On 2018-05-02 20:24, Eric Blake wrote: > On 04/21/2018 11:54 AM, Max Reitz wrote: >> This test case has been broken since 398e6ad014df261d (roughly half a >> year). qemu-img amend requires its output image to be R/W, so it opens >> it as such; the node is then turned into an read-only node automatically >> which is now accompanied by a warning, however. This warning has not >> been part of the reference output. >> > >> The reason nobody has noticed the breakage until now of course is the >> fact that nobody runs the iotests for nbd+bochs. There actually was >> never any reason to set the protocol to "nbd" but because that was >> technically correct; functionally it made no difference. So that is the >> first thing we are going to change: Make the protocol "file" instead so >> that people might actually notice breakage here. > > There's probably several tests that fail with an explicit pairing of nbd > as protocol + explicit format (I'm guilty of usually testing either the > format './check -qcow2' or the protocol './check -nbd', and not > combining the two). > >> So with this test now being a raw test, we can rework the amend test >> case to use raw instead. >> >> Signed-off-by: Max Reitz <mre...@redhat.com> >> --- >> tests/qemu-iotests/113 | 19 +++++++++---------- >> tests/qemu-iotests/113.out | 7 ++++--- >> 2 files changed, 13 insertions(+), 13 deletions(-) >> >> diff --git a/tests/qemu-iotests/113 b/tests/qemu-iotests/113 >> index 19b68b2727..4e09810905 100755 >> --- a/tests/qemu-iotests/113 >> +++ b/tests/qemu-iotests/113 >> @@ -38,16 +38,17 @@ trap "_cleanup; exit \$status" 0 1 2 3 15 >> . ./common.rc >> . ./common.filter >> -# We can only test one format here because we need its sample file >> -_supported_fmt bochs >> -_supported_proto nbd >> +# Some of these test cases use bochs, but others do use raw, so this >> +# is only half a lie. >> +_supported_fmt raw >> +_supported_proto file > > Naive question - can we pass a list to _supported_fmt/_supported_proto, > in which case the test runs under both elements of the list? That is, > I'm wondering if: > > _supported_fmt raw bochs > _supported_proto file nbd > > gives us any additional coverage (theoretically 4 times the testing > possibilities - but if the rest of the test hard-codes a particular > format or protocol then it might not actually execute a different > combination).
Well, anyone who's going to run any of these combinations is probably going to run raw+file as well. You could argue that if someone changes something about NBD, they might just run raw+nbd; but then again, this test doesn't really test nbd at all. In fact, it just tests qemu-img (we simply need a format that does not support image creation, but it's not like we actually do anything with that format). Same for bochs. (And I think anyone who wants to test qemu-img is indeed going to run the raw+file tests.) But see my reply to John, I think the best way would be if tests like these could be just run in "default" mode without forcing some format/protocol combination on them. Max
signature.asc
Description: OpenPGP digital signature
