Received-SPF: pass (google.com: domain of linux-kernel+bounces-82639-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Message-ID: <b2ab5439-361b-4160-4e73-171026ecc486@huawei.com>
Date: Tue, 27 Feb 2024 11:06:48 +0800
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Thunderbird/91.3.1
Subject: Re: [PATCH] scsi: libsas: Fix disk not being scanned in after being
 removed
Content-Language: en-CA
To: Jason Yan <yanaijie@huawei.com>, John Garry <john.g.garry@oracle.com>,
	<jejb@linux.ibm.com>, <martin.petersen@oracle.com>,
	<damien.lemoal@opensource.wdc.com>
CC: <linux-scsi@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
	<linuxarm@huawei.com>, <prime.zeng@hisilicon.com>,
	<chenxiang66@hisilicon.com>, <kangfenglong@huawei.com>
References: <20240221073159.29408-1-yangxingui@huawei.com>
 <d765d2c5-3451-4519-a5e1-9e8f28dcd6b3@oracle.com>
 <cdf241a7-8504-5b87-2d18-d2a971f6ebb9@huawei.com>
 <8dc34f04-943b-26fd-01bc-34fb98803503@huawei.com>
From: yangxingui <yangxingui@huawei.com>
In-Reply-To: <8dc34f04-943b-26fd-01bc-34fb98803503@huawei.com>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 8bit

Hi Jason,

On 2024/2/23 15:04, Jason Yan wrote:
> On 2024/2/23 12:04, yangxingui wrote:
>> Hi, John
>>
>> On 2024/2/22 20:41, John Garry wrote:
>>> On 21/02/2024 07:31, Xingui Yang wrote:
>>>> As of commit d8649fc1c5e4 ("scsi: libsas: Do discovery on empty PHY to
>>>> update PHY info"), do discovery will send a new SMP_DISCOVER and update
>>>> phy->phy_change_count. We found that if the disk is reconnected and phy
>>>> change_count changes at this time, the disk scanning process will 
>>>> not be
>>>> triggered.
>>>>
>>>> So update the PHY info with the last query results.
>>>>
>>>> Fixes: d8649fc1c5e4 ("scsi: libsas: Do discovery on empty PHY to 
>>>> update PHY info")
>>>> Signed-off-by: Xingui Yang <yangxingui@huawei.com>
>>>> ---
>>>>   drivers/scsi/libsas/sas_expander.c | 9 ++++-----
>>>>   1 file changed, 4 insertions(+), 5 deletions(-)
>>>>
>>>> diff --git a/drivers/scsi/libsas/sas_expander.c 
>>>> b/drivers/scsi/libsas/sas_expander.c
>>>> index a2204674b680..9563f5589948 100644
>>>> --- a/drivers/scsi/libsas/sas_expander.c
>>>> +++ b/drivers/scsi/libsas/sas_expander.c
>>>> @@ -1681,6 +1681,10 @@ int sas_get_phy_attached_dev(struct 
>>>> domain_device *dev, int phy_id,
>>>>           if (*type == 0)
>>>>               memset(sas_addr, 0, SAS_ADDR_SIZE);
>>>>       }
>>>> +
>>>> +    if ((SAS_ADDR(sas_addr) == 0) || (res == -ECOMM))
>>>> +        sas_set_ex_phy(dev, phy_id, disc_resp);
>>>> +
>>>>       kfree(disc_resp);
>>>>       return res;
>>>>   }
>>>> @@ -1972,11 +1976,6 @@ static int sas_rediscover_dev(struct 
>>>> domain_device *dev, int phy_id,
>>>>       if ((SAS_ADDR(sas_addr) == 0) || (res == -ECOMM)) {
>>>>           phy->phy_state = PHY_EMPTY;
>>>>           sas_unregister_devs_sas_addr(dev, phy_id, last);
>>>> -        /*
>>>> -         * Even though the PHY is empty, for convenience we discover
>>>> -         * the PHY to update the PHY info, like negotiated linkrate.
>>>> -         */
>>>> -        sas_ex_phy_discover(dev, phy_id);
>>>
>>> It would be nice to be able to call sas_set_ex_phy() here (instead of 
>>> sas_get_phy_attached_dev()), but I assume that you can't do that as 
>>> the disc_resp memory is not available.
>>>
>>> If we were to manually set the PHY info here instead, how would that 
>>> look?
>> Yes, I think it is indeed better to use sas_set_ex_phy, as you said, 
>> disc_resp memory is not available. Maybe we can use 
>> sas_get_phy_discover instead of sas_get_phy_attached_dev so we can use 
>> disc_resp?
> 
> Can we directly set phy->negotiated_linkrate = SAS_PHY_DISABLED here? 
> For an empty PHY the other variables means nothing, so why bother get 
> and update them?
The value of the negotiated link rate has two possible values ​​in the 
current processing branch: SAS_LINK_RATE_UNKNOWN and SAS_PHY_DISABLED, 
and both come from disc_resp. If we do not use disc_resp, but set a 
fixed value SAS_PHY_DISABLED for it, it may not be appropriate.

Thanks,
Xingui