Re: [PULL 8/8] migration: Add qtest for migration over RDMA

[Peter Xu <peterx@redhat.com>]

On Mon, Mar 10, 2025 at 08:33:14AM +0000, Zhijian Li (Fujitsu) wrote:
> Hi Stefan,
> 
> Copied to gitlab CI,
> 
> On 08/03/2025 16:42, Stefan Hajnoczi wrote:
> > On Sat, Mar 8, 2025 at 2:01 PM Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
> >>
> >> Hi,
> >>
> >> On 7/3/25 19:15, Fabiano Rosas wrote:
> >>> From: Li Zhijian <lizhijian@fujitsu.com>
> >>>
> >>> This qtest requires there is a RDMA(RoCE) link in the host.
> >>> In order to make the test work smoothly, introduce a
> >>> scripts/rdma-migration-helper.sh to
> >>> - setup a new Soft-RoCE(aka RXE) if it's root
> >>> - detect existing RoCE link
> >>>
> >>> Test will be skipped if there is no available RoCE link.
> >>
> >> Is it? Runing as user I'm getting:
> >>
> >>     RDMA ERROR: RDMA host is not set!
> > 
> > The CI is failing too:
> > https://gitlab.com/qemu-project/qemu/-/jobs/9350004599#L5590
> 
> Thanks for this info, unfortunately, there is no 'testlog.txt' in this gitlab-ci.

It has it.  Try look for "Job artifacts", then there're "Download" or
"Browse" for testlog.txt.  But there isn't much info.

# Start of rdma tests
# Running /aarch64/migration/precopy/rdma/plain
# Using machine type: virt-10.0
# starting QEMU: exec ./qemu-system-aarch64 -qtest unix:/tmp/qtest-1127030.sock -qtest-log /dev/null -chardev socket,path=/tmp/qtest-1127030.qmp,id=char0 -mon chardev=char0,mode=control -display none -audio none -accel kvm -accel tcg -machine virt-10.0,gic-version=3 -name source,debug-threads=on -m 150M  -serial file:/tmp/migration-test-R1OX22/src_serial -cpu max -kernel /tmp/migration-test-R1OX22/bootsect    -accel qtest
# starting QEMU: exec ./qemu-system-aarch64 -qtest unix:/tmp/qtest-1127030.sock -qtest-log /dev/null -chardev socket,path=/tmp/qtest-1127030.qmp,id=char0 -mon chardev=char0,mode=control -display none -audio none -accel kvm -accel tcg -machine virt-10.0,gic-version=3 -name target,debug-threads=on -m 150M  -serial file:/tmp/migration-test-R1OX22/dest_serial -incoming rdma::29200  -cpu max -kernel /tmp/migration-test-R1OX22/bootsect    -accel qtest
----------------------------------- stderr -----------------------------------
qemu-system-aarch64: -incoming rdma::29200: RDMA ERROR: RDMA host is not set!
Broken pipe
../tests/qtest/libqtest.c:199: kill_qemu() tried to terminate QEMU process but encountered exit status 1 (expected 0)

> 
> I learned that x86 runner worked well
> https://gitlab.com/qemu-project/qemu/-/jobs/9350004633
> 
> So I doubt this is aarch64 specific, but I don't have an aarch64 in hand.

I think it means it'll exit 0 even without ipv4 address in the script.  I
doubt whether we used to rely on:

  command -v rdma

But maybe that's available on the reproduced hosts, so it'll pass there.
OTOH, the script should fail the script if no avail ipv4 addr found.

To be explicit, the script does this:

  has_soft_rdma "$i" || return

So even if it failed to see the soft rdma and returned, IIUC
rdma_rxe_setup_detect() will still success.

Maybe it should be this instead?

  has_soft_rdma "$i" || exit -1

We could also sanity check the ipv4 address, e.g.:

  rdma_rxe_setup_detect | grep -Eo '^[0-9]{1,3}(\.[0-9]{1,3}){3}$'

-- 
Peter Xu