Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: md_raid: mdX_raid6 looping after sync_action "check" to "idle"
 transition
To:     Donald Buczek <buczek@molgen.mpg.de>, Song Liu <song@kernel.org>,
        linux-raid@vger.kernel.org,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        it+raid@molgen.mpg.de
References: <aa9567fd-38e1-7b9c-b3e1-dc2fdc055da5@molgen.mpg.de>
 <95fbd558-5e46-7a6a-43ac-bcc5ae8581db@cloud.ionos.com>
 <77244d60-1c2d-330e-71e6-4907d4dd65fc@molgen.mpg.de>
 <7c5438c7-2324-cc50-db4d-512587cb0ec9@molgen.mpg.de>
 <b289ae15-ff82-b36e-4be4-a1c8bbdbacd7@cloud.ionos.com>
 <37c158cb-f527-34f5-2482-cae138bc8b07@molgen.mpg.de>
 <efb8d47b-ab9b-bdb9-ee2f-fb1be66343b1@molgen.mpg.de>
 <55e30408-ac63-965f-769f-18be5fd5885c@molgen.mpg.de>
 <d95aa962-9750-c27c-639a-2362bdb32f41@cloud.ionos.com>
 <30576384-682c-c021-ff16-bebed8251365@molgen.mpg.de>
 <cdc0b03c-db53-35bc-2f75-93bbca0363b5@molgen.mpg.de>
 <bc342de0-98d2-1733-39cd-cc1999777ff3@molgen.mpg.de>
 <c3390ab0-d038-f1c3-5544-67ae9c8408b1@cloud.ionos.com>
 <a27c5a64-62bf-592c-e547-1e8e904e3c97@molgen.mpg.de>
 <6c7008df-942e-13b1-2e70-a058e96ab0e9@cloud.ionos.com>
 <12f09162-c92f-8fbb-8382-cba6188bfb29@molgen.mpg.de>
From:   Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
Message-ID: <6757d55d-ada8-9b7e-b7fd-2071fe905466@cloud.ionos.com>
Date:   Tue, 26 Jan 2021 15:06:01 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <12f09162-c92f-8fbb-8382-cba6188bfb29@molgen.mpg.de>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 1/26/21 13:58, Donald Buczek wrote:
> 
> 
>> Hmm, how about wake the waiter up in the while loop of raid5d?
>>
>> @@ -6520,6 +6532,11 @@ static void raid5d(struct md_thread *thread)
>>                          md_check_recovery(mddev);
>>                          spin_lock_irq(&conf->device_lock);
>>                  }
>> +
>> +               if ((atomic_read(&conf->active_stripes)
>> +                    < (conf->max_nr_stripes * 3 / 4) ||
>> +                    (test_bit(MD_RECOVERY_INTR, &mddev->recovery))))
>> +                       wake_up(&conf->wait_for_stripe);
>>          }
>>          pr_debug("%d stripes handled\n", handled);
> 
> Hmm... With this patch on top of your other one, we still have the basic 
> symptoms (md3_raid6 busy looping), but the sync thread is now hanging at
> 
>      root@sloth:~# cat /proc/$(pgrep md3_resync)/stack
>      [<0>] md_do_sync.cold+0x8ec/0x97c
>      [<0>] md_thread+0xab/0x160
>      [<0>] kthread+0x11b/0x140
>      [<0>] ret_from_fork+0x22/0x30
> 
> instead, which is 
> https://elixir.bootlin.com/linux/latest/source/drivers/md/md.c#L8963

Not sure why recovery_active is not zero, because it is set 0 before 
blk_start_plug, and raid5_sync_request returns 0 and skipped is also set 
to 1. Perhaps handle_stripe calls md_done_sync.

Could you double check the value of recovery_active? Or just don't wait 
if resync thread is interrupted.

wait_event(mddev->recovery_wait,
	   test_bit(MD_RECOVERY_INTR,&mddev->recovery) ||
	   !atomic_read(&mddev->recovery_active));

> And, unlike before, "md: md3: data-check interrupted." from the pr_info 
> two lines above appears in dmesg.

Yes, that is intentional since MD_RECOVERY_INTR is set by write idle.

Anyway, will try the script and investigate more about the issue.

Thanks,
Guoqing