hi, David,

On Wed, Sep 18, 2024 at 12:27:48PM +0100, David Howells wrote:
> David Howells <dhowells@redhat.com> wrote:
> 
> > Does this:
> > 
> > https://lore.kernel.org/linux-fsdevel/2280667.1726594254@warthog.procyon.org.uk/T/#u
> > 
> > 	[PATCH] cifs: Fix reversion of the iter in cifs_readv_receive()
> > 
> > help?
> 
> Actually, it probably won't.  The issue seems to be one I'm already trying to
> reproduce that Steve has flagged.
> 
> Can you tell me SMB server you're using?  Samba, ksmbd, Windows, Azure?  I'm
> guessing one of the first two.

we actually use local mount to simulate smb. I attached an output for details.

2024-09-11 23:30:58 mkdir -p /cifs/sda1
2024-09-11 23:30:58 timeout 5m mount -t cifs -o vers=2.0 -o user=root,password=pass //localhost/fs/sda1 /cifs/sda1
mount cifs success
2024-09-11 23:30:58 mkdir -p /cifs/sda2
2024-09-11 23:30:58 timeout 5m mount -t cifs -o vers=2.0 -o user=root,password=pass //localhost/fs/sda2 /cifs/sda2
mount cifs success
2024-09-11 23:30:59 mkdir -p /cifs/sda3
2024-09-11 23:30:59 timeout 5m mount -t cifs -o vers=2.0 -o user=root,password=pass //localhost/fs/sda3 /cifs/sda3
mount cifs success
2024-09-11 23:30:59 mkdir -p /cifs/sda4
2024-09-11 23:30:59 timeout 5m mount -t cifs -o vers=2.0 -o user=root,password=pass //localhost/fs/sda4 /cifs/sda4
mount cifs success


2024-09-11 23:31:00 mount /dev/sda1 /fs/sda1
2024-09-11 23:31:01 mkdir -p /smbv2//cifs/sda1
2024-09-11 23:31:01 export FSTYP=cifs
2024-09-11 23:31:01 export TEST_DEV=//localhost/fs/sda1
2024-09-11 23:31:01 export TEST_DIR=/smbv2//cifs/sda1
2024-09-11 23:31:01 export CIFS_MOUNT_OPTIONS=-ousername=root,password=pass,noperm,vers=2.0,mfsymlinks,actimeo=0
2024-09-11 23:31:01 sed "s:^:generic/:" //lkp/benchmarks/xfstests/tests/generic-group-07
2024-09-11 23:31:01 ./check -E tests/cifs/exclude.incompatible-smb2.txt -E tests/cifs/exclude.very-slow.txt generic/071 generic/072 generic/074 generic/075 generic/076 generic/078 generic/079


> 
> Also, will your reproducer really clobber four arbitrary partitions on sdb?

yeah, we setup dedicated hdd for tests on each test machine, e.g. for the
lkp-skl-d05 used in the test, it has:

nr_hdd_partitions: 4
hdd_partitions: /dev/disk/by-id/wwn-0x5000c50091e544de-part*

then in this 4HDD-ext4-smbv2-generic-group-07 test, also as in attached output

2024-09-11 23:26:17 wipefs -a --force /dev/sda1
/dev/sda1: 2 bytes were erased at offset 0x00000438 (ext4): 53 ef
2024-09-11 23:26:17 wipefs -a --force /dev/sda2
2024-09-11 23:26:17 wipefs -a --force /dev/sda3
2024-09-11 23:26:17 wipefs -a --force /dev/sda4
2024-09-11 23:26:17 mkfs -t ext4 -q -E lazy_itable_init=0,lazy_journal_init=0 -F /dev/sda1
2024-09-11 23:26:17 mkfs -t ext4 -q -E lazy_itable_init=0,lazy_journal_init=0 -F /dev/sda3
2024-09-11 23:26:17 mkfs -t ext4 -q -E lazy_itable_init=0,lazy_journal_init=0 -F /dev/sda2
2024-09-11 23:26:17 mkfs -t ext4 -q -E lazy_itable_init=0,lazy_journal_init=0 -F /dev/sda4


I also attached 074.full. KASAN issue occurs while this 074 test
in generic-group-07.

> 
> David
>