1. Patchew
  2. linux
  3. [PATCH 00/12] md: refacotor and some fixes related to sync_thread
  4. View patch

[PATCH 02/12] md: add a new enum type sync_action

Yu Kuai posted 12 patches 1 year, 8 months ago
There is a newer version of this series
[PATCH 02/12] md: add a new enum type sync_action
Posted by Yu Kuai 1 year, 8 months ago 
In order to make code related to sync_thread cleaner in following
patches, also add detail comment about each sync action. And also
prepare to remove the related recovery_flags in the fulture.

Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
 drivers/md/md.h | 57 ++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 56 insertions(+), 1 deletion(-)

diff --git a/drivers/md/md.h b/drivers/md/md.h
index 170412a65b63..6b9d9246f260 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -34,6 +34,61 @@
  */
 #define	MD_FAILFAST	(REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT)
 
+/* Status of sync thread. */
+enum sync_action {
+	/*
+	 * Represent by MD_RECOVERY_SYNC, start when:
+	 * 1) after assemble, sync data from first rdev to other copies, this
+	 * must be done first before other sync actions and will only execute
+	 * once;
+	 * 2) resize the array(notice that this is not reshape), sync data for
+	 * the new range;
+	 */
+	ACTION_RESYNC,
+	/*
+	 * Represent by MD_RECOVERY_RECOVER, start when:
+	 * 1) for new replacement, sync data based on the replace rdev or
+	 * available copies from other rdev;
+	 * 2) for new member disk while the array is degraded, sync data from
+	 * other rdev;
+	 * 3) reassemble after power failure or re-add a hot removed rdev, sync
+	 * data from first rdev to other copies based on bitmap;
+	 */
+	ACTION_RECOVER,
+	/*
+	 * Represent by MD_RECOVERY_SYNC | MD_RECOVERY_REQUESTED |
+	 * MD_RECOVERY_CHECK, start when user echo "check" to sysfs api
+	 * sync_action, used to check if data copies from differenct rdev are
+	 * the same. The number of mismatch sectors will be exported to user
+	 * by sysfs api mismatch_cnt;
+	 */
+	ACTION_CHECK,
+	/*
+	 * Represent by MD_RECOVERY_SYNC | MD_RECOVERY_REQUESTED, start when
+	 * user echo "repair" to sysfs api sync_action, usually paired with
+	 * ACTION_CHECK, used to force syncing data once user found that there
+	 * are inconsistent data,
+	 */
+	ACTION_REPAIR,
+	/*
+	 * Represent by MD_RECOVERY_RESHAPE, start when new member disk is added
+	 * to the conf, notice that this is different from spares or
+	 * replacement;
+	 */
+	ACTION_RESHAPE,
+	/*
+	 * Represent by MD_RECOVERY_FROZEN, can be set by sysfs api sync_action
+	 * or internal usage like setting the array read-only, will forbid above
+	 * actions.
+	 */
+	ACTION_FROZEN,
+	/*
+	 * All above actions don't match.
+	 */
+	ACTION_IDLE,
+	NR_SYNC_ACTIONS,
+};
+
 /*
  * The struct embedded in rdev is used to serialize IO.
  */
@@ -571,7 +626,7 @@ enum recovery_flags {
 	/* interrupted because io-error */
 	MD_RECOVERY_ERROR,
 
-	/* flags determines sync action */
+	/* flags determines sync action, see details in enum sync_action */
 
 	/* if just this flag is set, action is resync. */
 	MD_RECOVERY_SYNC,
-- 
2.39.2
Re: [PATCH 02/12] md: add a new enum type sync_action
Posted by Mariusz Tkaczyk 1 year, 8 months ago 
On Mon, 3 Jun 2024 20:58:05 +0800
Yu Kuai <yukuai3@huawei.com> wrote:

> In order to make code related to sync_thread cleaner in following
> patches, also add detail comment about each sync action. And also
> prepare to remove the related recovery_flags in the fulture.
> 
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
> ---
>  drivers/md/md.h | 57 ++++++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 56 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/md/md.h b/drivers/md/md.h
> index 170412a65b63..6b9d9246f260 100644
> --- a/drivers/md/md.h
> +++ b/drivers/md/md.h
> @@ -34,6 +34,61 @@
>   */
>  #define	MD_FAILFAST	(REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT)
>  
> +/* Status of sync thread. */
> +enum sync_action {
> +	/*
> +	 * Represent by MD_RECOVERY_SYNC, start when:
> +	 * 1) after assemble, sync data from first rdev to other copies, this
> +	 * must be done first before other sync actions and will only execute
> +	 * once;
> +	 * 2) resize the array(notice that this is not reshape), sync data
> for
> +	 * the new range;
> +	 */
> +	ACTION_RESYNC,
> +	/*
> +	 * Represent by MD_RECOVERY_RECOVER, start when:
> +	 * 1) for new replacement, sync data based on the replace rdev or
> +	 * available copies from other rdev;
> +	 * 2) for new member disk while the array is degraded, sync data from
> +	 * other rdev;
> +	 * 3) reassemble after power failure or re-add a hot removed rdev,
> sync
> +	 * data from first rdev to other copies based on bitmap;
> +	 */
> +	ACTION_RECOVER,
> +	/*
> +	 * Represent by MD_RECOVERY_SYNC | MD_RECOVERY_REQUESTED |
> +	 * MD_RECOVERY_CHECK, start when user echo "check" to sysfs api
> +	 * sync_action, used to check if data copies from differenct rdev are
> +	 * the same. The number of mismatch sectors will be exported to user
> +	 * by sysfs api mismatch_cnt;
> +	 */
> +	ACTION_CHECK,
> +	/*
> +	 * Represent by MD_RECOVERY_SYNC | MD_RECOVERY_REQUESTED, start when
> +	 * user echo "repair" to sysfs api sync_action, usually paired with
> +	 * ACTION_CHECK, used to force syncing data once user found that
> there
> +	 * are inconsistent data,
> +	 */
> +	ACTION_REPAIR,
> +	/*
> +	 * Represent by MD_RECOVERY_RESHAPE, start when new member disk is
> added
> +	 * to the conf, notice that this is different from spares or
> +	 * replacement;
> +	 */
> +	ACTION_RESHAPE,
> +	/*
> +	 * Represent by MD_RECOVERY_FROZEN, can be set by sysfs api
> sync_action
> +	 * or internal usage like setting the array read-only, will forbid
> above
> +	 * actions.
> +	 */
> +	ACTION_FROZEN,
> +	/*
> +	 * All above actions don't match.
> +	 */
> +	ACTION_IDLE,
> +	NR_SYNC_ACTIONS,
> +};

