On Tue, 2018-09-18 at 17:18 -0700, Omar Sandoval wrote: +AD4 On Tue, Sep 18, 2018 at 05:02:47PM -0700, Bart Van Assche wrote: +AD4 +AD4 On 9/18/18 4:24 PM, Omar Sandoval wrote: +AD4 +AD4 +AD4 On Tue, Sep 18, 2018 at 02:20:59PM -0700, Bart Van Assche wrote: +AD4 +AD4 +AD4 +AD4 Can you have a look at the updated master branch of +AD4 +AD4 +AD4 +AD4 https://github.com/bvanassche/blktests? That code should no longer fail if +AD4 +AD4 +AD4 +AD4 unloading the nvme kernel module fails. Please note that you will need +AD4 +AD4 +AD4 +AD4 kernel v4.18 to test these scripts - a KASAN complaint appears if I run +AD4 +AD4 +AD4 +AD4 these tests against kernel v4.19-rc4. +AD4 +AD4 +AD4 +AD4 +AD4 +AD4 Thanks, these pass now. Is it expected that my nvme device gets a +AD4 +AD4 +AD4 multipath device configured after running these tests? +AD4 +AD4 +AD4 +AD4 +AD4 +AD4 +ACQ lsblk +AD4 +AD4 +AD4 NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT +AD4 +AD4 +AD4 vda 254:0 0 16G 0 disk +AD4 +AD4 +AD4 +JRQlAA-vda1 254:1 0 16G 0 part / +AD4 +AD4 +AD4 vdb 254:16 0 8G 0 disk +AD4 +AD4 +AD4 vdc 254:32 0 8G 0 disk +AD4 +AD4 +AD4 vdd 254:48 0 8G 0 disk +AD4 +AD4 +AD4 nvme0n1 259:0 0 8G 0 disk +AD4 +AD4 +AD4 +JRQlAA-mpatha 253:0 0 8G 0 mpath +AD4 +AD4 +AD4 +AD4 No, all multipath devices that were created during a test should be removed +AD4 +AD4 before that test finishes. I will look into this. +AD4 +AD4 +AD4 +AD4 +AD4 Also, can you please fix: +AD4 +AD4 +AD4 +AD4 +AD4 +AD4 +AF8-have+AF8-kernel+AF8-option NVME+AF8-MULTIPATH +ACYAJg exit 1 +AD4 +AD4 +AD4 +AD4 +AD4 +AD4 to not exit on failure? It should use SKIP+AF8-REASON and return 1. You +AD4 +AD4 +AD4 might need to add something like +AF8-dont+AF8-have+AF8-kernel+AF8-option to properly +AD4 +AD4 +AD4 handle the case where the config is not found. +AD4 +AD4 +AD4 +AD4 OK, I will change this. +AD4 +AD4 +AD4 +AD4 +AD4 Side note which isn't a blocker for merging is that there's a lot of +AD4 +AD4 +AD4 duplicated code between these helpers and the srp helpers. How hard +AD4 +AD4 +AD4 would it be to refactor that? +AD4 +AD4 +AD4 +AD4 Are you perhaps referring to the code that is shared between the +AD4 +AD4 tests/srp/rc tests/nvmeof-mp/rc shell scripts? +AD4 +AD4 Yes, those. +AD4 +AD4 +AD4 The hardest part is probably +AD4 +AD4 to chose a location where to store these functions. Should I create a file +AD4 +AD4 with common code under common/, under tests/srp/, under tests/nvmeof-mp/ or +AD4 +AD4 perhaps somewhere else? +AD4 +AD4 Just put it under common. Hi Omar, All feedback mentioned above has been addressed. The following pull request has been updated: https://github.com/osandov/blktests/pull/33. Please let me know if you want me to post these patches on the linux-block mailing list. Note: neither the upstream kernel v4.18 nor v4.19-rc4 are stable enough to pass all nvmeof-mp tests if kernel debugging options like KASAN are enabled. Additionally, the NVMe device+AF8-add+AF8-disk() race condition often causes multipathd to refuse to consider /dev/nvme... devices. The output on my test setup is as follows (all tests pass): +ACM ./check -q nvmeof-mp nvmeof-mp/001 (Log in and log out) +AFs-passed+AF0 runtime 1.528s ... 1.909s nvmeof-mp/002 (File I/O on top of multipath concurrently with logout and login (mq)) +AFs passed+AF0-time 38.968s ... runtime 38.968s ... 38.571s nvmeof-mp/004 (File I/O on top of multipath concurrently with logout and login (sq-on- nvmeof-mp/004 (File I/O on top of multipath concurrently with logout and login (sq-on- mq)) +AFs-passed+AF0-38.632s ... runtime 38.632s ... 37.529s nvmeof-mp/005 (Direct I/O with large transfer sizes and bs+AD0-4M) +AFs-passed+AF0 runtime 13.382s ... 13.684s nvmeof-mp/006 (Direct I/O with large transfer sizes and bs+AD0-8M) +AFs-passed+AF0 runtime 13.511s ... 13.480s nvmeof-mp/009 (Buffered I/O with large transfer sizes and bs+AD0-4M) +AFs-passed+AF0 runtime 13.665s ... 13.763s nvmeof-mp/010 (Buffered I/O with large transfer sizes and bs+AD0-8M) +AFs-passed+AF0 runtime 13.442s ... 13.900s nvmeof-mp/011 (Block I/O on top of multipath concurrently with logout and login) +AFs-pass ed+AF0 runtime 37.988s ... runtime 37.988s ... 37.945s nvmeof-mp/012 (dm-mpath on top of multiple I/O schedulers) +AFs-passed+AF0 runtime 21.659s ... 21.733s Thanks, Bart.
