passt/test/migrate/iperf3_in6, branch master Plug A Simple Socket Transport test: Fix IPv6 address/prefix mismatch error 2025-11-27T21:20:18+00:00 Yumei Huang yuhuang@redhat.com 2025-11-26T03:21:47+00:00 391c15afedacbc4c5e01bfd05ef73be5ab1770ac For IPv6 addresses within the same scope group, Linux kernel preserves insertion order, where the first inserted address appears last. As a result, the ordering inside pasta differs from the host, causing a few tests to fail when they compare only the first address on each side. Fix this by checking that the guest’s first address matches any of the host’s addresses. Link: https://bugs.passt.top/show_bug.cgi?id=175 Signed-off-by: Yumei Huang <yuhuang@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Stefano Brivio <sbrivio@redhat.com> 
For IPv6 addresses within the same scope group, Linux kernel preserves
insertion order, where the first inserted address appears last. As a
result, the ordering inside pasta differs from the host, causing a few
tests to fail when they compare only the first address on each side.
Fix this by checking that the guest’s first address matches any of the
host’s addresses.

Link: https://bugs.passt.top/show_bug.cgi?id=175
Signed-off-by: Yumei Huang <yuhuang@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
netlink: Don't require address to be global, just not link local 2025-09-30T10:03:39+00:00 David Gibson david@gibson.dropbear.id.au 2025-09-30T05:49:11+00:00 a96a962d7e51582ea6aaf1f9180f1be7956f5d1f nl_addr_get() will only pick up global IPv6 addresses (RT_SCOPE_UNIVERSE). This is because link-local addresses aren't suitable. However site-local addresses (in the rare occassions they're used) would be fine for our purposes. In fact, anything wider than link scope is fine, so update the check to reflect that. Change the logic in the test scripts to match as well. Reported-by: Xun Gu <xugu@redhat.com> Link: https://bugs.passt.top/show_bug.cgi?id=122 Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Stefano Brivio <sbrivio@redhat.com> 
nl_addr_get() will only pick up global IPv6 addresses (RT_SCOPE_UNIVERSE).
This is because link-local addresses aren't suitable.  However site-local
addresses (in the rare occassions they're used) would be fine for our
purposes.  In fact, anything wider than link scope is fine, so update the
check to reflect that.

Change the logic in the test scripts to match as well.

Reported-by: Xun Gu <xugu@redhat.com>
Link: https://bugs.passt.top/show_bug.cgi?id=122
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
test: Add migration tests 2025-02-17T07:29:36+00:00 Stefano Brivio sbrivio@redhat.com 2025-02-13T12:14:14+00:00 a1e48a02ff3550eb7875a7df6726086e9b3a1213 PCAP=1 ./run migrate/bidirectional gives an overview of how the whole thing is working. Add 12 tests in total, checking basic functionality with and without flows in both directions, with and without sockets in half-closed states (both inbound and outbound), migration behaviour under traffic flood, under traffic flood with > 253 flows, and strict checking of sequences under flood with ramp patterns in both directions. These tests need preparation and teardown for each case, as we need to restore the source guest in its own context and pane before we can test again. Eventually, we could consider alternating source and target so that we don't need to restart from scratch every time, but that's beyond the scope of this initial test implementation. Trick: './run migrate/*' runs all the tests with preparation and teardown steps. Co-authored-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Stefano Brivio <sbrivio@redhat.com> 
PCAP=1 ./run migrate/bidirectional gives an overview of how the
whole thing is working.

Add 12 tests in total, checking basic functionality with and without
flows in both directions, with and without sockets in half-closed
states (both inbound and outbound), migration behaviour under traffic
flood, under traffic flood with > 253 flows, and strict checking of
sequences under flood with ramp patterns in both directions.

These tests need preparation and teardown for each case, as we need
to restore the source guest in its own context and pane before we can
test again. Eventually, we could consider alternating source and
target so that we don't need to restart from scratch every time, but
that's beyond the scope of this initial test implementation.

Trick: './run migrate/*' runs all the tests with preparation and
teardown steps.

Co-authored-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>