Re: [ata] 0568e61225: stress-ng.copy-file.ops_per_sec -15.0% regression

[Damien Le Moal]

On 2022/08/16 9:38, John Garry wrote:
> On 16/08/2022 16:42, Damien Le Moal wrote:
>> On 2022/08/16 3:35, John Garry wrote:
>>> On 16/08/2022 07:57, Oliver Sang wrote:
>>>>>> For me, a complete kernel log may help.
>>>>> and since only 1HDD, the output of the following would be helpful:
>>>>>
>>>>> /sys/block/sda/queue/max_sectors_kb
>>>>> /sys/block/sda/queue/max_hw_sectors_kb
>>>>>
>>>>> And for 5.19, if possible.
>>>> for commit
>>>> 0568e61225 ("ata: libata-scsi: cap ata_device->max_sectors according to shost->max_sectors")
>>>>
>>>> root@lkp-icl-2sp1 ~# cat /sys/block/sda/queue/max_sectors_kb
>>>> 512
>>>> root@lkp-icl-2sp1 ~# cat /sys/block/sda/queue/max_hw_sectors_kb
>>>> 512
>>>>
>>>> for both commit
>>>> 4cbfca5f77 ("scsi: scsi_transport_sas: cap shost opt_sectors according to DMA optimal limit")
>>>> and v5.19
>>>>
>>>> root@lkp-icl-2sp1 ~# cat /sys/block/sda/queue/max_sectors_kb
>>>> 1280
>>>> root@lkp-icl-2sp1 ~# cat /sys/block/sda/queue/max_hw_sectors_kb
>>>> 32767
>>>>
>>>
>>> thanks, I appreciate this.
>>>
>>>   From the dmesg, I see 2x SATA disks - I was under the impression that
>>> the system only has 1x.
>>>
>>> Anyway, both drives show LBA48, which means the large max hw sectors at
>>> 32767KB:
>>> [   31.129629][ T1146] ata6.00: 1562824368 sectors, multi 1: LBA48 NCQ
>>> (depth 32)
>>>
>>> So this is what I suspected: we are capped from the default shost max
>>> sectors (1024 sectors).
>>>
>>> This seems like the simplest fix for you:
>>>
>>> --- a/include/linux/libata.h
>>> +++ b/include/linux/libata.h
>>> @@ -1382,7 +1382,8 @@ extern const struct attribute_group
>>> *ata_common_sdev_groups[];
>>>          .proc_name              = drv_name,                     \
>>>          .slave_destroy          = ata_scsi_slave_destroy,       \
>>>          .bios_param             = ata_std_bios_param,           \
>>> -       .unlock_native_capacity = ata_scsi_unlock_native_capacity
>>> +       .unlock_native_capacity = ata_scsi_unlock_native_capacity,\
>>> +       .max_sectors = ATA_MAX_SECTORS_LBA48
>>
>> This is crazy large (65535 x 512 B sectors) and never result in that being
>> exposed as the actual max_sectors_kb since other limits will apply first
>> (mapping size).
> 
> Here is how I read values from above for max_sectors_kb and 
> max_hw_sectors_kb:
> 
> v5.19 + 0568e61225 : 512/512
> v5.19 + 0568e61225 + 4cbfca5f77 : 512/512
> v5.19: 1280/32767
> 
> They are want makes sense to me, at least.
> 
> Oliver, can you confirm this? Thanks!
> 
> On this basis, it appears that max_hw_sectors_kb is getting capped from 
> scsi default @ 1024 sectors by commit 0568e61225. If it were getting 
> capped by swiotlb mapping limit then that would give us 512 sectors - 
> this value is fixed.
> 
> So for my SHT change proposal I am just trying to revert to previous 
> behaviour in 5.19 - make max_hw_sectors_kb crazy big again.

I reread the entire thing and I think I got things reverted here. The perf
regression happens with the 512/512 settings, while the original 1280/32767
before your patches was OK. So is your patch defining the optimal mapping size
cause the reduction to 512/512. It would mean that for ATA, we need a sane
default mapping instead of SCSI default 1024 sectors. Now I understand your
proposed change using ATA_MAX_SECTORS_LBA48.

However, that would be correct only for LBA48 capable drives.
ata_dev_configure() already sets dev->max_sectors correctly according to the
drive type, capabilities and eventual quirks. So the problem comes from the
libata-scsi change:

dev->max_sectors = min(dev->max_sectors, sdev->host->max_sectors);

when sdev->host->max_sectors is 0 (not initialized). So maybe simply changing
this line to:

dev->max_sectors = min_not_zero(dev->max_sectors, sdev->host->max_sectors);

would do the trick ? Any particular adapter driver that needs a mapping cap on
the adpter max mapping size can still set sdev->host->max_sectors as needed, and
we keep the same defaults as before when it is not set. Thoughts ? Or am I
missing something else ?


> 
>>
>> The regression may come not from commands becoming tiny, but from the fact that
>> after the patch, max_sectors_kb is too large, 
> 
> I don't think it is, but need confirmation.
> 
>> causing a lot of overhead with
>> qemu swiotlb mapping and slowing down IO processing.
> 
>>
>> Above, it can be seen that we ed up with max_sectors_kb being 1280, which is the
>> default for most scsi disks (including ATA drives). That is normal. But before
>> that, it was 512, which likely better fits qemu swiotlb and does not generate
> 
> Again, I don't think this this is the case. Need confirmation.
> 
>> overhead. So the above fix will not change anything I think...
> 
> 
> Thanks,
> John


-- 
Damien Le Moal
Western Digital Research