> }
>> + rcu_read_unlock();
>> }
>>
>> static int linear_congested(void *data, int bits)
>> {
>> mddev_t *mddev = data;
>> - linear_conf_t *conf = mddev_to_conf(mddev);
>> + linear_conf_t *conf;
>> int i, ret = 0;
>>
>> + rcu_read_lock();
>> + conf = rcu_dereference(mddev->private);
>> +
>> for (i = 0; i < mddev->raid_disks && !ret ; i++) {
>> struct request_queue *q = bdev_get_queue(conf->disks[i].rdev->bdev);
>> ret |= bdi_congested(&q->backing_dev_info, bits);
>> }
>> +
>> + rcu_read_unlock();
>> return ret;
>> }
>>
>>
>> On Wed, Jun 17, 2009 at 12:05 PM, SandeepKsinha<sandeepksinha@xxxxxxxxx> wrote:
>>> Neil,
>>>
>>> On Mon, Jun 15, 2009 at 4:26 AM, Neil Brown<neilb@xxxxxxx> wrote:
>>>>
>>>> Hi,
>>>> Thanks for this patch, and sorry for the delay in reviewing it.
>>>>
>>>> I have a few issues:
>>>>
>>>> On Saturday June 6, sandeepksinha@xxxxxxxxx wrote:
>>>>> Signed-off-by: Sandeep K Sinha <sandeepksinha@xxxxxxxxx>
>>>>>
>>>>> Protecting mddev with barriers to avoid races.
>>>>
>>>> 1/ You need a lot more of an explanatory comment than this.
>>>> At least give some hint as to what the races are.
>>>> Give than the rcu primitives are used, it now makes sense to use
>>>> e.g. call_rcu to free the old 'conf'. That might reasonably be in a
>>>> separate patch, but the comment on this patch should at least at that
>>>> possibility.
>>>>>
>>>
>>> Sure. I shall do it for the final patch. I will also take care of this
>>> henceforth.
>>>
>>>>> diff --git a/drivers/md/linear.c b/drivers/md/linear.c
>>>>> index 9ad6ec4..a56095c 100644
>>>>> --- a/drivers/md/linear.c
>>>>> +++ b/drivers/md/linear.c
>>>>> @@ -28,9 +28,11 @@
>>>>> static inline dev_info_t *which_dev(mddev_t *mddev, sector_t sector)
>>>>> {
>>>>> int lo, mid, hi;
>>>>> - linear_conf_t *conf = mddev_to_conf(mddev);
>>>>> + linear_conf_t *conf;
>>>>>
>>>>> + rcu_read_lock();
>>>>> lo = 0;
>>>>> + conf = rcu_dereference(mddev->private);
>>>>> hi = mddev->raid_disks - 1;
>>>>>
>>>>
>>>> 2/ mddev->raid_disks should really be dereferenced before 'conf'.
>>>> Doing it the way you have done it, the 'raid_disks' value could be
>>>> larger than the value supported by the 'conf' so things could
>>>> go wrong.
>>>>
>>> Agreed. I hope you are referring to the case where a disk is in the
>>> process of being added to an array. Is that right ?
>>> Kindly confirm.
>>>>
>>>>> /*
>>>>> @@ -45,7 +47,7 @@ static inline dev_info_t *which_dev(mddev_t *mddev,
>>>>> sector_t sector)
>>>>> else
>>>>> lo = mid + 1;
>>>>> }
>>>>> -
>>>>> + rcu_read_unlock();
>>>>> return conf->disks + lo;
>>>>> }
>>>>
>>>> 3/ We are accessing conf->disks well after the rcu_lock has been released.
>>>> That is not exactly a problem with the code as it stands. But if
>>>> we do go ahead and free the old 'conf' with call_rcu, then this
>>>> becomes racy.
>>>> We should hold the rcu_read_lock for the entire time that we are
>>>> accessing the contents of 'conf'.
>>>>
>>> True.
>>>
>>>> That means we don't take rcu_read_lock in which_dev, but rather
>>>> take it in the two functions that call which_dev.
>>>>
>>>
>>> I have made that change.
