On Wed, 27 Mar 2024 13:26:45 +0100 Christian Schoenebeck <qemu_...@crudebyte.com> wrote:
> On Wednesday, March 27, 2024 12:28:17 PM CET Daniel Henrique Barboza wrote: > > On 3/27/24 07:14, Christian Schoenebeck wrote: > > > On Wednesday, March 27, 2024 10:33:27 AM CET Daniel Henrique Barboza > > > wrote: > > >> On 3/27/24 05:47, Christian Schoenebeck wrote: > > >>> On Tuesday, March 26, 2024 6:47:17 PM CET Daniel Henrique Barboza wrote: > > >>>> On 3/26/24 14:05, Greg Kurz wrote: > > >>>>> On Tue, 26 Mar 2024 10:26:04 -0300 > > >>>>> Daniel Henrique Barboza <dbarb...@ventanamicro.com> wrote: > > >>>>> > > >>>>>> The local 9p driver in virtio-9p-test.c its temporary dir right at > > >>>>>> the > > >>>>>> start of qos-test (via virtio_9p_create_local_test_dir()) and only > > >>>>>> deletes it after qos-test is finished (via > > >>>>>> virtio_9p_remove_local_test_dir()). > > >>>>>> > > >>>>>> This means that any qos-test machine that ends up running > > >>>>>> virtio-9p-test local > > >>>>>> tests more than once will end up re-using the same temp dir. This is > > >>>>>> what's happening in [1] after we introduced the riscv machine nodes: > > >>>>>> if > > >>>>>> we enable slow tests with the '-m slow' flag using > > >>>>>> qemu-system-riscv64, > > >>>>>> this is what happens: > > >>>>>> > > >>>>>> - a temp dir is created, e.g. qtest-9p-local-WZLDL2; > > >>>>>> > > >>>>>> - virtio-9p-device tests will run virtio-9p-test successfully; > > >>>>>> > > >>>>>> - virtio-9p-pci tests will run virtio-9p-test, and fail right at the > > >>>>>> first slow test at fs_create_dir() because the "01" file was > > >>>>>> already > > >>>>>> created by fs_create_dir() test when running with the > > >>>>>> virtio-9p-device. > > >>>>>> > > >>>>>> We can fix it by making every test clean up their changes in the > > >>>>>> filesystem after they're done. But we don't need every test either: > > >>>>>> what fs_create_file() does is already exercised in fs_unlinkat_dir(), > > >>>>>> i.e. a dir is created, verified to be created, and then removed. > > >>>>>> Fixing > > >>>>>> fs_create_file() would turn it into fs_unlikat_dir(), so we don't > > >>>>>> need > > >>>>>> both. The same theme follows every test in virtio-9p-test.c, where > > >>>>>> the > > >>>>>> 'unlikat' variant does the same thing the 'create' does but with some > > >>>>>> cleaning in the end. > > >>>>>> > > >>>>>> Consolide some tests as follows: > > >>>>>> > > >>>>>> - fs_create_dir() is removed. fs_unlinkat_dir() is renamed to > > >>>>>> fs_create_unlinkat_dir(); > > >>>>>> > > >>>>>> - fs_create_file() is removed. fs_unlinkat_file() is renamed to > > >>>>>> fs_create_unlinkat_file(). The "04" dir it uses is now being > > >>>>>> removed; > > >>>>>> > > >>>>>> - fs_symlink_file() is removed. fs_unlinkat_symlink() is renamed to > > >>>>>> fs_create_unlinkat_symlink(). Both "real_file" and the "06" dir > > >>>>>> it > > >>>>>> creates is now being removed. > > >>>>>> > > >>>>> > > >>>>> The change looks good functionally but it breaks the legitimate > > >>>>> assumption > > >>>>> that files "06/*" come from test #6 and so on... I think you should > > >>>>> consider > > >>>>> renumbering to avoid confusion when debugging logs. > > >>>>> > > >>>>> Since this will bring more hunks, please split this in enough > > >>>>> reviewable > > >>>>> patches. > > >>>> > > >>>> Fair enough. Let me cook a v2. Thanks, > > >>> > > >>> Wouldn't it be much simpler to just change the name of the temporary > > >>> directory, such that it contains the device name as well? Then these > > >>> tests > > >>> runs would run on independent directories and won't interfere with each > > >>> other > > >>> and that wouldn't need much changes I guess. > > >> > > >> That's true. If we were just trying to fix the issue then I would go > > >> with this > > >> approach since it's simpler. But given that we're also cutting half the > > >> tests while > > >> retaining the coverage I think this approach is worth the extra code. > > > > > > Well, I am actually not so keen into all those changes. These tests were > > > intentionally split, and yes with costs of a bit redundant (test case) > > > code. > > > But they were cleanly build up on each other, from fundamental > > > requirements > > > like whether it is possible to create a directory and file ... and then > > > the > > > subsequent tests would become more and more demanding. > > > > > > That way it was easier to review if somebody reports a test to fail, > > > because > > > you could immediately see whether the preceding fundamental tests > > > succeeded. > > > > The current test design is flawed. It's based on a premise that doesn't > > happen, i.e. > > a new temp dir will be created every time the test suit is executed. In > > reality the > > temp dir is created only once in the constructor of the test, at the start > > of qos-test > > (tests/qtest/qos-test.c, run_one_test()) and removed only once at the > > destructor > > at the end of the run. > > > > It's not possible to add a 'device name' in the created temp dir because > > we're too early > > in the process, the tests didn't start at that point. So, with the current > > temp dir design, > > the tests needs to clean themselves up after each run. > > > > Here's the alternatives I'm willing to go for: > > > > - what I just sent in v2; > > > > - add cleanups in all existing tests. We can keep all of them, but the > > 'create' tests > > will be carbon copies of the 'unlinkat' tests but with different names. Can > > be done; > > > > - if we really want the tests untouched we can rework how the 'temp dir' is > > created/deleted. > > The test dir will be created and removed after each test via the 'before' > > callback. To be > > honest this seems like the best approach we can take, aside from what I did > > in v2, and > > it's on par with how tests like vhost-user-test.c works. > > Yeah, the latter sounds like the best solution to me, too. > +1 > Don't get me wrong, I didn't want to burden you with more work. It's really > just that I think that restructuring all test cases is contra productive. > > If you want I can also look into that. Just let me know. > > Thanks! > > /Christian > > -- Greg