linux-kernel - Re: [PATCH 4/5] scsi: libsas: check the ata device status by ata_dev

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <5BA1BA68.2050900@huawei.com>
Date:   Wed, 19 Sep 2018 10:54:32 +0800
From:   Jason Yan <yanaijie@...wei.com>
To:     John Garry <john.garry@...wei.com>, <martin.petersen@...cle.com>,
        <jejb@...ux.vnet.ibm.com>
CC:     <linux-scsi@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <zhaohongjiang@...wei.com>, <hare@...e.com>,
        <dan.j.williams@...el.com>, <jthumshirn@...e.de>, <hch@....de>,
        <huangdaode@...ilicon.com>, <chenxiang66@...ilicon.com>,
        <miaoxie@...wei.com>, Ewan Milne <emilne@...hat.com>,
        Tomas Henzl <thenzl@...hat.com>, <linux-ide@...r.kernel.org>
Subject: Re: [PATCH 4/5] scsi: libsas: check the ata device status by
 ata_dev_enabled()



On 2018/9/18 21:54, John Garry wrote:
> +
>
> On 12/09/2018 09:29, Jason Yan wrote:
>> When ata device IDENTIFY failed, the ata device status is
>> ATA_DEV_UNKNOWN. The libata reported like:
>>
>> [113518.620433] ata5.00: qc timeout (cmd 0xec)
>> [113518.653646] ata5.00: failed to IDENTIFY (I/O error, err_mask=0x4)
>>
>> But libsas verifies the device status by ata_dev_disabled(), which
>> skiped ATA_DEV_UNKNOWN. This will make libsas think the ata device
>
> /s/skiped/skipped/
>

OK, thanks.

>> probing succeed the device cannot be actually brought up. And even the
>> new bcast of this device will be considered as flutter and will not
>> probe this device again.
>>
>> Change ata_dev_disabled() to !ata_dev_enabled() so that libsas can
>> deal with this if the ata device probe failed. New bcasts can let us
>> try to probe the device again and bring it up if it is fine to
>> IDENTIFY.
>>
>> Tested-by: Zhou Yupeng <zhouyupeng1@...wei.com>
>> Signed-off-by: Jason Yan <yanaijie@...wei.com>
>
> Reviewed-by: John Garry <john.garry@...wei.com>
>
>> CC: John Garry <john.garry@...wei.com>
>> CC: Johannes Thumshirn <jthumshirn@...e.de>
>> CC: Ewan Milne <emilne@...hat.com>
>> CC: Christoph Hellwig <hch@....de>
>> CC: Tomas Henzl <thenzl@...hat.com>
>> CC: Dan Williams <dan.j.williams@...el.com>
>> CC: Hannes Reinecke <hare@...e.com>
>> ---
>>  drivers/scsi/libsas/sas_ata.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/scsi/libsas/sas_ata.c
>> b/drivers/scsi/libsas/sas_ata.c
>> index 64a958a99f6a..4f6cdf53e913 100644
>> --- a/drivers/scsi/libsas/sas_ata.c
>> +++ b/drivers/scsi/libsas/sas_ata.c
>> @@ -654,7 +654,7 @@ void sas_probe_sata(struct asd_sas_port *port)
>>          /* if libata could not bring the link up, don't surface
>>           * the device
>>           */
>> -        if (ata_dev_disabled(sas_to_ata_dev(dev)))
>> +        if (!ata_dev_enabled(sas_to_ata_dev(dev)))
>
> I do wonder if ata_dev_disabled() needs to be updated to cover
> ATA_DEV_UNKNOWN also or even instead of this change?
>

We cannot do this now because this will make the ata eh process wrong.

>>              sas_fail_probe(dev, __func__, -ENODEV);
>>      }
>>
>>
>
>
>
> .
>