linux-kernel - Re: [ata] 0568e61225: stress-ng.copy-file.ops_per

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]

Date:   Fri, 12 Aug 2022 11:27:20 -0700
From:   Damien Le Moal <damien.lemoal@...nsource.wdc.com>
To:     John Garry <john.garry@...wei.com>,
        Oliver Sang <oliver.sang@...el.com>
Cc:     Christoph Hellwig <hch@....de>,
        "Martin K. Petersen" <martin.petersen@...cle.com>,
        LKML <linux-kernel@...r.kernel.org>,
        Linux Memory Management List <linux-mm@...ck.org>,
        linux-ide@...r.kernel.org, lkp@...ts.01.org, lkp@...el.com,
        ying.huang@...el.com, feng.tang@...el.com,
        zhengjun.xing@...ux.intel.com, fengwei.yin@...el.com
Subject: Re: [ata] 0568e61225: stress-ng.copy-file.ops_per_sec -15.0%
 regression

On 2022/08/12 10:17, John Garry wrote:
> On 12/08/2022 16:41, Damien Le Moal wrote:
>>>> we noticed the commit is already in mainline now, and in our tests, there is
>>>> still similar regression and also on other platforms.
>>>> could you guide us how to check "which host driver is being used for this
>>>> test"? hope to supply some useful information.
>>>>
>>> For me, a complete kernel log may help.
>> I had a look yesterday with my test rig. I did not see any difference in the
>> default max_sectors_kb values for various drives between 5.18 and 5.19 (current
>> linus tree). The test machine has 2 AHCI adapters: Intel and Marvell. Both use
>> the regular AHCI driver. I have another rig with different ATA adapters but it
>> is powered down and I am traveling... So cannot test that right now.
>>
> 
> FWIW, on QEMU I get a difference for IDE disk for ata_piix host.
> 
> Interestingly ata dev max_sectors kb also gets capped from 32MB (LBA48) 
> -> 256KB due to swiotlb max mapping size. (It would be capped by shost 
> default max sectors 512KB without that swiotlb limit). I assume capping 
> due to swiotlb limit is not occuring on Oliver's machine.

Yes, I was suspecting that we may be seeing a difference for anything that is
not AHCI, e.g. with other drivers.

But that seems to be the correct thing to do, no ? How was this working before
without applying the swiotlb limit ?

> 
> thanks,
> John
> 
> [    1.497233] ata7: found unknown device (class 0)
> [    1.498341] ata7.00: ATA-7: QEMU HARDDISK, 2.5+, max UDMA/100
> [    1.499030] ata7.00: 209716 sectors, multi 16: LBA48
> [    1.623795] ata2: SATA link down (SStatus 0 SControl 300)
> [    1.624633] ata1: SATA link down (SStatus 0 SControl 300)
> [    1.633395] ata6: SATA link down (SStatus 0 SControl 300)
> [    1.634200] ata5: SATA link down (SStatus 0 SControl 300)
> [    1.635094] ata4: SATA link down (SStatus 0 SControl 300)
> [    1.635887] ata3: SATA link down (SStatus 0 SControl 300)
> [    1.636748] scsi 6:0:0:0: Direct-Access     ATA      QEMU HARDDISK 
>   2.5+ PQ: 0 ANSI: 5
> [    1.641298] sd 6:0:0:0: [sda] 209716 512-byte logical blocks: (107 
> MB/102 MiB)
> [    1.642188] sd 6:0:0:0: [sda] Write Protect is off
> [    1.642770] sd 6:0:0:0: [sda] Mode Sense: 00 3a 00 00
> [    1.642783] sd 6:0:0:0: [sda] Write cache: enabled, read cache: 
> enabled, doesn't support DPO or FUA
> [    1.644149] sd 6:0:0:0: [sda] Preferred minimum I/O size 512 bytes
> [    1.645142] sd 6:0:0:0: Attached scsi generic sg0 type 0
> [    1.655145] sd 6:0:0:0: [sda] Attached SCSI disk


-- 
Damien Le Moal
Western Digital Research