linux-kernel - Re: [PATCH -next v4 0/7] md: make rdev addition and removal independent from daemon thread

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] [day] [month] [year] [list]

Message-ID: <2832ee84-4365-d35d-1ae4-fa260a1beb9f@huaweicloud.com>
Date:   Fri, 25 Aug 2023 14:44:31 +0800
From:   Yu Kuai <yukuai1@...weicloud.com>
To:     Song Liu <song@...nel.org>, Yu Kuai <yukuai1@...weicloud.com>
Cc:     xni@...hat.com, linux-raid@...r.kernel.org,
        linux-kernel@...r.kernel.org, yi.zhang@...wei.com,
        yangerkun@...wei.com, "yukuai (C)" <yukuai3@...wei.com>
Subject: Re: [PATCH -next v4 0/7] md: make rdev addition and removal
 independent from daemon thread

Hi,

在 2023/08/25 14:39, Song Liu 写道:
> On Thu, Aug 24, 2023 at 8:20 PM Yu Kuai <yukuai1@...weicloud.com> wrote:
>>
>> From: Yu Kuai <yukuai3@...wei.com>
>>
>> Changes in v4:
>>   - add some review tag;
>>   - add comments to make code more readadble for patch 4,6;
>>   - rework patch 7 a litter;
> 
> Applied v4 to md-next. But this set won't go into 6.6.
> 
Thanks! I'll work on rebasing the next huge patchset.

Kuai

> Thanks,
> Song
> 
>>
>> Changes in v3:
>>   - rename md_choose_sync_direction() to md_choose_sync_action() in patch 2;
>>   - fix an error in patch 3;
>>   - add flush_work(&mddev->sync_work) while change read-only array to
>>   read-write;
>>
>> Changes in v2:
>>   - remove patch 1 from v1 and some related patches, those patches will
>>   be sent later when rcu protection for rdev is removed.
>>   - add patch 2.
>>
>> This is the third patchset to do some preparatory work to synchronize
>> io with array reconfiguration.
>>
>> 1) The first patchset refactor 'active_io', make sure that mddev_suspend()
>> will wait for io to be done. [1]
>>
>> 2) The second patchset remove 'quiesce' callback from mddev_suspend(), so
>> that mddev_suspend() doesn't rely on 'quiesce' callback is registered,
>> and can be used for all personalites; [2]
>>
>> 3) This patchset make array reconfiguration independent from daemon thread,
>> and synchronize it with io will be much easier because io may rely on
>> daemon thread to be done.
>>
>> More patchset on the way!
>>
>> Yu Kuai (7):
>>    md: use separate work_struct for md_start_sync()
>>    md: factor out a helper to choose sync action from md_check_recovery()
>>    md: delay choosing sync action to md_start_sync()
>>    md: factor out a helper rdev_removeable() from remove_and_add_spares()
>>    md: factor out a helper rdev_is_spare() from remove_and_add_spares()
>>    md: factor out a helper rdev_addable() from remove_and_add_spares()
>>    md: delay remove_and_add_spares() for read only array to
>>      md_start_sync()
>>
>>   drivers/md/md.c | 308 +++++++++++++++++++++++++++++++++---------------
>>   drivers/md/md.h |   5 +-
>>   2 files changed, 218 insertions(+), 95 deletions(-)
>>
>> --
>> 2.39.2
>>
> .
>