Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 30 Mar 2022 13:21:14 +0200
From:   Andrea Righi <andrea.righi@canonical.com>
To:     John Garry <john.garry@huawei.com>
Cc:     Ming Lei <ming.lei@redhat.com>,
        Martin Wilck <martin.wilck@suse.com>,
        Bart Van Assche <bvanassche@acm.org>,
        "James E.J. Bottomley" <jejb@linux.ibm.com>,
        "Martin K. Petersen" <martin.petersen@oracle.com>,
        linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: filesystem corruption with "scsi: core: Reallocate device's
 budget map on queue depth change"
Message-ID: <YkQ9KoKb+VK06zXi@arighi-desktop>
References: <YkQsumJ3lgGsagd2@arighi-desktop>
 <f7bacce8-b5e5-3ef1-e116-584c01533f69@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <f7bacce8-b5e5-3ef1-e116-584c01533f69@huawei.com>
Precedence: bulk

On Wed, Mar 30, 2022 at 11:38:02AM +0100, John Garry wrote:
> On 30/03/2022 11:11, Andrea Righi wrote:
> > Hello,
> > 
> > after this commit I'm experiencing some filesystem corruptions at boot
> > on a power9 box with an aacraid controller.
> > 
> > At the moment I'm running a 5.15.30 kernel; when the filesystem is
> > mounted at boot I see the following errors in the console:
> > 
> > Begin: Will now check root file system ... fsck from util-linux 2.36.1
> > [/usr/sbin/fsck.ext4 (1) -- /dev/sda2] fsck.ext4 -a -C0 /dev/sda2
> > root: clean, 99646/122101760 files, 11187342/488376336 blocks
> > done.
> > [    4.636613] sd 0:2:0:0: [sda] tag#257 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.636655] sd 0:2:0:0: [sda] tag#257 CDB: Read(10) 28 00 00 00 4c 10 00 00 08 00
> > [    4.636689] blk_update_request: I/O error, dev sda, sector 19472 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.636734] sd 0:2:0:0: [sda] tag#258 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.636772] sd 0:2:0:0: [sda] tag#258 CDB: Read(10) 28 00 00 00 4c 18 00 00 08 00
> > [    4.636796] blk_update_request: I/O error, dev sda, sector 19480 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.636840] sd 0:2:0:0: [sda] tag#260 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.636877] sd 0:2:0:0: [sda] tag#260 CDB: Read(10) 28 00 00 00 4c 28 00 00 08 00
> > [    4.636901] blk_update_request: I/O error, dev sda, sector 19496 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.636944] sd 0:2:0:0: [sda] tag#259 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.636971] sd 0:2:0:0: [sda] tag#259 CDB: Read(10) 28 00 00 00 4c 20 00 00 08 00
> > [    4.637005] blk_update_request: I/O error, dev sda, sector 19488 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.637049] sd 0:2:0:0: [sda] tag#262 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.637085] sd 0:2:0:0: [sda] tag#262 CDB: Read(10) 28 00 00 00 4c 38 00 00 08 00
> > [    4.637118] blk_update_request: I/O error, dev sda, sector 19512 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.637161] sd 0:2:0:0: [sda] tag#264 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.637197] sd 0:2:0:0: [sda] tag#264 CDB: Read(10) 28 00 00 00 4c 48 00 00 08 00
> > [    4.637221] blk_update_request: I/O error, dev sda, sector 19528 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.637270] sd 0:2:0:0: [sda] tag#284 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.637306] sd 0:2:0:0: [sda] tag#284 CDB: Read(10) 28 00 00 00 4c e8 00 00 08 00
> > [    4.637332] blk_update_request: I/O error, dev sda, sector 19688 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.637375] sd 0:2:0:0: [sda] tag#286 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.637411] sd 0:2:0:0: [sda] tag#286 CDB: Read(10) 28 00 00 00 4c f8 00 00 08 00
> > [    4.637444] blk_update_request: I/O error, dev sda, sector 19704 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.637481] blk_update_request: I/O error, dev sda, sector 19664 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.637485] sd 0:2:0:0: [sda] tag#282 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.637487] sd 0:2:0:0: [sda] tag#287 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
> > [    4.637491] sd 0:2:0:0: [sda] tag#287 CDB: Read(10) 28 00 00 00 4d 00 00 00 08 00
> > [    4.637491] sd 0:2:0:0: [sda] tag#282 CDB: Read(10) 28 00 00 00 4c d8 00 00 08 00
> > [    4.637494] blk_update_request: I/O error, dev sda, sector 19672 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
> > [    4.747771] EXT4-fs (sda2): mounted filesystem with ordered data mode. Opts: (null). Quota mode: none.
> > 
> > If I reboot multiple times fsck requires a manual fix and I get dropped
> > to the initramfs shell. Some times the filesystem gets corrupted and I
> > need to redeploy the box.
> > 
> > If I use the same kernel with this commit reverted I can reboot as many
> > times as I want without any failure:
> > 
> >   813c6871f76b ("scsi: core: Reallocate device's budget map on queue depth change")
> 
> I would not have thought that this causes possible corruption.
> 
> > 
> > For now I've just reverted the commit, but I'll try to add some
> > debugging and collect more info.
> > 
> > Let me know if there's any specific test that you want me to try.
> > 
> 
> Please try this:
> https://lore.kernel.org/linux-scsi/yq1ee2kumrh.fsf@ca-mkp.ca.oracle.com/T/#t
> 
> It never made 5.17, which I would have hoped for.

Thanks John! It looks like this one is actually fixing the problem.
I rebooted multiple times and I didn't get any I/O error or corruption.

If you want you can add my:

Tested-by: Andrea Righi <andrea.righi@canonical.com>