On 01/06/2015 07:33 PM, Huang Ying wrote:
> FYI, we noticed the below changes on
>
> commit f2298c0403b0dfcaef637eba0c02c4a06d7a25ab ("null_blk: multi queue aware block test driver")
>
>
> testbox/testcase/testparams: vm-kbuild-yocto-i386/boot/1
>
> 320ae51feed5c2f1  f2298c0403b0dfcaef637eba0c
> ----------------  --------------------------
>         fail:runs  %reproduction    fail:runs
>             |             |             |
>             :10         100%          10:10    kmsg.VFS:could_not_find_a_valid_V7_on_nullb1
>             :10         100%          10:10    kmsg.VFS:could_not_find_a_valid_V7_on_nullb0
>            1:10         -10%            :10    kmsg.vm86_32:could_not_access_userspace_vm86_info
>
> vm-kbuild-yocto-i386: qemu-system-i386 -enable-kvm
> Memory: 320M
>
> [    1.971241] mtip32xx Version 1.2.6os3
> [    1.971311] blk-mq: CPU -> queue map
> [    1.971312]   CPU 0 -> Queue 0
> [    1.971314]   CPU 1 -> Queue 0
> [    1.971628]  nullb0: unknown partition table
> [    1.971748] blk-mq: CPU -> queue map
> [    1.971749]   CPU 0 -> Queue 0
> [    1.971751]   CPU 1 -> Queue 0
> [    1.972023]  nullb1: unknown partition table
> [    1.972096] null: module loaded
> [    1.972131] ibmasm: IBM ASM Service Processor Driver version 1.0 loaded
>
> I think this is at least confusing for the end user.

Sorry, not sure I follow at all. If you're expecting the null block 
driver to actually store data, then yes, you are going to have a bad 
time. It's a test tool for storage stack development.

-- 
Jens Axboe