Re: [PATCH v2 2/2] scsi: sd: Rework asynchronous resume support

[Sergey Shtylyov <s.shtylyov@omp.ru>]

Hello!

On 8/16/22 11:21 PM, Bart Van Assche wrote:

>> On Mon, Aug 15, 2022 at 3:49 PM Bart Van Assche <bvanassche@acm.org> wrote:
>>> Would it be possible to share the output of the command below? That
>>> should reveal which ATA driver is active on the test setup.
>>>
>>> find /sys -name proc_name | xargs grep -aH .
>>
>> /sys/devices/platform/soc/ee300000.sata/ata1/host0/scsi_host/host0/proc_name:sata_rcar
> 
> Thanks Geert for the help. Although I already posted a revert, I'm still trying to
> root-cause this issue. Do you perhaps know whether sata_rcar controllers support NCQ

   They don't. :-)

> and if so, what queue depth these controllers support? I think that information is available in sysfs. Here is an example for a VM:
> 
> # (cd /sys/class/scsi_device && for a in */device/*/*/ncq_prio_enable; do p=${a%/ncq_prio_enable}; grep -qi ata $p/inquiry || continue; grep -aH . $p/{queue_depth,ncq*}; done)
> 2:0:0:0/device/driver/2:0:0:0/queue_depth:32
> 2:0:0:0/device/driver/2:0:0:0/ncq_prio_enable:0
> 2:0:0:0/device/driver/2:0:0:0/ncq_prio_supported:0
> 2:0:0:0/device/generic/device/queue_depth:32
> 2:0:0:0/device/generic/device/ncq_prio_enable:0
> 2:0:0:0/device/generic/device/ncq_prio_supported:0
> 6:0:0:1/device/driver/2:0:0:0/queue_depth:32
> 6:0:0:1/device/driver/2:0:0:0/ncq_prio_enable:0
> 6:0:0:1/device/driver/2:0:0:0/ncq_prio_supported:0
> 
> Thanks,
> 
> Bart.

MBR, Sergey