>>>>>
>>>>> @@ -86,36 +88,49 @@ static int linear_mergeable_bvec(struct request_queue *q,
>>>>> static void linear_unplug(struct request_queue *q)
>>>>> {
>>>>> mddev_t *mddev = q->queuedata;
>>>>> - linear_conf_t *conf = mddev_to_conf(mddev);
>>>>> + linear_conf_t *conf;
>>>>> int i;
>>>>>
>>>>> + rcu_read_lock();
>>>>> + conf = rcu_dereference(mddev->private);
>>>>> +
>>>>> for (i=0; i < mddev->raid_disks; i++) {
>>>>> struct request_queue *r_queue = bdev_get_queue(conf->disks[i].rdev->bdev);
>>>>> blk_unplug(r_queue);
>>>>> }
>>>>> + rcu_read_unlock();
>>>>> }
>>>>>
>>>>> static int linear_congested(void *data, int bits)
>>>>> {
>>>>> mddev_t *mddev = data;
>>>>> - linear_conf_t *conf = mddev_to_conf(mddev);
>>>>> + linear_conf_t *conf;
>>>>> int i, ret = 0;
>>>>>
>>>>> + rcu_read_lock();
>>>>> + conf = rcu_dereference(mddev->private);
>>>>> +
>>>>> for (i = 0; i < mddev->raid_disks && !ret ; i++) {
>>>>> struct request_queue *q = bdev_get_queue(conf->disks[i].rdev->bdev);
>>>>> ret |= bdi_congested(&q->backing_dev_info, bits);
>>>>> }
>>>>> +
>>>>> + rcu_read_unlock();
>>>>> return ret;
>>>>> }
>>>>>
>>>>> static sector_t linear_size(mddev_t *mddev, sector_t sectors, int raid_disks)
>>>>> {
>>>>> - linear_conf_t *conf = mddev_to_conf(mddev);
>>>>> -
>>>>> + linear_conf_t *conf;
>>>>> + sector_t array_sectors;
>>>>> + rcu_read_lock();
>>>>> + conf = rcu_dereference(mddev->private);
>>>>> WARN_ONCE(sectors || raid_disks,
>>>>> "%s does not support generic reshape\n", __func__);
>>>>> -
>>>>> - return conf->array_sectors;
>>>>> + array_sectors = conf->array_sectors;
>>>>> + rcu_read_unlock();
>>>>> +
>>>>> + return array_sectors;
>>>>> }
>>>>>
>>>>> static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
>>>>> @@ -215,15 +230,14 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
>>>>> return -EINVAL;
>>>>>
>>>>> rdev->raid_disk = rdev->saved_raid_disk;
>>>>> -
>>>>> - newconf = linear_conf(mddev,mddev->raid_disks+1);
>>>>> + newconf = linear_conf(mddev,mddev->raid_disks + 1);
>>>>>
>>>>> if (!newconf)
>>>>> return -ENOMEM;
>>>>>
>>>>> newconf->prev = mddev_to_conf(mddev);
>>>>> - mddev->private = newconf;
>>>>> mddev->raid_disks++;
>>>>> + rcu_assign_pointer(mddev->private,newconf);
>>>>> md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
>>>>> set_capacity(mddev->gendisk, mddev->array_sectors);
>>>>> return 0;
>>>>> @@ -231,14 +245,17 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
>>>>>
>>>>> static int linear_stop (mddev_t *mddev)
>>>>> {
>>>>> - linear_conf_t *conf = mddev_to_conf(mddev);
>>>>> -
>>>>> + linear_conf_t *conf;
>>>>> +
>>>>> + rcu_read_lock();
>>>>> + conf = rcu_dereference(mddev->private);
>>>>> blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
>>>>> do {
>>>>> linear_conf_t *t = conf->prev;
>>>>> kfree(conf);
>>>>> conf = t;
>>>>> } while (conf);
>>>>> + rcu_read_unlock();
>>>>>
>>>>
>>>> 4/ We don't need the rcu protection here as we hold ->reconfig_mutex
>>>> both in linear_add and linear_stop, so they cannot race.
>>>> Adding a comment to this effect might be a good idea though.
>>>>
>>>
>>> Fine. Shall do this as well.
>>>
>>> The new patch will follow soon.
>>>
>>>> Thanks,
>>>>
>>>> NeilBrown
>>>>
>>>>
>>>>> return 0;
>>>>> }
>>>>> --
>>>>> Regards,
>>>>> Sandeep.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> “To learn is to change. Education is a process that changes the learner.”
>>>>
>>>
>>>
>>>
>>> --
>>> Regards,
>>> Sandeep.
>>>
>>>
>>>
>>>
>>>
>>>
>>> “To learn is to change. Education is a process that changes the learner.”
>>>
>>
>>
>>
>> --
>> Regards,
>> Sandeep.
>>
>>
>>
>>
>>
>>
>> “To learn is to change. Education is a process that changes the learner.”
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>
>
>
>
> --
> -- Sujit K M
>
--
Regards,
Sandeep.
“To learn is to change. Education is a process that changes the learner.”
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
