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: [PATCH] nvme: reject the ns when the block size is smaller than a
 sector
To:     Johannes Thumshirn <Johannes.Thumshirn@wdc.com>,
        Li Feng <fengli@smartx.com>,
        "martin.petersen@oracle.com" <martin.petersen@oracle.com>,
        Keith Busch <kbusch@kernel.org>, Jens Axboe <axboe@fb.com>,
        Christoph Hellwig <hch@lst.de>,
        "open list:NVM EXPRESS DRIVER" <linux-nvme@lists.infradead.org>,
        open list <linux-kernel@vger.kernel.org>
Cc:     "lifeng1519@gmail.com" <lifeng1519@gmail.com>
References: <20210113160621.98615-1-fengli@smartx.com>
 <SN4PR0401MB35985C08E5FFCDF0F5817A9D9BA90@SN4PR0401MB3598.namprd04.prod.outlook.com>
From:   Sagi Grimberg <sagi@grimberg.me>
Message-ID: <87698545-de26-c0ca-01e9-4dc2ddcacc80@grimberg.me>
Date:   Wed, 13 Jan 2021 14:12:59 -0800
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: <SN4PR0401MB35985C08E5FFCDF0F5817A9D9BA90@SN4PR0401MB3598.namprd04.prod.outlook.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk


>> The nvme spec(1.4a, figure 248) says:
>> "A value smaller than 9 (i.e., 512 bytes) is not supported."
>>
>> Signed-off-by: Li Feng <fengli@smartx.com>
>> ---
>>   drivers/nvme/host/core.c | 6 ++++++
>>   1 file changed, 6 insertions(+)
>>
>> diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
>> index f320273fc672..1f02e6e49a05 100644
>> --- a/drivers/nvme/host/core.c
>> +++ b/drivers/nvme/host/core.c
>> @@ -2161,6 +2161,12 @@ static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_id_ns *id)
>>   
>>   	blk_mq_freeze_queue(ns->disk->queue);
>>   	ns->lba_shift = id->lbaf[lbaf].ds;
>> +	if (ns->lba_shift < 9) {
>> +		pr_warn("%s: bad lba_shift(%d)\n", ns->disk->disk_name, ns->lba_shift);
>> +		ret = -1;

Meaningful errno would be useful. Also better to use dev_warn

>> +		goto out_unfreeze;
>> +	}
>> +
>>   	nvme_set_queue_limits(ns->ctrl, ns->queue);
>>   
>>   	if (ns->head->ids.csi == NVME_CSI_ZNS) {
>>
> 
> 
> But this only catches a physical block size < 512 for NVMe, not any other block device.
> 
> Please fix it for the general case in blk_queue_physical_block_size().

We actually call that later and would probably be better to check here..