I like if counter is keep in same style as rest enum values, like ACTION_COUNT.

Anyway LGTM.

Mariusz
Re: [PATCH 02/12] md: add a new enum type sync_action
Posted by Yu Kuai 1 year, 8 months ago 
Hi,

在 2024/06/11 16:31, Mariusz Tkaczyk 写道:
> On Mon, 3 Jun 2024 20:58:05 +0800
> Yu Kuai <yukuai3@huawei.com> wrote:
> 
>> In order to make code related to sync_thread cleaner in following
>> patches, also add detail comment about each sync action. And also
>> prepare to remove the related recovery_flags in the fulture.
>>
>> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
>> ---
>>   drivers/md/md.h | 57 ++++++++++++++++++++++++++++++++++++++++++++++++-
>>   1 file changed, 56 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/md/md.h b/drivers/md/md.h
>> index 170412a65b63..6b9d9246f260 100644
>> --- a/drivers/md/md.h
>> +++ b/drivers/md/md.h
>> @@ -34,6 +34,61 @@
>>    */
>>   #define	MD_FAILFAST	(REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT)
>>   
>> +/* Status of sync thread. */
>> +enum sync_action {
>> +	/*
>> +	 * Represent by MD_RECOVERY_SYNC, start when:
>> +	 * 1) after assemble, sync data from first rdev to other copies, this
>> +	 * must be done first before other sync actions and will only execute
>> +	 * once;
>> +	 * 2) resize the array(notice that this is not reshape), sync data
>> for
>> +	 * the new range;
>> +	 */
>> +	ACTION_RESYNC,
>> +	/*
>> +	 * Represent by MD_RECOVERY_RECOVER, start when:
>> +	 * 1) for new replacement, sync data based on the replace rdev or
>> +	 * available copies from other rdev;
>> +	 * 2) for new member disk while the array is degraded, sync data from
>> +	 * other rdev;
>> +	 * 3) reassemble after power failure or re-add a hot removed rdev,
>> sync
>> +	 * data from first rdev to other copies based on bitmap;
>> +	 */
>> +	ACTION_RECOVER,
>> +	/*
>> +	 * Represent by MD_RECOVERY_SYNC | MD_RECOVERY_REQUESTED |
>> +	 * MD_RECOVERY_CHECK, start when user echo "check" to sysfs api
>> +	 * sync_action, used to check if data copies from differenct rdev are
>> +	 * the same. The number of mismatch sectors will be exported to user
>> +	 * by sysfs api mismatch_cnt;
>> +	 */
>> +	ACTION_CHECK,
>> +	/*
>> +	 * Represent by MD_RECOVERY_SYNC | MD_RECOVERY_REQUESTED, start when
>> +	 * user echo "repair" to sysfs api sync_action, usually paired with
>> +	 * ACTION_CHECK, used to force syncing data once user found that
>> there
>> +	 * are inconsistent data,
>> +	 */
>> +	ACTION_REPAIR,
>> +	/*
>> +	 * Represent by MD_RECOVERY_RESHAPE, start when new member disk is
>> added
>> +	 * to the conf, notice that this is different from spares or
>> +	 * replacement;
>> +	 */
>> +	ACTION_RESHAPE,
>> +	/*
>> +	 * Represent by MD_RECOVERY_FROZEN, can be set by sysfs api
>> sync_action
>> +	 * or internal usage like setting the array read-only, will forbid
>> above
>> +	 * actions.
>> +	 */
>> +	ACTION_FROZEN,
>> +	/*
>> +	 * All above actions don't match.
>> +	 */
>> +	ACTION_IDLE,
>> +	NR_SYNC_ACTIONS,
>> +};
> 
> I like if counter is keep in same style as rest enum values, like ACTION_COUNT.

Thanks for the review, however, I didn't find this style "xxx_COUNT" in
md code, AFAIK, "NR_xxx" style is used more, for example:

enum stat_group {
         STAT_READ,
         STAT_WRITE,
         STAT_DISCARD,
         STAT_FLUSH,

         NR_STAT_GROUPS
};

Just to let you know, I'll keep this style in the next version. :)

Thanks,
Kuai
> 
> Anyway LGTM.
> 
> Mariusz
> .
>

Patchew 2.1-devel-b67e4c2

© 2016 - 2026 Red Hat, Inc.