Description of problem:

CI is flaky because of a panic during TestGatewayAPI/testGatewayAPIDNS    

The error was spotted once at https://prow.ci.openshift.org/view/gs/test-platform-results/pr-logs/pull/openshift_cluster-ingress-operator/1300/pull-ci-openshift-cluster-ingress-operator-master-e2e-aws-operator/1985814081684115456

The error is due to, on cluster-ingress-operator/test/e2e/gateway_api_test.go:570, an attempt to use a shadowed version of "gateway" variable, that is null because of an error that happened.

As a suggestion to fix:

• Make the operation of getting the gateway retry (as we did on other places) to avoid network flakes
• Do not try to read the gateway name from a possible null variable