I also think we should probably link samba (or kernel if others prefer) bugzillas for most of the test case failures. Especially those 1) which fail on current smb3 and 2) which don't require posix extensions to implement (e.g. don't require case sensitivity) and 3) which could conceivably should be able to run on a network file system (not failures on tests which require a block device for example) On Tue, May 29, 2018 at 7:56 PM, Ronnie Sahlberg <lsahlber@xxxxxxxxxx> wrote: > All, > > Please have a look at > https://wiki.samba.org/index.php/Xfstesting-cifs > > I have updated it a bit to make it easier to get started with running xfstests for cifs. > Especially for people that are new to either cifs or xfstests. > > > Xfstests with cifs is somewhat challenging because there are a lot of tests that do not pass. > The reasons for this varies and can be anything from things that have not yet been implemented, > things that will never be possible to do over smb2+ (we dont care about smb1 :-) ) due to semantics > to downright bugs that needs to be fixed. > > > To handle this and be able to run clean regression test runs I have added exclusion files to the wiki, and > also command lines for how to download and install them in your xfstest tree. > > For example: > wget https://wiki.samba.org/images/9/9f/Xfstests.exclude.incompatible-smb2.txt -O tests/cifs/exclude.incompatible-smb2 > which is used to download the exclusion file for smb2. > > The benefit of exclusion files is that it allows us to run a clean regression test of only the tests that we know shoud work. > The benefit of having these files on the samba wiki is that it means that we can all start working with the same list of tests > instead of as now where I suspect we all have slightly different private lists. > > Lets make the list of tests we use (or exclude) a canonical set of tests and let it live in a public place like the wiki > so we all work on the same data. > > > My idea is further that we should work on the exclusion files and document WHY each test fails. This can be done easily by adding > a #-comment to the line for the test and then just re-uploading the file. > > I plan to spend some time now and then to triage the failing tests and update the exclusion lists. > Would be awesome if others want to join in too. > > > What do you think? > > > regards > ronnie sahlberg > -- > To unsubscribe from this list: send the line "unsubscribe linux-cifs" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html -- Thanks, Steve -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
