Backport a patch to make some tests pass on Fedora Rawhide

f38
Timothy Redaelli 7 years ago
parent 799caffd96
commit 143fb14264

@ -64,6 +64,7 @@ ExclusiveArch: x86_64 aarch64 ppc64le s390x
# ovs-patches # ovs-patches
# OVS (including OVN) backports (0 - 300) # OVS (including OVN) backports (0 - 300)
Patch0: ovs-dev-ofproto-macros-Ignore-Dropped-log-messages-in-check_logs..patch
Patch10: 0001-ofproto-dpif-Delete-system-tunnel-interface-when-rem.patch Patch10: 0001-ofproto-dpif-Delete-system-tunnel-interface-when-rem.patch
Patch41: 0002-netdev-tc-offloads-Add-support-for-IP-fragmentation.patch Patch41: 0002-netdev-tc-offloads-Add-support-for-IP-fragmentation.patch
@ -323,7 +324,7 @@ rm -f $RPM_BUILD_ROOT/%{_bindir}/ovn-docker-overlay-driver \
%check %check
%if %{with check} %if %{with check}
if make check; then :; if make check TESTSUITEFLAGS='%{_smp_mflags}' RECHECK=yes; then :;
else else
cat tests/testsuite.log cat tests/testsuite.log
exit 1 exit 1
@ -629,7 +630,7 @@ chown -R openvswitch:openvswitch /etc/openvswitch
%changelog %changelog
* Tue Jun 19 2018 Timothy Redaelli <tredaelli@redhat.com> - 2.9.2-1 * Tue Jun 19 2018 Timothy Redaelli <tredaelli@redhat.com> - 2.9.2-1
- Update to OVS 2.9.2 - Update to OVS 2.9.2
- Do not run "make check" simultaneously. - Backport a patch to make some tests pass on Fedora Rawhide
* Tue Jun 19 2018 Miro Hrončok <mhroncok@redhat.com> - 2.9.1-2 * Tue Jun 19 2018 Miro Hrončok <mhroncok@redhat.com> - 2.9.1-2
- Rebuilt for Python 3.7 - Rebuilt for Python 3.7

@ -0,0 +1,54 @@
From patchwork Tue Jul 3 18:32:18 2018
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: [ovs-dev] ofproto-macros: Ignore "Dropped # log messages" in
check_logs.
X-Patchwork-Submitter: Ben Pfaff <blp@ovn.org>
X-Patchwork-Id: 938851
Message-Id: <20180703183218.32329-1-blp@ovn.org>
To: dev@openvswitch.org
Cc: Ben Pfaff <blp@ovn.org>
Date: Tue, 3 Jul 2018 11:32:18 -0700
From: Ben Pfaff <blp@ovn.org>
List-Id: <ovs-dev.openvswitch.org>
check_logs ignores some log messages, but it wasn't smart enough to ignore
the messages that said that the ignored messages had been rate-limited.
This fixes the problem.
It's OK to ignore all rate-limiting messages because they only appear if at
least one message was not rate-limited, which check_logs will catch anyway.
Reported-by: Timothy Redaelli <tredaelli@redhat.com>
Reported-at: https://mail.openvswitch.org/pipermail/ovs-discuss/2018-July/046978.html
Signed-off-by: Ben Pfaff <blp@ovn.org>
Tested-By: Timothy Redaelli <tredaelli@redhat.com>
---
tests/ofproto-macros.at | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/tests/ofproto-macros.at b/tests/ofproto-macros.at
index 7388a20a2236..2a56ae6e2f3e 100644
--- a/tests/ofproto-macros.at
+++ b/tests/ofproto-macros.at
@@ -400,6 +400,11 @@ check_logs () {
# all "connection reset" warning logs for similar reasons (either EPIPE or
# ECONNRESET can be returned on a send depending on whether the peer had
# unconsumed data when it closed the socket).
+ #
+ # We also ignore "Dropped # log messages..." messages. Otherwise, even if
+ # we ignore the messages that were rate-limited, we can end up failing just
+ # because of the announcement that rate-limiting happened (and in a racy,
+ # timing-dependent way, too).
sed -n "$1
/reset by peer/d
/Broken pipe/d
@@ -408,6 +413,7 @@ check_logs () {
/timeval.*disk: [[0-9]]* reads, [[0-9]]* writes/d
/timeval.*context switches: [[0-9]]* voluntary, [[0-9]]* involuntary/d
/ovs_rcu.*blocked [[0-9]]* ms waiting for .* to quiesce/d
+/Dropped [[0-9]]* log messages/d
/|WARN|/p
/|ERR|/p
/|EMER|/p" ${logs}
Loading…
Cancel
Save